o

\cbv-~

 

.A.‘rc ‘ v. .n.

a

“why" u

...<-. -IU'

”turn...” :3 H . . LE5} 14.5.?

 

 
 

. . . v
:1 : .5 ME: . V . . .1»,H:vm.r:

 

It?" .125.

 

\If I I I - v
- .

LIBRARIES
MICHIGAN STATE UNIVERSITY
( l EAST LANSING, MICH 48824-1048

53107325?»

This is to certify that the
dissertation entitled

FROM STEPS AND LANES TO THE NASCAR CIRCUIT:

TEACHERS’ RESPONSES TO PROFESSIONAL PAY

presented by

Debbi Cathy Ham's

has been accepted towards fulfillment
of the requirements for the

PhD . degree in Teacher Education

 

 

AMI.

\\Major Professor’s Signature
6 Hal $00<

 

Date

MSU is an Affirmative Action/Equal Opportunity Institution

 

'01-.-_)-—-.-v-.—-- ---'-4-»-

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

Mt“:

 

@GI 1078 (2606

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 cJClRC/DateDue.indd-p.15

FROM STEPS AND LANES TO THE NASCAR CIRCUIT:
TEACHERS’ RESPONSES TO PROFESSIONAL PAY

By

Debbi Cathy Harris

A DISSERTATION
Submitted to
Michigan State University

in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Teacher Education

2005

Abstract

FROM STEPS AND LANES TO THE NASCAR CIRCUIT:
TEACHERS’ RESPONSES TO PROFESSIONAL PAY

By

Debbi Cathy Harris
This study examines the voluntary Minneapolis Professional Pay Plan, and teachers’
responses to it. Alternative compensation systems such as this have been proposed as
mechanisms for improving teacher performance, yet attempts to implement these systems
have rarely been successful. The incentives created by these systems are intended to
align teachers” goals with the goals of their employers, consistent with principal agent
theory. Using the lens provided by this theory, I focus on some of the obstacles and
tradeoffs faced by the Minneapolis Public Schools as they introduce Professional Pay. I
conducted a mixed method study, using both qualitative and quantitative methods and
inductive and deductive inferential processes. Interviews with middle school teachers,
school administrators, and district personnel and a survey of middle school teachers are
the main data sources. Four primary obstacles to the policy’s success emerge from the
data: teachers’ limited understanding of the policy, their unwillingness to surrender their
professional autonomy in order to earn financial rewards, their limited interest in
financial incentives, and their perception that the district is not credibly committed to the
policy. These obstacles are largely ignored in current conceptualizations of principal
agent theory, leading me to propose several elaborations to current theory. These include
a greater focus on the impact credible commitment has on principal and agent behavior,

explicit attention to issues of understanding, relaxation of the excess capacity assumption,

and greater attention to the role professionalism and other non-pecuniary factors play in
decision-making. These elaborations enhance our understanding of the complex ways

teachers respond to incentives.

Copyright by
DEBBI CATHY HARRIS
2005

This work is dedicated to God,
“I can do all things through Christ who strengthens me " (Philippians 4:13),
and to my family for their constant love and support.

Table of Contents

ListofTablesvu
List of Figures ...................................................................................... ix

Chapter One
Dollars, Desires, and Disasters: Analyzing Teacher Incentive Policies .................... 1

Chapter Two
Methods ........................................................................................... 5 1

Chapter Three
A Portrait of Professional Pay ................................................................. 76

Chapter Four

Meaningful Change or a Flash in the Pan?
Credible Commitment and the Minneapolis Professional Pay Plan ..................... 125

Chapter Five
Riding Out the Storm: Teachers’ Responses to Professional Pay... .. . . . . . . . . .. . . . 168

ChapterSix
RevisitingAgencyTheory.....................................................................23O

Appendix]
InstrumentsZSZ

Appendix II
A Rasch Analysis of Three Measures of Teacher Perception:

Administrators, Colleagues, and District283

AppendixIII
MinnesotaZOOl Statute 122a.4l3-415......................................................303

Appendix IV
The 2003-2005 Teacher’s Contract for Minneapolis Public Schools... ... . . . ...... ...308

References315

vi

List of Tables

Table 2.1
Sample Description and Professional Pay Enrollment — Interviews ....................... 62

Table 2.2
Sample Description and Professional Pay Enrollment — Surveys ......................... 64

Table 3.1
Selected Salaries From the 2001-2002 Minneapolis
Steps and Lanes Salary Schedule .............................................................. 79

Table 3.2
Comparing District and Union Literature to the 2003-2005 Teachers’ Contract ........ 99

Table 4.1
Completeness of Information and Perception of Credible Commitment
Among Interviewed Teachers ................................................................. 149

Table4.2
Accuracy of Information and Perception of Credible Commitment
Among Survey Respondents..................................................................150

Table 4.3
Enrollment Status and Perception of Credible Commitment
Among Interviewed Teachers ................................................................. 151

Table4.4
Enrollment Status and Perception of Credible Commitment
Among SurveyRespondents..................................................................154

Table 4.5
Stability of District Policies and Perception of Credible Commitment
Among Interviewed TeacherSISS

Table4.6
Stability of District Policies and Perception of Credible Commitment
Among Survey Respondents..................................................................156

TableS.1
Participation in Rewarded Activmesl74

Table 5.2
Survey Respondents’ Understanding of Plan Components ............................... 200

vii

Table 5.3

Interviewed Teachers’ Reasons For Lack of Active Participation ....................... 221

Table 5.4

Surveyed Teachers’ Reasons For Non-enrollment. .. .

Table AII.l

Perception ofAdministrators Items

Table AII.2

Person Measure Summary for Administrator Items .

Table AII.3

Perception of Colleagues Items ...........................................

Table AII.4

Person Measure Summary for Colleague Items. .. ....

Table AlI.5

PerceptionofDistrictItems

Table AII.6

PersonMeasure Summary forDistrict Items

viii

.....................

...223

...287

....290

291

.....295

...297

....301

List of Figures

Figure AII.l

Scree Plot of Eigenvalues for Administrator Items ......................

Figure A112

Hierarchy ofEndorsability for Administrator Items... ... . . . . .. . . . .. .

Figure AII.3

Scree Plot ofEigenvalues for Colleague Items. .. .............,.........

Figure AII.4

HierarchyofEndorsabilityforColleague Items

Figure AII.5

Scree Plot of Eigenvalues for District Items... ... ... . . . . .. ... .

Figure A116

HierarchyofEndorsabilityforDistrict Items

.................... 288

...289

....292

....294

...299

...300

Chapter One

Dollars, Desires, and Disasters: Analyzing Teacher Incentive Policies

What is the “best” way to pay teachers? Should we continue paying teachers
based only on their experience and education, or does this merely reward mediocrity?
Would it be better to base teachers’ pay on their performance in the classroom or their
students’ learning, or would this undermine cooperation among colleagues and encourage
an unhealthy level of teaching to the test? There are no easy answers to these questions,
and my study is designed to help us understand why. My vehicle for clarifying the
effects of compensation systems on teacher behavior will be an examination of middle
school teachers’ responses to an alternative compensation policy, the Minneapolis
Professional Pay Plan (Professional Pay).

My interests in this topic are threefold. First, I was often frustrated by salary
schedules during my own teaching career, since my salary and that of my colleagues did
not seem to have any relationship to either our effort or our performance. I wondered if
there was a better way. I am certainly not alone in asking this question. There is general
agreement that the uniform salary schedule — the bureaucratic compensation system used
in most schools — is imperfect: it provides few financial incentives for excellence, and the
rewarded characteristics — experience and attainment of advanced degrees — are not well
aligned with the characteristics of high quality teachers (Goldhaber, 2002; Rivkin,
Hanushek, & Kain, 2000). However, while the problems with the uniform salary

schedule are fairly obvious, it is not clear that alternatives will fare any better. There is

still much to learn about the effect that alternative compensation systems have on
teachers’ performance.

My second interest is theoretical in nature. Policy-makers often use an economic
framework, specifically a simplified version of agency theory, when considering
alternative compensation plans. The reasoning works something like this:

Currently, we reward teachers for longevity and for completing advanced degrees

that may not positively affect their teaching and, by extension, their students.

Instead, it makes more sense to reward teachers for outcomes or actions that have

a positive effect on students, such as raising student achievement or

demonstrating high performance in the classroom. This argument is seductive; its

logic appears unassailable.

Unfortunately, alternative compensation policies have rarely lived up to this
promise. The logic used by many policymakers is oversimplified since it fails to
recognize the complex interplay of incentives embedded within any compensation
system, an interplay that has long been recognized within the fields of economics and
public policy. The more sophisticated rendition of agency theory developed by
economists allows us to understand the likely consequences of various compensation
systems.

When I began this research, I believed agency theory would prove a valuable
analytical tool, but I wondered if an economic model could capture all of the dynamics
operating within a school environment. Economic models are often atomistic in their
focus on individuals; agency theory is no exception. While the theory’s focus on the

individual is valuable, teachers do not operate in isolation. Their work lives take place

within institutional and political contexts that may have a powerful affect on teachers’
responses to policy.

In addition, agency theory presumes that actors understand what they need to do
in order to receive rewards and are able to take the necessary action. Thus, problems of
imperfect understanding and limited capacity are largely ignored by the theory. Some
teachers may need opportunities to learn how to improve before they can do so; ignoring
the possibility of limited capacity may hinder effective implementation.

One final assumption that is often made by agency theory is that financial
inducements are a strong motivator. However, other rewards, such as professional
autonomy and improved student learning, may provide teachers with stronger
inducements than the promised financial rewards.

My questioning of these assumptions brings me to my third interest, which is
advancing our empirical understanding of teacher compensation policy. The issue of
teacher compensation frequently appears on the educational policy agenda, for policy-
makers and school administrators often presume that the current system is ineffective and
that alternatives will result in improved teacher performance. Unfortunately, these
alternatives rarely live up to their promise and careful empirical study may help us
understand why this occurs. By exploring one alternative compensation policy —
Professional Pay — using the agency framework, I hope to deepen our understanding of
teacher compensation policy.

At the same time, I hope to use the Minneapolis case to refine and elaborate
agency theory. As discussed above, several agency assumptions may be erroneous within

the school context. By empirically testing both agency theory and its assumptions, this

study becomes part of a cycle whereby theory is developed, tested, refined, and retested,
gaining conceptual sophistication and empirical usefulness over time.
Agency Theory: A Framework for Analysis

Agency theory has its roots in conflict between individuals. As long ago as 1776,
Adam Smith recognized that company managers could not always be expected to work as
hard as stockholders might desire. He noted that “such an unremitting exertion of
vigilance and attention, as cannot long be expected from the directors of a joint stock
company” (Smith, 1776, quoted in Jensen, 1994, p. 460) was required in order for a
company to remain viable in the competitive marketplace.

Early economists recognized the existence of conflicting goals between
employers and employees, but they were largely unable to explain why employment
arrangements existed at all in the real world. Given the problems inherent'in the
employee/employer relationship, we would expect vertical integration of tasks to be
managed through the use of short-term contracts between suppliers (the equivalent of
employees in our discussion) and consumers (employers). Coase’s theory of the firm
(1937) shed light on this dilemma. Coase realized that even imperfect
employee/employer relationships would be preferable to short-term contracts under
certain conditions. The reason is that there are costs involved in open market
transactions. Separate negotiation of each transaction can eat up enormous amounts of
time, both suppliers and consumers may be uncomfortable with the uncertainty involved
in constant renegotiation, and exchanges made outside the firm are generally taxable

while those within the firm are not. When these market costs exceed the costs associated

with the inefficiency of employee/employer relationships, it is economically desirable to
establish those relationships.

Recognition of the rationale for employment helps us understand why
employment exists, but it does not illuminate the underlying conflicts involved. The
development of agency theory helps us understand why conflicts exist whenever two
individuals enter into an employment contract. The underlying problem, understood by
Adam Smith and refined by later economists such as Arrow (1985) and Alchian and
Demsetz (1972), is that employers and employees do not always share the same goals.
An employer may hope that his or her department will produce 50 game boards each day
to complement the 50 sets of game pieces completed in another department, but an
employee may find that level of production tiring and unpleasant. The manager of a
social service agency may want cases handled as expeditiously as possible, but the social
workers he or she employs may place a higher premium on improving their clients’ lives,
even if this proves quite time-consuming. Employment contracts that outline employers’
expectations are used to minimize these conflicts.

Employment contracts attempt to align agents’ goals with those of the principals
who employ them.1 In our case, teachers are the agents and incentive contracts are used
to encourage them to behave in ways consistent with employers’ preferences. Faced with
incentives, agents can reasonably be expected to respond. These responses are

sometimes compatible with the principal’s goals, but often not.

 

’ The agency theory “principal” is the entity with whom the agent enters into an
employment contract. In a traditional public school system, the school principal is not
the contracting entity and would not be considered the principal. Throughout this text,
the word “principal” is used to refer to the contracting entity. School principals will be
referred to as such or as school administrators.

Agency theory is predicated on the idea that people will change their behavior to
increase their chances of receiving rewards or decrease the likelihood of undesirable
outcomes. Parents implicitly understand this when they slap a toddler’s hand as she
reaches towards a stove burner — the child will associate touching the stove with the
uncomfortable slap, changing her behavior so as to avoid both the burner and the slap.

That behavior will change in response to incentives is the key assumption of
agency theory, but it is not the only one. Several other assumptions are made as well.
The first relates to capacity. Agents must either have excess technical or effort capacity.
Technical capacity relates to an individual’s ability to work “smarter.” In order for
people to change their behavior, they must be capable of doing so. If we expect agents to
improve their performance, they must possess the knowledge and skills necessary to do a
better job or the time necessary to learn new skills. If a car salesman with no expertise in
automobile repair is offered a financial incentive to assist in the repair shop when the
sales floor is not busy, he is unlikely to be an asset to the repair shop unless he also
receives training in mechanics.

Effort capacity relates to the employee’s ability to work harder. For example, a
piece rate worker on a factory job often has a limited ability to perform her job more
efficiently. Certain steps must be followed, in a certain order, or the line ceases to
function. In order to improve her performance, the piece rate worker must work faster;
she must increase her effort. The assumption that excess technical or effort capacity
exists, and the implications of its violation, will be discussed in greater detail in Chapters

3 and 5.

Another assumption underlying the theory is that the employment contract is not
subject to multiple interpretations and that agents completely understand it. We assume
that there is a single, monolithic incentive contract and that agents understand what they
are being asked to do and the associated rewards. The gap between this assumption and
the reality of the policy under study will become clear in later chapters.

Most agency theorists have assumed that agents are motivated solely by economic
self-interest. Ethical motivation is largely ignored. Empirical evidence does not support
this assumption. Most employment contracts are simple and incomplete and employee
behavior is largely unmonitored (Stevens & Thevaranjan, 2003). This state of affairs
begins to make sense when ethical motivation is incorporated into the theory (Arrow,
1985). There is evidence that agents will weigh their preference for greater reward/effort
ratios against a preference for providing a “fair” or reasonable effort for the payment they
receive (Stevens & Thevaranjan, 2003). In other words, shirking has a cost to the agent if
it makes the agent feel that she has behaved unethically. In certain professions, agents
may also begin to internalize social norms that prohibit shirking. The physician/patient
relationship has been presented as an example of this (Arrow, 1985). Patients do not
require their physicians to sign complex contracts before agreeing to pay for their care,
yet patients are generally satisfied with the results and physicians rarely engage in grossly
negligent behavior. This lack of shirking on the part of most physicians may be at least
partially due to ethical norms instilled in them as part of their professional training. We
will see in Chapter 5 that teachers may be another group whose work levels are often

motivated by ethical, as well as economic, considerations.

Finally, we assume that the policy will continue in its present form into the
foreseeable future and that promised rewards will be paid. When this is the case, credible
commitment exists: the agent can trust the principal to appropriately reward her changes
in behavior and, in the case of long-term rewards, that these rewards will persist for the
life of the employment contract. Furthermore, under our assumption of complete
understanding, we assume that agents are aware of that fact. The reasonableness of the
credible commitment assumption has been explicitly questioned in the scholarly literature
related to agency theory. This literature, and its implications for Professional Pay, will be
examined in more detail in Chapter 4.

Even when the above assumptions are met, it is still difficult to design an
employment contract that perfectly aligns agents’ goals with those of the principal. This
dilemma is largely due to the nature of most jobs and associated control problems. Two
types of control problems — hidden action and hidden information — make it very difficult
for a principal to design an incentive system free of perverse incentives for undesirable
agent responses.

The problem of hidden action arises when the principal cannot accurately
ascertain the agent’s effort level. This arises when the principal cannot directly observe
the agent’s actions and when the effect of those actions is partially dependent on factors
beyond the agent’s control (Arrow, 1985). These conditions are both met in many
employment situations. It is usually impractical for the principal to monitor what the
agent is doing at all times. If the principal is constantly looking over the agent’s
shoulder, it tends to defeat the purpose of having agents in the first place. This occurs at

all levels of the system. Stockholders cannot observe what the CEO of a company is

doing at all times and the manager of a restaurant cannot simultaneously observe the
cashier, line cooks and sous chefs, wait and bus staff, and dishroom workers. This
problem also occurs when principals (or their designated monitors — often the school
principal) attempt to observe teachers’ actions. Teaching is often work done in relative
isolation. At best, the school principal or other monitor will only be able to observe a
small portion of the teacher’s actions since a great deal of teaching takes place behind
classroom doors, with no adult observers present.

This lack of observation would not be a problem if outcomes were completely
dependent on the agent’s effort but, as with our first condition, this is virtually never the
case. Output can rarely be considered a perfect substitute for effort because there is
almost always a random element to output. The effort of a corporate CEO is not
perfectly correlated with stock price, the outcome of interest to shareholders. Changing
stock market conditions, consumer demand, and regulatory environments all affect stock
price and are beyond the control of the CEO. An unpleasant customer may complain
about the performance of an exemplary waiter and a soufflé may fail to rise in humid
weather despite the chef’s best efforts.

Principals who wish to correlate teachers’ efforts and outcomes face a similar
problem. Even if we are willing to assume that student achievement is the critical
outcome of interest, teachers have limited control over that achievement. Some portion
of student achievement can be attributed to an individual teacher’s effort but a large part
of achievement may be determined by factors as varied as the student’s socioeconomic

status and prior learning. When students do not achieve, it is difficult for the principal to

be certain that the problem has to do with poor teacher effort rather than these other
factors.

It is also difficult to ascertain an individual agent’s effort level when outcomes are
the result of group effort. A student’s ability to achieve is the result of the efforts of all
the student’s past and present teachers. It is very difficult to assign credit or blame to a
particular teacher in these situations. This problem is recognized within the field, and
scholars and policy-makers have attempted to design and implement measurement
systems that isolate the contributions individual teachers make to student learning (Lee &
Weimer, 2002; Sanders & Horn, 1998; Webster & Menro, 1997). Unfortunately, these
systems are far from perfect. There is evidence that even the most sophisticated attempts
to isolate teachers’ contributions to student achievement suffer from measurement errors
due to score volatility (Kane & Staiger, 2002; Rogosa, 1999) and a lack of comparability
across grade levels (Martineau & Plank, 2004).

The problem of hidden action makes it nearly impossible to design an incentive
contract that perfectly aligns the agent’s goals with those of the principal and the problem
of hidden information only serves to exacerbate the problem. Hidden information relates
to the agent’s incentive to misrepresent or withhold information from the agent if doing
so may benefit the agent. Production workers may restrict their output and then attempt
to convince management that they are working at capacity. In one classic study, a
coworker described a colleague, Ray, as “one of the real heroes of the department”
because, when the time-study man arrived to set Ray’s production quota, “he’d jump
around the machine like a monkey on a string, with the sweat just pouring off him. . .I

never did see Ray sweat a drop when he was actually running a job; he was always about

10

40 pounds overweight, the laziest guy I ever did see” (Whyte, 1955, p. 17). Ray worked
very hard to convince management that he simply couldn’t work any harder than he
already was. The dishroom staff may attempt to convince management that they need
additional staff to handle the onslaught of dishes that appear at the lunch hour not
because they are incapable of managing the additional load but because doing so is
unpleasant. Teachers may fail to reveal the full extent to which they can raise student
achievement in an effort to avoid burdensome performance expectations.

Principals can avoid some, but not all, of the problems of hidden action and
hidden information through monitoring. The downfall to this response is that monitoring
is an expensive activity. The principal must balance the desire to obtain information
about agents with the cost of doing so.

Control problems and monitoring costs will affect various teacher incentive
systems in different ways. The remainder of this chapter will use agency theory as a
framework for analyzing incentive systems. Teacher compensation systems will be
categorized in several different ways and the implications of placement within each
category will be analyzed. Theoretical expectations about the likely effects of the current
bureaucratic teacher compensation system, performance (behavior)—based teacher
compensation systems, and outcome-based teacher compensation systems will be
discussed. Relevant empirical work will be noted as well. Following that, the
implications of programs that are piece rate will be compared to those that are

tournaments, and individual and group reward systems will be compared.

11

A Brief History of Teacher Merit Pay

Beginning in 1862, England established a system of teacher compensation known
as “payment by results.” Under this system, teacher salaries were dependent on student
attendance and on the number of students passing examinations. This would be typical of
merit pay programs for the next hundred years or so. They generally linked teacher
salaries to student outcomes in a relatively straightforward manner and little attempt was
made to consider differences among student populations when calculating teacher
rewards. Opinions about payment by results varied but many criticisms voiced in the
18005 echo current concerns about merit pay for teachers. Criticisms included concerns
that it was impossible to capture everything that matters about teaching with a test, that
test results were unstable from year to year due to random fluctuations unrelated to
teachers’ efforts, that tests and testing conditions were not always consistent, that merit
pay encouraged teachers to focus on certain students and certain material, that teachers
had an incentive to cheat, and that teachers tried to move to schools with wealthier
students since these students are more likely to pass examinations regardless of the
quality of instruction received (Coltham, 1972; Rapple, 1992). There was little evidence
that the program had its intended effect: encouraging teachers to perform at higher levels.
In fact, some evidence suggests that the program encouraged teachers to behave in ways
that were inconsistent with quality teaching. One teacher even wrote that, “When one of
my backward boys died of bronchitis a few weeks back I felt a measure of relief; for his
death would make one failure less” (Coltham, 1972, p. 26).

Despite these criticisms, payment by results persisted for thirty-five years. The

program was eventually dropped, however, due to a combination of factors. One of these

12

was dissatisfaction with the program on the part of school inspectors. As the years
passed and the inspectors responsible for program implementation retired, they often
voiced their criticisms. One inspector described payment by results as “an ingenious
instrument for arresting the mental growth of the child” (Coltham, 1972, p. 24), while
another complained that one byproduct of payment by results was that “Children are
passing out of the state-aided schools of to-day by thousands without having gained a
love of learning, in fact with a positive dislike of acquiring knowledge” (Rapple, 1992, p.
307). Other factors contributing to the policy’s demise were that teachers, a group
largely opposed to the program, were beginning to organize and gain power over their
working conditions, and that the curriculum and pedagogy rewarded by the program were
becoming unpopular with both educational leaders and the public.

Attempts to link teacher pay and performance were launched in the United States
early in the twentieth century. During the progressive era, there was great interest among
educational administrators in tying teacher salaries to their performance, consistent with
the focus on efficiency that dominated both the public and private sectors during that
time. Ellwood Cubberly and other leading educators advocated the use of merit pay
plans on the grounds that they would increase system efficiency and help attract and
retain the best teachers (S. M. Johnson, 1984). As was true of the Victorian period, these
plans generally tied teacher pay to student performance. By 1918, almost half of
surveyed U.S. school districts had instituted some form of merit pay for teachers (S. M.
Johnson, 1986) but these programs were usually short-lived. Plans were often
contentious, pitting teachers, administrators, civic associations, and organized labor

against each other. An attempt in 1917 to introduce merit pay in St. Paul, Minnesota

13

caused such discord that it “polarized the city” (Thomas & Moran, 1992). The district
superintendent was an enthusiastic supporter of merit pay and the issue was covered
extensively by the media. Both the local teachers union and teachers themselves
generally opposed the idea while many local school administrators expressed support.
The Housewives League was opposed but 18 out of 22 women’s civics groups favored
the idea. At the height of the controversy, when 14 teachers and principals had just been
fired because of their vocal opposition to merit pay, the superintendent resigned to accept
3 hi gher-paying position in Buffalo and the idea was abandoned.

St. Paul was not the only district that abandoned the idea of merit pay. By 1928,
enthusiasm had abated, with fewer than 20% of US. districts reporting using merit pay
(Mumane & Cohen, 1986). In 1939 one scholar noted that districts were “discarding all
attempts to reward teachers according to subjective ratings of efficiency” (Elsbree, p.
449). Elsbree explained to his contemporaries that, “the harmful effects on teacher
morale of applying rating devices appear to more than offset the good that is
derived,. . .since rating tends to accentuate rivalry it is, therefore, regarded as a destructive
device to use in rewarding teachers, [and]. . .the total contribution of an individual teacher
to the development of a particularly child cannot be measured accurately” (Elsbree,
1939, p. 450). Again, the criticisms of merit pay he outlined resonate with contemporary
concerns: teacher disapproval, destruction of cooperative spirit, and the difficulty of
developing and implementing valid and reliable measures of performance.

Interest in merit pay was rekindled in 1957 when Sputnik’s launching revived
concerns about the quality of America’s schools. At that time, roughly 10% of districts

began using merit pay but the majority of programs survived five or fewer years

14

(Murnane & Cohen, 1986). Only one-third of surveyed school districts that had a merit
pay plan in 1959 still had the plan ten years later (Porwoll, 1979).

Interest in merit pay revived again in the 19803. The frightening portrait of
American education portrayed in A Nation at Risk left policy makers searching for ways
to improve schools. President Reagan suggested that, “Teachers should be paid and
promoted on the basis of their merit and competence. Hard-eamed tax dollars should
encourage the best. They have no business rewarding incompetence and mediocrity” (S.
M. Johnson, 1984, p. 175), and by 1985, twenty five states had mandated the
development of incentive pay programs for teachers (Frase, 1992). For example,
Houston implemented the Second Mile Plan which provided teachers with financial
incentives for entering teaching, as well as for high student test scores, low absenteeism,
and teaching in geographic or subject shortage areas. During the first two years of the
program, evaluators found conflicting results: measured factors - mainly student test
scores and teacher attendance —- had improved but teachers generally reported that the
Second Mile Plan had not affected their behavior (L. M. Miller & Say, 1982). By the
early 19905, many programs, including the Houston plan, had disappeared.

Merit pay programs tend to be short—lived. A 1978 Education Research Service
survey found that 183 school districts had experimented with merit pay plans for an
average of six years before abandoning their programs (Frase, 1992) and that one-third of
the plans had survived two years or less (Guernsey, 1986). The most common reasons
for discontinuation were problems in conducting evaluations, administrative difficulties,
teacher resistance, inadequate funds, and inadequate measurement instruments (Calhoun

& Protheroe, 1983). When a similar survey was administered in 1983, the reasons given

15

for plan discontinuation were similar (Calhoun & Protheroe, 1983). A more recent study
found that 75 percent of the merit plans a group of researchers studied in 1983 had been
discontinued by 1993 (Hatry, Greiner, & Ashford, 1994).

In districts where merit pay plans persist, the plans have often evolved to become
merit pay programs in name only. They change so that extra pay is tied to tasks outside
classroom instruction, extra pay is awarded to virtually everyone, participation is
voluntary with only a few teachers quietly participating, and teachers participate in plan
design in meaningful ways (Mumane & Cohen, 1986). With the exception of teacher
participation, these characteristics allow merit plans to persist by changing them into
something other than merit pay plans.

Recent Teacher Incentive Programs

Merit pay has again attracted the attention of both policy makers and academics,
this time in the form of both performance-based and outcome-based pay. Earlier
programs, as well as some of the more recent ones, tended to be outcome-based. In
outcome-based programs, teachers receive additional pay for increased student
performance. These outcomes are usually weighted heavily towards student test scores
but other measures, such as attendance and school leaving, may be included as well.
Performance-based pay programs are those in which teachers are paid for demonstrated
behaviors, such as the use of a specific pedagogical technique. In some of these
programs, teachers are also rewarded for attending specific professional development
opportunities.

The Consortium for Policy Research in Education (CPRE) teacher compensation

group, headed by Allan Odden, has led recent efforts to educate policy makers about

16

alternatives to the uniform salary schedule. Odden and his colleagues have been
particularly enthusiastic about the potential of performance-based pay to improve
teaching, and CPRE has provided the practitioner community with both hands-on
assistance in developing plans and with the theoretically-supported rationale for doing so
(see Odden & Kelley, 1997). Several districts, including Douglas County, CO,
Steamboat Springs, CO, and Minneapolis, MN (the district that is the focus of this study)
have adopted or are in the process of phasing in merit pay plans that have been heavily
influenced by the CPRE work. In addition, CPRE is conducting ongoing research on
several merit pay plans to document their effects. This work promises to both enrich our
scholarly understanding of alternative compensation and provide policy-makers with
information that will assist them in designing future compensation programs.

Despite the recent attention given to performance-based pay, much of the
published empirical work on merit pay focuses on outcome-based pay. The reasons for
this are practical in nature. Several outcome-based plans are stable, having been around
for a while. In contrast, most performance-based programs are relatively new, so there
has not been time to ascertain and document their impact. Hybrid plans that include
elements of both performance- and outcome-based compensation are also being used in
some districts but, as with performance-based plans, these programs are too recent to be
well represented in the current body of empirical research literature.

South Carolina initiated an outcome-based incentive program in 1984. The
program rewarded schools that met criteria for student achievement (test score gains) and
teacher and student attendance. The rewards were not distributed to teachers; they were

used to supplement school funds, so the program is not technically a merit pay program.

17

It is still worth studying, however, since schools typically used their reward money to
purchase items such as teaching supplies and room air conditioners that teachers were
enthusiastic about receiving. When Richards and Sheu (1992) studied the South Carolina
program, they were particularly interested in the distribution of rewards and the effect
that different measurement systems would have had on award distribution. They found
that South Carolina’s system tended to reward wealthier schools despite the fact that
schools were grouped into bands in an attempt to prevent that from happening. When
they simulated several alternatives to the existing banding program, they found that while
the favored groups changed, each alternative favored one type of school over another.
While increasingly sophisticated measurement techniques are constantly being developed
to avoid these problems, it will be very difficult to tell when bias has been completely
eliminated since teachers and students are not randomly distributed throughout the
system.

When researchers study outcome-based programs, they often find that programs
have a positive effect on rewarded outcomes. Whether this is unequivocally good news,
however, is not entirely clear. For example, improved student test scores are only a cause
for celebration if they reflect increased student learning, not if they reflect increased
gaming of the reward system by teachers. The Dallas Independent School District began
a program to reward teachers for student performance in 1991. This group incentive
program rewarded all of the teachers in a school for the gains made by the school’s
students. Gains were calculated using a sophisticated value-added methodology that
attempts to remove the effects of socioeconomic status and other factors outside of the

teacher’s and school’s control (Webster & Menro, 1997). Ladd (1999) analyzed the

18

impact of this outcome-based compensation system on student achievement and found
that Dallas students did seem to be improving relative to the children in comparable
Texas districts. While this is encouraging, the author cautions that the sources of test
score improvements were not investigated in her study. It is unclear whether the
improvement actually reflects increased student learning and whether there were any
undesirable consequences of the program outside of its favorable impact on test scores.

Eberts, Hollenbeck, and Stone (2002) found that improvement on measured
outcome characteristics did not necessarily reflect improved student learning. When the
teachers in a Michigan alternative high school were offered a bonus for each student who
completed their courses, the student course completion rate did improve. If the
underlying purpose of the program had been course completion, that would have been
fine, but the true goal was improved student learning. During the same period that course
completion was improving, student attendance in class was actually declining. Students
were being convinced not to drop classes (the rewarded measure) but they were not being
convinced to come to class. Also, there was anecdotal evidence that teachers were
diluting their coursework expectations to encourage students to remain enrolled in their
classes.

Much recent research on merit pay has its intellectual roots in expectancy theory,
a theory of motivation based on the premise that three conditions — expectancy,
instrumentality, and valence — must be met to motivate individuals to change their
behavior. The first condition, expectancy, is based on the employee’s expectation that a
change in his or her effort will affect measured performance. Let us suppose that a

salesman will receive a cruise for every 20 cars he sells. Expectancy is high if the

19

salesman believes that his selling effort, rather than random luck or some other factor,
will influence his sales and that he is capable of selling 20 cars within the designated
period. For instrumentality to be high, the employee must feel that a change in measured
performance will result in the receipt of the incentive. Our salesman must believe that
the dealership will actually provide the cruise and not find an excuse to wiggle out of the
obligation. Valence refers to an incentive’s attractiveness. Valence is not limited to the
stated reward; consequences ranging from pride in a job well done to increased
exhaustion will all affect valence. Our salesman must find the idea of going on a cruise
attractive. He might find the idea of being seen as a “winner” attractive as well but might
not be looking forward to working longer hours in order to sell more cars.

As is true with any framework, the use of expectancy theory tends to highlight
certain aspects of a merit pay program — its effect on motivation in this case - while
providing us with fewer insights about others. Expectancy theory has its disciplinary
roots in human resources — an area that has traditionally focused on issues of motivation.
Agency theory, the framework that will be applied in my analysis, is less helpful than
expectancy theory at unpacking the motivational aspects of merit pay but, as will be
demonstrated in later sections, it is quite helpful when we are interested in unpacking the
perverse incentives and unintended consequences embedded in virtually every pay plan.
Agency theory’s roots are in economics, a field that has tended to focus less on broad
aspects of motivation and more on self-interested behavior by individuals who seek to
maximize their economic utility. Each of the two strands of research, with their different
foci, will contribute to our overall understanding of the effect compensation has on

behavior.

20

Several published studies using expectancy theory have looked at outcome-based
group incentives. In these programs, a school’s teachers receive additional pay based on
aggregated school-wide student test scores. One interesting finding that emerged in
studies of the Kentucky and Charlotte-Mecklenberg, NC plans is that teachers said they
were motivated to improve by the promise of increased student achievement, the positive
recognition the program provided, or by the fear of sanctions but that the bonus money
itself was not particularly motivating (Heneman, 1998; Heneman & Milanowski, 1999;
Kelley, 1999). When surveyed, teachers tended to respond that they would rather have
rewards allocated to school improvement than teacher bonuses (Kelley, 1999). This is
consistent with claims that bonus money was not particularly motivating but inconsistent
with teacher’s actions when given control over reward allocations. Under the Kentucky
program, each school’s teachers voted on whether to earmark rewards for school
improvement or for teacher bonuses. The teachers in 98 percent of rewarded schools
voted to earmark at least some of the funds for teacher bonuses (Kelley, 1999). The
reasons for this apparently contradictory behavior are not entirely clear but it is possible
that teachers want to present themselves as altruistic and immune to the lure of more
money but, when faced with a concrete choice, they reveal their positively valent
attitudes towards increased pay.

Surveyed teachers worried about their ability to affect student achievement in
ways that would be captured by the assessment system. While they felt that it was
generally appropriate to hold them accountable for student achievement goals, they
expressed concern about being penalized for outside factors, such as lack of parental

support, that were beyond their control (Heneman, 1998). Over half of teachers felt that

21

hard work would not be enough to allow them to earn rewards (Kelley, Heneman, &
Milanowski, 2002). They worried that they did not have the skills, particularly
pedagogical skills, to math rewarded levels (Heneman, 1998; Kelley & Protsik, 1997)
and that program administrators would continuously raise expectations over time. Even
in schools that consistently received rewards in the past, roughly one-third of teachers
were unsure whether cooperative hard work would allow them to receive future rewards
(Milanowski, 1999). Teachers were also concerned about whether the established goals
for individual schools were fair (Kelley & Finnigan, 2003).

Teachers say that incentive programs motivate them to work harder (Heneman,
1998). They spend more time on tasks related to teaching (Kelley, 1999) and take steps
to align their teaching with the recommended curriculum (Kelley & Protsik, 1997). One
problem with the program’s success in focusing teacher effort is that it may narrow the
breadth of these efforts. Some Kentucky teachers explained that they were concentrating
on higher order thinking skills (which were heavily weighted in the assessment program)
to the exclusion of basic skills and that they were deflecting thoughtful student questions
so that they could cover all of the tested material (Kelley, 1999).

Many teachers complain that incentive systems increase stress. Part of this
increased stress is the result of increased effort levels (Heneman, 1998; Heneman &
Milanowski, 1999; Kelley, 1999; Kelley etal., 2002); teachers are working longer and
harder. Another source of stress is the sanctions embedded in many programs (Kelley et
al., 2002). While these can be direct sanctions, such as the reconstitution of schools that

consistently fail to meet expectations, indirect sanctions such as the negative publicity

22

that results from the failure to appear of the list of rewarded schools cause a great deal of
stress as well.

Current programs may lead to improved student learning since teachers report
working harder to earn incentives but none of the studied programs can be declared
unqualified successes. Increased teacher effort holds some potential for improving
student learning but the increased stress that accompanies greater effort is troubling.
Teachers may also lack the capacity, particularly the pedagogical skills, to make their
increased efforts effective. In addition, there is some evidence that program designers
receive exactly what they ask for. This is desirable when teachers keep broad program
goals in mind while working to improve performance on measured indicators, but there is
some evidence that gaming occurs as well — teachers concentrate on measured indicators
even when this focus in inconsistent with exemplary teaching and learning.

The remainder of this chapter will discuss the implications that agency theory has
for a general analysis of teacher incentive systems. This theoretical frame will highlight
the tradeoffs and perverse incentives that are tangential in much of the current empirical
work on teacher compensation.

A Typology of Teacher Compensation Systems

When considering teacher compensation, three main types of systems are
generally discussed. These are usually referred to as bureaucratic systems, performance
(behavior)-based systems, and outcome-based systems. Bureaucratic systems are often
found in unionized professions; hours worked and tenure generally play a large role in
compensation under these systems. In a performance-based system, some portion of

salary is tied to observable behaviors by the agent, such as the number of cold calls

23

conducted by a salesperson or a teacher’s demonstration of a specific pedagogical
technique. As the name implies, outcome-based systems tie compensation to results,
such as the value of goods sold by a salesperson or students’ test scores. The remainder
of this section discusses each conceptualization through the agency lens.

Bureaucratic Teacher Compensation Systems

The current uniform salary schedule used in approximately 95% of public school
districts (Goldhaber, 2001) is a bureaucratic compensation system. Teachers are
rewarded for years of teaching experience and for the attainment of advanced degrees.
The salary schedule applies to all teachers in a district, regardless of subject or grade
level, and is often negotiated by union representatives. When policy-makers discuss
teacher compensation alternatives, they tend to have the most sophisticated understanding
of the bureaucratic compensation system. This may be because the consequences of this
compensation choice play out in our schools on a daily basis. Policy-makers are often
familiar with the shortcomings of this system and many of their criticisms are valid.
While this system is often criticized for its failure to reward high levels of teacher effort
or achievement, it has several important advantages as well.

One benefit of bureaucratic compensation systems is that teachers have no
financial incentive to direct their attention to certain students while ignoring others or to
adopt specific pedagogical techniques whether or not they are effective with a particular
group of children. As we shall see, some alternative compensation systems may
financially encourage teachers to focus their attention on certain groups of students since
increased performance by these children is likely to increase the teacher’s pay. Other

alternatives may provide teachers with financial incentives to teach in certain ways. This

24

is not a problem if the rewarded pedagogy will result in improved teaching and learning
but this may not always be the case.

The uniform salary schedule also allows teachers to take risks with their pedagogy
without facing financial risks at the same time. Teachers can take chances and learn new,
ultimately more effective ways of teaching even if performance drops off for a while
without worrying about a sudden drop in salary while they are mastering the new
pedagogy-

Teachers believe that the uniform salary schedule is objective (Odden & Kelley,
1997). This is not a trivial benefit. If the compensation system is perceived as
capricious, the principal will be forced to spend a great deal of time and effort dealing
with complaints and low morale.

Another benefit is that bureaucratic compensation systems require minimal
monitoring (Odden & Kelley, 1997). It is easy to determine a teacher’s years of
experience, particularly when a teacher remains in the same school for many years, and
teachers can be required to submit a certified transcript when they apply for salary credit
based on the attainment of additional degrees. The costs of obtaining the information
necessary to effectively administer a bureaucratic compensation system are low relative
to the costs associated with alternatives.

Another advantage is predictability of salary expenditures. Unless union
negotiations force an unexpected change in the salary schedule, the principal can predict
anticipated teacher salary outlays with a high degree of accuracy. Most areas of the
country base school revenues on property taxes or state foundation allowances and have a

limited ability to rapidly increase school revenues. The principal may be willing to

25

forego the benefits of incentive programs in order to maintain fixed, predictable salary
costs.

A final advantage is that bureaucratic compensation systems give teachers no
financial incentive to hide weak performance. Teachers can reveal weaknesses and ask
for assistance without being concerned that their pay will suffer as a result of their
honesty.

The main problem with this type of compensation system is that there is no
financial incentive for teachers to maximize their work effort. Salary is based only on
experience and education; performance is not considered. While other forces may
encourage teachers to improve their performance, the compensation system provides no
incentives to do so. Since there is no link between performance and pay, the principal
has limited leverage to force teachers to improve their performance. If a teacher knows
what needs to be done and chooses not to do it, the principal’s only means of punishment
are teacher dismissal (which would require demonstration of a gross failure to perform
duty in the case of a tenured teacher) or relatively soft punishments such as placement in
the classroom next to the gymnasium or assignment to the worst possible bus duty. The
effects of punishments such as terrible bus duty are further diminished since union rules
would probably prohibit this type of principal behavior so the punishment could not be
explicitly linked to the agent’s poor performance.

High quality teachers may feel unappreciated and unrewarded in bureaucratic
compensation systems. While high quality teachers may feel intrinsic satisfaction and
receive positive recognition from peers, parents, and supervisors, these messages will

conflict with the message of mediocrity contained in their paychecks. The fact that low

26

performers in their district receive the same compensation as high performers may
discourage high quality teachers from continuing to persist in the exhausting, often
frustrating work of exemplary teaching.

Another problem with this schedule is that it creates adverse selection — it does
not necessarily attract desired people to teaching. In the retail sales sector, bureaucratic
compensation (pay based on hours worked) is more likely to be used than pay based on
sales productivity when there is little discretion in the job and when managers view the
sales position as . .just a job - just a way for people to earn some extra money. It’s not
a career.” (Eisenhardt, 1988, p. 504). This pattern in retail sales is consistent with the
predictions of agency theory and suggests that bureaucratic compensation may not be
attracting desired candidates to teaching, a professional career that requires a great deal of
discretionary decision making and where it is hoped that high performers will remain in
the system for long periods. High performers have an incentive to choose business
careers that will pay a premium for their talents and a disincentive to choose fields such
as teaching that will not financially compensate them for excellence.

Performance (Behavior)-Based Teacher Compensation Systems

Performance-based compensation systems are intended to overcome some of the
disadvantages of bureaucratic compensation systems outlined but simultaneously
introduce a new set of difficulties. In this type of compensation system, some portion of
a teacher’s salary is tied to the teacher’s performance. This is often achieved through the
use of classroom observations of the teacher’s behavior but can also be linked to
demonstrated acquisition of a desired skill set. The public understanding of performance-

based compensation often appears to be quite thin when compared to their familiarity

27

(and accompanying contempt) for the uniform salary schedule. Policy-makers and the
public rarely consider the problems of hidden action and hidden information and they
also tend to assume that any reasonably astute observer should be able to easily and
inexpensively identify quality teaching or the skills necessary to be a high quality teacher.

One important benefit of performance-based compensation is that it provides
teachers with financial incentives to improve their teaching skills. This incentive system
may be particularly motivating for teachers whose performance is close to the cut score
performance levels for additional pay. PGA golfers are more likely to increase their
effort levels as their scores move closer to the next payoff level in a tournament
(Ehrenberg & Bognanno, 1990). This is particularly true as the financial payoffs between
levels increase.

Another advantage of this type of compensation system is that it allows principals
to reward differential teacher performance without the confounding influences, such as
student background, that can make outcome-based compensation difficult to justify.
While the perceived objectivity of systems varies, there is the potential for performance-
based systems designed so that it is easy for teachers to understand why pay differentials
exist and what they need to do to receive higher pay.

Performance—based compensation may also encourage high quality teachers to
remain in teaching while reducing teaching’s attractiveness to low performers. High
performers will be financially rewarded for their increased efforts while poor performers
may decide that other fields will provide them with higher salaries when they fail to

receive performance bonuses.

28

One problem with performance-based compensation is that it can be difficult to
connect measurable behaviors to quality teaching. The ultimate goal of principals is to
encourage teachers to be high performers. While a great deal is known about teaching
and learning, a certain amount of mystery and magic still remain. By one estimate, only
about 3% of a teacher’s contribution to student achievement can be explained by skills
that are easy to measure (Goldhaber, 2001). The remaining 97% is attributable to
qualities, such as enthusiasm, that are not measurable and for which good proxies are not
available.

To make matters worse, there is no one teaching style or set of skills that is
clearly superior to others. Some traditional teachers do a marvelous job of educating
students and so do some constructivist teachers. These styles require different skills and
look very different in practice. Designing a performance-based teacher compensation
system that rewards all of the different manifestations of excellent teaching is an
impossible task.

While the task of linking performance measures to quality teaching is quite
difficult, recent research has begun to shed some light on this dilemma. Traditional
teacher evaluation systems have generally done a poor job of identifying high quality
teachers but subject-specific, carefully designed evaluations can substantially improve the
strength of the relationship between teacher performance scores and student achievement
(Porter, Youngs, & Odden, 2001). When examining four districts with teacher evaluation
systems that were considered exemplars, researchers did find that the evaluation system
did a better job of identifying high quality teachers (identified as teachers whose student

test gains during the year exceeded expectations) than the experience measure rewarded

29

under bureaucratic compensation systems (Milanowski, Kimball, & White, 2004; White,
2004). In addition, a positive, significant relationship existed between overall
performance scores and student achievement (Gallagher, 2004; Milanowski, 2004;
Milanowski et al., 2004; White, 2004). Unfortunately, the strength of the relationship
varied substantially depending on the grade and subject taught (Milanowski et al., 2004)
and which subject area’s student test scores were used (Gallagher, 2004; White, 2004).

A further problem arose when researchers parsed out the components of the
overall performance score. In Cincinnati, planning and professionalism domain scores
were more highly correlated with student achievement than instructional domain scores
(Milanowski, 2004). This leads us back to the difficulty of measuring what matters.
While instructional ability is clearly an important factor, and one that we want to
encourage teachers to improve, our proxies for it are so poor that planning and
professionalism appear to have a greater effect on student achievement than the
instruction students receive. A final problem is that even the exemplary performance
measures examined in these studies failed to identify teachers who were adept at closing
the achievement gap within their classrooms as superior to those whose performance
maintained or increased the gap - a shortcoming that could have serious implications for
equity (Borman & Kimball, 2004)

Performance-based teacher compensation systems are vulnerable to abuse by
teachers. Teachers may try to take advantage of the system by concentrating only on
measured performance attributes and ignoring equally important, but unrewarded,
behaviors. Teachers are encouraged to focus their efforts on measured behaviors under

these systems. Given the complexity of teaching, performance-based systems cannot

30

assess teachers on the full range of desired behaviors and still be administratively
feasible. Teachers have a financial incentive to spend excessive amounts of time on
measured behaviors. This is not to say that teachers will ignore anything that is not
measured by the system but the attention that they pay to non-rewarded behaviors is in
spite of the compensation system, rather than because of it. This tendency to
overemphasize rewarded behaviors has been observed in other groups such as NFL
quarterbacks (C. Brown, 1990) and computer programmers (Prendergast, 1999). Given
that much of good teaching remains unmeasurable, the quality of teaching could
conceivably deteriorate even as measured performance improves.

Teachers may also concentrate on rewarded behaviors only in the context of an
evaluation (Firestone, 1994). Most of teaching takes place behind closed doors; a teacher
could choose to do the difficult work of high quality teaching only when being observed
by an evaluator.

Finally, teachers may try to manipulate their class makeup to increase the
proportion of easy-to-teach students in an attempt to make their teaching appear smooth
and effective. High school teachers might encourage challenging students to transfer to a
different class or to leave school altogether. Teachers with a particular gift for working
with challenging populations might be discouraged from doing so. For example, a
teacher who was particularly effective at working with mainstreamed emotionally
conflicted children might feel that it would be financially prudent to minimize the
number of these children placed in his or her classroom since any disruptive behavior

might negatively affect the performance evaluation.

31

Another problem with performance-based teacher compensation is that the
principal has a perverse incentive to decrease professional development opportunities for
teachers (Firestone, 1994). As mentioned previously, school districts have a limited
ability to fund rising salaries. While a district may want all of its teachers to improve, it
may not be able to afford the increased demand for professional development that allows
them to do so or to pay for the resulting performance-based salary increases.

School principals have often used existing low-stakes evaluations as a tool to
encourage teachers and to build a team (Mumane & Cohen, 1986). Teachers are
sometimes given high evaluations and then encouraged to live up to those evaluations.
Favorable evaluations are also used to foster collegiality and feelings of administrative
support. If evaluations are tied to salaries, inflated ratings will push salaries to
unacceptably high levels. School principals and other evaluators will face pressure to
keep average ratings at or below a certain affordable level. This will undermine the
traditional motivational uses of teacher evaluations.

Performance-based teacher compensation systems require more monitoring than
bureaucratic systems. Someone must evaluate teacher performance and this usually
involves multiple classroom observations by a trained evaluator. The costs of these
evaluations, in terms of training evaluators and time spent doing the evaluations and
documenting compensation decisions, can be considerable. Another evaluation cost is
the time that teachers may spend attempting to influence the evaluator. It is very difficult
to reduce teaching to a set of objective measures - the technology of teaching is unclear
and appropriate behaviors are context dependent. In addition, the evaluator can only

observe a small slice of the teacher’s performance, and employees are particularly likely

32

to try to influence evaluators under these conditions (Milgrom & Roberts, 1988). Rather
than trying to improve their teaching, a daunting task, teachers may spend large amounts
of time on the evaluator’s pet projects or convincing the evaluator that the observed
lesson was just the result of a bad day and does not reflect typical performance. The use
of multiple evaluators can minimize this problem but it raises monitoring costs at the
same time it reduces influence costs.

Unlike bureaucratic systems, performance-based systems may discourage teachers
from revealing their weaknesses and seeking assistance (S. M. Johnson, 1986). Teachers
who are struggling may try to hide their poor performance so that it does not negatively
affect their pay.

Outcome-Based Teacher Compensation Systems

Given the difficulty of measuring teacher performance and the imperfect
relationship between measurable performance and outcomes, outcome-based teacher
compensation systems can be an attractive alternative but they are not an ideal one. As
with bureaucratic and performance-based systems, there are advantages and
disadvantages to using an outcome-based system. The public understanding of outcome-
based compensation tends to be oversimplified. As discussed earlier, teacher
performance is not the sole determinant of student achievement and disentangling the
influence of other factors is a Herculean task. Despite this, outcome-based systems have
a simple appeal — if we want teachers to improve student achievement, why not reward
them for doing so? Scholars and educators who urge caution are sometimes dismissed as

obstructionist preservers of the status quo and the arguments made on both sides often

33

ignore the interplay of benefits and problems created when outcome-based systems are
adopted. A more thoughtful analysis is in order.

One benefit of this type of system is that it focuses attention on results rather than
performance. As mentioned above, the technology of teaching is somewhat unclear.
There is no agreed upon best method for teaching all children; a focus on results allows
teachers to use their professional expertise to decide the best way to reach particular
students.

Outcome-based compensation’s objectivity may be attractive to teachers.
Teachers are often familiar with the tests and other criteria (such as student attendance)
being used to evaluate them and will appreciate that the evaluating entity cannot show
favoritism when matching compensation to performance under this system.

Another advantage of outcome-based systems is that, unlike performance-based
measures, they encourage teachers to seek assistance in weak areas. These systems do
not penalize poor performance, only poor results. Teachers will be able to discuss their
shortcomings and work with colleagues and administrators on solutions since doing so
will enhance, rather than lessen, the chances of receiving incentive pay.

Another benefit is that holding teachers responsible for student learning makes
sense to the public. Since taxpayers fund our public schools, it is important that they
support any changes to the omnipresent uniform salary schedule.

Properly designed outcome-based compensation systems may encourage desirable
candidates to enter and remain in teaching. While the link between effort and reward is
not as clear as with performance-based systems, it is an improvement over the practically

nonexistent link in bureaucratic teacher compensation systems. Candidates who believe

34

that they have the potential to be very effective teachers and teachers who are effective
and receive incentive pay have an incentive to teach under outcome-based systems. This
predicted increase in the quality of the work force is empirically verified in a study of
auto glass workers (Lazear, 1996). When the Safelite Glass Corporation switched from a
bureaucratic pay system to piece rates, more able employees tended to remain with the
firm while unproductive employees left. Newly recruited employees under the piece rate
system also tended to be more able than newly recruited employees under the
bureaucratic pay system had been. The quality of the company’s work force improved as
a result of the move to outcome-based compensation.

One huge problem with outcome-based compensation is that it may exacerbate
the existing maldistribution of teachers in the system. Under the current system, affluent
suburban districts often offer higher base salaries than their inner city and rural
counterparts. Tying pay to student performance will not rectify this problem and may
make it even worse. Students in low SES schools often have a more difficult time
achieving school success than their more affluent contemporaries. If pay is tied to
absolute measures of student achievement, teachers have a strong financial incentive to
transfer to affluent schools since they are more likely to receive achievement bonuses in
these schools.

In an attempt to avoid this problem, value-added achievement measures can be
used instead of absolute measures. With a value-added system, perverse incentives to
move to more affluent schools are largely avoided. One potential problem with this is
that value-added systems may institutionalize low expectations for certain students since

past performance is used as a predictor of future potential. Another problem is that

35

teachers can only control their own performance in a given school year and there are
factors affecting success each year that are outside of their control. Is it fair to sanction a
teacher for poor performance if he or she put forth a strong effort but did not receive
enough textbooks from the central office that year? Is it the teacher’s fault if the
curriculum mandated by the state is culturally irrelevant within the local community and
parents do not support their efforts to teach it? In business, sales departments often
blame poor quality for low sales while production departments blame halfhearted sales
efforts (G. J. Miller, 1992). Teachers may blame poor achievement gains on factors
outside of their control and feel that even the value-added system is unfair.

Another potential problem with the value-added solution is the complexity of
most value-added systems. Teachers are unlikely to trust a compensation system they do
not understand (Mumane & Cohen, 1986) and value-added systems are generally quite
difficult to understand. Additionally, even if teachers trust the system, they may have
trouble understanding the complicated relationship between their performance and the
resulting gain score. This makes it difficult for them to improve their performance.

As with performance-based systems, the cost of administering and monitoring
outcome—based systems is higher than the cost of bureaucratic systems. Student
outcomes must be matched to particular teachers and testing conditions must be similar
across classrooms if comparisons are to be made. Procedures for handling transient
students and those with special challenges such as limited English proficiency must be
developed and communicated. The design and maintenance of the system is often a

complex task, particularly in the case of value-added systems. The cost of administering

36

and monitoring the incentive system may reduce the total money available for teacher
salaries.

In addition to these costs, salary expenses will rise if the program is effective.
Teachers will work harder, student achievement will increase, and teachers will earn
incentive pay. The combined effect of higher salary expenses and higher monitoring
costs may be more of a financial burden than districts are willing or able to tolerate.

Teachers have an incentive to discourage certain types of students from remaining
in their classes under an outcome-based incentive system (S. M. Johnson, 1986) just as
they did with the performance-based system discussed previously. If students are
unlikely to be high achievers, teachers may discourage them from enrolling in their class.
Teachers may also spend time and effort trying to influence school administrators to give
them “good” students. While teachers may already engage in some of this behavior for
other reasons, outcome-based compensation gives teachers a financial incentive to do so
as well.

Besides productive time lost while teachers are attempting to influence
administrators, teachers’ attempts to curry favor with the administrators responsible for
student-teacher assignments may undermine collegiality in a school as well. Teachers are
unlikely to work well with their peers when they feel that those peers are trying to steal
all of the “good” students for their own classes.

Helping efforts among teachers may diminish under outcome-based compensation
systems (Firestone, 1994). Teachers may refuse to “waste” time helping each other when
doing so takes away from time they could spend improving their own results. In a study

of physicians in group practice, teamwork (measured as the frequency with which

37

physicians discuss another group member’s case in an advisory capacity) was less likely
to occur when group revenues were divided according to individual case revenue than
when revenues were split evenly among the group’s physicians (Encinosa, Gaynor, &
Rebitzer, 1997).

This type of system also gives teachers financial incentives to concentrate only on
the knowledge needed for the current year’s assessments rather than spending time
introducing ideas that will not be tested until the following year. Education is
cumulative; concepts may be introduced one year but not tested until the child has been
exposed to them several times. The fourth grade teacher’s compensation will suffer if the
third grade teacher prepares students quite well for the third grade test but leaves them
unprepared to succeed in the fourth grade. This is hardly a recipe for improved teamwork
and collegiality.

Outcome-based systems tend to encourage agents to take a short-term view of
improvement. In one empirical study, it was found that mutual fund managers change
their behavior after interim return figures are released to the public (K. C. Brown,
Harlow, & Starks, 1996). If a fund shows an above average interim return, managers act
very conservatively for the remainder of the year. If the fund has below average interim
returns, managers take much greater than normal risks for the remainder of the year.
Once the new year begins, both groups revert to their normal behavior until the following
year’s interim return figures are released. Navy recruiters tend to engage in behaviors
such as high school visits that will improve long-term recruitment efforts early in their
measurement cycle but, if they are close to a reward cutoff, they will engage in short-

term recruitment efforts as the end date for rewards approaches (Asch, 1990). Outcome-

38

based pay may encourage teachers to take actions that maximize their short—term chances
of receiving merit pay each year even if those actions are not the ones that maximize their
long-term growth as excellent teachers or their students’ learning.

The reliability of outcome-based measures is another problem. If a single test
score is used to measure teacher performance, volatility due to sampling variation (the
characteristics of a particular cohort of children) and one-time factors (such as a flu
epidemic) casts doubt on the reliability of results (Kane & Staiger, 2002). Increasing the
number of tests included in the compensation system lessens this problem but adds to the
cost of program administration. If multiple years of data are included, teachers may be
unwilling to expend effort on performance improvements that will not be rewarded for
several years. This pattern was observed in Navy recruiters: as the period between
rewards lengthened, effort was reduced (Asch, 1990). Extending the period between
rewards may be a particular problem in education where merit pay plans have tended to
be short-lived — teachers may hesitate to expend effort when the merit pay plan may not
even survive the measurement cycle.

Measures used in outcome-based compensation systems tend to be quite limited.
Attention is often focused on academic achievement as measured by standardized tests.
Agency theory predicts that what gets rewarded will get done and other things may be
ignored. There is some evidence that employees in the public sector may be particularly
likely to focus their attention on rewarded outcomes even when the amount of the reward
is small (Courty & Marschke, 1997). Teachers may focus on the rewarded outcomes
simply because they provide them with clear, measurable objectives in an otherwise

confusing environment. If policymakers only care about standardized test results, a

39

system that heavily rewards teachers for test results is fine. This is unlikely to be true,
however; schools have historically had multiple, conflicting goals. The tendency of
outcome-based compensation systems to target effort towards a narrow set of goals may
explain why these systems have been used so infrequently in the public sector. Most
public agencies have numerous competing goals, and hi gh-powered incentive programs
do not work well in this type of environment (Courty & Marschke, 1997; Dixit, 1997).

Policy makers may add a variety of measures to the system in an attempt to
correct for this problem. As was a potential problem with value-added outcome
measures, the use of multiple measures may confuse teachers and make them distrust the
system. Another problem with increasing the complexity of the system is that it may give
teachers more opportunity to use the system to their own advantage than simpler systems.
As the complexity of executive compensation programs increases, executives find
opportunities to manipulate the system so that they are paid extremely well when
performance is outstanding but are still relatively well paid when they perform badly
(T osi & Gomez-Mejia, 1989). Teachers may discover ways to manipulate complex
systems as well, particularly when they are involved in system design.

When teachers are rewarded for high absolute student performance, they have a
financial incentive to focus on students whose achievement is relatively easy to improve.
They may underserve gifted students since these children will have relatively high
achievement even without a great deal of teacher attention. In addition, it is very hard to
move children from the 97‘h percentile in achievement to the 99’h percentile. Children

with disabilities may also be ignored, either because they are excluded from the tested

40

population or because the achievement levels of non-disabled children are easier to
increase.

A final disadvantage is that these systems give teachers a financial incentive to
focus strongly on tested items and even to cheat on the test. Cheating can take many
forms, from outright changing of student answers to more subtle means such as
encouraging absenteeism on testing day for certain students. While encouraging students
to cheat would never be desirable, the consequences of focusing on tested items are
somewhat less clear. If the tests are well designed, aligned with the curriculum, and
include attention to higher order thinking skills, problems are minimized. On the other
hand, if a test’s main virtue is that it is inexpensive, rather than excellent, the
disadvantages of teaching to the test are obvious.

It may be possible to minimize teachers’ tendencies to focus disproportionate
attention on certain students or portions of the curriculum through careful monitoring.
Besides the cost involved in increased monitoring, there is some evidence that monitoring
is particularly problematic in the public sector. It is very difficult to minimize the
discretion of street level bureaucrats such as teachers since rules cannot describe the full
range of decisions that must be made (Lipsky, 1983). This means that monitoring in
many parts of the public sector must be both close to particular situations and sensitive to
changing conditions. In a study of the Federal Job Training Program, these conditions
were not met and monitoring was ineffective (Courty & Marschke, 1997). Even when
monitoring problems are recognized, the presence of multiple stakeholders with

conflicting goals and entrenched bureaucracy within the public sector make it very

41

difficult to get consensus about what should be done to improve the assessment system
(Courty & Marschke, 1997).
Piece Rate and Tournament Incentive Systems

Alternatives to bureaucratic teacher compensation systems offer teachers the
opportunity to earn higher pay through attainment of certain performance or outcome
goals. The link between the measure used and increased pay can be established in two
ways. The first of these — piece rate — is when a reward is earned through attainment of
an absolute, fixed goal. In these systems, teachers are told what performance or outcome
level they must reach in order to receive rewards prior to the beginning of the
measurement period. The second — tournament — is when rewards are based on the
relative ranking of all the teachers in the system. Teachers are told what percentage of
top performers will receive rewards prior to the beginning of the measurement period.

The main advantage of a piece rate system is that it provides teachers with a clear
target to be reached. In a properly designed piece rate system, the teacher’s reward is
dependent only on his or her own ability and effort. The teacher has no financial
incentive to undermine the work of colleagues since the number of winners in this system
is limitless.

One disadvantage of piece rate systems is that they make it difficult for the
principal to keep salary expenditures at predetermined levels. If earned bonuses exceed
projections, whether because of higher than expected effort or random fluctuations,
administrators may face a budget shortfall. As mentioned, school districts have limited
control over their ability to meet unexpected expenses so this problem provides a strong

disincentive for the use of piece rate systems. One study of merit pay systems found that

42

almost 17 percent of districts had eliminated their merit pay program because the
programs were too expensive (S. M. Johnson, 1986). More recently, a centrally
mandated performance pay system in the United Kingdom was virtually eliminated after
only a few years, largely due to complaints about the higher salary costs being home by
local schools under the program (Wragg, Haynes, Wragg, & Chamberlin, 2004).

Another disadvantage of piece rate systems is that teachers may collectively
refuse to meet bonus criteria in an attempt to get the standard for rewards lowered. This
is a common problem when piece rates are used in the factory setting (Lawler, 1990;
Milgrom & Roberts, 1990; G. J. Miller, 1992; Whyte, 1955).

Tournaments avoid both of the problems mentioned in the discussion of piece
rates. A tournament allows the principal to determine the amount of incentive pay to be
rewarded and set the cutoff in the rankings so that it matches available funds.
Tournaments also provide few incentives for voluntary work restrictions. There is no
standard to lower; rewards are simply based on relative rankings. High performers have
an incentive to convince lower performers to restrict their output so they can easily
maintain their high relative rankings but lower performers have no financial incentive to
agree to do so.

Tournaments may also protect agents, teachers in this case, from the effects of
random shocks that affect the entire group (K. C. Brown etal., 1996; Sappington, 1991).
For example, suppose that a testing company renorrns the test used by a district to assess
teacher outcomes and that it is now more difficult to achieve results above the national
norm. In a tournament system, the same proportion of teachers is still eligible for

rewards even if overall performance dips. In a piece rate system, the renorrning would

43

make it more difficult for teachers to receive rewards through no fault of the teachers
themselves.

However, tournaments provide a strong disincentive for cooperative behavior
among coworkers (Drago & Garvey, 1998). This is a serious disadvantage in a
profession such as teaching where cooperation and teamwork are highly valued attributes.
Beyond the disincentive for cooperation, tournaments provide coworkers with an
incentive to sabotage each other’s work. Every teacher who feels that he or she might
have a chance of earning a reward has a financial incentive to undermine the work of
colleagues. The system is a zero-sum game designed to have winners and losers and few
people like to lose.

Piece rate and tournament systems may be attractive to different types of people.
Tournaments may appeal to people who enjoy competition while piece rates may attract
those who value collegiality. Teaching has historically attracted people with a relatively
collegial disposition (Lortie, 1975); this might change if the use of tournaments increases.
Individual - and Group-Based Reward Systems

Teachers can receive rewards based on their individual efforts or they can be
rewarded for the efforts of a larger group, generally the school. In practice, group-based
reward systems have generally been outcome-based systems but this does not have to be
the case. Performance goals could be set for a group of teachers and rewards based on
the group’s attainment of those goals.

A major strength of individual rewards is that they provide high performers with a

strong incentive to remain in teaching and low performers with a strong incentive to

leave. The impact of one’s performance is not muted by that of colleagues; individuals
are held financially accountable for their own teaching.

On the negative side, individual rewards do not provide financial incentives for
helping colleagues or performing tasks such as hall duty that are necessary for the smooth
functioning of the school but provide few individual benefits to the teacher. In fact, quite
the reverse is true, because agents are financially better off if they focus only on directly
rewarded tasks rather than those where benefits accrue to others, such as assisting novices
(Posner, 2000).

In order for individual-based rewards to be possible, the principal must develop a
scheme for allocating credit to individuals. In an activity such as teaching where
everything from prior academic achievement to the discipline policies of a child’s other
teachers can affect both the individual teacher’s ability to teach and his or her students’
ability to learn, it is difficult to do this in a simple yet fair manner.

Group-based rewards avoid this problem by explicitly recognizing the team nature
of school production. The school’s teachers are rewarded for their collective effort.
Collective results are often easier to measure and monitor than individual results, so
group—based systems are generally less costly to administer than their individual-based
counterparts. While group-based systems may lessen the measurement problems in an
evaluation system, they do not eliminate them. Problems of reliability do not disappear
(Milanowski, 1999), and problems of appropriate allocation of credit and blame are not
entirely avoided with the new unit of analysis since the effects of the district, prior

schools, parents, and community must still be separated from school effects.

45

Group rewards provide an incentive for teachers to help their colleagues since
financial rewards are dependent upon the entire group’s performance. They also provide
an incentive for teachers to participate in activities that benefit the school as a whole even
if the benefit to the individual teacher is small. While group rewards do not have the
perverse incentives to undermine colleagues that are often embedded in individual reward
systems, Malen (1999) cautions that group reward systems with perverse incentives for
cheating or for gaming the system may not promote collegiality in expected ways. As
she puts it, “Deception and deceit are not compelling rallying points for teacher
collaboration. Rather, they may be sources of division and derision that become every bit
as troublesome as the faculty strains that occurred under individual-based merit pay” (p.
389).

Unfortunately, group-based reward systems may also provide high performing
teachers with incentives to leave low performing schools where they are often needed the
most. An individual teacher has a limited ability to raise the group’s average so all
teachers would prefer to go to a high performing school. The teachers who are most
likely to have the opportunity to do so are the high performers. Group-based reward
systems may worsen the current, inequitable distribution of high quality teachers among
schools.

Group-based systems provide an incentive for individuals to free ride on the
efforts of colleagues. An individual teacher can put forth a minimal effort and still
receive the financial reward as long as his or her colleagues refrain from the same
behavior. On the other hand, if the teacher does not believe that his or her colleagues will

refrain from free riding, that teacher still has no incentive to work hard since the free

46

riding of colleagues will prevent receipt of the reward. Either way, each individual has
an incentive to put forth minimal effort. Agency theory would predict that higher quality
workers would have an incentive to monitor the performance of their lower quality
colleagues and encourage them to improve but empirical evidence is mixed (Prendergast,
1999; Welbourne & Gomez-Mejia, 1995).

Group-based systems also provide high quality teachers with few incentives to
work to capacity since they will only receive a small portion of the reward for their effort;
most of the reward will be distributed to other group members. A study of service
representatives for a discount brokerage firm found that the introduction of a group
reward system increased the productivity of low performance workers but the
performance of high quality workers remained static or declined (Hansen, 1997).

Under a group-based system, high quality teachers may decide to leave teaching
altogether and move to an industry with individual rewards (as long as they believe that
the skills that make them a high quality teacher will transfer to other occupations). Low
quality teachers will have an incentive to remain in teaching since they can receive a
financial reward that is derived from the work of their colleagues. Group-based systems
are also more likely to appeal to individuals who value collegiality while individual
rewards may be more attractive to those with a more competitive nature.

Conclusion

There is no perfect compensation system for teachers. Incentive programs may
change teacher behavior but their behavior may not change in the intended ways.
Alternatives to the current compensation system often provide teachers with perverse

incentives to respond in undesirable ways. These perverse incentives are not the result of

47

poorly designed alternatives; they are inherent in any compensation system where the
agent has information not available to the principal, where there is a stochastic element to
output, and where the principal cannot directly observe the agent’s actions. All these
conditions are present in teaching and provide teachers with opportunities to game the
compensation system.

Pitfalls to performance-based systems include the difficulty of specifying
desirable behaviors, performance that complies with the letter rather than the spirit of the
performance goals, and the high cost of monitoring teacher behavior. Outcome-based
systems fare no better. Specifications are still difficult to develop since policy-makers
have multiple goals, teachers still have an incentive to comply with the letter rather than
the spirit of goals, and monitoring costs are still higher than the cost of the current
compensation system. Piece rate systems make it difficult to control costs and teachers
may collectively restrict output but tournaments sabotage cooperation. Individual
incentive systems destroy collegiality but group incentives encourage free riding. All the
alternatives will tend to attract and retain teachers who are different than the ones who
find the current system appealing but it is not entirely clear that these will necessarily be
better teachers. The details of the problems vary among the alternatives but the basic
difficulty remains — it is impossible to design a compensation system that does not have
undesirable consequences.

The fact that alternatives to the current system are flawed does not imply that
bureaucratic compensation is desirable. It too has its drawbacks. The most serious of
these are that it does not provide teachers with a financial incentive to maximize their

effort and that it may discourage desirable candidates from selecting and remaining in

48

teaching. Many teachers, administrators, and policy makers are comfortable with this
system since its use is so prevalent but maintenance of the status quo only makes sense if
the status quo is working tolerably well and there are those who would argue that it is not.

Agency theory allows us to consider alternatives to the status quo in a cautious,
even pessimistic, manner. While this theory, taken from the dismal science of
economics, may not help us find the perfect system or to feel particularly optimistic about
the potential for incentive systems to improve performance, it does allow policy makers
to consider the current system and its alternatives with eyes wide open. By providing a
clear picture of the strengths and the limitations of various options, an analysis grounded
in agency theory helps policy makers choose teacher compensation systems that are
compatible with the goals and the idiosyncrasies of their own situations.

The chapters that follow focus on one alternative compensation system, the
Minneapolis Professional Pay Plan. Minneapolis Public Schools recently introduced this
voluntary alternative to their traditional bureaucratic compensation system. Under
Professional Pay, teachers do not receive salary credit for gaining experience; rather, they
earn credit in a variety of ways, including participation in certain professional
development activities and meeting school improvement goals. Within our agency
framework, this design would lead us to expect particular responses. In Chapter 3, we
will look at the history of Professional Pay, the shape it has presently taken, and our
agency-derived expectations of its effect. Chapter 4 considers the likelihood that
Minneapolis will be able to credibly commit to the policy and the effect that has on

teacher behavior. Chapter 5 examines teachers’ responses to Professional Pay and the

49

implications those responses have for improving teacher quality within the district while

Chapter 6 concludes.

50

Chapter Two

Methods

In this chapter I describe and explain the research methods used in this study.
First, the research design is discussed. The next three sections discuss instrument
construction, sampling, and data collection methods, respectively. Following that, I
discuss the data analysis techniques used in the interview and survey portions of the study
as well as the techniques used to combine the two. I conclude with a discussion of
several challenges I faced and the study’s limitations.

Research Design

In this study, I hoped to begin unpacking the “black box” of teacher response to
alternative compensation systems. I concentrated on one particular alternative
compensation policy, the Minneapolis Professional Pay Plan, trying to gain a better
understanding of teachers’ responses and the underlying reasons. Once I had established
this goal, I took a pragmatic approach to research design, considering which research
design would provide the greatest opportunity to answer my particular research questions,
rather than becoming bogged down in disputes over the general superiority of specific
designs (R. B. Johnson & Onwuegbuzie, 2004).

I chose a mixed method study. In a mixed method study, the researcher uses both
qualitative and quantitative methods and inferential processes are both inductive and
deductive (Tashakkori & Teddlie, 1998). I drew on two main sources of data: interviews
and a survey. The choice of a combination of qualitative and quantitative data collection

methods allowed me to build on the strengths of each. As Miles and Huberman (1994)

51

put it, “at bottom, we have to face the fact that numbers and words are both needed if we
are to understand the world” (p. 40).

The interviews allowed me to tap into respondents’ contextually embedded
perspectives and ask more nuanced questions about details. Respondents were able to
use their own language to explain their concepts of the policy and their responses to it.
The open-ended nature of the questions also provided me with opportunities to be
surprised. Respondents were not limited to a list of likely responses or those predicted by
theory; indeed, the participants often nominated factors I had not previously considered.

The surveys enhanced the generalizability of my interview data beyond the small
subset of teachers with whom I spoke. In addition, the highly structured nature of the
survey allowed me to systematically measure potentially relevant factors.

I used both inductive and deductive inferential processes, moving inductively
from my observations to theory, deductively from theory to predictions, and then
returning to observational data as a means of testing my predictions. By repeatedly
moving through this cycle — both within and across each type of data — I hoped to
elaborate upon existing theory in ways that remained true to respondents’ experiences.

This study is a single case study with multiple sites within the case. The use of a
single case, Minneapolis, allowed me to develop a rich understanding of teachers’
responses to one particular policy, Professional Pay, and the district context within which
those teachers work. Teachers do not respond to policy in a vacuum. They have at least
some knowledge or beliefs about the history, political environment, and bureaucracy of

their district. By limiting my study to one district, I was able to collect data independent

52

of the teachers about the system within which teachers made their response decisions and
of the types of opportunities they have had to learn about the policy.

Within the case, I further limited my sampling to middle schools, for reasons
conceptual and practical. I was planning to conduct interviews in three sites and hoped
that I would be able to visit sites with varying levels of policy enrollment to determine
how school contexts might shape enrollment. By remaining within one level of the
system - middle schools - I would not have to try to disentangle the effect of school level
from other differences in the school context, a nearly impossible task with only three
schools in the interview sample. My other reason was more practical in nature. I would
need to gain the trust of respondents quickly during my brief visits and my own middle
school teaching experience provided me with an avenue to do so.

Instrument Development
Interview Protocol Development

Interviews were conducted using a structured interview protocol (see Appendix I).
Structured instruments were used to minimize the collection of unnecessary data, reduce
unintentional researcher bias, and allow for comparability across sites (M. B. Miles &
Huberman, 1994). The teacher protocol included questions about the teacher’s
background, participation in and assessment of Professional Pay, relationship with
colleagues and the school principal, and past experiences with central office policies and
personnel. Before writing the protocol, I developed a series of testable hypotheses that
were consistent with agency theory. Next, I examined the findings of previous
compensation researchers in the K-12 and other sectors. This second step allowed me to

nominate additional factors, such as understanding, that would not generally be

53

considered critical to an agency analysis but that might affect teachers’ responses. I then
developed a series of questions that would allow me to test my hypotheses. For example,
teachers were asked about previous interactions with central office personnel since the
credible commitment literature indicated that those experiences might affect teachers’
policy perceptions.

Once I had developed my initial interview protocols, I piloted the instrument with
two former teachers, one of whom had previously participated in a merit pay program.
This allowed me to improve the instrument’s clarity and to check the interview length
(this was of particular concern since most interviews would need to be completed within
a teacher’s 45 minute break period). The instrument was modified based on the pilot.
Following that, I conducted interviews at my first school site, Washington Middle
School.2

While the instrument worked reasonably well during those interviews, when I
returned from that site visit, I met with a faculty advisor and debriefed about the strengths
and limitations of the protocol, reflecting on what questions needed additional prompts,
and what questions produced little helpful information. Based on that analysis, I made
additional improvements before the next set of interviews. In a few cases, I changed the
language used in a question so that it was more similar to the language used by my initial
respondents. In other cases, I took advantage of the insights gained from my initial
analysis of the first round of interviews to probe some areas, such as current professional
development activities, more deeply while reducing the emphasis on other, less

promising areas.

 

2 All school names are pseudonyms.

54

I also developed structured interview protocols for use when interviewing school
principals and district implementers (see Appendix I). The school principal interviews
provided information about whether teachers’ perceptions were consistent with the school
principal’s. They also provided me with additional information about the school and
district environment. District implementer interviews helped me understand the district’s
political and institutional environment, as well as policy particulars. District
implementers were asked about the current policy as well as its history and future. They
were also asked about their perceptions of its effect, as well as about the effects of state
and local actors on the district environment.

Survey Development

The survey instrument was designed after I completed the interviews and had
performed my initial analysis (see Appendix I). This provided me with a set of
conceptual ideas, gleaned from the interviews, that could be incorporated into the survey
(M. B. Miles & Huberman, 1984). As a result, I had three sources to draw from when
designing the survey: agency theory predictions, the findings of previous compensation
researchers, and my own initial findings. Based on these sources, I again developed a
series of testable hypotheses.

Next, I developed a series of survey items to test my hypotheses. When doing so,
I tried to match my items to those used by previous researchers whenever possible. For
example, I was able to use items in the Department of Education’s Schools and Staffing
Survey (SASS) to determine respondents’ perceptions of the school administrator. Using
this approach, rather than creating an entirely new set of items, had two advantages. The

first is that the items had already been field tested and found to provide appropriate

55

measures of the underlying construct of interest. The second is that using the same items
as other researchers improves the usefulness of the full body of survey work, allowing for
later comparisons of responses across different teacher populations.

Survey items echoed the main sections of the interview protocol. There were
items about teachers’ knowledge about and opinion of Professional Pay, relationships
with central office and school personnel, past experiences with district policies, and
demographic information. Enrolled teachers were asked about their responses to
Professional Pay. After developing a draft set of items, I mapped them back onto my
testable hypotheses, in effect, doing a conceptual inventory. This helped ensure that
survey items provided sufficient information to test each hypothesis, while allowing me
to eliminate unnecessary or redundant items.

I conducted my surveys online, choosing this format for several reasons. The first
is that the relative novelty of the online format may increase response rates (Dillman,
2000). The online format also eliminated the confusion generally caused by skipped
sections in paper surveys. Respondents often become confused when they are asked to
skip portions of a survey (Dillman, 2000), but the online survey automatically directs
respondents to relevant sections based on previous responses. Other advantages are that
the online format considerably lowers mailing and printing costs, and that data are
immediately entered into a database, eliminating the arduous task of entering data.

One potential disadvantage of the online format is that respondents must be able
to manage the technology. MPS teachers receive school and district information through
e-mail on a regular basis and are expected to respond to e-mails from parents as well.

Given this, it seemed unlikely that many potential respondents would be unable to open

56

the survey link and complete the survey. Nonetheless, a few teachers may have been
discouraged. Another potential disadvantage is that computers are idiosyncratic and what
appears to look like a well designed survey on one computer may be gibberish on
another. To minimize this, I used an online survey publication tool, SurveyMonkey,
which is designed to work on a variety of platforms, including older computers. In
addition, I asked colleagues with a variety of computers and browsers to open the survey
to make sure that they had no difficulty clicking on responses and moving from page to
page.

The completed survey was piloted with six MPS teacher volunteers. Three were
elementary teachers; three taught high school. Half were enrolled in Professional Pay,
half not. I asked the teachers for feedback on both the ease of survey completion and the
content and language of the items themselves. Based on their responses, a few minor
changes were made.

Sampling
Interview Sampling Frame

During the two year data collection period, MPS operated eight middle schools. I
conducted interviews in three of those schools. By limiting my study to three schools,
rather than interviewing small numbers of teachers at a large number of schools, I was
able to learn about the leadership and culture within each school. Whyte (1955) found
that corporate and departmental culture strongly affected workers’ motivation to respond
to incentive pay. Other researchers have found that school culture plays a critical role in

shaping school responses to reform (McLaughlin & Talbert, 2001). School culture varied

57

across my sites, helping me identify potential sources of variation that mediate response
and understanding.

Since MPS preferences precluded random sampling (there were several schools
where MPS restricted access) or sampling of a high, moderate, and low enrollment school
(the district wanted me to speak with as many enrollees as possible and recommended
against interviewing in low enrollment schools), I conducted interviews in schools with
high, low, and moderate student achievement. I suspected that teachers3 in low achieving
schools might view their ability to earn rewards differently than colleagues in high
achieving schools and this design allowed me to look for those differences.

As previously noted, the first set of interviews took place at Washington Middle
School during the spring of 2003. Washington is low achieving relative to other MPS
middle schools and has an above average percentage of English Language Learners. A
relatively high proportion of students qualify for free and reduced lunch at Washington as
well. This site was selected based on Jay Ritterson’s recommendation. Ritterson, the
district employee responsible for the policy’s implementation during the 2002-2003
school year, recommended Washington as it had an above average enrollment rate,
allowing me to reach many enrollees in one location, and because he believed that
Washington’s teachers would be willing to cooperate with an outside researcher.

The following fall, interviews were conducted in two more middle schools,
Maclaven and Pressley. Maclaven has above average student achievement while Pressley

has moderate achievement (using district norms as the basis for comparison). Maclaven

 

3 All licensed school personnel are eligible for Professional Pay and are included in this
study. The majority of these licensed personnel are teachers and the entire group will be
referred to as teachers throughout in the interest of simplicity.

58

has an above average enrollment rate while Pressley is about average; again, this reflects
district preferences. One school has an above average percentage of English Language
Learners, while the other is below average. As was true of Washington, Maclaven and
Pressley serve relatively high poverty populations. Only two district schools were
relatively low in poverty and the district was not comfortable with my using either of
them for interviews (one is included in the survey sample, however).

Survey Sampling Frame

I initially contacted seven out of eight MPS middle school principals to ask for
permission to survey their teachers. The eighth principal was not contacted as the school
had recently been reconstituted and the district did not want the staff disturbed. Five
principals agreed, including the three principals in schools where interviews had been
conducted. The two additional schools increased the sample’s representativeness since
one was a relatively low poverty school and both had below average enrollment rates.
Within each school, all teachers were invited to participate in the survey.

One potential problem with the survey sampling frame is that the district and
principals restricted the sample. If the principals who chose not to allow participation are
systematically different than the principals who agreed to the study and if these
differences affect teacher response to Professional Pay, this may skew results.

Data Collection
Interviewing
I spent three days in Washington, two and a half days in Pressley, and two days in

Maclaven. Prior to my arrival, each school’s principal announced my upcoming visit

59

during a staff meeting and made copies of the interview protocols and consent forms
available.

The majority of the interviews were conducted in the teachers’ lounge or a comer
of the library. A few teachers wanted more privacy; those interviews were conducted in
empty classrooms. In two cases, teachers asked me to conduct interviews in their
classrooms while students were watching educational videos.

Teachers self-selected into the study, but I did encourage participation. Initially,
teachers were suspicious of my intent, worrying that I was a “district mole” as one
teacher put it -- a researcher who had been planted by the district to gain confidential
information — or that I would portray them as lazy or unintelligent. The fact that I was
affiliated with an out-of—state university and had no ties to the district outside of my
interest in Professional Pay helped assuage the first fear. My own experience as a middle
school teacher helped with the second. Many teachers told me that they decided to talk to
me because I was “one of them,” rather than a hostile outsider. As I gained the trust of a
few teachers, I asked them to encourage their colleagues to participate as well. Several
teachers noted that they agreed to talk to me because of the encouragement they received
from a trusted colleague. As a result, some groups within the school may have been more
heavily sampled than others.

At Maclaven, the principal included the news of my arrival during morning
announcements, telling teachers my location and asking enrollees to report for a brief
meeting before school that morning, during which he encouraged them to speak with me.
At Washington, teachers were reminded of my interest in interviewing them during

morning announcements midway through my visit.

60

While I attempted to interview all interested teachers, there were several cases
where I already had interviews scheduled during a teacher’s available times. This was a
particular problem at Pressley, since teachers there seemed particularly hesitant to agree
to be interviewed early in my visit. When I returned home from the field visit, I asked
Pressley’s principal to forward an e-mail to the staff asking them to contact me if they
were interested in being interviewed. Three teachers responded and I interviewed them
over the telephone.

With the exception of one teacher who was uncomfortable with taping, all of the
interviews were audio-taped. In addition, I took some handwritten notes during the
interviews. I tried to keep these to a minimum as the teachers tended to focus on reading
my notes, rather than the interview itself. Whenever there was a break between
interviews, I would write down my immediate thoughts and impressions as well as any
irregularities, such as the fire drill that interrupted one interview.

Forty-nine teachers were interviewed, twenty-three in the spring and the
remainder in the fall. From 24 to 54 percent of the teachers in sampled schools agreed to
be interviewed, for an average of 37 percent (Table 2.1). Some of this variation may be
explained by the fact that I was able to spend the most time in the smallest of the three

schools and that school had a higher proportion of teachers interviewed as a result.

61

Table 2.1: Sample Description and Professional Pay Enrollment — Interviews

School Percentage of Percentage of Percentage of
teachers interviewed teachers enrolled in enrolled teachers in
in school school sample

Pressley 24 % 50 % 50 %

Maclaven 38 % 40 % 36 %

Washington 54 % 73 % 48 %

Sample Average 37 % 54 % 48 %

As mentioned above, I spent varying amounts of time in each school (three, two
and a half, and two days) and the number of interviews in each school reflects those ’
differences (23, 14, and 9 plus 3 telephone interviews). Quite a few teachers commented
during later interviews that they agreed to talk to me because a colleague recommended
that they do so. This word of mouth advertising resulted in the number of interviews per
day increasing each day that I was in a school.

Another factor may have been the collegiality within each school. The number of
interviews was largest in the school with the highest overall collegiality score (as
measured by the survey) and lowest in the school with the lowest collegiality score. In
less collegial schools, teachers may have spent less time interacting and valued each
other’s recommendations less, moderating the impact of any word of mouth advertising.

Within the schools where interviews were conducted, the percentage of teachers
enrolled in Professional Pay varied from 40 to 73 percent, with an average enrollment
rate of 54 percent (Table 2.1), but a slightly lower percentage, 48 percent, of interviewed
teachers were enrolled. In two schools, school-wide enrollment rates and the enrollment

rates of teachers who agreed to be interviewed were very similar. In the school with the

62

highest enrollment rate, non-enrollees were overrepresented in the interviews. This is
somewhat surprising since we would expect enrollees to be more willing to discuss
Professional Pay. 1 emphasized that I was interested in everyone’s opinion, not just the
opinions of enrollees, so I may have inadvertently introduced this bias.

As already noted, I also interviewed each school principal in his or her office, as
well as the three person team responsible for the policy’s implementation in one
respondent’s office suite.

Data Collection Through Surveying

My online survey was administered during the spring of 2004. As potential
respondents may be hesitant to open e-mail from an unknown person (Dillman, 2000), I
initially contacted teachers by having prenotification letters placed in their school
mailboxes. This letter briefly described the project, provided each teacher with a unique
logon ID to be used when they accessed the survey system, and included my e-mail
address. I also informed teachers that respondents would be entered in a drawing for a
$50 Barnes and Noble gift certificate in an effort to boost my response rate.

Several days later, I sent each teacher a personalized e-mail reminding them of
my letter and providing them with the survey link. Dillman (2000) recommends that the
time between prenotification letters and the questionnaire be shortened from the normal
one week to a two or three day period when electronic mediums are used; this might
increase the likelihood that recipients will remember the letter, rather than deleting what
appears to be an unsolicited e-mail. He also recommends sending e-mails from an
individual address, rather than through a listserv to minimize confidentiality concerns and

increase response rate. I followed these recommendations. Non-respondents were sent

63

two reminder e-mails, encouraging their participation and resending their logon IDs and
the survey link.

A total of 90 usable surveys were returned. For medium effect sizes, this
provided power of approximately .91 with an alpha of .05 (Paul & Erdfelder, 1992). For
small effect sizes, the corresponding power is .16 (Paul & Erdfelder, 1992)" Since a
power of .80 is usually considered sufficient (J. Miles, 2004), this sample size allowed for
detection of effects that are medium or greater.

From 30 to 50 percent of the teachers in sampled schools responded, with an
average of 38 percent (Table 2.2). Response rates were higher in the schools where
interviews were conducted. It seems likely that this is because teachers in these schools,
whether they had participated in interviews or not, were somewhat familiar with the

study, increasing their comfort with the survey.

Table 2.2: Sample Description and Professional Pay Enrollment - Surveys

School Percentage of teachers Percentage of Percentage of
surveyed in school teachers enrolled in enrolled teachers in
school sample
Pressley 40 % 50 % 80 %
Maclaven 50 % 40 % 44 %
Washington 34 % 73 % 64 %
Phillips 30 % 34% 47%
Eason 32 % 20 % 12 %
Sample Average 38 % 41 % 48 %

 

4 Power is based on using a Chi-square contingency test to analyze difference among
subgroups, such as participants and non-participants.

64

The Professional Pay enrollment rate in the schools where surveys were
conducted was 41 percent and a slightly higher percentage of the teachers who responded
to the survey, 48 percent, were enrolled. Within individual schools, the percentage of
enrolled survey respondents was often quite different than the school-wide enrollment
average, with three schools having a higher percentage of enrolled respondents and two
having a lower percentage. There did not seem to be a particular pattern to this.
Inclusion in the interview sample, school demographics, and the school-wide enrollment
rate were all unrelated to the direction of survey sampling skew within each school.
Written Artifacts

While the interviews and surveys were my primary data sources, I also collected
written artifacts, searching district and local union websites, district and union
publications such as newsletters, and the local newspapers for policy-relevant documents.
The implementation team also provided several documents, such as timelines, that had
been created largely for use within district headquarters.

Data Analysis
Interview Analysis

The first step in my analysis was transcribing interviews and adding my field
notes to the transcripts. Following that, I read the interviews, looking for patterns in
responses. Through this inductive process, I developed a thematic conceptual matrix (M.
B. Miles & Huberman, 1994) of perceived response factors. For example, teachers could
respond to one perceived response factor, prior professional commitments, by adding
additional rewarded commitments, by substituting rewarded ones for financially

unrewarded commitments, or by choosing not to participate in rewarded activities at all.

65

The resulting matrix helped reduce and display the data, while also sending me in new
conceptual directions. Several factors that appeared important to respondents (e.g.,
reluctance to substitute rewarded offerings for teachers’ own professional development
agendas) had not been highlighted in my initial framework. This sent me back to
examine relevant literature, allowing me to better conceptualize my observations. After
reviewing the relevant literature on teachers’ professional development decisions, I began
to conceptualize teachers’ reluctance to trade their previous professional development
agendas for the rewarded one in terms of professional autonomy.

I then returned to the transcripts, searching for evidence confirming or casting
doubt upon my emerging conceptual matrix as well as my initial hypotheses. This
deductive process helped me refine my developing framework while also highlighting the
complicated nature of the responses. In many cases, for instance, a teacher’s story
supported or refuted my hypotheses but these simplified classifications were unsatisfying
as I sought to understand the influence of and interactions among various factors. For
example, almost half of interviewed teachers cited insufficient time as a factor in their
lack of active participation. For many of these teachers, the effect of this perceived lack
of time was exacerbated by a belief that rewarded offerings were not professionally
worthwhile — teachers believed that they could not take on additional activities and were
also unwilling to substitute activities. Looking at the percentage of teachers citing time
while ignoring the overlap with those who had doubts about the value of rewarded
activities could lead to ineffective policy recommendations. In sum, coding a response in

one way or another reduced the response in ways that obscured important information.

66

To help with this dilemma, I sketched a series of small case studies highlighting
the messy, complex nature of response. In these short cases, I constructed a narrative
around each person’s response, attempting to capture the array of factors that they
described, as well as the logic of their decision-making process. For example in
explaining Annette’s case, I began by constructing a picture of her beliefs about teaching.
Annette took her own professional development seriously and thought carefully about
what opportunities were worthwhile and how to incorporate what she was learning into
her practice. I then described Annette’s impressions of the way Professional Pay did (and
did not) fit into her vision of professional learning.

After writing the cases, I analyzed them for further insights, particularly
relationships that I might not have seen in the larger data set. For example, Annette’s
story demonstrated that time pressures and professional autonomy can work together to
depress response.

Another part of my analysis was the construction of a series of analytic memos.
These memos were based on my conceptual matrix and helped flesh out my emerging
analysis. In some cases, the memos supported my initial analysis while in others, my
ideas changed as my understanding deepened. One early memo looked at the concerns
teachers expressed during the pilot interviews. When I began the memo, I knew that
issues of “fairness” had been mentioned repeatedly but I did not have a rigorous
conception of what teachers meant by fair. As I began to describe the phenomenon and
look for confirming and disconfirrning evidence in the transcripts, I began to see that
fairness, as used by respondents, was not one factor. Some teachers talked about the

unfairness of teacher evaluations as a pay mechanism. These teachers were actually

67

expressing a lack of understanding since teacher evaluation is not a plan component.
Other teachers talked about the possibility that promised rewards would never be paid —
they had little short-term trust in the district. Others expressed fears that rewarded
opportunities would be available to some teachers and not others — they had equity
concerns. Writing the memo allowed me to move away from the ambiguity of fairness
and develop a more clearly conceptualized set of factors.

While this narrative is linear, my analysis was not. I moved back and forth
between the different procedures, picking up and sometimes discarding potentially
fruitful ideas along the way. While many of the threads that ran through my interview
analyses are included in this dissertation, some are not. The story of response I tell in this
dissertation captures some of the complexity of the Professional Pay Plan, but other
aspects will be highlighted in future writing.

As I moved through this iterative process, I tried to attend carefully to issues of
validity - descriptive, interpretive, and theoretical (Maxwell, 1992). By taping,
transcribing, and repeatedly listening to the interviews, I maximized the accuracy of my
reporting, thereby increasing descriptive validity. Interpretive validity is strengthened
when a researcher tries to base understanding on respondents’ perspectives rather than
imposing his or her own preconceived notions onto the situation. “Pure” induction is, of
course, not possible, given the conceptual frames that human observers carry into all
inquiry (Popper, 1963). I strengthened interpretive validity through my inductive reading
of the data and by constantly returning to teachers’ own accounts of events when trying
to make sense of their stories. To strengthen theoretical validity, the researcher must

consider whether her conceptual explanations would appear reasonable to members of the

68

community being studied. As Cusick (1983) notes, the field researcher should strive to
“unravel and explain the complexity of the events so that others who share similar
circumstances may find ways to express and understand their world” (p. 143). After
conceptualizing my observations, I returned once again to the data to see if the ideas I
was espousing echoed the ideas, if not the language, expressed by respondents. I also had
the opportunity to discuss my emerging hypotheses with district implementers, checking
to see whether my conceptualization made sense to them.
Survey Analysis

The surveys allowed me to systematically measure factors that appeared critical
based on existing theory and my emerging understanding of the interview data. Before
beginning my analysis, I developed item clusters related to constructs of interest. This
was a relatively straightforward process, as I had mapped survey items onto my
conceptual framework during the design phase. So, for example, teachers’ prior
experiences with district policies were explored using four survey items related to
agreement with, stability of, and consistency of district policies. Another cluster of nine
items related to teachers’ understanding. In three cases, I wanted to construct a new
variable to appropriately capture the contribution relevant items made to the construct of
interest. I used a Rasch Rating Scale Model to develop scaled variables for Perception of
Administrators, Perception of Colleagues, and Perception of District (see Appendix II for
details).

I also had to address the question of weighting, since enrollees were slightly
overrepresented among survey respondents. Statistical tests were performed using both

weighted and unweighted data; results were not significantly different. Weighted results

69

are reported throughout the analysis, as these are generally considered more accurate
reflections of the underlying population.

I analyzed responses both as an aggregated body of information and within
particular groups - such as participants/nonparticipants — using appropriate statistical
techniques. To begin my analysis, I ran a series of frequency distributions and crosstabs,
providing an overview of teacher response and an emerging understanding of
relationships among variables. I compared findings for various subgroups, such as
enrolled/non-enrolled and individual schools. As with the interviews, patterns began to
emerge which I tested for significance. I also used the survey data to test the robustness
of my interview findings.

Putting It All Together

The interview and survey data complemented each other, with interview data
providing more detail and complexity, and survey data enhancing generalizability and
allowing statistical analysis. Twenty-four respondents completed both interviews and
surveys. These were spread among the three schools in roughly equal numbers, with 10
not enrolled and the remaining 14 enrollees. This overlap provided me with the
opportunity to check both the consistency of teachers’ stories and the convergent validity
of the interview and survey instruments. Stories were generally consistent and, when
teachers’ responses changed over time, these changes were in expected directions. For
example, several teachers appeared to have learned more about the policy between the
interview and the survey and these changes were more pronounced with the earlier

interviews.

70

Both instruments appeared to be measuring similar underlying constructs. For
example, my impression of policy understanding based on the interviews was similar to
responses to understanding items in the survey. Convergent validity, on its own, does not
imply content validity — it is possible that I was actually measuring something other than
the construct of interest in both cases - but it does support using the survey data to
generalize from interview findings as the same constructs appear to be measured in both
cases.

Once I had established that both instruments seemed to be functioning in similar
ways, I began integrating the data. I searched for places where the two methods yielded
similar findings and those where linkages were weak. So, for example, I found that both
methods produced similar findings about time pressure’s effect on enrollment but that
teacher reports of future plans for active participation were inconsistent between the two
methods. Throughout this process, I kept returning to the data, seeking to disqualify
hypotheses where support was weak. The written artifacts further rounded out the
analysis. I compared the stories told by teachers, district personnel, and school principals
to the evidence contained within the written artifacts. In many places, the written
artifacts confirmed what I was told while, in others, they illuminated sources of confusion
among respondents.

I used the three sources of data — surveys, interviews, and written artifacts — to
check for consistency. Johnson and Onwuegbuzie (2004) note that, “If findings are
corroborated across different approaches then greater confidence can be held in the
singular conclusion; if the findings conflict then the researcher has greater knowledge and

can modify interpretations and conclusions accordingly” (p. 19). When interview and

71

survey results converged, this triangulation strengthened my findings. When
contradictions arose, I revisited my data and relevant literature to seek out more robust
explanations

Challenges and Limitations

I faced several challenges during this study, the most problematic of which were
selection bias and the limitations of self-reported data. There were teachers in each
school who were unwilling to be interviewed. There also may have been interested
teachers who were so heavily scheduled with parent and team conferences during my
short visit that they were unable to find 45 free minutes to talk. If either of these groups
differed from respondents across measured constructs, their omission introduced bias. As
mentioned earlier, respondents often encouraged their colleagues to agree to interviews.
There is evidence that different professional and social communities exist within schools
(McLaughlin & Talbert, 2001) and my interviews probably overrepresented some of
these communities while missing others.

There may be selection bias in the survey data as well. Nonrespondents may have
differed in meaningful ways from respondents and the electronic medium used may have
encouraged some teachers while intimidating their less technologically savvy colleagues.
I did take steps to moderate the effect of selection bias, such as weighting survey data by
enrollment status and performing separate analyses of various subgroups. These steps
helped avoid disproportionate representation, at least along some obvious parameters.

Virtually all of the data used in this study was self-reported. Respondents may
have deliberately misled me in some cases. Teachers may have been hesitant to attribute

financial motivation to their actions and instead attributed those actions to a concern for

72

student welfare. This may be a particular problem in a profession such as teaching where
there is normative pressure to place student welfare above other considerations (Lortie,
1975). In other cases, teachers may have believed they were telling me the truth but
suffered from a lack of self-knowledge. For example, teachers may have honestly have
believed that they were working extremely long hours and could not add another thing to
their busy schedules when, in reality, work that went home with them each evening was
rarely touched and they enjoyed a great deal of leisure time. To minimize these
problems, I looked for evidence outside of my study supporting or refuting my findings.

My research design limited my study in two important ways. I chose to conduct a
single case study, allowing for a thicker understanding of the forces at work within that
case, but introducing limitations, among them that the lack of a comparison group leads
to an absence of controls, the process of being studied may affect findings while the
“comparison” group (the status quo) does not undergo study, the group being studied may
differ from the status quo group for reasons that have nothing to do with the policy under
study, and the group being studied may be responding to policies and contextual factors
other than the ones being studied (Campbell & Stanley, 1966). Campbell (1975) suggests
that his own earlier criticisms of single case studies may have been a bit strong and
further suggests that these limitations can be minimized by considering the predictions
generated by the theory as a body of evidence and retaining the theory only if most of
these predictions prove to be true for most of the observed subjects (Campbell, 1975). If
the researcher holds her theoretical expectations lightly, she will be less likely to lend

credence only to that evidence which confirms her predictions. When assessing the

73

appropriateness of my theoretically grounded predictions, I remained open to surprises
and actively looked for disconfirming evidence.

A further limitation is that the treatment under study is unique. Policy particulars
are likely to condition responses in significant ways and these particulars vary widely
across contexts, with Professional Pay looking very different from alternative
compensation policies in places such as Denver, CO and North Carolina. In isolation,
Professional Pay’s uniqueness places tight constraints on what this study teaches us about
teacher response but the study does not exist in isolation. Researchers are slowly
developing a body of evidence about responses to various compensation policies which,
while they individually share the limitations of my own work, begin to provide us with
more generalizable insights when considered together.

Finally, the most significant limitation of the study is that I gathered no data
concerning teacher effectiveness. Professional Pay’s proponents hope that it will
improve teacher practice, yet I was unable to test whether active participation influenced
practice. While teachers who had participated in rewarded activities generally reported
little change to their practice, self-reported data is problematic. It is possible that practice
changed in ways that teachers were unable or unwilling to recognize and I am
underestimating the policy’s effect.

Another related problem is the policy’s attractiveness to high quality teachers — a
group that the district would like to retain. If active, enthusiastic participants tended to be
better teachers, paying those teachers more and possibly increasing their retention rate
makes sense, whether or not their practice changes as a result of rewarded activities.

Unfortunately, I was unable to systematically assess teacher quality. Student

74

achievement data was not available at the teacher level and some of the school principals
were uncomfortable evaluating teacher performance. Asking teachers about their
colleagues was problematic as well, since teachers rarely observe each other’s practice
and questions about colleagues’ performance might have undermined the trust that was
critical to an effective interview. Extensive classroom observation of practice was
beyond the scope of this study. In sum, further work will be needed to develop
professionally responsible and respectful ways to obtain information on teacher

performance before the policy’s effect on practice and retention can be determined.

75

Chapter Three

A Portrait of Professional Pay

In this chapter, I describe the Minneapolis Professional Pay Plan. Like many
policies, it is impossible to reduce Professional Pay to a static, monolithic representation.
Before we can fully understand the policy, we first need to understand the history leading
up to its adoption. The first section of this chapter will provide a brief history of the
district’s decision to offer Professional Pay. Following that, I describe the policy
drawing on the information embedded in various policy representations, including the
enabling legislation, district and union literature, the teacher’s contract, newspaper
articles, and discussions with the district team responsible for implementation. As is the
case with all policies, “what” the policy “is” is neither simple nor straightforward, as
these sources are not all consistent in content or emphasis. Finally, I locate the policy
within the typology described in Chapter 1 and discuss the agency predictions arising
from that placement.

Two major themes emerge from the resulting portrait of Professional Pay. The
first theme is one of complexity. The policy is complicated and constantly evolving.
Again, this is often the case with education policy — districts must try to satisfy a diverse,
ever-changing group of stakeholders both within and outside of the district — but the
complicated policies that result are difficult for teachers to understand. In addition,
complexity and shifting parameters allow stakeholders to interpret the policy in varied
ways, depending of their beliefs and areas of expertise. This discretion may increase the

local relevance of the implemented policy but may also further increase teacher

76

confusion. The second theme relates to the district’s recognition of the need to provide
teachers with a mechanism to increase their technical capacity if teaching quality is to
improve. The district understands that working harder may not be enough; teachers may
also need to learn how to direct their efforts in ways that will help students.

A Brief History

Minneapolis is a large city of approximately 380,000 residents. St. Paul, a city of
280,000, directly abuts Minneapolis, making the urban area even larger. According to
the US. Census Bureau, the median Minneapolis household income in 1999 was $37,974
with approximately 67 percent of employed residents working in management,
professional, sales, and office occupations. Almost 20 percent of Minneapolis’ families
with children live below the poverty line, with seven percent of families with children
earning less than $10,000. While most (85 percent) of the city’s residents who are 25
years of age or older have graduated from high school, a much smaller percentage, 37
percent, have graduated from college. Nineteen percent of residents speak a language
other than English at home and 14.5 percent of residents were born outside of the United
States. In 2000, there were 73,117 children between the ages of five and 19 living in the
city.

Minneapolis Public Schools (MPS) serves approximately 47,500 students, two-
thirds of whom are eligible for free and reduced lunch and more than 11,000 of whom are
English language learners. The system has approximately 8,300 employees, 4,603 of
whom were teachers in the 2002-2003 school year. Per pupil spending was $10,226
during 2002-2003 with the state providing $7,439 of that amount (Facts you should

know, 2003). In comparison, national per pupil spending was $7,396 during the 2000-

77

2001 school year and Minnesota was close to the national average at $7,645 (St. John,
2003). While spending is substantially above the state and national averages, large, urban
districts with substantial percentages of English Language Learners and low income
students do typically spend more than these averages. During the 2002-2003 school year,
average spending for school districts with 40,000 to 99,999 students, 20.8 percent or
greater of students eligible for free or reduced lunch, and a large central city location was
$8,533 (Public school district finance peer search, 2004). Lower spending districts in
this peer group were often located in the South (El Paso, TX spent $6,889 per student)
while spending in demographically similar Midwestern districts was closer to
Minneapolis’ spending levels (Indianapolis, IN and Columbus, OH spent $11,184 and
$9,607, respectively).

Average student achievement is low relative to other districts in the state; fewer
than half of the district’s students passed the mathematics, reading, or writing portions of
the state test in contrast to a statewide average of 72 percent (White, 2003). Again, this
pattern is typical of high poverty, urban districts. Recently, the NAEP mathematics and
reading examinations were administered to eighth graders in ten urban districts. Average
achievement in these districts was significantly below the national average. Almost half
— 49 percent — of urban students were below basic in mathematics achievement compared
to 33 percent of students nationally (Results of the first NAEP trial urban district
assessment in mathematics, 2004). Reading results were similar; 41 percent of urban
eighth graders were below basic in mathematics achievement compared to 28 percent of
eighth graders nationally (Results of the first NAEP trial urban district assessment in

reading, 2004).

78

Prior to the implementation of Professional Pay, all of the district’s teachers were

paid according to the steps and lanes schedule. This schedule, similar to the ones used in

most school districts, rewards teachers for years of experience and the attainment of

advanced degrees. Table 3.1 is a streamlined version of the salary schedule used for the

2001-2002 school year. The actual salary schedule has more steps (changes based on

years of experience) and lanes (changes based on educational attainment) than this

truncated version but the basic pattern holds - salaries increase with experience and

educational attainment.

Table 3.1: Selected Salaries From the 2001-2002 Minneapolis Steps and Lanes Salary

Schedule

Step (Years of

Experience)
1

3
5
8
15
20

30

Around five years ago, a small group of district administrators and teachers began

BA
$29,521
$32,567
$36,066
$44,363
$45,363
$46,863

$49,363

BA plus 15

credits
$30,179

$33,033
$36,579
$45,033
$46,033
$47,533

$50,033

Lane (Educational Attainment)

MA
$32,051
$35,550
$39,002
$45,879
$56,695
$58,195

$60,695

MA plus 15

credits
$32,997

$36,566
$40, 301
$47,049
$59,392
$60,892

$63,392

informally discussing the idea of offering an alternative salary schedule. While the

concept appealed to the discussion group, one group member commented, “it was very

79

hard to get your hand around how to actually make it work.” Given the difficulty of
turning abstract ideas into a concrete reality, attention to the issue waned for a while as
other work responsibilities intervened.

By 1998, a small group of people committed to the idea of alternative
compensation created the more formal Continuous Improvement Compensation
Committee, or CIC, to work on the issue again. The key problem, in their eyes, was
reframing the pay system to make it more “meaningful.” They hoped to reward and
encourage excellence and were unsure whether the uniform salary schedule was an
effective way to do so. CIC’s membership was fluid and, at times, included members of
the Human Resource Department, the (now defunct) Policy and Planning Department, the
PDP (teacher evaluation system) coordinator, the lead mentor, the superintendent, local
union leadership, the union lobbyist, the union’s executive board, school administrators,
local business leaders, community members, and classroom teachers. Various
subcommittees spun off to look at particular issues related to alternative compensation
and work continued for two years.

The Minneapolis Federation of Teachers (MFI’) was involved in and supportive
of CIC’s efforts from the beginning. The long-time president of the MFI’, Louise
Sundlin, participated in the AFT ’3 Alternative Compensation Task Force and the Teacher
Union Reform Network (TURN). The Alternative Compensation Task Force arose in
response to compensation discussions that were taking place nationally and at the state
and local levels. Rather than simply react to initiatives once they arose, national AFI‘
leadership decided to examine both the traditional uniform salary schedule and

alternatives so that they could be proactive on the issue of alternative compensation. In

80

April, the AFT approved a resolution arising from the group’s work stating that the AFT
"believes it is time to explore viable, fair and educationally sound teacher compensation
options that will raise salaries while contributing to efforts already under way to assure
hi gh-quality, well-prepared teachers for all students" (Real incentives for professionals,
2001).

TURN was started in 1998 with the goal of restructuring teachers’ unions so that
they are better able to increase student achievement. T wenty-one local unions are
members of TURN, including Cincinnati and Denver, both of which have also been
involved in alternative compensation policies. TURN emphasizes cooperative, rather
than adversarial, relationships among the union, its teacher-members, and administrators
(Teacher Union Reform Network). Both of these experiences, involvement in the
Alternative Compensation Task Force and TURN, increased Ms. Sundlin’s expertise in
the area of alternative compensation and her willingness to work with the district on
developing an alternative to steps and lanes. She was a willing and welcome collaborator
as CIC worked to develop a plan.

Over time, CIC went through several name changes, including Standards-based
Pay and Pay for Performance, and it was eventually decided that Professional Pay, a term
borrowed from the AFT’s Alternative Compensation Task force, was an appropriate
moniker for the nascent policy.

Meanwhile, the governor and state legislature were becoming interested in
alternative compensation as well. This was not entirely coincidental; the MFT’s lobbyist
had been discussing some of the district’s ideas with state officials for some time. In

2000, Governor Ventura began promoting the idea of state funding for alternative

81

compensation and the state House K-12 Education Finance Committee began considering
the issue. When legislation was drafted in the spring of 2001, legislators and the
governor drew heavily on MPS and MFI' personnel for guidance. The final statute does
not mirror the district’s plans perfectly. For example, MPS and MFT had not planned for
the new policy to completely eliminate experience credit (‘steps’) while the legislation
requires it to do so to be eligible for funding. Despite differences such as this, the
legislation does provide a framework within which the district can easily work.

In 2001, the state legislature passed a bill to fund “restructured” teacher
compensation systems within the state. The enabling legislation states,

A restructured teacher compensation system is established under

subdivision 2 to provide incentives for teachers to improve their

knowledge and skills and for school districts to recruit and retain highly

qualified teachers, and to support teachers’ roles in improving students’

educational achievement. (Minnesota 2001 Statute 122a.413-415).

The legislature further requires that the system must not be based on years of
service, must “encourage teachers’ continuous improvement in content knowledge,
pedagogy, and use of best practices,” and that the district must “implement an objective
evaluation system, including classroom observation, that is aligned with the district’s or
the site’s educational improvement plan.” Appendix III contains the full text of the
relevant statutes.

In response to this funding opportunity, the district quickly introduced
Professional Pay for the 2002-2003 school year. In order to qualify, a district had to

demonstrate union and district collaboration, implement an alternative to steps and lanes

82

that met the criteria outlined above, and sign up at least 25 percent of the teacher
bargaining unit for the new plan by July 3, 2002. Professional Pay easily met the first
criteria but the remaining two presented a challenge. The district submitted an initial
plan to the Minnesota Department of Education early in 2002 and had the full application
ready in March. The submitted plan was loosely specified. For example, while the plan
talked about paying teachers for participating in and demonstrating implementation of
targeted professional development opportunities, neither the types of opportunities that
would be offered nor the criteria for demonstration were clearly specified. Despite this
lack of specificity, the state accepted it for funding. Only the 25 percent participation
hurdle remained before the district qualified for the funds. The rather vague plan details
were quickly presented to teachers and, as an incentive to participate, teachers who
signed up were promised $650 bridge payments as long as they joined the plan by June
30, 2002. Over fourteen hundred teachers, or about 30% of the district total, signed up by
the deadline and the district received $5.1 million in program funding for the 2001-2002
and 2002-2003 school years.

While Minneapolis was not the sole recipient of the state’s restructured
compensation funding, they did receive the largest share of it. The work that had been
done within the district prior to the legislation’s passage put them in a strong position to
quickly pull together a successful application. In addition, the district’s and union’s
informal input into the legislative design process helped ensure that the final legislation
would be consistent with the type of alternative compensation that the district was

already laying the groundwork for through CIC.

83

As we have seen, this policy developed out of years of discussion involving
multiple stakeholders. Various parties including legislators, union officials, district
representatives, and classroom teachers had input that helped shape the policy. We will
see in the next section that the result is a policy that is both loosely specified and
complex. We have also seen that this policy arose out of conversations centered on
compensation. It was designed to use compensation as a tool to improve teacher quality.
An assumption underlying the legislation is that money motivates good teachers to
remain and inspires all teachers to improve. We shall see later in this chapter that not all
stakeholders see compensation as the heart of the policy and, in Chapter 5, that teachers
participate, or choose not to participate, in the policy for both financial and non-financial
reasons.

Policy Representations: Written Artifacts

There is no single document that is the definitive source of information about
Professional Pay. This description draws upon the enabling legislation, district literature,
union literature, the 2003-2005 teachers’ contract, and newspaper articles. Each text
represents a different face of the policy. The enabling legislation provides boundaries
that must be respected if funding is to continue. It provides a set of parameters within
which MPS must work. The district and union literature provide information about the
district’s hopes and plans for Professional Pay; the contract outlines the district’s
obligations. The district and union would like this policy to provide teachers with
multiple avenues to increased compensation and the literature they have produced reflects
this. Some aspects of the policy that are outlined in the literature are operational while

others provide glimpses of their long—term hopes. The contract is much more

84

conservative. The terms outlined in the contract provide information only about aspects
which are guaranteed to remain in place for at least two years and that will not evolve
during that period. Finally, the newspaper articles represent the public face of
Professional Pay.

Setting the Boundaries: Enabling Legislation

As mentioned previously, MP8 is able to offer Professional Pay because it is
being funded through a state grant authorized by Minnesota 2001 Statute 122a. 413-415.
In order to continue receiving funding, MPS must work within the parameters established
by the legislation. As already noted, the legislation states that funded policies must
provide teachers with incentives to improve their skills and must support teachers’ efforts
to improve student outcomes. The legislation goes on to specify particular criteria that
must be met. In a complete break with steps and lanes, the new policy cannot include
years of service as a salary determinant. The policy must encourage teachers to improve
their content knowledge, pedagogical skills, and use of “best practices.” In addition, it
must include an “objective evaluation system, including classroom observation.” The
final criterion relevant to our discussion is that the policy must provide “career
advancement options for teachers retaining primary roles in student instruction.”

While these criteria provide the district with quite a bit of latitude, we will see in a
moment that whether or not Professional Pay actually falls within the stated parameters is
debatable. The current policy uses an action research requirement to evaluate teachers,
rather than classroom observation. When this legislation was being crafted, scholars and
consultants were focusing on linking teacher pay and classroom performance. Allan

Odden and Carolyn Kelley, leading scholars in the field of alternative teacher

85

compensation, had released a book advocating knowledge- and skills-based pay that
includes a classroom performance component (Odden & Kelley, 1997). This book draws
upon academic research but is written in a style that is accessible to policy-makers as
well as scholars. Through their affiliation with the Consortium for Policy Research in
Education (CPRE), Odden, Kelley, and their colleagues were also offering reports,
conferences, and consultations to policy-makers. Charlotte Danielson, a consultant
affiliated with Educational Testing Service, was also promoting her “framework for
teaching” as an evaluative tool for assessing teacher performance. In order to use the
framework, a series of teacher behaviors are observed within the classroom setting.
These assessments can then be used to evaluate teacher performance in an alternative
compensation system. In addition to her consulting work with various school districts
interested in alternative compensation, Danielson released a book and various supporting
materials (Danielson, 1996). In this environment, it seems likely that Minnesota
legislators wanted approved programs to base at least some portion of teacher pay on
actual classroom observation, rather than relying on action research as a proxy for
observation.

In addition, the legislation calls for career advancement options to be available yet
the only career advancement option available for classroom teachers is the increased pay
they receive. It is unclear whether this is all the legislature had in mind when they
included this requirement; it seems likely that they were looking for a policy that
included career ladders such as those that were briefly popular in the 1980s and are
beginning to reappear in districts such as Denver, Colorado. The Milken Family

Foundation was promoting their Teacher Advancement Program (TAP) when the

86

legislation was written. One of the five core principles underlying TAP is that teachers
should be provided with multiple career paths, such as becoming mentor or master
teachers, allowing teachers to increase their salaries while remaining in the classroom.
As is true of CPRE, the Milken Family Foundation was promoting their ideas to policy-
makers throughout the country when the legislature was considering alternative
compensation.

One underlying premise of the legislation is the belief that teacher learning is a
necessary condition for teachers’ improvement. This may seem like a straightforward
idea but it is one that is often ignored when compensation policy is designed. For
example, policies that pay teachers for test scores assume that teachers are able to raise
these scores if they are provided with sufficient incentive to do so. The assumption is
often made that teachers are capable of both working harder and working smarter - they
know what to do but it takes incentives, most often financial ones, to motivate them to do
so.

This assumption—that agents are capable of increasing their production (in our
case, that teachers can increase student learning}—is not limited to education policy; it is
widespread. Corporate gainsharing and commission sales systems are based on the
assumption that ability is present and motivation, particularly financial motivation, is
both necessary and sufficient to increase production. The agency framework assumes
that sufficient technical capacity exists and that, given adequate incentives, performance
will improve. This may be a reasonable assumption in some cases, such as on an
assembly line. Once the job is learned, it may be possible to speed up the line by paying

employees on a piece goods basis. Applying this assumption to complex jobs, such as

87

teaching, is less reasonable. There are undoubtedly teachers who are simply lazy. These
teachers may know how to improve student achievement and be choosing to do what is
easy, rather than what they know will help students.

There is ample evidence, however, that some teachers lack professional
knowledge; this lack of knowledge might compromise their ability to improve their
teaching. Research on elementary mathematics teachers, for example, suggests that many
teachers possess neither the content knowledge nor pedagogical skills to teach
mathematics for understanding (Ball, 1990a; Ball, Lubienski, & Mewbom, 2001; Ma,
1999; Wilson, Floden, & Ferrini-Mundy, 2002). Approximately one-fifth of secondary
science and social studies teachers have neither a major nor a minor in those subjects or
related fields and the problem is even worse for secondary mathematics and English
teachers (Ingersoll, 2001). Other scholars note the significant differences between
teachers and their students. Sometimes called the “demographic divide,” many scholars
argue that teachers need to be prepared with new knowledge of diversity and how to
teach diverse student populations effectively (Cochran-Smith, 2004; Delpit, 1988;
Ladson-Billings, 1999; Seiler, 2001). It seems unreasonable to expect these
underqualified teachers to improve their teaching without providing them with
opportunities to learn how to do so.

It appears that the legislature recognized this when they designed Minnesota 2001
Statute 122a. 413—415. While the legislation’s ultimate goal is to use financial incentives
to improve student learning, the legislation does so by encouraging and rewarding teacher
learning. Rather than rewarding teachers only when they improve student outcomes (a

very difficult thing to measure), this policy rewards teachers for learning how to do so

88

and applying that learning. We shall see in the next section that MPS carries this one step
further, by building a program that not only rewards teachers for taking steps to increase
their technical capacity but that actually attempts to provide them with opportunities to
do so.

A final point that should be noted is that, while the legislature recognized the role
of teacher learning in raising teacher quality, this legislation was primarily about
compensation. Increased compensation was assumed to be the key motivator for
improvement. It does not appear that the legislation was intended to serve primarily as a
mechanism for professional development; rather, the professional development was in
place because increased capacity is a necessary precondition before new compensation
monies can be allocated wisely.

Ideals and Hopes: District and Union Literature

MPS had been discussing an alternative compensation system for several years.
The legislation discussed above allowed them to turn those ideas into reality. One way
that the resulting policy is represented is in literature that the district and union make
available to teachers and the public, both in print and online. A full portrait of the policy
cannot be found by consulting one coherent, definitive document; rather, one must
connect the dots between various documents released over time. As mentioned, once the
enabling legislation was in place, a plan was quickly submitted to the state and approved.
This plan was vague and continued to evolve even as it was being considered for
approval.

Once MPS received tentative approval from the state, teachers needed to learn

about the policy quickly since the district needed 25 percent participation to qualify for

89

funds. Informational meetings were held, particularly with union stewards, and literature
was disseminated through mailings, monthly newsletters, and the union website. This
section will discuss the policy as represented by this literature. For many teachers,
written information would have been the most accessible way to learn about the policy.
Some interviewed teachers mention that they had also been invited to or attended
informational meetings; however, these took place outside of the regular school day and
were infrequent so they were not a viable opportunity to learn about the policy for the
majority of teachers. In addition, several teachers who had attended meetings
complained during interviews that they had left the meetings confused. We shall see that
this policy is quite complex; it is not surprising that teachers would have difficulty
understanding it based on an after-school meeting. Even teachers who attend meetings
will probably still need to refer to written documents for clarification.

The literature outlines both underlying ideas and specific components. While
emphasis varies, the body of literature is internally consistent for the most part. The
policy was changing as literature was published so details do change but the fundamental
ideas and main components remain the same throughout. The policy is based on three
fundamental ideas: (1) teachers should be rewarded for increasing their skill in
improving student learning; (2) focused professional development can assist teachers in
increasing their skills; and (3) no teacher should be penalized for choosing to participate
in the voluntary plan (Ritterson, 2003).

Under the plan, teachers are rewarded in various ways. Some of these were
operational during the 2002-2003 school year (the first year of the program), others

became operational during the 2003-2004 school year (when components were written

90

into the new teacher’s contract), and others are more long-term in nature. The rewards
fall into four categories: results-based incentives, continuous improvement compensation,
capacity building awards, and extended professional responsibility compensation.

Results-based incentives are based on district- and school-wide achievement
(primarily measured in terms of test scores). District literature promised each participant
$1,000 if district goals were met. Most district goals were met during the 2002-2003
school year and participants will receive $600. School-wide achievement awards vary,
depending on the size of the school, and Professional Pay participants at each qualifying
school vote on whether to distribute the funds as teacher bonuses or use them for another
purpose.

Continuous improvement compensation pays teachers for college credits and
degrees; however, the compensation formula is different than under the old “steps and
lanes” schedule. In addition, district and union literature states that teachers must be able
to demonstrate that the courses are relevant to their work. Teachers earn $1,000 per ten
college credits up to a maximum of 60 credits, $2,000 for a masters (plus credits), $4,000
for a second masters or education specialist degree (plus credits), and $6,000 for a
doctorate (plus credits). Teachers are paid $3,000 for the attainment of National Board
for Professional Teaching Standards (NBPTS) certification. Continuous improvement
compensation also pays teachers $1,000 for every three Minneapolis Public Schools
University (MPSU) credits they earn. All of these amounts will be added to the teacher’s
base salary. In light of the problems some states are having with ongoing funding of

NBPTS certified teachers (Salzer, 2003; Teacher quality bulletin, 2004), there was some

91

concern about adding rewards to a teacher’s base, particularly the MPSU 1-2-3 rewards,
but all of the written plan documentation treats these amounts as ongoing.

Comparing the two compensation alternatives, the steps and lanes schedule pays a
teacher with one year of experience an additional $2,825 annually for attainment of a
master’s degree and the credits associated with it (starting with a Bachelor’s degree as the
basis for comparison). If a teacher has five years of experience, he or she will earn an
additional $2,994 annually for the master’s degree. Under Professional Pay, these
teachers would have approximately $5,000 added to their annual pay. This calculation
assumes that the teacher earns 30 college credits while obtaining the master’s degree.
This is the minimum number of credits that a teacher must take at St. Thomas University
(a local university) to obtain a master’s degree in education. While teachers under
Professional Pay no longer qualify for increases based on experience, or steps, teachers
actually receive more for degree attainment under the new plan than they do with
traditional lane changes.

One important difference between the two salary schedules is that district
literature regarding advanced degrees mentions that “classes must be relevant to the
teacher’s wor ” (Ritterson, 2003) to qualify for Professional Pay rewards. The steps and
lanes schedule does not include this restriction.

The $3,000 teachers receive for attaining NBPTS certification is also larger than
the amount paid under steps and lanes. Under steps and lanes, teachers who become
NBPT S certified will move one lane to the right of their current placement. This amount

varies but never exceeds $2,000.

92

Additional pay for successful completion of MPSU courses is not available to
teachers on steps and lanes.

MPSU courses are offered by the district at no cost to the teacher. To earn three
credits, a teacher must attend 10 hours of MPSU professional development, demonstrate
implementation of the strategies taught during the 10 hour training (videotaped lessons,
etc.), and demonstrate professional reflection on the strategies (study groups,
demonstration lesson, etc.). Teachers can take up to two MPSU courses per year and
earn credit for up to 27 courses during their careers. These courses are generally referred
to as MPSU 1-2-3. The district did not offer any MPSU 1-2-3 classes during 2002-2003
but plans to offer at least two in 2003-2004. ENVoYTM, the first MPSU 1-2-3 class, is
already underway with 183 participants. Math literacy will be offered this year as well
and several other courses are in the planning stages.

The district team currently responsible for implementing Professional Pay
(hereafter referred to as “the team”) hopes to eventually offer MPSU 1-2-3 credit for
school based professional development as well. This would increase the options
available to teachers and other eligible personnel and allow building personnel to
concentrate on issues that are meaningful within that particular school’s context. Plans
for this are still in the early stages and this is not mentioned in the literature.

Capacity building awards reward teachers for gaining skills that the district
deems critical. Plan participants are paid stipends of $500 for attending and
implementing “critical skill set” professional development offerings. The 2002-2003
offering was “Improving Instruction through Analysis of the Data.” There will not be

any critical skill sets offered during the 2003—2004 school year. According to the team,

93

they were told to concentrate only on MPSU course offerings this year because the
district wanted to see how the implementation and professional reflection components of
MPSU 1-2-3 worked in practice. There are also longer term plans to reimburse teachers
for the costs associated with certification in high need areas, such as special education.

There is not a strong conceptual distinction between MPSU courses and critical
skill sets. Both components center around professional development and both require
teachers to demonstrate that they are implementing what they have learned. The
difference between the two is a difference of degree, rather than kind. MPSU courses are
longer in duration (ENVoYTM included five required class sessions while “Improving
Instruction through Analysis of the Data” only required one class session) and require
more extensive demonstration of implementation and reflection. Reflecting the greater
demands of MPSU courses, teachers receive increases to their base salaries for successful
completion of these courses; smaller, one-time stipends are used to reward teachers for
critical skill set completion.

Extended professional responsibility compensation pays teachers for taking on
responsibilities beyond their instructional duties. Defining responsibilities and
reasonable compensation for this component has proven to be difficult, thus, it is not
clear when this part of the plan will be fully implemented. The literature consistently
includes this component but provides no detail as to amounts or qualifying
responsibilities. During the interviews, some teachers expressed the belief that particular
activities they participated in would be rewarded but there is nothing in the literature
supporting that level of specificity. During 2004-2005, the district will begin establishing

criteria to reward Math and Literacy Demonstration Teachers from the Leaming

94

Partnership (a district initiative that is operated outside of Professional Pay) as a first step
towards extended professional responsibility compensation. The team expressed
concerns that, “you also have to be careful because you can create all this stuff and then if
the money goes away and the district doesn’t build it in, then it [the compensation]
stops... And if you do that to teachers, you’re going to be in trouble. . . [teachers will say,]
‘That’s fine, then I’m not doing the coordination anymore either.”’ Right now, most of
the responsibilities that would be rewarded are done on a volunteer basis. Moving to paid
positions in an environment of unstable funding carries the risk that the district will be
unable to find volunteers if funding is cut.

Given that this policy is constantly evolving, the literature is surprisingly
consistent. One theme that appears throughout is that the policy is focused on teacher
learning. While results-based incentives and extended professional responsibility
compensation are explained in the literature, the literature mainly focuses on continuous
improvement compensation and capacity building awards. These are mentioned more
frequently and explained in far greater detail than the other two components. These are
the components that both directly reward teacher learning and that provide teachers with
additional opportunities to learn (through MPSU 1-2-3 courses and critical skill sets).
The ultimate end result the district desires in increased student learning but the district
recognizes that teacher learning is necessary for this to occur. Towards this end, the
district both rewards this learning when it occurs and provides teachers with the means
necessary to learn in the form of professional development opportunities.

Another idea that emerges from the district and union literature is that the policy

is centered on compensation. As is true of the enabling legislation, the district recognizes

95

that teacher learning must take place before teacher quality can improve so the district
has components in place to make learning opportunities available to teachers but this is,
first and foremost, about paying teachers in new and, in the district’s eyes, smarter ways.
It is assumed that the promise of increased pay will motivate teachers but the district also
recognizes that teacher learning may need to take place before this increased motivation
can result in better teaching.

Another notable feature of the policy is its complexity. While this complexity
provides teachers with multiple ways to increase both their professional skills and their
compensation, it also provides them with multiple ways to become confused. Simplistic
merit pay programs have a long history of failure and MPS tried to address this by
designing a nuanced compensation plan that provides teachers with a variety of ways to
increase both their pay and their teaching skills. Unfortunately, this comes with a price.
The policy is confusing. We shall see in Chapter 5 that some teachers simply ignore the
policy, professing not to understand it, while others harbor misconceptions that either
turn them against the policy or give them false hope that they will be rewarded for
activities that are, in fact, not rewarded.

The lack of a single definitive document describing the policy increases the
potential for teacher confusion as well. Various documents describe and explain the
policy and these documents do not refer teachers to a master document providing the
“last word” on the policy. (In fact, no such document exists.) While these documents
form a fairly consistent body of literature, plan features are highlighted more in some

documents than others and timelines for implementation are rarely clear.

96

Tempering the Ideals: The 2003-2005 Teacher’s Contract

At the time it was adopted, contract negotiations had begun for the 2003-2005
school years but the implementation timeline required the policy to be put into place
during the 2002-2003 school year in order to receive funds. During the first round of
interviews, teachers had very little information about the proposed contract. At the time
of the second set of interviews and the survey, teachers were operating without a contract
and the 2003-2005 contract was largely in its current form and awaiting ratification. The
draft contract was readily available for teachers to review at that time.

The 2003-2005 Teacher’s Contract (relevant sections of which are included in
Appendix IV) outlines the district’s contractual obligations. As we will see, these
obligations are considerably fewer than those outlined in district and union literature.

Results-based incentives based on district- and school-wide achievement are not
mentioned in the 2003-2005 Teacher’s Contract. The district was able to pay these
rewards without including them in previous contract language so their exclusion in the
2003-2005 contract language does not mean that the district will not pay these rewards,
just that they are not contractually obligated to do so.

Continuous improvement compensation is included in the contract. Rewards for
academic coursework, advanced degrees, NBPTS certification, and MPSU 1-2-3 credits
will all be added to the teacher’s base salary. The contract does not include the restriction
that academic coursework must be relevant to the teacher’s work in order to qualify for
rewards. According to one team member, teachers rarely earn advanced degrees that are
not relevant to their work so including this restriction in the contract was unnecessary at

this time.

97

Capacity building awards are also not mentioned in the contract. As with results-
based incentives, this allows the district to offer these rewards but does not obligate them
to do so. Critical skill sets and reimbursement of the costs associated with certification in
high need areas are not mentioned in the contract. The absence of language regarding
capacity building awards does not mean that they cannot be rewarded. According to the
team responsible for implementing professional pay, the district has paid these rewards
through the use of memorandums in the past and plans to continue to do so in the future.

The same is true of extended professional responsibility compensation. While not
included in the 2004-2005 contract, the team responsible for implementing Professional
Pay plans to implement extended professional responsibility compensation during this
contract period. At a minimum, they hope to reward Math and Literacy Demonstration
Teachers from the Leaming Partnership and PDP (teacher evaluation system)
coordinators during the current contract period.

The contract also establishes a beginning salary for teachers who sign up for
Professional Pay that is equal to their current salary under steps and lanes. Any
additional monies earned will be in addition to this beginning salary.

If the program is discontinued, teachers are to receive the higher of their final
Professional Pay salary or the step and lane salary corresponding to their education and
experience at the time of discontinuation.

A teacher who looked at the contract to learn about the policy would come away
with a very different view than someone who examined the literature discussed above but
there is overlap between the two documents (Table 3.2). In both union and district

literature and in the teacher’s contract, it is obvious that Professional Pay differs

98

significantly from steps and lanes. Both sources make it clear that additional years of

experience, the lynchpin of steps and lanes, are not rewarded. It is also clear from both

sources that teachers will be rewarded for professional development. Rewards for the

attainment of advanced degrees, NBPTS certification, and completion of MPSU 1-2-3

courses are mentioned in both sources.

Table 3.2: Comparing District and Union Literature to the 2003-2005 Teachers’ Contract

Plan Component

District- and school-wide
achievement rewards
Compensation for college
credits and degrees
NBPTS certification
MPSU 1-2-3

Critical skill sets

Reimbursement of certification
costs in high need areas

Extended professional
responsibility

Provisions for plan
discontinuation

District and Union Literature

Included

Included - only rewarded if
“relevant to the teacher’s
wor ”

Included

Included

Included

Included

Included

Not mentioned

2003-2005 Teachers’
Contract

Not mentioned

Included — rewards all
credits and degrees

Included
Included
Not mentioned

Not mentioned

Not mentioned

Included

There are significant differences between the district’s contractual obligations and

the implied promises in district and union literature, however. A teacher looking at the

contract would only learn that he or she would not earn step increases for experience, that

99

the magnitude of lane changes and the NBPT S certification award would change, and that
MPSU 1-2-3 courses would be rewarded. The literature also discusses rewarding
participants for meeting school and district goals, attending critical skill sets, and taking
on extended professional responsibilities. The district is not contractually obligated to do
any of these. Should financial difficulties or philosophical and political differences
reduce Professional Pay’s popularity and feasibility, the district can renege on the
promises made in the literature without legal ramifications. The implementation team
intends to keep the implied promises of the literature but, as will be discussed in Chapter
4, there may be forces at work within and outside the district making this difficult to do.

The possibility of plan discontinuation is not mentioned in any of the district or
union literature that I obtained. In contrast, the contract is quite specific about the impact
that a discontinuation decision would have on an individual teacher’s pay. This
difference may have arisen because both the district and union are interested in the
program’s success. They may feel that mentioning the possibility of plan discontinuation
in informational literature will dissuade prospective participants. Teachers often see
policies come and go; bringing the possibility of plan discontinuation to their attention
may encourage them to dismiss Professional Pay as another briefly fashionable policy
that can be safely ignored.

Another possibility is simply timing. The school year was well underway before
the current teacher’s contract was ratified. Most of the literature was produced prior to
contract ratification and the district and union may have been exercising caution about
making promises about plan discontinuation that would not find their way into the final

contract. This explanation, while possible, is somewhat implausible given the district’s

100

and union’s willingness to produce documents discussing rewards such as extended
professional responsibility awards that are not mentioned in the final contract.

With the exception of the discontinuation protections outlined in the contract, the
contract often promises less than district and union literature. The contract outlines the
district’s obligations; the literature, its aspirations. While the district and union may fully
intend to implement the program as it is presented in the literature, they do not have to do
so. As we will see in chapter 4, this ability to renege on promises made in the literature
may affect teacher’s responses.

The omission of certain components from the contract is not necessarily unusual
and may actually improve the policy’s chances for success. The prohibitive costs of
writing and enforcing a contract outlining every aspect of employment would make it
undesirable to do so (Jensen, 1994) and, in almost all cases, is not even a possibility. In
order to write a complete contract, both parties would have to be aware of all possible
future situations and the appropriate response to those situations so that appropriate
responses and rewards could be specified. Given that this is virtually never the case,
contracts, particularly employment contracts, are normally incomplete (Milgrom &
Roberts, 1992). The 2003-2005 Teacher’s Contract is no exception; it would be
impossible to fully specify all aspects of a teacher’s job. Looking more specifically at an
example related to Professional Pay, district personnel cannot specify which Professional
Pay courses will be offered to teachers over the next few years as a part of their
contractual obligation. Program popularity, the availability of trained facilitators, and
shifting professional development needs all affect course offerings and are unknown at

the time the contract is written.

101

In addition to the difficulty of fully specifying the employment contract, it may be
undesirable to do so even if it was possible. Using memorandums, rather than contract
language, to implement some policy components allows the district and the union to
make modifications and improvements (Conley, Muncey, & Gould, 2002). Since there
are few successful models of alternative compensation to follow, the ability to adapt and
modify proposed policies may be particularly important to their long term success. By
maintaining discretion over Professional Pay components, the district and union are able
to improve the program over time.

This benefit must be weighed against the possibility that limited contractual
obligations will lead teachers to conclude that the policy’s promises will not be kept and
that the policy itself may disappear. In an environment where teachers believe that the
district will live up to informal, as well as contractual, obligations, this may not be a
problem. This is not the case in MP8, however. We shall see in the next chapter that
many MPS teachers are concerned about the district’s ability and willingness to live up to
the literature’s promises and they are pessimistic about the policy’s long-term
persistence.

The differences between the teacher’s contract and district and union literature
may also foster confusion. Given the complexity of the policy, the potential for
misunderstanding is high even when written literature is consistent. When a legal
document, such as the teacher’s contract, appears to be describing a truncated form of the
policy, this will only increase misunderstanding. Teachers may be unsure of what to

believe — the promises embedded in the literature or those in the contract.

102

One message that is consistent between the literature and the contract is the
emphasis on teacher learning. Professional development is highlighted in the contract.
Every component mentioned in the contract relates to teacher learning while components
that are not as tightly coupled to professional development, such as extended professional
responsibility and district- and school-wide achievement rewards, are not part of the
contract. In addition, the contract explicitly states the importance of teacher learning in
the section related to Professional Pay. The contract states, “Teachers are encouraged to
continuously improve their professional skills and knowledge through coursework and
other professional development activities which enhance their instructional skills and
help them meet the needs of the students they teach.”

As is true of the enabling legislation and the district and union literature, the
contract presents Professional Pay as a compensation policy. The focus in the contract is
on the policy’s effect on teacher pay. While the links between MPSU1-2-3 courses and
increased pay are outlined, the policy is placed within the contract sections relating to
pay, rather than professional development. In addition, outlined obligations are financial
in nature; for example, the district is required to offer a specified financial reward
($1,000) for MPSU 1-2-3 course completion but they are not obligated to offer a
minimum number of courses each year.

The Public Face: Newspaper Articles

A final written representation of the policy can be found in the news media. An
analysis of the archives from the dominant local newspaper, the Star-Tribune, reveals
scant coverage of the pay plan. During 2002, there were only two articles in the

newspaper that primarily focused on Professional Pay. One appeared shortly before

103

teachers voted to approve the plan, the other shortly afterward. There are other articles
that mention Professional Pay or the enabling state legislation, but these are primarily
concerned with other issues and mention the policy only in passing.

During my site visits, all three schools had a copy of the current edition of the
Star-Tribune in the teachers’ lounge each day. Teachers often flipped through the paper
during breaks, stopping to read articles of interest. This was the only newspaper I saw in
the schools, although others may have been available. The articles that appear provide
little information to assist teachers in evaluating Professional Pay. In the article
appearing shortly after teachers voted to approve the plan, it is reported that Professional
Pay “would allow them to collect bonuses and raises for acquiring and using certain skills
believed to boost student achievement” (Shah, 2002). The article goes on to explain that
the district and union will collaboratively decide what teacher practices and training
would be rewarded. It is also noted that teachers will have to demonstrate that they are
using what they have learned through classroom observations (this is not the case in the
final plan). This is consistent with the policy’s emphasis on teacher learning and with the
other written artifacts discussed above but provides only vague information.

While the articles link the policy to teacher learning, thereby reinforcing that
message, they do little to minimize any confusion that teachers might be experiencing.
They are short, ignoring some aspects of the policy altogether while providing only an
overview of the rest. In addition, one article notes that, “District leaders hope to
implement the new system fully for all teachers in the next two years” (Shah, 2002,
emphasis added). This statement is inconsistent with district and union assurances that

the policy would remain voluntary and that there are no plans to eliminate steps and

104

lanes. There is nothing in the literature made available to teachers or in the contract to
indicate that full participation is a goal. The article may lead some teachers to conclude
that Professional Pay will lead to the elimination of steps and lanes. During the
interviews, some teachers did have this misconception and, in all cases, they did not want
to see this occur.

One idea resonating through all these written representations is that Professional
Pay is first and foremost a compensation policy. From the name of the policy itself
through the detailed explanations of the impact the policy will have on teacher pay, the
primary focus is changing the way that teachers are compensated. While the district
recognizes the need to increase technical capacity before teachers can respond to
increased pay, compensation lies at the policy’s core. We shall see in the next section
that local implementers may disagree with this focus.

Policy Representations: The Local Implementers

Written policy artifacts are only one source of information about policies.
Teachers also learn about policies from their own and colleagues’ experiences and from
school and district administrators. When data collection occurred for this study, this
policy was so new that few teachers had been able to learn much from their colleagues’
experiences. The limited experiences that teachers have had with the policy and shared
with colleagues will be discussed in Chapter 5.

School administrators are largely uninvolved with the policy. None of the
principals interviewed knew a great deal about the policy or felt that they had an active

role to play in its enactment or success. The teachers agreed. None of the interviewed

105

teachers mentioned any role for school administrators in their participation decision or in
the implementation of Professional Pay.

District personnel, however, play an active role in dissemination of policy
information and its implementation. Our portrait of this policy is a charcoal outline at
this point. We know what the boundaries of the policy are but not how it is enacted. It is
the district implementers who breathe life into it and pull some parts of the policy into the
sunlight while casting others in shadow.

Professional Pay is administered by Teacher and Instructional Services (TIS), not
Human Resources. This placement highlights the district’s early commitment to using
Professional Pay as a tool for teacher growth rather than a merit pay emphasis. T18 is
populated by individuals who are largely committed to teacher growth and professional
development. Many of them have extensive experience as classroom teachers and
teachers are not a group that has historically embraced merit pay. Unlike human
resources managers, they are not trained as specialists in compensation or personnel
management but are quite comfortable helping teachers develop their professional skills.

This TIS placement is consistent with the beliefs of previous district leadership.
Previous leaders indicated that Professional Pay’s core is making professional
development available to teachers so that they can improve, not linking rewards to
student performance. There is some evidence that the new interim district superintendent
is much more results-oriented. As one respondent noted, “Our superintendent is afraid
that if we take all this money from the state, from the taxpayers, and we don’t show
results, meaning all the test scores go way up, that we will have wasted the taxpayers’

dollars and that then that will be shame on Minneapolis.” Despite his more results-

106

oriented focus, the new superintendent has not attempted to change the makeup of the
implementation team or the policy’s administrative home within TIS.

The implementation team espouses preferences we might expect from personnel
whose expertise is in teacher growth. As one team member stated, “Professional Pay is
not really about that legislative model of paying teachers to do a certain thing. It’s much
more about setting up a system that allows teachers and encourages teachers to develop
professionally. And the pay incentive itself is your way into that.” The other team
members substantively agreed with this emphasis on teacher growth and relative de-
emphasis on compensation. While the governor and state legislature expressed interest in
funding an alternative compensation system, the team is interested in funding high
quality professional development. They see the additional compensation provided to
teachers as a desirable, but not necessary, plan component. I

One of the early leaders of Professional Pay was Jay Ritterson. In the year prior
to implementation, Mr. Ritterson had been co-chair of the Professional Pay design team
while on leave to the Minneapolis Federation of Teachers. When the district was ready to
begin plan implementation, two facilitator positions, one full- and one half-time, were
created within TIS. During the initial year of implementation, 2002-2003, Jay Ritterson
was brought back into the district as a full-time Professional Pay facilitator. This position
was not considered administrative; Mr. Ritterson was classified as a Teacher on Special
Assignment or TOSA. While they report to the district office, rather than a school
principal, TOSA’s remain under the teacher’s contract and see themselves as teachers,

rather than administrators. Ronna Locketz, another TOSA, was given a half time position

107

assisting Mr. Ritterson with Professional Pay implementation. The remaining half of her
time was spent as Lead Mentor for Career in Teaching.

After launching the plan in 2002-2003, Mr. Ritterson decided to return to the
classroom. Ms. Locketz moved into the full-time facilitator position as his replacement.
The district added a second half-time facilitator position and the two half-time positions
were filled by Linda Trevorrow and Lynn Nordgren. All three TOSA’s involved in
implementation have strong ties to teacher professional development. Both Ms.
Trevorrow and Ms. Nordgren work on Professional Pay half-time and work on issues
related to teacher professional development during the other half of their time.

The team envisions Professional Pay as a way to establish a system that allows and
encourages teachers to develop professionally. They are focused on teacher
professionalism and teacher growth; the compensation attached to the professional
development opportunities is secondary. As one team member noted, “If money becomes
the core, then we’ll lose on this.” They envision Professional Pay as one policy within a
larger framework designed to provide coherent professional development within MPS.
One team member said that, “Our idea is to embed it in the whole.” The whole is the
district-wide system of supports and training to encourage teachers to develop as
professionals. In other words, the plan is not just a way to pay teachers in innovative
ways. The team sees the plan’s core as professional development; the teacher
compensation money, while desirable, is just the mechanism through which professional
development will be offered and funded.

The team is concerned about making sure that the professional development

offerings are of high quality and that instructors are well trained. They worry that, with

108

limited resources (discussed in Chapter 4), there may not be enough money to pay for
high quality offerings and compensate teachers. The legislation does not address the
issue of financing any aspects of the program other than teacher compensation and there
is virtually no discretionary money in the district to fund professional development. The
team is trying to find grant recipients within the district who are willing to work with
them on this problem. One team member said, “We’ve been trying to get somebody that
has a lot of money on a grant to budge, to make a decision.” Funding constraints will be
discussed in more detail in Chapter 4 but the team’s concerns about funding high quality
offerings are consistent with their emphasis on the policy as a vehicle for professional
development, rather than a merit pay program.

While the preferences and backgrounds of policy implementers are always
important, it may be a particularly salient issue in this case. As mentioned earlier, the
legislation to fund Professional Pay was passed during 2001 and proposals needed to be
submitted early in 2002 in order to qualify for funding. Mr. Ritterson notes that, at the
time the plan was approved, it was “as yet only vaguely outlined” (Ritterson, 2003). The
approved policy was partially underdetermined and not all clearly articulated.
Professional Pay is based on a written policy that was still incomplete during initial
implementation. This lack of specificity gave plan implementers, in collaboration with
district and union personnel, quite a bit of latitude when fleshing out plan details.

Another factor increasing implementer discretion in this case is the complex
nature of the plan. It is difficult for casual observers to fully understand it. One team
member recalled that, “Our previous superintendent, sitting at the table all last year. . .,

was having difficulty understanding all of the components.” By all accounts, the

109

previous superintendent was interested in Professional Pay and was attentive during
discussions. Despite her efforts, she was finding it difficult to remember and understand
all the details of this complicated policy without investing large amounts of time in it -
time that she simply didn’t have. In this situation, her only options were to spend a great
deal of time on Professional Pay or to trust that the implementers were behaving
appropriately. Given that the first option was unworkable, she chose to trust the team.

Local implementers see this policy as one component of a district-wide
professional development framework. Other stakeholders may see it differently — they
may be more focused on compensation reform. Since crafting and implementation has
been an iterative process, the enacted policy reflects both emphases. This policy is
presented as a new compensation system but it both rewards teacher learning and
provides a vehicle for new learning opportunities. The resulting policy is complex and
subject to interpretation. When a policy trickles down to implementers, there is always
room for interpretation but this may be particularly true when the policy is complicated.
For example, various stakeholders interpreted Title I regulations in different ways during
the initial years of implementation (Bailey & Mosher, 1968; Murphy, 1971; Peterson,
Rabe, & Wong, 1986). These interpretations were often influenced by stakeholders’
personal beliefs about the role of the federal government in education and by their
personal areas of expertise. When Title I was originally written, restrictions on the use of
funds were vague and subject to interpretation. Local implementers took advantage of
this vagueness to implement the program in ways that were consistent with their beliefs
and that made administration manageable given local expertise, or the lack thereof.

Following this period and in response to varying local interpretation, the program’s

110

regulatory requirements were considerably tightened and enforcement increased. We see
something similar to this early history occurring in Minneapolis. The enabling legislation
sets loose parameters for grant eligibility. The district and union administrations have
designed a policy that (more or less) falls within the stated parameters but is also
sensitive to local context and needs. The local implementation team has further
interpreted the policy in ways that are compatible with their area of influence and
expertise. The implementation team believes teacher learning is the key to improved
student learning and, furthermore, their particular expertise is in professional
development. As a result, they tend to interpret the policy in ways that focus on
professional development, rather than compensation. It remains to be seen whether the
legislature will limit district discretion through increased regulation and enforcement as
the program matures; we will see in the next chapter that the long-term prospects for
Professional Pay are somewhat dubious so it is unclear whether maturation will occur or
if this policy, like so many education policies, will simply become a footnote in the
district’s history.

All stakeholders agree that this policy should grow, reward, and retain teachers in
ways that will benefit students. They also agree that both professional growth and
compensation are important mechanisms to reach this goal. They differ in the theory of
action that they assume underlies this particular policy. The implementation team sees
professional growth as the primary mechanism to improve teaching. The state
legislation emphasizes the use of compensation as a mechanism to improve classroom

teaching while recognizing the need for professional growth if this is to occur.

111

Both the district’s CIC committee and the state legislature were motivated by a
belief that changing the pay structures used to reward teachers could spur improvement.
Their primary emphasis was on rewarding teachers for certain behaviors and outcomes
and they hoped that money would motivate teachers to change their behavior in positive
ways and encourage high performers to remain with the district.

While the implementation team would not disagree with this idea, the innovative
pay structure is not their primary motivation for supporting the policy. They are
motivated by a desire to make high quality professional development available to
teachers. This is reflected in their concerns about finding the funds to develop quality
offerings.

Professional Pay’s placement within TIS reflects the district’s desire to use the
policy as a tool promoting teacher’s professional growth. This may have been a savvy
decision since the program’s success in actually improving student achievement may well
hinge on the quality of the professional development offerings. The reward structure
itself, once established, requires little specialized knowledge to implement. The same is
true of the district and school performance awards. These plan components are based on
goals that are set at the district and school levels and administering them is
straightforward. The MPSU 1-2-3 offerings, the heart of the program, are a different
story. This component’s success is very sensitive to the expertise of the team putting
together the course offerings. The same is true of the critical skill sets. These are the
components that require the most manpower to implement and the most specialized

knowledge. T18 is the home of most of this knowledge within MPS. Having an

112

implementation team with expertise in professional development makes sense since this
is where program administration requires the most specialized knowledge.

The placement within TIS also reinforces the message in the written artifacts that
Professional Pay is focused on teacher learning. Interviewed teachers talk about calling
TOSA’s with curricular questions and for assistance with their own professional
development. They associate Human Resources with compensation questions, not TIS,
so the separation of the policy from Human Resources subtly deemphasizes its initial
roots in compensation.

The placement within TIS embeds the policy in the existing district professional
growth agenda. As discussed above, the implementation team hopes the policy will
become part of the existing framework and make more high quality growth opportunities
available to teachers. They downplay the compensation aspects of the policy. This
stands in contrast to the written artifacts. While the literature, contract, and newspaper
articles all talk about the policy’s ties to teacher learning, they present it as a
compensation policy, first and foremost. This is not the case with the implementation
team. They are not particularly engaged in compensation and discuss Professional Pay
solely in terms of teacher learning.

What is Professional Pay .7

Professional Pay is not a static, well articulated program. It is still very much a
work in process and this affects its implementation. As a vaguely formulated, evolving
policy, it is subject to multiple interpretations. Stakeholders with varied preferences
selectively interpret the policy’s intent and the enacted policy is sometimes inconsistent

with the written mandate. Since Professional Pay’s enabling legislation, district and

113

union literature, and contractual obligations look somewhat different from each other in
substance and emphasis, stakeholders can each support their own view by referring to the
artifact that most closely mirrors their interpretations.

The different emphases found in various policy representations allows the policy
to act like a hologram; what you see depends on what angle you approach it from. The
legislation satisfies those who seek innovation, district and union literature highlights the
opportunity for teachers to simultaneously increase their pay and their professional
learning, the contract reassures those who worry the district may be promising too much
by limiting its legal obligations, and the implementers see this as a way to encourage both
teachers and the district to invest in teacher professional development.

Initial discussions at the state and district level focused on compensation as a tool
to develop, reward, and retain good teachers. This was the framework that the legislature
seems to have had in mind when they passed the enabling legislation and appropriated
funds. The enacted policy is much more focused on professional development as a tool
to grow and retain good teachers. Rewards are a secondary benefit, not the primary
motivator, in the enacted policy. This can largely be explained by the policy’s vague
specifications and complexity and the room this leaves for interpretation. The
implementation team has been able to shape this policy to serve their own and district
professional development goals while satisfying state funders that compensation is being
rethought as well.

The policy’s complexity provides room for interpretation. Understanding its
nuances is not a job for the casual observer. Mildly interested stakeholders may latch

onto components that resonate for them while devoting little time to understanding the

114

whole. For example, the school and district performance awards are outcome driven
while the MPSU 1-2-3 rewards are more developmental in nature. The current governor,
a proponent of outcome driven rewards, might focus on the first component while the
implementation team might focus on the second. When resources are allocated, some
components might be more generously funded than others, depending on stakeholder
preferences.

The implementation team focuses on the professional development components.
While they administer the complete policy, most of their time and attention is devoted to
components that are consistent with their beliefs about improving teaching. The
teacher’s contract supports them in this interpretation since the contractually specified
components are all focused on professional development. There are two potential
problems with this emphasis. The first is that the close ties between rewards and
particular professional development offerings will only result in improved teaching if the
professional development is of high quality. We shall see in Chapters 4 and 5 that there
are forces working against implementers’ plans to develop high quality offerings, among
them limited resources for developing and implementing rewarded offerings.

Exacerbatin g that problem is a reward structure that encourages teachers to
quickly move from one rewarded offering to another. Teachers need to make their own
arrangements if they would like to continue the learning that begins when they attend a
district course but Professional Pay has no incentives for sustained follow-up and may
even provide teachers with a disincentive to do so. Teachers cannot increase their pay if
they attend a course twice and there are no courses that are designed to build on skills

learned in previous courses. The only way teachers can increase their pay is to

115

continuously take additional courses. Since professional development time is limited by
the hours in a day if nothing else, a teacher must choose between spending time working
with colleagues or alone on incorporating what was learned in a course into his or her
practice (an activity that carries no financial rewards) and taking another course that
introduces completely new material (a rewarded activity). This discourages teachers
from becoming deeply engaged in any particular offering.

The second potential problem is one of sustainability. We will see in the next
chapter that there are strong forces for change within the policy environment. There is
personnel churn at the higher levels of district leadership and political forces that are
pushing the state towards an emphasis on compensation based on student outcomes. The
policy’s complexity would allow for a changed emphasis — more money could be shifted
to the results-based incentive part of the plan — but this might leave teachers who enrolled
in a plan that appeared to emphasize professional development feeling angry or betrayed.
This problem will be discussed in more detail in Chapter 4.

Complexity allows the policy to offer something for everyone and allows
selective interpretation of its intent by stakeholders but it also increases the potential for
confusion and misunderstandings among the teachers who are expected to respond in
meaningful ways. We will see in Chapter 5 that both lack of understanding and
misconceptions are common. Some teachers complain that they do not understand what
the policy is asking them to do while others harbor misconceptions about the policy’s
intent and structure. Some teachers believe compensation is tied to test scores while

others believe that participation on school committees is rewarded and still others think

116

that professional development is rewarded. All of these components are included in the
policy yet none of these individual pieces provides a complete picture of the policy.

The confusion that follows from complexity may be exacerbated by the somewhat
different messages embedded in the various written artifacts and the implementers’
interpretations of those artifacts. When policy instruments such as the written artifacts
discussed above are inconsistent, confusion often results. In a study of California
mathematics reforms, it was found that the use of coherent instruments and consistent
policy communication improved the chances that teachers understood the policy’s
underlying intent and would reform their teaching in desired ways (Cohen & Hill, 2001).
The mathematics reforms proposed in California asked teachers to rethink their
assumptions about what mathematics is and how it should be taught. When teachers
received conflicting messages about the policy, they were more likely to misinterpret this
ambitious goal as something more modest, such as increasing the use of manipulatives or
adopting a new text. When messages were consistent, they appeared more likely to
understand that the reforms entailed a radical restructuring of their mathematics
instruction, rather than reducing the policy to a set of new tasks.

Minneapolis teachers are being asked to rethink the role of compensation in their
professional lives but are receiving inconsistent messages about what this means. A
district newsletter might emphasize the critical skill set offerings while a memo discusses
eligibility for results-based incentives. District literature stresses the voluntary nature of
participation while a newspaper article implies that the policy might replace the

traditional salary schedule within two years. These varied messages can be quite

117

confusing to teachers who have little time to invest in learning about the policy and will
further the confusion created by its complex nature.

Professional Pay remains a hologram. While primarily designed as a
compensation policy, it can also be viewed as a vehicle for professional development.
Past compensation policies attempting to increase teacher performance through the use of
financial incentives were often unsuccessful when they failed to recognize the need to
simultaneously increase teacher capacity. Recognizing this, MPS included a substantial
mechanism to increase capacity in their alternative compensation system. This, in turn,
resulted in an increasingly complex policy. While complexity is not an inherently
negative characteristic, it does increase the potential for confusion. Teachers who are
trying to find out about Professional Pay find that there are several components, each of
which is individually somewhat complex, and that these are evolving over time.
Exacerbating this problem is the fact that complex, evolving policies are subject to
multiple interpretations. The state, district and union, popular press, and local
implementers all appear to have slightly different interpretations of its details and intent.
No one document or person appears to be the definitive authority on the policy and
teachers may be left to wonder what they are being asked to do and what promises are
really being made.

Predictions: How Might Teachers Respond?

As discussed in Chapter 1, agency theory allows us to make predictions regarding
likely responses to particular compensation systems. For example, we saw that group-
based rewards may inspire greater helping behaviors among colleagues but also provide

an incentive for free riding while individual-based rewards would tend to have the

118

opposite effect. Placing Professional Pay within the typology presented in Chapter 1 will
allow us to make testable predictions about teachers’ likely responses. Reality is
generally messier than theoretical representations and this is certainly the case here. The
complex nature and varied representations of this policy make neat categorization
difficult. Nevertheless, we can generally place plan components and the overall plan
within the typology.

Agency theory tells us that teachers will weigh the financial benefits gained
against the costs, primarily time and effort, of making the necessary changes. Since
Professional Pay is voluntary, we would predict that teachers who do not feel that the
benefits outweigh costs will remain on steps and lanes. We would also expect most
enrolled teachers to actively participate, thereby earning rewards. If teachers do not plan
to be active participants, they are financially better off remaining on steps and lanes,
where they periodically receive automatic salary increases.5 While teachers who are on a
“frozen” area of the steps and lanes schedule (more experienced teachers do not receive
automatic annual increases) will not lose anything by enrolling, they are still unlikely to
enroll because agents are generally risk averse. Professional Pay has no track record in
MPS, so it would be considered a somewhat risky choice, discouraging enrollment
among teachers expecting stagnant salaries under either policy.

Professional Pay primarily rewards individual teacher actions. The primary
components — continuous improvement compensation (payment for advanced degrees,

academic coursework, NBPTS certification, and successful completion of MPSU 1-2-3

 

5 We will ignore the district- and school-wide performance awards for the moment.
While enrollees may automatically qualify for these rewards, we shall see in Chapter 5
that teachers often have little or no information about them.

119

offerings) and capacity building awards (successful completion of critical skill sets) - are
based on individual teacher actions, rather than group participation. In addition, they are
equally available to all teachers completing the component; teachers do not have to
compete for a fixed number of rewards. The primary components are performance-
based: teachers are evaluated based on their behavior rather than outcomes. While the
requirement that teachers be successful in obtaining degrees and pass academic
coursework introduces an element of outcome-based compensation into these offerings,
they are primarily performance-based since the failure rate for these endeavors is low.
There is a higher failure rate for NBPTS certification, but far fewer teachers participate in
this activity than in academic coursework. While the action research requirements for
MPSU 1-2-3 and critical skill sets were designed to be outcome measures, they largely
serve as placeholders. Teachers simply have to demonstrate the ability to use what they
have learned on one project, rather than show that they have incorporated new skills into
their ongoing practice. The primary components would therefore be generally
characterized as individual, piece rate, and performance-based.

The other operational component is the results-based incentive. This rewards
school and district level attainment of specified outcomes. All MPS teachers who enroll
in Professional Pay are eligible for these rewards. This component would be
characterized as group, piece rate, and outcome-based.

The final component is extended professional responsibility compensation.
Individuals are rewarded for their own extended professional responsibilities and are
unaffected by colleagues’ choices when rewards are determined, making this an

individual reward, but the limited number of responsibilities to be rewarded when this

120

becomes operational will introduce a tournament element. For example, only a limited
number of testing coordinators are named, even though many teachers contribute to the
successful management of testing within the school. Should the district decide to reward
testing coordinators, it is entirely possible that two teachers in a school will compete for
the named position. It does not appear that the district intends to require that teachers
demonstrate that they were effective in their extended responsibilities, but rather that they
hold the position for one year. This lack of any measurement, whether of performance or
outcomes, makes this a bureaucratic component. In summary, extended professional
responsibility compensation is individual, tournament, and bureaucratic.

Overall, we would probably characterize Professional Pay as an individual, piece
rate, and performance-based compensation system since continuous improvement
compensation and capacity building awards provide teachers with the bulk of their
opportunities to earn rewards and are emphasized both in available literature and by the
implementation team. This characterization would be tempered by the group and
outcome-based nature of the results-based incentive, particularly for teachers whose
understanding of the policy emphasized this component. Since extended professional
responsibility compensation is not emphasized in the literature or by the implementation
team and was not an operational component when data was collected, this component
probably had a minimal affect on teachers’ perceptions.

As discussed in Chapter 1, we would expect an individual system to promote a
high level of individual performance while discouraging teachers from helping
colleagues. In addition, there is little incentive for a teacher to devote time to activities

that benefit group performance. Since teachers have no incentive to work for the benefit

121

of the entire group or to help struggling colleagues, we would predict that collegiality
would decline with the introduction of Professional Pay.

In addition to being largely individual, rewards also tend to be piece rate,
rewarding every teacher who completes activities. There are no incentives for teachers to
undermine colleagues’ efforts since the number of “winners” under this system is
limitless. To some degree, collaboration might even be encouraged, as it may be easier
for teachers to complete the work required in an academic course or action research
project if they work together. This may help to offset the disincentives for collegiality
that are inherent in individual rewards.

Teachers do have an incentive under a piece rate system to work together to have
the standards for rewards lowered. For example, groups of teachers might argue that the
action research requirement for MPSU 1-2-3 rewards is onerous and should be
eliminated. Because the policy is quite new and the majority of MP8 teachers are not
enrolled, we would not expect to see this happening yet.

The greatest problem with piece rate rewards is that the principal has little control
over total salary expenditures. If large numbers of teachers choose to participate in
rewarded activities, they must all be rewarded, not just the most successful ones in the
group. We would predict that MPS might attempt to prevent this problem by limiting
opportunities for teachers to participate in rewarded activities or, if salaries become too
high, by eliminating Professional Pay. This chapter discussed the limitations the
teacher’s contract places on the district’s obligation to offer rewarded activities and these
limitations are consistent with our prediction. We would also predict that the actual

offerings would be sparse and that the program might prove to be short-lived. Past

122

attempts at piece rate merit pay have often been eliminated as ever-increasing numbers of
teachers qualified for rewards and salary expenses rose (S. M. Johnson, 1986; Wragg et
al., 2004).

The bulk of rewarded activities are performance-based. Teachers are rewarded
for participating in certain learning activities and we would expect participation in these
activities to increase. At the same time, we would expect participation in non-rewarded
activities to decrease since this would allow teachers to earn more without increasing
their total effort levels. Teachers might substitute rewarded offerings for non-rewarded
professional development activities or they might decrease the time they spend on
activities such as after-school tutoring or serving on unrewarded school committees. The
current system, with placeholders for outcome measures, does not provide any financial
incentive for teachers to incorporate what they have learned into their teaching. Once the
rewarded activity is completed, we would predict that teachers would quickly move on to
another rewarded activity, largely ignoring what they have learned unless the new skills
will allow them to decrease their overall effort.

Performance-based measures also provide the principal with an incentive to limit
opportunities, just as piece rate measures do. If teachers participate in large numbers of
rewarded offerings, their salaries will increase to financially and politically unsustainable
levels. To prevent this, the principal has an incentive to limit the availability of rewarded
opportunities and, as cost increases, to eliminate the program altogether.

In Chapter 4, I discuss Professional Pay from the principals’ standpoint in greater
detail, providing empirical evidence that generally supports our predictions regarding

principal behavior. Professional Pay also provides teachers with incentives to change

123

their behavior. The bulk of incentives encourage teachers to focus on activities that
benefit themselves, rather than the greater school community. They are also encouraged
to participate in as many rewarded activities as possible in order to earn rewards and to
substitute rewarded activities for non-rewarded ones. They have no financial incentive to
incorporate the resulting learning into their practice and there is some disincentive, as
doing so would take away from time that could be spent participating in another rewarded
activity. In Chapter 5, we will examine teachers’ responses to Professional Pay. We
shall see that factors outside of customary agency analysis may be playing a large role in
teachers’ responses, limiting, but not eliminating, the usefulness of our theoretical

predictions.

124

Chapter Four
Meaningful Change or a Flash in the Pan?

Credible Commitment and the Minneapolis Professional Pay Plan

This chapter examines the likelihood that Professional Pay will become a
permanent fixture in Minneapolis Public Schools (MPS) and the implications this has for
successful implementation. There are factors at work, such as shifting political priorities
and personnel turnover at the top levels of district leadership, that make it unlikely that
Professional Pay will persist in its current form. In addition, teacher’s beliefs about the
district’s commitment to this policy are examined. If teachers believe a policy will not
endure, they have little incentive to change their practice in ways that are consistent with
the policy’s long-term goals and are better off taking advantage of any short-term
opportunities created by the policy while avoiding the often difficult work of meaningful
long-term change. We will see that Minneapolis teachers are often aware of the district’s
inability to make a credible commitment and do not expect Professional Pay to be a long-
terrn program.

Innovative compensation systems for teachers have rarely met with success.
While there are many causes for the demise of these programs, one common reason is
that financial and political support is not sustained over time. California promised
teachers performance bonuses in 1999 but these bonuses were never paid. Iowa’s
statewide experiment in pay for performance died before it was ever born due to
insufficient funding (Reid, 2003). Kentucky and Georgia are considering the elimination

of annual bonuses to teachers with National Board Certification (Salzer, 2003; Teacher

125

quality bulletin, 2004). Philadelphia has discontinued its performance pay plan due to the
program’s high cost ("Big-city districts scrap reward-based systems of evaluating
teachers," 2003). In 2000, the central government mandated a performance pay system
for all 24,000 schools in the United Kingdom. In response to funding concerns at the
local school level, the central government virtually eliminated the program in 2004

(W ragg et al., 2004). There are exceptions - despite the larger than expected price tag,
North Carolina’s governor and legislators continue to fund their teacher bonus program in
an environment of fiscal shortfalls (Silberrnan, 2003) — but they are rare.

MPS may have difficulty continuing to offer Professional Pay and convincing
teachers that Professional Pay will have staying power. Changing leadership at the state
and local levels and tenuous funding make it unlikely that teachers will believe that the
district has a sustained commitment to the program. Without this sustained commitment,
teachers have limited financial incentives to do the hard work necessary to earn rewards.

While this chapter will look only at the Minneapolis case, it has implications for
education policy more generally. The broader question that this chapter begins to
examine is whether a public school district can credibly commit to policies that, in order
to be effective, require a long term financial and political commitment. The answer to
this question may very well be a resounding, “No.” Education has a reputation for
constant policy change; educational mandates come and go before it is really possible for
meaningful change to take place (Cohen, 1988; Cuban, 1993; Tyack & Cuban, 1995).

Reforms initiated in the 19505 tended to focus on excellence and the need for
American scientists and engineers to remain intellectually competitive with their Soviet

counterparts (Cuban, 1993; Dow, 1991). During the 19605, attention shifted to reforms

126

designed to improve the quality of education received by poor and minority children
(Murphy, 1971). As reformers shifted their attention away from issues of excellence
towards a focus on equality and access, many of the curricular mandates adopted during
the 19505 were abandoned (Clowse, 1981; Dow, 1991). Teachers who had embraced
these reforms and invested time in learning how to teach the new curriculums found
themselves understanding the reforms just in time to see them replaced. The open
classrooms created during the 19705 provide another example of the shifting nature of
educational mandates. In an attempt to facilitate ‘informal education,’ large open spaces
were created in many schools to allow students and teachers to observe and participate in
activities across classroom boundaries. By the early 19805, walls were beginning to
reappear in many of these spaces as the luster of ‘informal education’ faded (Cuban,
1993).

Ideas in education come and go and each new wave is often accompanied by a set
of mandates designed to move education in new directions. This shifting policy focus
does not simply reflect an inability of educators to make up their minds; it may be a result
of the political and institutional characteristics of public school districts. While this
chapter will look only at MPS, the shifting political priorities and changes in leadership
that affect this district are not unique — they are found in many school districts. This may
make it very difficult for school districts to effectively enact policies that require a
sustained commitment to change.

Credible Commitment: A Framework for Analysis
Hidden action, hidden information, and monitoring costs, discussed in Chapter 1,

are not the only obstacles to effective use of incentives. A perceived lack of credible

127

commitment also makes it unlikely that an agent will respond in desired ways. As
discussed in Chapter 1, the agency problem arises in the first place because the principal6
cannot be certain of the agent’s effort level. Sometimes, the agent will not work very
hard but, because of factors such as market conditions or luck, produces high output
despite low effort. At other times, the agent will work extremely hard but circumstances
will work against her and output will be low. We could avoid the agency problem
altogether if we forced agents to reveal how hard they are capable of working and to
work at efficient effort levels. The principal would not have to worry about the agent’s
effort level — the employment contract would ensure appropriate effort. We shall see in a
moment that eliminating the agency problem would not eliminate all of the incentives for
undesirable behavior; the principal has incentives to undermine the system through a lack
of credible commitment even when the agent is working efficiently.

As long as we make a few heroic assumptions along the way, it is theoretically, if
not practically, possible to develop a compensation system that forces agents to reveal
how hard they are capable of working and to work at efficient effort levels. An
incentive-compatible system forces agents to reveal formerly hidden information - how
hard they are capable of working - by providing a strong financial incentive to do so. A
joint forcing contract establishes a minimum output (equal to the expected output when
the agent is working at efficient effort levels) which must be reached before the principal
is obligated to pay the agent for her work. This forces agents to work at efficient effort

levels, thereby eliminating the problem of hidden action.

 

6 As mentioned in Chapter 1, the agency theory “principal” is the entity with whom the
agent enters into an employment contract.

128

A simple example will serve to illustrate how this scenario works. Our principal,
Steve, owns a car dealership. Maria is one of Steve’s salespeople. In a joint forcing
contract, Maria will be paid only if her output is sufficient; otherwise, she will receive no
compensation. Her compensation will be exactly equal to the minimum that she will
accept to work at the desired effort level. Let us suppose that Maria is willing to work at
the desired effort level for $1,000 per month.7 Furthermore, assume that she is capable of
selling ten cars each month if she works at the desired effort level. Steve does not have
this information but he can use an incentive-compatible mechanism to discover it. We
will assume that Steve only sells one type of car, the Zippy, and that the amount he pays
the manufacturer for each Zippy is unknown to Maria. When interviewing Maria for the
sales position, Steve tells her that she will be hired if the cost of her monthly salary plus
the cost of the total number of Zippys she sells each month will leave him with a 15
percent profit. Maria has an incentive to tell Steve the truth about how many cars she can
sell each month and exactly how much money she requires to do so (again, we assume
that Maria actually knows this — all of these simplifying assumptions are a large part of
the reason why this is possible from a theoretical, rather than practical, standpoint). If
Maria overstates the required compensation or understates the number of cars she can sell
in an average month, Steve may decide not to hire her because his profit margin will be
insufficient. If Maria understates her compensation or promises to sell too many cars, she
will not be receiving adequate rewards for her effort and would be better off working
elsewhere. So Maria should be willing to reveal the truth — the system is incentive-

compatible.

 

7 This amount would be determined by factors such as Maria’s taste for work, her
financial obligations, and alternative labor market opportunities.

129

Once Steve has determined that Maria is capable of selling ten cars each month
and that she needs an incentive of $1,000 to do so, he will need to design a joint forcing
contract to force her to actually do so. Assume that she works at the desired effort level,
she will sell ten cars 90 percent of the time; the remaining 10 percent of the time, poor
market conditions will make her unable to sell ten cars. If Steve offers her $1,111 per
month as long as she sells 10 cars and $0 if she sells fewer, Maria will make an average
of $1,000 per month if she works at the desired effort level.8 She is unlikely to accept
this, however, since she has bills to pay each month that make her averse to the risk of
receiving no compensation during some months. We shall assume that an extra $200 will
compensate her for this risk. Maria will thus be willing to accept a contract that pays her
$1,311 each month that she sells 10 cars and $0 each month that she does not. The
resulting contract will be Pareto efficient. This means that no partner in the exchange can
be made better off without making at least one other individual worse off.

It is theoretically possible to design these kinds of contracts but we virtually never
see them in practice. Our principal would have to be willing to spend a great deal of time
and money ascertaining a reasonable output requirement for any typically complex job.
In addition, the risk premium the principal would have to pay the agent to accept an all-
or-nothing contract would probably be quite high. Agents who work strictly on
commissions are almost always paid for each successful transaction. Agents working on
commission can work less hard and still earn some money, which is problematic for the
principal, but the agent is not forced to absorb all of the risk involved in the random

element of output, saving the principal from prohibitively high risk premiums. Despite

 

8 ($1,111 x .9) + ($0 X .1) = ~$1,000

130

the impractical nature of the incentive-compatible joint forcing contract described above,
we will assume that one exists for the moment to simplify our discussion of credible
commitment.

We have assumed away the problems of hidden information, hidden action, and
monitoring costs. To further simplify our story, we will also assume away the problem of
multiple goals. In our above example, we assumed that all Steve cares about is the
number of cars Maria sells. In reality, he would probably also care about her selling
tactics since a “hard sell” might result in short term sales but hurt the prospects for long
term business. He would also be upset if she attempted to steal customers from the other
salespeople he hired.

Even under these idealized assumptions, the incentive system still provides an
incentive for undesirable behavior but, in this case, the principal would be the one with
an incentive to undermine the system. The root of the problem is the fact that Pareto
efficient schemes are not budget balancing. That is, any incentive mechanism that is
Pareto efficient will produce a residual profit or 1055 (G. J. Miller, 1992) because each
agent must receive marginal returns equal to the last unit of effort expended. Problems
arise when this residual is absorbed by the principal (or someone who can influence the
principal’s behavior). The principal has an incentive to manipulate the incentive system
to maximize the residual (or, in the case of schools, minimize salary costs) (G. J. Miller
& Hammond, 1994). If the principal does not succumb to this temptation and appears
unlikely to do so in the future, he or she is demonstrating a credible commitment.

In the above example, we noted that Steve requires a 15 percent profit margin.

Maria is told this but she does not know what Steve must pay for each Zippy. Once

131

Maria has accepted employment and would have to incur both actual and psychological
costs to look for another position, Steve has an incentive to tell Maria that the price of
Zippys has increased (even if they have not) to see if he can get her to accept a slightly
lower salary. If he can convince her to do so, the scheme is no longer Pareto efficient
(Maria’s well being has declined even though Steve’s has increased) but Steve is
individually better off because there is more residual available for him to absorb (his
profit margin might increase to 18 percent).

If agents believe that the principal may manipulate the system to maximize the
residual, all of the problems of hidden action and hidden information reappear. Agents
will not trust the principal to honor the terms of their agreement. Once again, agents will
have an incentive to shirk. If the principal and agent only interacted for one period and
information about their interaction would not become known to others, neither party
would have an incentive to behave in a trustworthy fashion — each would attempt to
maximize their gain from the one-time employment transaction. While this does
occasionally occur, principals and agents generally have multiple transactions with each
other; they play repeated games. Even when the game is not repeated between that
particular principal and agent, information about the principal’s behavior (and, to a lesser
extent, the agent’s behavior) during that game is often available to future partners. It is
only when a game is repeated that players may have an incentive to behave credibly.

The folk theorem tells us that, in a repeated games situation, there is more than
one player strategy that can result in a sustainable equilibrium. While one possible
equilibrium is credible behavior by both parties, that is not the only likely outcome (G. J.

Miller, 1992). Any equilibrium that is better than the worst possible scenario is

132

sustainable, including principal reneging and the resulting breakdown of agent trust.
There are two factors that will increase the likelihood the principal will make a credible
commitment to refrain from manipulating the incentive system: when external constraints
such as legal restitution make it costly to renege on agreements and when the principal
values a trustworthy reputation.

The principal is more likely to behave credibly when there are external constraints
that force this behavior. External constraints are laws that are designed to “tie the king’s
hands” (North & Weingast, 1989). For example, if the principal decides to renege on a
negotiated labor contract, the agent may be able to successfully sue for damages. Laws
signal the principal that a decision to renege on an agreement may prove costly and thus
may serve as a signal to the agent that the principal is less likely to renege. However,
while external constraints can encourage trustworthy behavior by the principal, they
cannot guarantee it. If it were easy to observe and verify the agent’s actions and the
principal’s response, the agency problem would not exist in the first place. External
constraints, while helpful, cannot force credible behavior in a situation where there are
incomplete contracts. There are aspects of employment contracts that are not specified in
advance and legislation is of limited usefulness in these cases.

The second factor making the principal less likely to renege is when the principal
values a trustworthy reputation and is concerned about maintaining a reputation as a
desirable employer. Let us suppose that the principal pays the agent wages that are less
than the value of the agent’s marginal product early in the agent’s career and pays wages
that are above the value of the marginal product later in the agent’s career. This is a

strategy that is used by some principals to induce higher levels of effort early in the

133

agent’s career (thereby minimizing some of the typical agency problems). The principal
has an incentive to promise higher pay late in the agent’s career and then renege on the
agreement. Doing so may not be in the principal’s best interest, however, because the
principal’s reputation will be damaged. Future agents, aware of the damaged reputation,
will refuse to enter into contracts that pay them less than their marginal value of
production early in their careers. Exploitation of current agents may force the principal
to pay a wage premium in order to entice future agents to accept employment (Kreps,
1990; Williamson, 1985). The principal will no longer have use of this tool to encourage
higher effort levels and will be worse off, in the long run, than if the principal had not
reneged on the initial agreement. Reputations, once damaged, are very difficult to
rebuild, so any short-sighted reneging on the part of the principal will have long term
consequences for future transactions (Tirole, 1996).

Agents are aware of the presence of externally imposed constraints and of the
principal’s reputation. The former would be publicly available information and the latter
is revealed by the trustworthiness (or lack of it) demonstrated by the principal in other
interactions experienced or heard about by the agent. The only time that the agent is
better off behaving in a trustworthy fashion is if the principal does so as well. The
principal’s credible behavior is a necessary, but not sufficient, condition for the agent to
behave credibly.

In an idealized case, the principal is completely trustworthy and the agent
recognizes this. This scenario maximizes the residuals available to the principal as long
as the agent behaves credibly as well. As long as the agent recognizes the principal’s

trustworthiness, the agent has an incentive to behave credibly since the agent knows that

134

the rewards for doing so will be appropriate — that the he or she will receive marginal
returns equal to the last unit of effort expended.

It has been argued that in a situation where the agent and principal can trust each
other completely, a contract is not even required since it simply adds the unnecessary
costs of designing, monitoring, and bonding the contract (reducing the residual by the
amount of those costs) to a perfectly trustworthy relationship (Al-Naj jar & Casadesus-
Masanell, 2001). In the real world, trust is relative, not absolute, and some sort of
contract is generally necessary even in a trusting relationship. In addition, this argument
requires several heroic assumptions that must be met for the agent to be considered
completely trustworthy. For trust in the principal to be a sufficient condition for credible
behavior on the part of the agent, either the principal would have to be able to perfectly
ascertain the agent’s effort level (rendering the agency problem obsolete) or the principal
would have to design a perfect incentive-compatible, joint forcing contract so that the
agent was not tempted to shirk (which, as discussed above, is theoretically, but probably
not practically, possible). This leaves us with credible commitment, once again, as a
necessary but not sufficient condition for credible behavior on the part of the agent.

When credible commitment is considered as an aspect of agency theory, an
historical analysis becomes necessary to make sensible predictions regarding agent
responses. If the agent believes that trust in the principal will be rewarded, the agent will
be more likely to behave credibly as well — we will be more likely to achieve a trust-
honor equilibrium. The greater the agent’s misgivings about the principal’s
trustworthiness, the less likely a rational agent will be to behave in a trustworthy manner.

Since the agent’s perception of the principal’s trustworthiness is conditioned on previous

135

experiences, we must consider these previous experiences when predicting likely agent
responses.

Another, and more intractable, problem is that principals in the public sector have
limited power over actions that may affect their reputations. Public institutions do not
control their own destinies. Bureaucrats in these institutions may spend years developing
a reputation as a trustworthy principal by behaving in a credible manner and then find
they are unable to honor commitments when their budgets are slashed by the legislature.
Voters in a school district may freeze or roll back property taxes and district officials may
be unable to honor commitments through no fault of their own. New state or federal
mandates may be at odds with the promises that district administrators have made to their
teachers and they may be forced to break those promises or face stiff sanctions. We will
see that a well intentioned district may be unable to credibly commit. The problem is not
a lack of good-hearted people; it may, to at least some degree, be an inescapable dilemma
for public agencies.

This chapter looks at Professional Pay from the agent’s perspective, rather than
the principal’s. The principal wants the agent to behave in certain ways but the agent
also expects certain behaviors from the principal if the agent is to have an incentive to
respond in those ways. The agent must believe that the principal will honor promises
made and not change the rules in midcourse. We will see MPS may be unable to credibly

commit to Professional Pay’s continuation and that teachers are aware of these problems.

136

Normal Chaos: The District Context

It is very difficult for a public agency to demonstrate credible commitment and
MPS is no exception. There are few external constraints forcing public agencies to honor
their commitments and people within the agency often have limited control over the
agency’s reputation.

External constraints

Legally, MP8 is only obligated to continue offering Professional Pay through the
2004-2005 school year. The 2003-2005 Teacher’s Contract, ratified January, 2004, states
that:

The standard salary schedules and Professional Pay Plan are a part of a teacher's

continuing contract as outlined in this Agreement (Article 1, Section C. 1). Pending

continued, adequate Minnesota State funding for the plan and beginning prior to

June 30, 2005, the Union and the District shall renegotiate continuation of the

Professional Pay Plan.

Should the district decide not to continue the plan, it can simply decide that the
state funding, even if it still exists, is not ‘adequate.’ In this way, the district can legally
renege on its promises to teachers.

For example, teachers are told that $1,000 annual payments for the successful
completion of MPSU 1-2-3 courses will be paid as long as the teacher chooses to remain
in Professional Pay. Should the plan be eliminated, teachers will receive an altered
payment stream. They will receive either their Professional Pay salary at the time of plan
termination or be placed back on steps and lanes, whichever is higher. While this

protects them in the short term, teachers will no longer have the ability to increase their

137

salaries by continuing to participate in rewarded activities if the plan is terminated.
Under the current contract, there is nothing to stop the district from freezing the pay of
former participants until they gain enough additional years of experience or new
academic degrees to make the salary due to them under steps and lanes equal the amount
that they were earning under Professional Pay. In that case, the salary gains that teachers
have made through plan participation will slowly disappear.

At the time of the interviews, the 2003-2005 contract was under negotiation.
During the first set of interviews, there was great uncertainty about the new contract. At
the time of the second set of interviews and the surveys, the proposed contract appeared
likely to be ratified but this was not certain. In all cases, uncertainty existed and teachers
did not know what would happen to their salaries if Professional Pay was eliminated.
While they knew that Professional Pay was a part of the negotiation process and
generally expected the payment amounts in the contract to match promised amounts (i.e.,
$1,000 for successful completion of an MPSU 1-2-3 course), interviewed teachers did not
seem to have any clear expectations about how their salaries would be affected if the plan
was discontinued.

The short term of the district’s contractual obligation is not the only aspect of the
contract limiting the district’s obligation to behave credibly. Another factor, discussed in
Chapter 3, is the conservative nature of the contractual obligations. The district is not
obligated to offer a minimum number of rewarded professional development
opportunities each year and it is not contractually obligated to pay results-based
incentives or extended professional responsibility compensation at all. While there is no

indication that the district plans to do so, it could offer rewards only for advanced degrees

138

and NBPT S certification (albeit at a higher level than under steps and lanes) in any given
year and still be within the parameters of the contract. Since teachers do not receive
experience credit under Professional Pay, a given teacher could wind up making less
under Professional Pay than under steps and lanes if this occurred. During the 2002-2003
school year when the first set of interviews took place, one critical skill set was offered
(this is the component that has a one-time payment of $500) and no MPSU 1-2-3 courses
were offered (the $1,000 reward from this component is added to a teacher’s base salary).
The following year when the second interviews and surveys took place, no critical skill
sets were offered and one MPSU 1-2-3 course was offered. As we will see in Chapter 5,
this very limited menu of offerings does not appeal to all enrolled teachers and some of
them chose to forego rewards rather than participate in an activity they did not see as
professionally worthwhile.

While the salary schedule for the traditional steps and lanes is also renegotiated at
two year intervals, we shall see in the next section that reputational constraints may come
into play here, increasing the district’s credible commitment to maintaining that
schedule.

Control over reputation: Changing players and unstable fimding

Two factors make it very difficult for MPS to control its own reputation; a
constantly changing set of decisionmakers within and outside the district and the district’s
limited control over its financial future.

MPS has experienced instability in the superintendency. During the initial stages
of Professional Pay discussions, the district was led by a superintendent, Peter

Hutchinson, who advocated the “new business model.” One respondent explained that he

139

believed “it was his job to help and be collaborative” and that new ideas just “flew” in
this environment.

As the district came closer to implementing Professional Pay, Hutchinson was
replaced by Carol Johnson. According to the implementation team (the group of three
district employees responsible for managing Professional Pay), Johnson was supportive
but, by nature, she wanted to think carefully about every decision. Late in her tenure,
eroding economic conditions resulted in district budget cuts. These cuts were unpopular
and, according to some respondents, eroded some of the trust that had been created
between the union and the district under Hutchinson’s leadership.

During her tenure, Johnson hired David Jennings as chief operating officer.
Johnson left MPS shortly after I conducted the first set of interviews and Jennings was
named interim superintendent. He was serving in that capacity when the second set of
interviews and the surveys were conducted. Jennings’ background includes positions
with the Greater Minneapolis Chamber of Commerce, the Minnesota Department of
Commerce, and as a Minnesota legislator. It does not appear that he had any experience
in educational administration prior to his position in Johnson’s administration. Some
respondents indicated that management-union trust continues to erode under Jennings’
leadership.

In addition, Jennings may not share his predecessor’s goals. Previous leaders
indicated that Professional Pay’s core objective is making professional development
available to teachers, not linking rewards to performance. There is some evidence that
Jennings is much more results-oriented. As one respondent from the second round of

interviews notes, “Our superintendent is afraid that if we take all this money from the

140

state, from the taxpayers, and we don’t show results, meaning all the test scores go way
up, that we will have wasted the taxpayers’ dollars and then that will be shame on
Minneapolis.” Respondents feel that Jennings would prefer a pay plan that tied
compensation to outcomes, specifically test results, rather than the more process-oriented
Professional Pay.

At the time of data collection, another superintendency shift was expected.
Shortly after his appointment as interim superintendent, Jennings was selected by the
school board as the new permanent superintendent. Following extensive community
debate about both his credentials and the selection process itself, he resigned the
permanent position two weeks after accepting it. He announced that he was no longer
interested in the permanent position but would remain as interim superintendent until a
replacement could be found. The school board can leave Jennings in the interim position
for as long as three years; during the data collection period, it was not clear when a
permanent replacement would be named.

It should be noted that another superintendency shift took place after the data
were collected. Thandiwe Peebles, the former superintendent of the Cleveland Municipal
School District, accepted the permanent superintendency post, replacing Jennings.
Jennings has since left MPS altogether.

Other local personnel changes have affected support for Professional Pay as well.
The former director of Policy and Planning (a department that no longer exists) was quite
influential with the superintendent and actively supported Professional Pay. The former
director of human resources was also very supportive and had participated in discussions

from the beginning. Both of these men have left MP8. The human resource department

141

has experienced three leadership changes in the last few years. The shifting leadership
makes it difficult to maintain Professional Pay support since each leader has his or her
individual priorities and policy beliefs.

Professional Pay is housed within the Department of Teacher Instructional
Services (T IS). The three people responsible for plan implementation report to the
director of this department. Two of them are assigned to Professional Pay on a part-time
basis and must oversee other programs as well. The TIS department has experienced the
same personnel churn as the rest of the system: TIS currently has its sixth director in
eight years. The style and priorities of these leaders has varied. Yet another new director
of TIS was recently named and the department was awaiting her arrival at the time of the
second round of interviews.

The constant turnover within MPS makes it very difficult to sustain local support
for Professional Pay. Different leaders have different visions of Professional Pay and
differing commitment levels. The implementation team cannot be certain that new
personnel will remain in their positions long enough for it to be worth it to try to educate
them about the plan. The implementation team must constantly reeducate district
leadership, a frustrating task that consumes a great deal of their time.

Recently, there have been superintendency changes at the state level as well. In
January, 2003, Governor Tim Pawlenty appointed Cheri Yecke as the new Minnesota
Commissioner of Education. Yecke is seen as quite conservative and, according to one
respondent, “now they have this vendetta about Minneapolis. They would just as soon not
have us do pro pay.” According to another respondent, the new commissioner is “from a

very small town, Maple River, which is ultra conservative and all the people that she

142

works with are from Maple River, the committees and everything, so everything has
turned back. Even the standards, they undid the [academic learning] standards and made
them extremely conservative again.” Whether this characterization is entirely true or not,
the perception at the local level is that support for Professional Pay has eroded within the
Minnesota Department of Education as a result of the leadership change.

While teachers, administrators, and the implementation team did not appear to
anticipate another turnover in the superintendency at the time of the interviews, another
change has occurred since the data were collected. The Minnesota Senate failed to
confirm Yecke’s appointment in May, 2004 and Alice Seagren was appointed to the post
in July. According to the local press, Seagren has a “different style but same philosophy”
as her predecessor (Pugmire, 2004).

Even if problems of changing local and state education leadership were overcome,
the district might still find itself unable to credibly commit to the plan’s continuation. As
mentioned earlier, Professional Pay is funded through a short-term state grant. The
continuation of this funding depends on legislative action. Recent political shifts may
decrease the chance for continued legislative and executive support.

Minnesota has been moving in a more politically conservative direction in recent
years. The current governor, Pawlenty, is substantially more conservative that his
predecessor, Jesse Ventura. Both leaders have been interested in educational issues but
have had very different ideas about what works.

Ventura, Minnesota’s previous governor, was a vocal proponent of the statute that
is currently funding Professional Pay. While he wanted to see teacher compensation tied

to performance, he was relatively flexible about appropriate mechanisms to do so. When

143

crafting the statute’s details, the state legislators at the time were receptive to input from
the MPS and the Minneapolis Federation of Teachers. The union leadership has been
very stable in Minneapolis and local union leaders and lobbyists have been able to
develop sustained relationships with legislators and their staff. In addition, the
collaborative, supportive relationship enjoyed by the district and the union further
strengthened their ability to collectively influence legislative action. According to one
respondent, “Minneapolis takes a really big lead on a lot of things and then other school
districts may pick up on it or not but the state kind of looks to Minneapolis.”

This situation has changed with Governor Pawlenty’s election. Many respondents
believe that the new governor would like to see teacher salaries based on test scores, an
option that is unpalatable to many teachers. Pawlenty proposed a “super teacher” plan
that would pay teachers with high student test scores in high poverty schools bonuses of
up to $40,000. He said that his proposal, “opens the door to the whole discussion about
how we pay teachers,” (Walsh, 2003) and his plan is centered on rewarding teachers for
test scores, not for professional development. He told one reporter that while, “he
understands that the Minneapolis schools cannot control the poverty of their families or
the readiness of their new students to learn, focusing funding on results could have a
dramatic effect” (Walsh, 2003).

Since the interviews took place, Pawlenty has retreated somewhat from his “super
teacher” proposal due to a lack of legislative support and instead has initiated a
“performance-based teacher pay system” that is being piloted during the 2004-2005
school year in three Minneapolis elementary schools and in the Waseca school district, a

four school district in the Southern part of the state (Draper, 2004; Salisbury, 2004). The

144

new policy is being funded by a state grant and Minneapolis will receive $2.6 million
annually to fund the plan. This new policy is modeled after the TAP program being used
in districts such as Denver and does not compensate teachers for participation in
professional development activities and their application. It rewards teachers for test
score gains within their school and within their classrooms and for receiving favorable
evaluations of their teaching. In addition, teachers are rewarded for acting as mentors
and teacher coaches. When discussing the new policy, Pawlenty noted that, “We want to
move more toward a performance-pay plan, and this is a step in the right direction”
(Draper, 2004).

The govemor’s displeasure with the current state of affairs at MPS does not seem
to be limited to Professional Pay. The governor pointed to MPS when discussing districts
that have “unacceptable outcomes” despite increased education funding (Walsh, 2003)
and, according to respondents, does not support many of the district’s efforts. One
respondent believes, “we have a change in governor this last year and a change in some
of the legislators and so now they’re kind of looking at Minneapolis differently. ..,
negatively. They think that we sucked all of the money out of education,” leaving very
little for other Minnesota districts. Another commented, “I don’t think they wanted to
give it [the grant money to fund Professional Pay] to us. I honestly don’t think they
wanted Minneapolis to have that money. I think that there is an attitude out there that
Minneapolis gets everything.” If the new state leadership does indeed view MPS as
wasteful, this is likely to jeopardize continued state funding of the program. This is
particularly true in the current economic environment; Minnesota, like most states, is

struggling to balance its budget and looking for ways to cut costs

145

It would be very difficult for MPS to fund Professional Pay without assistance
from the state. The district has already had to resort to large cuts as a result of their
current budgetary shortfall. Things are unlikely to improve quickly. Unlike most
districts, federal and state categorical grants and grants from private foundations are the
single largest funding source in the MPS budget. The current economic downturn makes
it unlikely that the size and availability of grants will increase in the short term. In
addition, the nature of grant based funding makes it difficult to use categorical funds to
support Professional Pay. These grants are from a variety of sources (state, various
foundations, federal, etc.) each having different goals and spending restrictions. This
makes it difficult for the district to fund unsupported initiatives under the best of
circumstances. Professional Pay cannot be supported without funding above and beyond
that needed to sustain steps and lanes. Two of the fundamental premises of the policy are
that no teacher will make less under Professional Pay than he or she would have under
steps and lanes and that the district will offer high quality professional development as
part of the plan. In order to remain true to these promises, the district will have to spend
money beyond that required by steps and lanes.

As mentioned in the previous section, reputational factors may increase teachers’
perceptions of the district’s credible commitment where the traditional steps and lanes
salary schedule is concerned. School districts, including Minneapolis, have a long
tradition of only minimal tinkering with steps and lanes. Salary levels and the size of
increments may change somewhat but very few districts have ever abandoned steps and
lanes altogether and most of these have reinstituted it after experimenting with

alternatives. While MPS could abandon steps and lanes, its history of a commitment to

146

maintaining this schedule makes it less likely that teachers will worry about this
possibility. In addition, the Minneapolis Federation of Teachers supports the
maintenance of steps and lanes as an option for teachers. While the district could act
unilaterally, it has a history of cooperation with the union that makes this unlikely.

MPS is trying to implement a policy that its target audience may already be
dismissing. They are hoping that Professional Pay will encourage teachers to learn new
skills, use these skills in the classroom, and then teach their colleagues as well. This is
slow, hard work and teachers may not find the benefits of doing so to be worth the cost
when the financial rewards for doing so are unlikely to outlive Professional Pay.

The next section will look at teacher’s perceptions of the district’s credible
commitment. Depending on whether teachers are aware of the difficulty in sustaining
this policy, they may respond in different ways. The opportunities to earn rewards under
Professional Pay require hard work on the part of teachers but, in most cases, the rewards
for putting in that hard work are added to the teacher’s base salary for the life of the
program. The perceived effort/reward ratio will be very different for a teacher who
expects Professional Pay to be a short-lived policy than for a teacher who expects to reap
rewards for the remainder of his or her MPS teaching career.

Wary and Worried: The Teacher Response

Teachers have serious doubts about the future of Professional Pay. Over half of
interviewed teachers believe that the program will not survive much beyond the 2004-
2005 school year. Fewer than 15 percent of interviewed teachers believe that the
program will still be in place in several years and the remainder are unsure what to

expect. Surveyed teachers are also generally pessimistic. When asked whether they

147

agreed that, “The district will still be offering Professional Pay in five years,” 39 percent
of teachers generally agreed with the statement, and only two percent strongly agreed.
Only 24 percent of teachers generally agreed that, “The state will continue to fund
Professional Pay after the initial grant runs out,” and no respondents completely agreed
with the statement. Twenty-three percent of teachers generally agreed that, “The district
will continue funding Professional Pay after the initial state grant runs out,” and one
respondent completely agreed with the statement. Looking at the funding statements
together, only 27 percent of teachers generally or completely agreed with at least one
statement.

In order to examine whether various factors affected teacher’s perceptions of
credible commitment, I constructed contingency tables. Interviewed teachers are
categorized by the completeness of their information about the program9 and their
perceptions of the district’s credible commitment10 to Professional Pay. The resulting

variables are subjective and the reader should note that I assigned values based on

 

9 The completeness of information variable was based on teachers’ responses to several
questions. The first asked them to tell me what they knew about the program. Teachers
often mentioned both elements that were part of Professional Pay and misconceptions
about rewarded behaviors. In a set of prompts following this, I mentioned specific plan
components and asked them if they were aware of these components and how they
worked. Responses were categorized so that roughly half of respondents had more
complete and accurate understanding than the average and roughly half had less.

’0 The perception of credible commitment variable was also based on teachers’ responses.
While talking about plan components and their opinions of and responses to Professional
Pay, respondents sometimes volunteered their predictions about the district’s future
behavior regarding the plan. After discussing Professional Pay for a while, teachers were
specifically asked what they thought the plan would look like in a few years. This
information was then combined to create a perceived credible commitment variable. A
positive perception indicates that respondents believe the plan will persist and will not be
altered in undesirable ways that hurt teachers. A negative perception indicates that
respondents believe that the plan will be discontinued or will be changed in ways that the
respondent found undesirable.

148

responses that were not always consistent or straightforward. The reader should also
guard against assuming a constant scale. Two positive responses do not necessarily
offset two negative responses - the positive responses may have been extremely
enthusiastic while the negative responses may have been mildly negative. That said, the
tables do provide one way to reduce and simplify the data into a more manageable form
so that patterns can be observed.

Not surprisingly, the less teachers know about Professional Pay, the less likely
they are to express a strong opinion about the district’s credible commitment to
Professional Pay (Table 4.1). Relatively uninformed teachers are often unsure how
current funding works or who is spearheading the program at the district and state levels.
Several noted that, “I really haven’t given it much thought” so they felt unable to

comment on the future.

Table 4.1: Completeness of Information and Perception of Credible Commitment Among
Interviewed Teachers

Perception of District’s Credible Commitment

Teacher’s Information Positive Negative Unsure
Below average 4% 52% 43%
Average 33% 67% 0%
Above average 18% 64% 18%
Total 13% 58% 1 29%

Surveyed teachers were asked whether seven statements about plan components

were true or false in order to assess their knowledge. Teachers with five correct responses

149

are categorized as having average information while those with a smaller or larger
number of correct responses are categorized as having below or above average
information.11 Teachers were also asked the degree to which they agreed with the two
funding statements mentioned above; “The district will continue funding Professional
Pay after the initial state grant runs out,” and “The state will continue funding
Professional Pay after the initial grant runs out.” The four point Likert scale used for
these items forced a choice; teachers could not select “unsure” as a response. Teachers
who either generally or completely agreed with either of the first two statements are
characterized as having a positive perception of funding commitment. Only 27, 29, and
29 percent of respondents with below, average, or above average information,
respectively have a positive perception of continued funding (Table 4.2). These
differences are not statistically significant and reflect widespread pessimism about future
funding.

Table 4.2: Accuracy of Information and Perception of Credible Commitment Among
Survey Respondents

Funding Perception Perception of Policy Continuation
Teacher’s Information Positive Negative Positive Negative
Below average 27% 73% 36% 64%
Average 29% 71% 42% 58%
Above average 29% 71% 57% 43%
Total 28% 72% 43% 57%

 

’1 The average number of correct responses was 4.8 with a mode of five correct
responses. If teachers did not respond to all seven items, they were excluded from the
analysis.

150

Teachers were also asked whether they agree with the statement, “The district will
still be offering Professional Pay in five years.” Those who either generally or completely
agreed with this statement are characterized as having a positive perception of policy
continuance. The percentage of respondents with below, average, or above average
information having a positive continuation perception are 36, 42, and 57, respectively.
While it appears that teachers with more accurate information may have a more positive
perception of policy continuation, differences are not statistically significant.

Interviewees who are enrolled in Professional Pay are more likely to have a
negative perception of the district’s credible commitment than respondents who are not
enrolled (Table 4.3). This is inconsistent with theoretical expectations. We would expect
respondents with more positive expectations to be more likely to enroll than respondents
with negative expectations because the longer expected life of the policy would increase

the reward/effort ratio for participating.

Table 4.3: Enrollment Status and Perception of Credible Commitment Among
Interviewed Teachers

Perception of District’s Credible Commitment

Teacher’s Enrollment Status Positive Negative Unsure
Enrolled 9% 74% 17%
Not Enrolled 17% 46% 38%
Total 13% 60% 28%

There are two reasonable explanations for these empirical results. The first is that

enrollees may simply be trying to qualify for short-term benefits. Several respondents

151

mentioned that they were going to take advantage of “cheap credits” — the professional
development provided by the district to Professional Pay participants at no charge — for as
long as the program lasted. Others said that they were frozen on the steps and lanes
schedule (experience pay increases are not annual for more experienced teachers) and so
had nothing to lose by signing up for a year or two. In addition, there was a bridge
payment offered at the program’s inception to encourage enrollment. Since interviews
were conducted early in the policy’s history, enrollment may have been motivated by the
bridge payment and some of these participants may later move back to steps and lanes.
Finally, the reward for critical skill set participation is not added to base pay - they are
short term rewards — and the full payment is received shortly after successful completion
of the critical skill set. Some enrolled teachers may be taking advantage of short-term
opportunities even though they do not believe that the program will stay in place for very
long.

Another possibility is that the interview data are driven by the fact that non-
enrollees are more likely to be unsure whether the district is credibly committed or not.
This is not the result of greater plan knowledge on the part of enrollees (correlation = .22)
and may reflect the fact that enrollees have a financial stake in Professional Pay and have
thought more about the program’s future. Non-enrollees cite reasons including an
inability to commit to additional professional development due to time constraints and
impending retirement when discussing why they are not interested in the program. Since
these teachers have already made the decision not to enroll based on reasons unrelated to
the policy’s persistence, they have little reason to spend time thinking about the district’s

commitment. In contrast, enrollees may have learned more about the program after

152

enrolling (because they had a financial stake in doing so) and concluded that it was
unlikely to persist.

As mentioned earlier, survey respondents were forced to make a positive or
negative choice on items related to credible commitment. When “unsure” is no longer an
option, we see the expected patterns (Table 4.4). Enrollees are more likely to have a
positive perception of funding and policy continuation than non-enrollees. Differences
are statistically significant (Phi = .204, (1:041 and Phi = .254, a=.01 for funding and
policy continuation, respectively) and the magnitude of the relationships would be
characterized as important.12 While this pattern reflects the theoretically expected
behavior — teachers are more likely to participate in the policy when they believe credible
commitment exists — it may also reflect enrollees’ reluctance to express pessimism.
Teachers who are enrolled may simply be hoping that the policy will endure and survey
responses may reflect their unwillingness to admit fears that they may have enrolled in a
program with no long-term benefits. As with the interview results, the relationship may
also be influenced by timing — the survey was conducted the year after the bridge
payment was offered - or enrollees may be trying to take advantage of short-term

opportunities to increase their salaries.

 

’2 Phi values below .2 indicate a negligible relationship, values from .2 to .5 are
characterized as important, and those greater than .5 are considered very strong.

153

Table 4.4: Enrollment Status and Perception of Credible Commitment Among Survey
Respondents

Funding Perception Perception of Policy Continuation
Enrollment Status Positive Negative Positive Negative
Enrolled 37% 63% 56% 44%
Not Enrolled 19% 81% 31% 70%
Total 27% 73% 41% 59%

The district’s past actions may also affect credible commitment. Teachers who
observed instability in past district policies may be more likely to expect Professional Pay
to be short-lived. It appears that this may be the case (Table 4.5). Sixty percent of
interviewed respondents feel that district policies are not stable and these teachers are
more likely to believe that the district is not credibly committed to Professional Pay.

This finding is not unique to Minneapolis. A study of another alternative compensation
program, Pennsylvania’s Lead Teacher Program, also noted the residual effects of
constantly changing policy directives (Ceroni & Garman, 1994). Teachers told stories of
devoting time and energy to policies that were later abandoned. Over time, they learned
to ignore new policies and became increasingly cynical about the odds a new policy
would persist. In other words, past experiences with unstable policies negatively affected

their perception of credible commitment regarding newly introduced policies.

154

Table 4.5: Stability of District Policies and Perception of Credible Commitment Among
Interviewed Teachers

Perception of District’s Credible Commitment

Policy Stability Positive Negative Unsure
Not stable 14% 66% 21%
Mixed/unsure 0% 20% 80%
Stable 14% 57% 28%
Total 13% 58% 29%

The relationship between perceptions of district policy stability and credible
commitment was particularly pronounced among teachers who were interviewed during
the 2003-2004 school year.13 Between these interviews and those that took place the
preceding school year, the state and district superintendents were replaced and local
teacher’s contract negotiations stalled. Following this upheaval, teachers were more
pessimistic about the district’s credible commitment to Professional Pay (74 percent
compared to 50 percent the prior year) whether they generally found policies stable or
not. Teachers appeared to be weighing the district’s recent poor behavior more heavily
than actions taken in the past when thinking about credible commitment. This is
consistent with Tirole’s (1996) and Miller’s (1992) observations that the principal’s
reputation is fragile and can be easily destroyed by any perceived untrustworthy
behavior.

The surveys, which were also conducted during the 2002-2003 school year, show

a pattern similar to that of the later interviews. Seventy-nine percent of surveyed

 

’3 Detailed breakout available upon request from the author.

155

respondents disagreed with the item that stated, “District policies remain stable over
time,” and teachers who felt that policies were not stable were more likely to believe that
the district and/or state would discontinue funding and that the policy would no longer be
in place in five years (Table 4.6). The relationship between perceptions of stability in
district policies and continued funding was important and statistically significant (Phi =
.249, a=.014) while that between stability and policy continuation shows a similar pattern

but is not significant (Phi = .166, (1:.099).

 

Table 4.6: Stability of District Policies and Perception of Credible Commitment Among
Survey Respondents

 

 

 

 

 

 

Funding Perception Perception of Policy Continuation
Policy Stability Positive Negative Positive Negative
Stable 48% 52% 63% 37%
Not stable 21% 79% 43% 57%
Total 27% 74% 59% 41%

 

 

 

 

 

These empirical results are consistent with our theoretical expectation that the
stability of past district policies affects future expectations. We might also expect
expectations to be colored by a respondent’s overall perception of MPS. Teachers who
have positive perceptions of the district might be more likely to expect the district to
honor its commitments. This hypothesis was tested using the survey data. The survey
contained ten items, including the stability statement above, that were combined to create

a perception of district variable.” The correlation between a positive perception of the

 

’4 Appendix 11 provides more detail on the creation of this variable and the perception of
colleagues and perception of administrator variables.

156

 

district and positive expectations regarding future funding was strong and statistically
significant (Point biserial correlation = .256, a=.012) as was the correlation between a
positive perception of the district and positive expectations about policy continuation
(Point biserial correlation = .362, (1:.000). Teachers’ perceptions of the district may
affect their views of the district’s valuation of a trustworthy reputation. Since principals
who value a trustworthy reputation are seen as less likely to renege on agreements and
more likely to credibly commit, respondents with positive perceptions of the district are
more likely to believe that credible commitment to the policy exists.

Another possible explanation is that some respondents are simply more optimistic
than others. They may tend to use the positive end of the rating scale more often and to
view colleagues, administrators at the school level, and district level personnel and
decisions in a positive light. This type of respondent might be optimistic about credible
commitment as well. To test this possibility, the funding and policy continuation
variables were compared to the perception of colleague and perception of administrator
variables. In all cases, correlations were small in magnitude and statistically
insignificant. While general optimism may play a role in the correlation between
perceptions of credible commitment and of the district, it is not driving the results.

Many teachers are pessimistic about Professional Pay’s future and virtually none
express unbridled optimism. The majority of respondents believe that the policy will
have a short life and even fewer are confident about future funding. This pessimism
appears unrelated to understanding. While we will see in Chapter 5 that a lack of
understanding has implications for implementation, it does not appear to have a

significant effect on perceptions of credible commitment. The same is true of enrollment

157

status. Respondents’ enrollment decisions are only loosely related to their perceptions of
credible commitment. This may be because some of the plan’s rewards, such as the
bridge payment offered at its inception and payment for participation in critical skill sets,
are short-term in nature. Some enrollees may have based their decision on these short-
terrn rewards, rather than activities with longer reward streams such as attainment of
advanced degrees and completion of MPSU 1-2-3 courses.

Teachers’ perceptions of past policy stability and their overall perception of the
district are both correlated with perceptions of credible commitment. While it is possible
that these relationships are the result of an omitted variable, the empirical findings are
consistent with theoretical expectations and do not appear to be the result of response
bias. It appears that both past experiences with district policy stability and respondents’
overall perceptions of the district’s reputation affect perceptions of credible commitment.

Why Are Teachers Concerned? Perceived Obstacles to Credible Commitment

During the interviews, teachers who were pessimistic about the district’s credible
commitment to Professional Pay were asked about the specific reasons for their
pessimism. Reasons varied but many of them center on a lack of political will to
continue the program at the local and state levels and on funding constraints. These
reasons are similar to the challenges discussed in chapter 3; they are sensible reflections
of the MPS context.
lack of political will

Respondents perceive that the political will necessary to sustain the program is
lacking. At the local level, they feel that support for policies comes and goes and that

Professional Pay will be no different. MPS teachers are not unique in their perception

158

that policy is often unstable as a result of shifting political power. In a study of urban
teachers, it was found that teachers were angry about “the steady, perplexing changes
they experience. Veteran teachers live through a constant barrage of ‘restructuring’ and
‘reform’ that arrives with each new wave of power, either in the central office of the
school system or in the city or state in which they happen to teach” (Nieto, 2003, p. 65).

The complaints of Minneapolis teachers are similar to those in Nieto’s study. One
interviewed teacher noted, “It seems like every year or few years, what we are attempting
goes out the door and we try something else,” while another commented that, “I kind of
feel like it’s going to be a fad. . .and with all of these cuts, Professional Pay might be lost
in the shuffle, too.” Teachers doubt that Professional Pay will receive long-term support
since policies have been rapidly abandoned in the past.

Teachers believe that the changing cast of players at the district level exacerbates
this policy churn. One said, “I think it will be just like everything else, just kind of fall
away by the wayside and then we’re on to something different because somebody else
has thought of some big plan that’s going to make things wonderful.” Another felt that,
“Somebody will come in with a brand new idea,” while still another said that:

...then all of a sudden, someone looks at it and says, ‘Well, this isn’t working.

Whose idea was this?’ Because we have a short history, we have a short attention

span. Someone says, ‘Let’s get rid of it and do something else because we’re

better than the other people.’ I mean, that’s just the way it goes.
They believe that current district leadership is likely to be transitory and new leaders will
discontinue Professional Pay. Newly hired district administrators often introduce new

policies. While this is not an altogether negative behavior, particularly if old policies

159

were ineffective, it increases teacher’s pessimism about the district’s credible
commitment.

Political changes at the state level increase pessimism as well. As mentioned
earlier, teachers do not believe the new governor supports Professional Pay and several
mentioned his Superteacher plan, a proposal to pay a limited number of “superteachers”
of core subjects as much as $100,000, as a potential replacement for Professional Pay.
They worried that, . .with the change in superintendents, the change in different things
that are going on, budget cuts, with Pawlenty, the way he’s running the budget now,
unless there’s something at the end of the rainbow, I don’t, I don’t really see that that’s
going to happen,” and that, “I don’t think this plan will go far because of politics.
Pawlenty’s master teacher thing will win out.”

They are no more optimistic about the odds that plan continuation will receive
legislative support. They do not believe that voters will be supportive of a plan that
guarantees no downside risk for teachers — “If it involves increasing the teacher pay, the
average, then the taxpayers are going to scream. . .I can’t see taxpayers paying more than
they are now” — and feel that current legislators will not support the plan’s continuation
without public support. One respondent noted, “The state thing will dry up I suspect.
Things change with election cycles and how education is looked at will change.”
Another respondent believed that, “[plan continuation] would be up to the
politicians. . .and considering the atmosphere out there, I doubt it very much. I think this
may be it,” because “they [legislators] do not want teachers to make more money. . .I
don’t think they’re interested in teachers getting paid more.” This perceived lack of

support at the state level, combined with the history of policy and personnel churn at the

160

local level, results in a pessimistic view of the district’s desire and ability to credibly
commit.

Teachers do not see strong support for Professional Pay at either local or state
levels. At the local level, they feel that policies are generally not sustained over time and
that personnel changes fuel this short-term outlook. On the state level, they feel that
teachers, particularly Minneapolis teachers, have fallen from grace and there is no
political support for programs benefiting teachers.

Funding constraints

A lack of political will at the state level might not disturb teachers if they believed
the district would and could continue the program without state funding. As mentioned
in the previous section, teachers are unsure if the district has the will to do so, given its
history of policy instability and personnel churn. Many teachers also question whether
the district has the financial ability to continue funding Professional Pay once the state
grant runs out.

When asked whether “The district will continue funding Professional Pay after
the initial state grant runs out,” and "The state will continue funding Professional Pay
after the initial grant runs out,” only 15 percent of respondents expressed agreement. Not
one survey respondent strongly agreed with either funding statement while almost one
quarter strongly disagreed with both statements. Some surveyed teachers were also
suspicious about the district’s intentions or ability to honor their current financial
commitment. When non-enrolled teachers were asked to what extent: “I don’t believe
that the district will actually pay us what we have earned,” affected their decision, 24.4

percent responded, “to some extent,” and 15.6 percent responded, “to a very great

161

extent.” These teachers do not believe that the will and the ability to fund current
obligations are both present.

During the interviews, teachers expressed skepticism regarding future funding as
well. When asked about the program’s future, one teacher responded, “In the near future,
I see it disappearing or going way back just because of money and the way the economy
is and everything.” MP8 is currently experiencing a budget shortfall and one teacher was
unsure whether even current salary obligations could be met: “There are a lot of questions
as to whether there was actually money to pay people. Even in steps and lanes.”
Teachers do not believe that the district has the ability to fund Professional Pay without
assistance from the state.

Between lack of local political will and funding constraints, teachers generally
believe, as one teacher puts it, “It’ll go away. . .with the finances the way they are. . .Our
district does a number of things, they were great ideas but they can’t follow through with
it so I can’t imagine how they would follow through with this.” Teachers do not believe
that the district has made a credible commitment to Professional Pay. Past experiences, a
lack of political support, personnel churn at the local level, and fiscal pressures all play a
role in their belief that the Professional Pay will be a fleeting feature of the policy
landscape.

Skepticism regarding funding was not limited to teachers who believed that the
policy will disappear over the next few years. Several enrolled teachers who reported
being unsure about Professional Pay’s prospects over the next several years indicated that
their uncertainty was partly the result of financial concerns. When asked whether she

thought Professional Pay would still be around in several years, one teacher said, “You

162

know, that’s really hard to answer because I have no idea what the money situation is
going to be like.” She indicated that money would be a factor in the district’s ability to
continue the policy but did not feel she understood the district’s finances well enough to
make a judgment about the likelihood of continued funding.

Not all teachers felt that the district would be unable to fund the program.
Eighteen percent of survey respondents generally agreed that the district would fund the
program after the state grant ran out. During the interviews, six teachers expressed a
belief that the district would fund the program after the state grant period because of its
positive effect on teachers and students. One teacher commented, “I think that probably
the district, after a couple of years. . .they’d be seeing benefits from it in teachers. . .in
student achievement. . .I think that they would pick it up. I think it’d be in their best
interest to pick it up and help. Depending what the budget looks like.” Another said, “As
much as they could, I think they would [continue funding Professional Pay], if it’s
popular enough.” Even these optimists express some concern about the district’s ability
to fund the program. Another teacher was blunter. He speculated that, “It [state funding]
will vanish. Outright, overnight, just vanish and then the individual districts will be
responsible for picking it up somewhere. So another damn program will be cut.” He told
me that he believes the district will receive negative publicity and a hand slap from the
state if they use state funds to establish the program and then abandon it. He worries that
the district might eliminate other, more educationally valuable programs in order to
continue funding Professional Pay. This subset of teachers is more hopeful about plan
continuation than the majority of respondents but they are still concerned about the

district’s ability to pay for the plan in the long-term.

163

Conclusion

Professional Pay may turn out to be a short-lived innovation. The political will
and financial resources to sustain the program appear to be lacking. Teachers are aware
of the district’s inability to credibly commit and this may affect their responses. Some
teachers may be reluctant to participate in what they perceive as a short-term policy.
Other teachers may be choosing to participate but doing so selectively. They may engage
in rewarded activities that are quite similar to what they would have done anyway and
avoid the difficult work of meaningful change. There is some evidence from the
interviews that teachers may be engaging in these behaviors; this will be discussed in
more detail in the next chapter. If the district is unable to credibly commit to
Professional Pay and to persuade teachers of their commitment, the program is unlikely
to work in the intended ways. Teaching quality and the performance of students are
unlikely to improve under these circumstances and the fault will not lie with the policy
details themselves but with the political and institutional characteristics of the district
within which they are embedded. Without credible commitment, Professional Pay is
likely to become yet another example of a policy that ripples only the surface of the water
while leaving the strong currents underneath the surface untouched.

There is evidence that MPS cannot credibly commit to Professional Pay and that
teachers are aware of this problem. This would not be particularly disturbing if this
problem were unique to Minneapolis but this may not be the case. All of the factors that
make it difficult for MPS to credibly commit are likely to occur in other districts. In a
democracy, shifting political support for policies is the norm, rather than the exception,

and all school districts are dependent on political support for their programs to survive.

164

Even in districts where a greater proportion of funds come from local property taxes
rather than categorical grants from the state, school board members are answerable to
taxpayers. School districts do not control their own destinies; their fates are decided by
politicians who may or may not understand the effect that policy churn has on people
within the educational system.

School districts may face a greater problem in credibly committing than other
public agencies. While all public agencies are dependent on government patronage, few
face as much scrutiny as schools. When problems arise in other agencies, we rarely find
mothers and children marching on the capital steps or these problems being front page
news. Education has a high political profile and it is tempting for politicians to make
their mark by introducing a new wrinkle in the education policy arena. In addition,
politicians may feel more comfortable dictating policies in education than in many other
agencies. New York’s Mayor Bloomberg feels competent to dictate a reading program to
be used in all the city’s schools yet how many politicians would feel comfortable
dictating the training procedures that a fire department should use? In this kind of
environment, with shifting political preferences and new players feeling that they can and
should make their mark on schools, it is difficult for a school district to sustain policies
over time.

Inconsistent state support is not the only problem facing MPS as it tries to
credibly commit to Professional Pay. Local personnel churn is a problem as well. As
with inconsistent state support, the local personnel churn observed in Minneapolis is not
unique. Few urban school districts have superintendents who remain for long periods of

time. The average urban school superintendent remains with a district for an average of

165

five to six years before being replaced (Blom et al., 2002; Yee & Cuban, 1996). When
all districts, not just urban ones, are studied, median tenure increases to 7.5 years (N atkin,
2002). While MPS has had much higher turnover in recent years, even this five- to six-
year average for urban school districts would be a relatively short time frame if a teacher
is concerned about whether a compensation policy will persist over his or her entire
career. New superintendents and district administrators come in with their own ideas and
with programs that rarely outlive the new leadership, so teachers’ concerns about
changing superintendents is probably justified.

School districts in general and MPS in particular may be unable to credibly
commit. This inability does not mean that new policies will never replace the status quo
but does make it more complicated to make them work in the intended ways. An
awareness of these problems is helpful as policies are developed. External constraints
can be used to tie the king’s hands and force states and districts to honor the promises
they have made. MPS tied its own hands somewhat when it agreed that teachers would
receive their Professional Pay rewards if the plan were discontinued. The initial funding
cycle for new policies can be lengthened to lessen the chance that a policy will be
eliminated before it has become institutionalized. MPS does not have this tool at its
disposal; the grant based nature of the program means that the legislature, not the district,
controls the length of the funding cycle. Either of these actions, external constraints or
lengthened funding cycles, would increase the district’s ability to credibly commit to a
program, and thereby enhance teachers’ perceptions of this commitment.

Policy-makers often complain about the difficulty of changing teachers’ behavior.

This is not surprising in an environment where policies come and go and teachers have

166

few guarantees that their efforts to change will be rewarded. If policy-makers cannot
credibly commit to a policy, teachers will have little motivation to make long-term
changes to their practice. Change is slow and difficult work, and it is tempting for
teachers to ignore the policy de jour since the next regime change is likely to, once again,
change the rules of the game. If school districts are unable to credibly commit, the most
carefully designed policies may be doomed to failure. There is evidence that MP8 is
unable to credibly commit to long-term policies due to the political and institutional
environment within which they operate. The factors that make it difficult for MPS to
credibly commit — shifting political support for policies, unstable leadership, and tenuous
funding — are all common problems faced by many districts, particularly those in urban
areas. The inability of school districts to credibly commit has been largely ignored in the
literature but may be a significant factor in the failure of many policies to positively
affect practice. This work begins to examine the role of credible commitment but much

work remains to be done before we can examine its full policy impact.

167

Chapter Five
Riding Out the Storm:

Teachers’ Responses to Professional Pay

In this chapter I examine teachers’ responses to Professional Pay and the reasons
they provide for those responses. First, we will look at teachers’ enrollment choices in
this voluntary program. We will see that the majority of MPS teachers are not currently
enrolled and many enrollees do not appear to be active participants. The second section
of this chapter will unpack the reasons for this muted policy response. The third section
will place those reasons within the agency theory context, proposing theoretical
elaborations that will enhance our understanding of teachers’ responses to compensation
systems.

Four main factors affecting teacher response will be explored. The first relates to
the issue of capacity that arose in Chapter 3. We will see that the district’s efforts to
address limited technical capacity by tying incentives to professional development may
be ineffective in many cases. We will further extend the question of excess capacity to
consider whether the assumption that teachers can work harder than they already do is a
reasonable one. The policy’s inclusion of professional development opportunities
implicitly assumes teachers can spend more time learning than they already are. For
reasons that are both professional and personal, this does not hold for all teachers.

The second reason, also foreshadowed in Chapter 3, concerns the agency
assumption that consistent, complete policy understanding exists. This assumption

appears to be violated in our case.

168

The third question concerns the usefulness of financial rewards as a primary
incentive for teachers. While agency theory allows for a range of incentives, it is
generally deployed as justification for using financial incentives to change behavior. It
appears that, given current reward levels, incentives other than financial rewards may be
more powerful for teachers.

The fourth issue is related to credible commitment. This chapter will not
explicitly revisit Chapter 4’s findings but the reader should keep them in mind. Most
teachers do not believe Professional Pay will be a long-term policy. Skeptical about the
district’s commitment, informants worry that the district will be unable to fund the policy
once the initial grant period is over. This perception may color teachers’ responses when
they are questioned about other aspects of the policy, such as the desirability of rewarded
professional development opportunities. While underlying concerns about credible
commitment will not be addressed in this chapter, one aspect — trust - will be discussed.
Several teachers expressed concerns about the likelihood the policy will be administered
fairly and promised rewards paid. These concerns are reminiscent of, but distinct from,
concerns about longer-term credible commitment.

Enrollment Choices

Before unpacking teachers’ responses, it is helpful to look at their options and at
the kinds of teachers selecting each option. Teachers can respond to the voluntary
Professional Pay plan in one of three ways. They can decide not to enroll in the program
and remain in steps and lanes; they can enroll in the program but not participate in any
rewarded activities; or they can enroll in the program and earn additional pay by

participating in rewarded activities. Teachers who participate in rewarded activities

169

choose whether or not to modify their practice in response to those activities. All of these
teacher choices affect the policy’s results.
The Enrollment Decision

At the time the research was conducted, roughly 36 percent of MPS teachers were
enrolled in Professional Pay. A slightly higher percentage, 40 percent, of the district’s
middle school teachers had enrolled. There was a great deal of variation in the
percentage of enrolled teachers across middle schools, ranging from 20 to 73 percent.
Enrollment does not appear related to school achievement or to the school’s percentage
of English Language Learners or children eligible for free or reduced lunch. As
discussed in Chapter 2, the weighted surveys appear to provide a reasonable sample of
the district’s middle school teachers; thus, the more detailed information on participation
which follows — information which is unavailable for the district population as a whole —
is based on these samples.

Teachers with more than ten years of teaching experience are significantly more
likely to enroll in Professional Pay than less experienced teachers (50.9 percent compared
to 32.0 percent, Chi-square=3.796, a =.051). This may be because the salaries of more
experienced teachers are frozen at various points on the traditional steps and lanes
schedule. Once teachers have more than ten years of experience, they move up a step on
the salary schedule approximately every five years, rather than annually. When teachers
compare the increase they will automatically receive on steps and lanes to the rewards
they believe they will earn on Professional Pay, teachers in these frozen areas have a

greater — at least financial - incentive to participate. Another possibility is that more

170

experienced teachers, who tend to be older, may have fewer family obligations (e.g.,
young children) making it easier for them to participate in rewarded activities.

Teaching assignment and enrollment are strongly correlated (Phi =.381, o. =.001).
Only 24.5 percent of core subject teachers are enrolled in Professional Pay, while 65.7
percent of classroom teachers who do not teach core subjects and 42.0 percent of eligible
personnel who are not classroom teachers are enrolled. This pattern appears strange at
first, but we will see that eligible personnel whose work is outside the classroom setting
may be unlikely to participate because rewarded offerings are less relevant to their work.
It is less clear why teachers of core subjects (reading, English, social studies, science, and
mathematics) are reluctant to enroll. Teachers of core subjects may be reluctant because
they are interested in subject-specific professional development (all rewarded offerings
thus far have been generic). Teachers of core subjects are also more likely to say that
their non-enrollment decision was influenced to a very great extent by a belief that, “The
extra money wasn’t worth the effort required to earn it” (25.6 percent compared to 5.0
percent).

Gender does not appear to play a role in the enrollment decision: 43 percent of
surveyed females and 40 percent of surveyed males were enrolled, an insignificant
difference.

Teachers with more positive perceptions of their administrators are also more
likely to be enrolled (Point-biserial correlation = .214, a =.029). Since perceptions of
school administrators are not correlated with school site to any significant degree, this
correlation does not appear to be an artifact of different working conditions across

schools.

171

Professional Pay is administered by district (rather than school) administrators,
but there is no significant correlation between perception of district and the enrollment
decision. This is perplexing but one possible explanation is that the small number of
teachers with strongly positive perceptions of the district made it difficult to discern
differences between enrolled and non-enrolled teachers. Another possible explanation is
that the items related to district perception covered a broad range, including perceptions
of personnel, policy stability, and agreement with policy decisions.

Responses to the item, “I trust district personnel to implement Professional Pay in
a fair, evenhanded manner,” are significantly correlated with the enrollment decision
(Point-biserial correlation = .210, a =.033). This makes sense — teachers who trust the
district to administer the policy appropriately are more likely to enroll even if they do not
always agree with district policies or find particular personnel easy to deal with. When
trust is not present, the teacher may minimize interactions with the untrustworthy party
(T schannen-Moran & Hoy, 2000) — in this case the district — by avoiding participation in
policies that are subject to district discretion. Many teachers view Professional Pay as
subject to district discretion so, if they trust the district, it makes sense that they would be
more likely to enroll.

Forms of Participation

Once teachers decide to enroll, they have a further decision to make: the form of
their participation. They can be token participants who enroll but then do not participate
in rewarded activities or they can actively attempt to earn rewards. A little over half,
thirteen out of twenty-three, of the enrolled interviewees reported some form of active

participation. At the time of the interviews, active participation could take two forms:

172

participation in rewarded district course offerings or earning additional academic credits
and certifications.

Twelve (out of 13) active participants mention enrolling in district courses that
qualified for rewards. Seven teachers reported participating in the critical skill set on
“Improving Instruction Through the Use of Data” (Table 5.1). For the 22 teachers
interviewed during the 2002-2003 school year, this course was the only qualifying class
that had been offered at the time of the interview. The twenty-six teachers interviewed
the following spring would have also had the opportunity to enroll in the first MPSU 1-2-
3 class, ENVom. Four teachers were ENVoYTM participants, two had also taken the
critical skill set the previous year. When asked if they had done anything that qualified
for additional pay, three teachers reported taking other district classes that they believed
qualified for Professional Pay. While these classes did not actually qualify for rewards,
they are included as a form of active participation since these respondents believed that

they were part of Professional Pay.

173

Table 5.1: Participation in Rewarded Activities

Form of participation # of teachers
No action as of yet 10
Participation in critical skill set 7
Participation in MPSU 1-2-3 class 4 *

Participation in district classes that did not qualify for rewards but 3
that respondents believed would qualify for rewards

Finish master’s degree, begin working towards NBPTS 1
certification

* Two of these teachers had also participated in the critical skill set. Twenty-two of the
forty-e1 ght interviews took place before any MPSU 1-2-3 offerings were available.

Only one teacher reported participating in activities outside of the district course
offerings. This teacher was finishing his master’s degree by the end of the school year
and planned to begin working towards NBPT S certification the following year. He
understood that these activities qualified for rewards and this influenced his enrollment
decision.

The fact that only one teacher reported working towards advanced degrees and
NBPT S certification as his form of participation is not surprising given the study’s time
frame. Since advanced degrees and NBPT S certification are both more involved
undertakings than district coursework, some teachers may have been reluctant to commit
to these activities so early in the policy’s implementation. Concerns about the policy’s
persistence, discussed in Chapter 4, may discourage participation in longer-term
activities. As the program matures, participation in these longer-term activities may

become more frequent. The short time frame from the policy’s inception to the

174

interviews is also undercounting teachers who are slowly working towards degrees but
have not taken any classes since enrolling. Two teachers who were working towards
advanced degrees mentioned that they were looking forward to being rewarded for future
classes under Professional Pay but have not had time to take any academic classes since
enrolling.

Altogether, thirteen enrolled teachers report participating in a rewarded activity.
Almost all had taken some form of workshop that they believed qualifies for rewards.
While this still leaves only a little more than half the enrolled sample actively
participating, these numbers may be misleading. Several factors may be decreasing the
proportion of participants who have participated in rewarded activities.

First, the district had only offered a few courses that were specifically designed
for enrollees and advertised as such when the interviews were conducted. Also, some
program components, such as rewards for taking on extended professional
responsibilities, were described in the literature available to teachers as they made their
enrollment decision, but were not operational when interviews were conducted. Teachers
who found non-operational components appealing may be planning to qualify for
additional pay through these components in the future.

In addition, there is uneven understanding among enrollees about what counts for
Professional Pay and about course availability, a topic that I will discuss later in this
chapter; some participants may have been waiting for more information before signing up
for rewarded activities. Finally, the district offered a bridge payment to teachers who
agreed to enroll in the program during the summer of 2002. This may have encouraged

some teachers to sign up who are uninterested in active participation.

175

In the survey, teachers were asked, “to what extent have you or do you plan to
(within the next year) engage in professional development activities that qualify for
Professional Pay?” The overwhelming majority of enrolled teachers (88.2 percent) report
to some or a very great extent with a substantial minority (30.2 percent) reporting a very
great extent. This stands in contrast to the interview finding that only 56.5 percent of
enrollees had actually participated in rewarded activities and supports the idea that more
enrollees plan to take activities than have been able to thus far. In addition, almost half
(48.8 percent) of surveyed enrollees agree that they have or plan within the next year to
take university coursework in response to Professional Pay. While these percentages are
not as high as those who plan to participate in rewarded professional development
activities, the pattern holds. Enrolled teachers plan to participate more actively over the
next year than they have so far.

The policy’s ultimate goal is not merely to have teachers participate in rewarded
offerings; it is to have them use what they have learned to improve their practice.
Interviewed teachers were asked how they had used what they learned in their practice.
Of the 13 teachers who participated in rewarded activities, one reported no change to her
practice, seven reported very small changes, and five reported somewhat larger changes.
No one reported substantial changes. As one teacher notes, “I would say it probably
affected it [her practice], kind of like a small ripple instead of a huge ripple in a lake.”

Surveyed teachers were also asked about changes to their practice. When asked if
they planned to continue using their new skills after satisfying the program requirements,
69.8 percent of teachers generally agreed and 25.6 percent of teachers completely agreed.

In a similar manner, 85.7 percent agreed that they have already or plan within the next

176

year to change the way they teach in response to the policy. The survey did not attempt
to gather information about the nature of these changes but it seems reasonable to expect
them to be similar to the incremental changes reported during interviews.

This cursory reporting should not be interpreted as an analysis of actual changes
to practice. Changes were self-reported and teachers may have over- or under-reported.
No classroom observations took place and with the exception of Andy, whose story will
be told later in this chapter, none of the reported changes to practice were verified by
other school personnel. An accurate depiction of the policy’s effect on teacher practice
was beyond the scope of this project but would be an excellent topic for further study.

Why Is Meaning/id Participation Rare?

The majority of MPS teachers are not enrolled in Professional Pay and almost half
of interviewed enrollees reported that they had not participated in any rewarded activities.
This section will explore the reasons why so few teachers seem to be responding to this
policy. Issues of capacity, understanding, motivation, and trust all affect teachers’
responses and are discussed in this section.

Building Capacity or Missing the Mark?

MPS hoped to address capacity issues by linking additional pay to participation in
targeted professional development opportunities. The MPSU 1-2-3 offerings, critical
skill sets, and financial rewards for advanced academic coursework and NBPTS
certification all attempt to increase teachers’ expertise. Unfortunately, these attempts
may be missing their mark for many teachers. Interviewed teachers often indicated a
belief that they are better judges of their professional needs than district personnel. Many

teachers also indicated that, even when offerings are worthwhile, they don’t have the time

177

or the energy to take advantage of them. Both of these beliefs are formidable obstacles to
the district’s attempts to build capacity through Professional Pay.

Teachers are often reluctant to cede control over their professional development
choices. Some of these teachers are quite skeptical of the district’s ability to offer
worthwhile activities. Jean is one of these teachers.

Jean ’s story.

Jean15 does not teach in a classroom setting but, as a licensed employee, she is
eligible for Professional Pay.16 While she has only worked at Washington Middle School
for a few years, she has over 15 years of district experience. In contrast to many of her
gregarious peers, Jean tends to be soft spoken and thinks carefully before speaking.

Jean enjoys the time she spends with colleagues and often works with classroom
teachers to help meet the needs of specific students. She finds her colleagues in
Washington to be more engaged in professional development than at other schools she
has worked in and enjoys the opportunity to share new ideas with them. Jean believes the
school principal “has really supported people in terms of getting other training, paying
people to come in and do things after hours and whatever, or take other classes,” and that
this has contributed to the atmosphere of learning.

Jean does not generally participate in the same professional development
opportunities as Washington’s classroom teachers. Like many licensed employees who
are not teachers, she has to take specialized professional development in order to

maintain her license and has tended to concentrate on this. The press to take specialized

 

’5 All school names and names of respondents are pseudonyms.
’6 Licensed personnel include positions such as classroom teachers, media specialists,
speech language clinicians, TOSAs, and school social workers.

178

professional development has largely dissuaded her from enrolling in Professional Pay.
She worries that, if she enrolls, she will be “trying to sort of juggle two different systems
at once” — the licensure system and Professional Pay. She is willing to manage this
juggling act if it would help her grow as a professional but she feels that, “they [district
implementers] have not been able to identify the kinds of training that would be
applicable to all the people who are covered by it. . .It wasn’t clear to me that there were
going to be the professional offerings that would be helpful to me.” Jean is unwilling to
participate in rewarded offerings that will not help her grow professionally.

Jean also worries about the quality of the rewarded offerings. Several colleagues
at her own and other schools who participated in the critical skill set offering told her that
it was low quality and a waste of their time. Jean wants to be sure that her professional
development time is spent in high quality, relevant courses and is not sure that
Professional Pay supports those efforts.

Jean is unwilling to trade her own professional development agenda for the one
rewarded by the district. Jean is unconvinced that the district’s agenda will provide her
with the skills she needs to do her job well and the fact that the district offerings will not
help her meet licensure renewal requirements for her specialization supports this concern
— if her licensing board does not support the district’s agenda, why should she? Thus, for
specialists operating outside of classroom settings, the district’s efforts to increase
technical capacity may be missing the mark. The skills needed to help these specialists
work smarter may not be the ones covered in district offerings.

This frustration is not limited to licensed personnel who are not classroom

teachers. Non-classroom teachers, elective teachers, and teachers of core subjects all

179

voice similar concerns — a lack of relevance to their professional needs and poor quality.
Some of this is understandable. There are few funds available for developing and offering
meaningful rewarded opportunities. Most of the enabling grant is earmarked for teacher
salaries and professional development funds are quite limited throughout the district.
Faced with this dilemma, current offerings have tended to be somewhat general in nature
(data driven instruction, classroom management strategies) but largely oriented towards
classroom teachers. They may be too vague to appeal to classroom teachers focusing on
their particular content area or student population while being too specific to appeal to
personnel whose main responsibilities are outside the realm of the classroom.

Moreover, even teachers who believe that rewarded activities are probably
worthwhile are often reluctant — like Jean — to cede control over their professional
development. Consider Annette’s story.

Annette ’s story.

Annette is a young teacher at Pressley Middle School. She teaches several core
subjects and, as a result, has several preps each day. Annette appears to take her teaching
responsibilities quite seriously. In addition to her unusually heavy (for a middle school)
planning load, Annette often puts in time after school working with particular students, is
working on an advanced degree, participates in ongoing district professional development
opportunities on the weekends, and mentioned several workshops that she had attended
the previous summer. Most of these activities are subject-specific and, since Annette is
trying to improve her knowledge and pedagogical skills in all the subjects she teaches,
she is constantly weighing the relative merits of various professional opportunities.

Annette spoke repeatedly about the importance of reflecting on her practice:

180

I think whenever you’re reflecting on your teaching as a practice, whenever
you’re in a class or doing some development with it, or even just at home
joumaling, I think you’re always improving yourself as a teacher because you’re
looking at what you’re already doing and trying to make it better.

Annette tries to incorporate the best parts of what she is learning into her
teaching. She feels that the only way to determine what will be successful in her own
classroom is by trying things out and reflecting on her experience.

Annette’s interest in her own learning extends to her interactions with colleagues.
She has an excellent rapport with the teachers on her team and reports that they often
discuss curriculum, student needs, special education concerns, and the impact of various
programs and policies on their students. Her only frustration is that she has few
opportunities to discuss particular lessons and student work with colleagues in her school.
Although she has those opportunities in some workshops and institutes she attends,
Annette wishes there was time to participate in lesson study with the teachers in her own
school as well.

Annette occasionally calls on the expertise of school administrators and district
personnel when she is having difficulty and has been satisfied with the quality of the
assistance she has received. Despite this, she describes her interactions with the district
office as “frustrating”: “Just a lot of running around and go to this office and then go to
that office and, ‘No, you don’t go here. We moved here.’ And a lot of dead ends.”
Annette’s frustration was largely limited to interactions with the human resource

department — Annette relates a story about running around to four different offices to

181

solve a seemingly straightforward problem — but these experiences were sufficiently
aggravating to unfavorably color Annette’s general impression of district personnel.

Annette enrolled in Professional Pay because she was close to completing an
advanced degree and realized that participants receive a larger pay jump than the one
provided by the steps and lanes schedule. Beyond that, she is not particularly interested
in rewarded activities. She has a clear vision for her own professional learning and is not
interested in substituting rewarded activities for her current plans.

Given that she already stays late to work with students and participates in
professional deve10pment on the weekends and during the summer, Annette cannot add
more to her already crowded schedule. Annette sees this policy as another layer to be
added to her current professional development plans. Motivated to improve her practice
through professional development, Annette does not see the rewarded activities as a
particularly appropriate means to do so. Hence, the rewarded activities are an addition to,
rather than a replacement for, her current efforts. She hopes the district will arrange for
substitute coverage and offer some rewarded activities during the school day but, if that
does not happen, she does not see any way to add extra professional development to her
current load. She is already spending as much time as she can on job-related activities.

In addition to the class time that she would have to spend to earn rewards, Annette
is leery of the policy’s action research requirement. This is a bit surprising since Annette
is an active action researcher. When asked about this, she explained:

When you are taking classes and you are adapting them into your class work, it

becomes tedious and monotonous to record it, it becomes more red tape to get that

into the district to get your bonus. . .If I actually have to write up a research project

182

and submit that, then that’s going further than me just doing professional

development to improve my teaching in the classroom. And then I think that does

take away from the student learning because it takes away from your planning
time, it takes away from your organizational time. . .I think there are so many
teachers who are not in the Professional Pay plan who are obviously going to
professional development and. . .I think that was one of the reasons why they
chose not to [participate].

Annette conducts action research related to professional development coursework
and writes journal entries where she reflects on that research, but is unwilling to spend
additional time packaging this work for district review.

The main obstacles for her are an unwillingness to substitute rewarded offerings
for her preferred professional development choices and insufficient time to layer
additional activities onto to her already crowded schedule. Her professional development
decisions are primarily motivated by a desire to improve her teaching, not financial
rewards, and she does not see a tight coupling between her desire to seek out learning
opportunities and the policy’s focus on professional development. In addition to her
unwillingness to cede control over her professional development choices to the district,
Annette is unwilling to spend her precious professional development time documenting
her learning for the district’s edification.

Annette does not appear to be particularly motivated by the additional money
available to plan participants. While she signed up to increase the size of her reward for
attaining an additional degree, she was already close to earning the degree before the

policy was put into place, so she did not have to change her behavior in order to earn

183

more. Several teachers who have participated in rewarded offerings report similar
behavior; they would have participated in the activity whether or not there was a financial
incentive to do so. Annette also reports being aware that the plan pays rewards to
enrollees in schools that meet their quality performance goals but seemed quite
unconcerned about whether or not she actually received the money. While Annette would
prefer to make more money - the larger rewards for advanced degrees were her primary
reason for enrolling - most of her focus is on improving her professional skills, not
financial considerations.

Annette’s story reveals the complexity underlying each teacher’s story. While
discerning patterns in teacher’s responses is helpful as we seek to understand response, it
is important not to dismiss the messiness of individual stories. The “black box” of
decision-making can be simplified through the use of patterns but, if policy modifications
are to be effective, policy-makers must return to individual stories as well. For example,
Annette had mentioned during her interview that paid release time for professional
development would appeal to her. If policy-makers acted on this simple request while
ignoring the rest of Annette’s story, the change might be ineffective. Annette is
interested in subject-specific professional development and might be reluctant to leave
her students - even if there was a substitute — to attend more generic opportunities.

While the details may vary, the outline of Annette’s story is not unique.
Teachers’ stories are affected by their unique subject matter, personal circumstances, and
professional experiences but many shared her concerns regarding control over their

professional development agendas and time constraints.

184

One elective teacher, Paul, reported attending a diverse set of professional
development offerings largely on his own time and expense. He strongly believes that it
is his professional obligation to remain involved in continued learning:

It’s all about being a professional. . .It’s like going to professional conferences that

are within your . . . area of expertise. It would be great for the school to be able to

pay for those things for you to go to, but then you should also be a professional
and you should go to them on your own. You should read professional journals
just because you’re a professional. If you’re saying that teaching is a profession
and we’re using profession in whatever form of the word, then you need to be
professional.

Despite his enthusiasm for his own professional growth and willingness to invest
uncompensated time in both professional development activities and activities with
students, Paul decided against enrolling in Professional Pay. Well informed about the
plan, familiar with its operational components, and generally positive about it, Paul was
not willing to substitute the district’s choices for his own and does not have time to take
on new obligations.

Another perceived problem with rewarded offerings is a lack of integration with
current activities. MPS teachers are asked to respond to a variety of policies: literacy
initiatives, testing directives, and individual Professional Development Plans (PDP).
Even cursory attention to all of these separate, sometimes conflicting, policies eats up an
enormous amount of time especially since — consistent with the district’s belief that
capacity building is an important component of successful policy implementation - many

place their own professional development demands on teachers.

185

Two enrolled teachers mentioned that they were already taking a series of
unrewarded classes associated with grants their schools had received. Since none of the
rewarded offerings appear connected to what they are learning in those classes, the
offerings do not appeal to them. One of these teachers, Beth, commented, “if you’re
putting something else on the workload and have it required to go up on the pay scale, it
sure better be integrated with what you’re already doing because otherwise this is like
you can’t make yourself do anything beyond what you’re doing or you get crazy.” When
asked if she would consider signing up for a rewarded offering, she said, “If it was good
for my program, if I had the time, I would sign up for it. But as you get older, you think
you have the energy but you don’t have as much as you thought you did. . .Part of the
whole problem with this whole thing is where is the teacher going to get the time to do
that much extra work. . .So it’s the classic problem of too much to do.”

While Beth mentioned that she did not have the energy she had as a younger
teacher, she was still going far beyond the minimum professional development required
of her. She remains involved in her own professional growth despite the fact that, as a
seasoned teacher with over 20 years of experience, she faces few external pressures to do
so. The classes she was taking focused on a particular group of challenging students and
she felt the time spent improving her expertise in teaching these children was worthwhile.
She did not have additional time to spend on professional development unless it
dovetailed with those current activities. The rewarded offerings appeared disconnected
from her current focus and, therefore, distracting or irrelevant. Already working at

capacity, she was unwilling to take on more.

186

This perceived disconnect between Professional Pay and other professional
development may be discouraging teachers who are otherwise well-suited for the plan
from participating in rewarded activities. Annette, Jean, and many other teachers, both
enrollees and non-enrollees, who were not active participants were involved in
professional development activities; they were not trying to do the minimum needed to
get by. They are exactly the kinds of teachers the plan was designed to retain and reward
— teachers who work hard at improving their professional skills, often on their own time.
Yet they are failing to reap financial rewards under Professional Pay.

Professional autonomy.

MPS has tried to develop a policy that will reward teachers for increasing their
technical capacity while providing the means to do so. These teachers have already
developed their own plans to increase their technical capacity - sometimes collectively
on a grant with colleagues, sometimes on their own — and they are unwilling to let go of
these plans to make time for the district’s capacity building activities. They believe that
they are the best judges of what they need in order to become better teachers.

Survey respondents express similar sentiments. Seventy-one percent of non-
enrolled teachers agreed that a preference for making their own choices about
professional development was a factor in their decision. Just as teachers continue to
teach in ways they feel are beneficial to their students even when state mandates instruct
them to change these practices (Grimmett & Neufeld, 1994), MP8 teachers continue to
follow a professional development agenda that they feel will allow them to serve their

students better, even when this is not the rewarded agenda.

187

The district should not be surprised. The district’s own literature regarding
Professional Pay states:

Adults are autonomous and self-directed...Specifically, they must get

participants’ perspectives about what topics to cover and let them work on

projects that reflect their interests. . .Finally, they must show participants

how the class will help them reach their goals. . .Leaming has to be

applicable to their work or other responsibilities to be of value to them.

(Ritterson, 2003, emphasis in original)

The district understands that teachers are unlikely to unquestioningly embrace the
district’s professional deve10pment agenda, but does not appear to have the resources
necessary to solve this problem.

Teachers’ unwillingness to substitute the district’s professional development
agenda for their own can be interpreted in several ways. Teachers may be exercising what
Buchmann (1990) characterizes as “empty” personal freedom, clinging to their own
choices because they are more pleasant than district choices — less work or more focused
on topics they personally enjoy learning about. This interpretation is consistent with the
assumptions of agency theory since it assumes that the individual is attempting to
minimize work effort.

Another potential interpretation is that teachers are reluctant to give up their
professional autonomy. Professional autonomy is characterized by an unwillingness to
surrender control over events based on “fidelity to professional ends and to those people
who are actually one’s clients” (Buchmann, 1990, emphasis in original). In this scenario,

teachers place student needs above their own desire for additional pay and believe their

188

own professional development choices will serve students better than the district’s
choices. Most interviewees seemed more committed to professional autonomy than
personal freedom. When discussing their reluctance to substitute rewarded offerings,
they tended to bring the conversation around to the types of professional development
that will allow them to better serve their students. Some teachers refer to subject specific
pedagogy as an important skill to develop while others discuss their interest in learning
how to more effectively teach certain challenging student populations.

It is, of course, possible that teachers couched their preferences in these terms
because of normative pressures. It is more acceptable in education to justify an act in
terms of one’s students than oneself. However, my brief experiences with many of these
teachers led me to believe this is not the only reason. Interactions that I witnessed with
students and colleagues outside of the interview context were generally consistent with an
ethic of professionalism and caring. On several occasions, I overhead teachers discussing
a student’s academic and behavioral difficulties during their free period. Teachers who
were experiencing some success with the student shared their insights and advice with
their more frustrated colleagues. Another teacher, worried about a frequently absent
student, was asking colleagues what they knew about the student’s home situation since
the teacher was having difficulty reaching anyone by telephone. While humor and
exasperation often played a role in the interactions I witnessed, a sense of caring flowed
through almost all of them. Fidelity to one’s clients - students in this case — seemed
pervasive within the schools. Other researchers have found evidence that teachers are
critical consumers of professional development, choosing opportunities they believe will

be beneficial to their students (Cohen, 1990; Cohen & Hill, 2001; Wilson, 1990). While

189

these choices may not always be consistent with policy-makers’ preferences, they do
reflect fidelity to professional ends and to their students.

Time constraints.

Teachers’ unwillingness to substitute the district’s professional development
agenda for their own might not be a problem if teachers believed they were capable of
doing more. They could do both sets of activities. Agency theory generally assumes
agents are capable of increasing effort but these teachers do not believe that is the case.
Having already given up much of their leisure time, teachers talked about time spent in
professional development, participating in school and district committees, and working
with students and did not believe that they could put in greater effort than they already
were. This was particularly true of teachers in the highest achieving of the three
interview sites, Maclaven Middle School. At Maclaven, two-thirds of non-enrolled
interviewees cited a lack of time as the primary reason for their decision to remain on
steps and lanes and almost all of them recited a daunting list of current professional
obligations. For example, one teacher talked about working with students after school
several days a week, serving in a leadership role on a school committee, attending several
subject-specific out-of-town workshops and several local district professional
development offerings each year, and reading professional journals. These teachers are
overwhelmed by their current activities and see meaningful participation in Professional
Pay as an additional obligation, not as a way to be rewarded for their efforts.

Surveyed teachers agree insufficient time is an obstacle to meaningful
participation; indeed, lack of time was the most common obstacle to enrollment among

surveyed teachers. Seventy-five percent of surveyed non-enrollees agreed that “I’m

190

already doing as much as I can and couldn’t take on one more thing” was a factor in their
non—enrollment decision. Forty-seven percent agreed with this statement to a very great
extent.

MPS teachers are not unique in feeling time pressure. In Nieto’s (2003) work
with excellent urban teachers, she finds that, “In my work with teachers, this is the one
issue on which almost all agree — there simply is not enough time in the day to do all that
they need to do” (p. 125). Many teachers Nieto worked with recognized the need to
carve out time for learning, as well as the more obvious tasks such as grading and
planning. They found it extremely difficult to meet all of their professional obligations,
believing they had already maximized their effort.

There are, of course, some teachers who are not particularly involved in
professional development activities and are nevertheless unwilling to add more to their
workdays. These teachers cite personal, rather than professional obligations as the reason
why they cannot do more. Two respondents had young children and did not feel they
could increase their time away from their families. Another was having a difficult year
professionally; survival was consuming all of her energy. One teacher who talked about
a full plate of professional development activities also mentioned caring for an aging
parent. Personal preferences for limiting workdays may play a significant role in
decision making by other respondents as well. Interviewed teachers may have
overemphasized their professional obligations while minimizing their personal
preferences given the interview context.

Not all teachers are unwilling to take on additional tasks. Two participants saw

Professional Pay as another avenue to obtain high quality professional development,

191

enrolling because of the additional professional development opportunities made
available to them. Enrollees received priority when signing up for rewarded offerings
and they found this appealing:

[Participation] made good sense to me because I take advantage of the classes.

And I wanted to be able to take advantage of the classes. I knew if I was part of

the plan, not only would I get the monetary benefit of the plan, but I would also

be given some priority when I sign up for those classes. And they have filled up
pretty quickly.

These teachers are exceptions to the generally observed pattern. Teachers are
often unwilling to take on additional professional obligations, particularly when these
obligations are seen as professionally unrewarding or inconsistent with their own
professional development agendas or other district initiatives. Their vision of working
smarter is continuing the professional development agendas they have constructed for
themselves rather than incorporating the district’s rewarded offerings into that agenda, or
displacing their agenda for the district’s. One possible factor influencing these decisions
and perceptions might be scarcity of information or misinformation, issues that surfaced
in the interviews. The next section will discuss the role limited information and

misinformation plays in teachers’ responses.

Limited Information and Misinformation
Professional Pay is a fairly new option for teachers and many do not completely
understand this rather complicated plan. Some respondents were unaware of key plan

components and others harbored misconceptions. While the majority of teachers were

192

interested in learning about the plan, some have not received adequate opportunities to do
so; others suggested that they had received information but been too busy to read it; still
others have not sought out information because they had heard that pay was tied to test
scores and this idea did not appeal to them. A little over half of surveyed non-enrollees
indicate that they did not know enough about the program to be comfortable participating
in it. And almost all survey respondents, non-enrolled and enrolled, revealed an
incomplete or incorrect understanding of the policy. Let us first consider Bonnie’s story
as illustrative of how information problems may discourage participation.

Bonnie ’3 story.

Bonnie has been teaching a core subject for almost 20 years, all of them in
Minneapolis middle schools. She has been at her current school, Washington Middle
School, for several years. While Bonnie is interested in improving her practice, family
obligations make it difficult for her to attend sustained professional development. She
recently took several district courses to improve her teaching of English Language
Leamers, but found it difficult to find the time to attend the courses and incorporate what
she was learning into her teaching. She decided not to enroll in the final course in the
series because of those time constraints. Conversations with colleagues are the mainstay
of her current professional learning. She discusses strategies with other teachers and
reports that the shared district classes they have taken often provide common ground for
those conversations.

Money is also an obstacle for Bonnie. Her salary helps support her family and, as
her children approach college age, discretionary funds to pay for professional

development are not available. Bonnie is interested in increasing her salary but is not

193

certain that Professional Pay will help her do so. She is considering working towards her
master’s degree if she “can handle the tuition” and, at the time of the interview, believed
that this would only be rewarded if she remained in steps and lanes: “1 have to decide if
I’m going to do the Professional Pay or the masters or if I’m just going to be satisfied
with what I’m presently making.”

Despite Bonnie’s inability to invest a great deal of time or money in her own
professional development, she appears interested in her students’ welfare. During my
visit to her school, I was able to observe Bonnie during some of her breaks and lunch
periods. In addition to participating in the usual social chitchat of the lounge, Bonnie
initiated several conversations with colleagues about particular children who were either
experiencing academic difficulty or social problems. She is concerned about her
students, both as learners and as children who need nurturing, but competing demands at
home limit her ability to invest a great deal of outside time in formal professional
development.

Bonnie would like to earn more and feels that the steps and lanes schedule gives
her very few ways to do 50. She found the flexibility of Professional Pay attractive, for
she had heard that there are several ways to earn rewards under the plan. However, she
was unable to describe any of the rewarded activities although she remembers that
“certain kinds of committees are counted too” but feels that “you can run programs after
school and not necessarily become a better teacher,” so she is not interested in this type
of participation.

Bonnie recognizes that she knows very little about the plan’s specifics but family

obligations have made it difficult for her to attend informational meetings. Other than the

194

meeting notices, Bonnie has only seen one memo related to Professional Pay and, based
on that memo, Bonnie feels that, “it [the policy] all still seems very vague.” Until she
receives additional information, she has decided to remain under steps and lanes.

If Bonnie were better informed about Professional Pay, she might be interested in
enrolling. Some means to earn rewards, such as critical skills sets, do not require a large
time commitment and are offered free of charge to participants. Others, such as rewards
for meeting school and district performance goals, do not require any additional effort on
Bonnie’s part. While this might be problematic if policy-makers hope the policy will
improve the quality of teaching, it would make enrollment more appealing to Bonnie. In
addition, Bonnie believes that the attainment of advanced degrees, an option that she is
considering, is not rewarded when, in reality, the rewards for advanced degrees are
greater under Professional Pay than under steps and lanes. At the time of the interview,
there had only been a few opportunities for Bonnie to learn about the policy and she was
unable to take advantage of informational meetings because of family obligations.
Without more information, Bonnie was unwilling to take a perceived risk with her salary.

Bonnie was not alone. Several other teachers do not feel they understand the plan
well enough to enroll. One teacher is interested in the concept but reports that, “I don’t
even really know what it is.” Another teacher had trouble understanding the literature
she had received and said, “When I did get the stuff, it made no sense to me so I just
tossed it,” while another commented, “I didn’t fully understand it to the point where I
could tell right away that it would be better for me than what I have already.” These
teachers have received information but do not have the time and motivation to try to learn

about this complex plan.

195

Other teachers are enrolled in the plan but limited understanding is discouraging
their active participation. Three enrolled interviewees were unaware the district was
offering classes that qualified for Professional Pay. One teacher, Mark, does not realize
the policy is even focused on professional development: “I don’t see any support coming
from the system other than my checkbook getting fatter. So I think it’s a way to get the
conveyer running faster at the auto plant without having to do anything else. Pour in the
money.” As discussed in Chapter 3, this is exactly the scenario the district was trying to
avoid when the policy was designed. Despite the district’s efforts to design the policy so
that technical capacity is increased, Mark thinks of this as an effort to increase learning
(speeding up the conveyor) without giving him the tools to do so intelligently. He did
hear something about the critical skill set offering after it was over but he was unsure
what it had been about and exactly how it related to the policy. Given that the district
characterizes Professional Pay’s emphasis on professional development as ‘the heart of
the program,’ it appears that the district has not done a particularly good job of
communicating information to all the enrolled teachers.

These teachers feel that they don’t know very much about the plan. They all
report receiving information from the district or union but found the information
insufficient. Other teachers have misconceptions that make the plan unattractive to them.
These teachers think they understand certain plan components when, in fact, they do not.

Bonnie’s belief that advanced degrees are not rewarded is typical of this
confusion. Another teacher was also confused about whether or not the plan rewarded
advanced degrees. He was about to do a lane change and did not think Professional Pay

rewards advanced degrees, “50 I didn’t feel that I was going to benefit from doing it.

196

And beyond that, I didn’t have enough information. . .I opted to stay with what I knew.”
Both Bonnie and the other teacher, who was at a different school, mentioned that they did
not feel that they understand the plan and both believed that advanced degrees are not
rewarded.

Other teachers have the impression that rewards are primarily based on student
outcomes, teacher performance evaluations, or both. One teacher said, “The way I
remember it was that it was, I think a school had to set some goals about achievement.
And if you met those goals, you would be rewarded one way or the other.” He is unaware
of any rewards for professional development, the program’s core. Another commented,
“I’ve only heard that it’s some kind of reward system that would depend on achievement
of the students but I haven’t kept up to date on the specifics of it.” Uninterested in
rewards based on student outcomes, this teacher did not take advantage of opportunities
to learn more about the plan. In reality, student outcomes are a small plan component,
not the centerpiece as these respondents seem to believe. Interviewed teachers were
generally opposed to outcome-based compensation, believing that this would be unfair to
teachers (who have little control over who they teach) and would not support students
(since teachers would have a financial incentive to prefer students with certain
characteristics). Teachers who believed that Professional Pay is outcome-based were
often uninterested in actively seeking out information — they were predisposed to dismiss
the policy. If they learned more about the policy, they might be convinced that the
professional development aspects of it are in the best interest of students. Unfortunately,
this learning is unlikely to occur when teachers have limited information and the district

has a limited capacity to provide that information.

197

Some teachers also believed that teacher performance evaluation is part of the
plan although it is not included at all. They were uninterested in a plan of this type
because of concerns about whether performance evaluation can be done in an objective
manner: “It sounds like something that would be almost impossible to implement fairly.

I don’t know how or who would really be able to be responsible for deciding who gets
more money fairly.” Another noted, “I mean, it’s all just subjective. It all just depends
on who is doing the evaluation of the teacher and what that person’s particular views
are. . .I’m against Professional Pay in the way they do it at least.” Here, both teachers
were critical of a component that does not exist,

When another enrolled teacher was asked if she had done anything in response to
the plan, she answered, “No, not that I know of. I don’t know what I’m expected to do so
I haven’t done anything because of it [the plan].” She is so poorly informed that she does
not even know if she has participated in any rewarded activities. She is not alone.
Another teacher reports that she signed up without knowing very much because her
colleagues were positive about the policy. She particularly trusted one colleague: “She [a
colleague] encouraged us to also sign up so we signed up. We had no idea what we were
signing up for at the time.” At the time of the interview, this teacher still had only vague
notions about what she had signed up for. She has not made a conscious decision to
avoid rewarded activities; the decision is a result of limited information: “I haven’t done
anything as a plan participant because I had no idea what you were supposed to do.” As
with the three enrollees who were unaware of offered classes, her lack of active

participation was due to a lack of information.

198

Problems of limited information and misinformation may decrease with time.
The plan is still quite new and teachers have had limited opportunities to learn about it.
As more teachers participate in rewarded activities, they may begin to share their
experiences with colleagues. In addition, teachers will have more exposure to district and
union literature and presentations as time passes. This may lessen the problem of limited
information and misinformation but it is unlikely to disappear.

The policy had been in place for almost two years at the time of the survey and
over half of non-enrolled teachers still cited lack of information as a problem. When
asked to what extent various factors affected their participation decision, just over half of
non-enrolled respondents agreed to some or a very great extent that, “I don’t know
enough about the program to feel comfortable participating in it” was a factor. One-third
of respondents strongly endorsed that statement, agreeing that it influenced them to a
very great extent while an additional 18 percent agreed that it influenced their decision to
some extent.

In addition, misconconceptions were still common at the time of the survey.
Teachers were asked whether a series of statements was true or false. Table 5.2 reports
the percentage of respondents who had correct responses to each statement. The majority
of respondents incorrectly believe that Professional Pay does not reward advanced
degrees. While only a small number of interview respondents mention this as a decision
factor, teachers are unlikely to mention this as a drawback unless they are planning to
work on an advanced degree in the near future. When interviewed teachers were asked
specifically about this component, they often were unsure whether or not advanced

degrees were rewarded. The district’s focus on providing teachers with information

199

about MPSU 1-2-3 courses and critical skill sets may be leading teachers to conclude that

traditional university coursework is not rewarded.

Table 5.2: Survey Respondents’ Understanding of Plan Components

Statement

1. Professional Pay pays teachers more if they earn advanced
degrees.

2. Under Professional Pay, teachers receive additional pay for all the
professional development that they participate in.

3. Professional Pay pays teachers more if their students score well on
tests.

4. Professional Pay pays teachers more if they take on
responsibilities outside of teaching (i.e., testing coordinator).*

5. Only certain district professional development opportunities
qualify for professional pay.

6. Teachers must demonstrate that they use what they have
learned in their classrooms in order to receive professional pay
for workshops.

7. Some teachers who participate in Professional Pay may make less
money than they did before.

Percentage of

correct responses

35%

76%

71%

37%

87%

88%

86%

Note: For bolded items, “True” was the correct response. For all other items,

“False” was the correct response

* This component was planned, but not yet operational, at the time of the survey.

In addition, almost two-thirds of teachers did not believe that extended

professional responsibility compensation was a component. Enrolled teachers were

significantly more likely to harbor this misconception than their non-enrolled peers (77

200

percent compared to 53 percent, Chi-square=6.015, a =.014). This component is
mentioned in the available literature but enrolled teachers may concentrate on what is
currently rewarded. Since extended professional responsibility compensation was not
operational at the time of the survey, teachers were less likely to consider it a component.

While respondents were much more likely to answer the other questions correctly,
substantial numbers of teachers were still confused about whether teachers are paid for
high student test scores and whether all professional development is rewarded. Twenty
nine percent of teachers believed that enrolled teachers would receive higher salaries if
their students received higher test scores.

Almost one-quarter of respondents believed that all professional development is
eligible for rewards. Sixteen percent agreed with both items two and five despite the fact
that they cannot both be true. The items were both on the screen at the same time and
respondents had the opportunity to change their answers at any time during the survey.
These teachers may not have been reading the survey questions carefully or may have
found the wording confusing. Given the problems with illogical responses to this pair of
questions, it is difficult to be certain if incorrect responses to question two were the result
of misconceptions about the policy or poor survey design.

Some teachers, 14 percent, believe that enrolled teachers may see their salaries
diminish. A larger percentage of non-enrollees believe this than enrolled teachers (29
percent compared to 7 percent, Chi-square=2.718, a =.10). This misconception may be
discouraging some non-enrolled teachers from considering plan participation.

While problems of insufficient information and misinformation may diminish

over time, they were still quite common almost two years after the policy was put in

201

place. These problems appear to be depressing enrollment. As mentioned above, over
half of non-enrolled respondents agreed that insufficient information played a role in their
non-enrollment decision. Non-enrolled teachers do not necessarily believe the worst - 41
percent report being somewhat satisfied with Professional Pay and 66 percent agree it
should continue to be offered - but they were unwilling to enroll without more
information. Teachers who would be well served by Professional Pay may be choosing
to remain on steps and lanes because of limited or incorrect information and thereby
missing the opportunity to earn higher salaries and to further their professional growth
through district offerings. This may adversely affect their willingness to remain with
MPS. In addition, if they believe that the district is engaging in behavior (e.g., paying
teachers for student test scores) that they strongly disagree with, their opinion of the
district is likely to suffer, also making them more likely to leave.

Information problems among enrolled teachers may have unfortunate
consequences as well. The first of these is lost learning opportunities. Professional Pay
is predicated on the idea that increasing technical capacity is an important component of
better teaching. The district hopes to encourage teachers to participate in a package of
activities, including advanced degrees, specific professional development opportunities,
and extended professional responsibilities, that it believes will increase teachers’
technical capacity. If teachers are unaware of those opportunities, the policy’s full
potential to increase technical capacity may not be realized.

The second potential problem is that teachers may begin to suspect the worst.
Mark’s worry that Professional Pay was speeding up a conveyor belt is illustrative. In

the absence of information to the contrary, Mark assumes that the district focuses on

202

student outcomes when distributing rewards and that the district is unaware of the limited
technical capacity of teachers to improve. He is becoming disenchanted and suspicious
based largely on his lack of information. The district has designed a policy that largely
addresses Mark’s concerns but this message is not getting out -— not even to the enrollees.

Information problems did exhibit a pattern that may be helpful to implementers
attempting to provide teachers with opportunities to learn about the policy. In both
interviews and surveys, teachers who reported information problems tended to cluster in
certain schools. Further investigation of this clustering revealed no clear correlation
between the average collegiality, perceptions of school administrators, district
perceptions, school-wide Professional Pay enrollment rate, or school achievement level
and the school’s likelihood of having information problems. While these are not the only
within school factors that might be affecting information availability, they are likely
candidates. Since they do not seem to be related to information problems, factors
external to the school may be affecting information availability.

While the data from this study do not allow for an analysis of these outside
factors, it might be worthwhile for later researchers to look at the relationship between
learning opportunities made available to teachers in particular schools and the frequency
of information problems. Teachers in some schools may have had a greater number of
opportunities to learn about the policy if district and union literature was distributed in a
timely, accessible manner within their schools. District-wide informational meetings
may have been held at locations and times that were more convenient for teachers at
some schools than others and union stewards may have been better informed in some

locations. In addition, some learning opportunities may be more effective than others.

203

One study of the implementation of the 1985 Mathematics Framework in California
found that some opportunities to learn about the policy affected teachers’ ideas more than
others (Cohen & Hill, 2001) and this may be true in our case as well. MPS teachers can
learn about this policy in a variety of ways including literature available in schools, the
union website, informational meetings led by district or union personnel, and the
experiences of colleagues who are participants. Some learning opportunities may be
more effective at helping teachers learn about the policy than others.

In sum, the assumption that agent understanding is consistent and complete is
violated. Let us now consider another assumption, that of the strength of financial
motivators in changing behavior.

Money as a Primary Motivator

This policy, like all compensation policy, is predicated on the assumption that
money has a significant effect on employee behavior and decisions. While there is
undoubtedly some truth to this - it seems safe to assume that MPS teachers are not
motivated solely by non-pecuniary rewards —money may play a smaller role than
expected in changing teacher behavior. Virtually all interviewed teachers indicated that
they care about money, but only a small minority considered the policy’s promises of
additional pay a sufficient incentive for changing their behavior.

One of the most striking aspects of the interviews was how infrequently teachers
introduced money into the conversation, given that we were discussing a compensation
policy. Every teacher, even those who knew very little about Professional Pay, believed
that it is primarily a new way to pay teachers, not a professional development or

accountability policy. Despite knowing that Professional Pay is, as the name implies,

204

primarily about pay, they talked at length about the policy’s perceived impact (or lack
thereof) on their professional development agendas and on their students, but they had
little to say about the actual rewards. When directly asked about rewards, they virtually
all agreed that more money would be desirable but very few seemed particularly
interested in changing their behavior in order to earn it. Quite a few teachers mentioned
that the prospect of increased pay had motivated them to enroll but most of these same
teachers were not participating in any rewarded activities — when it came time to change
their behavior in order to earn rewards, they were not doing 50. Of those who had
participated in rewarded activities, only a third cited the prospect of increased pay as their
primary motivator. While agency theory makes the general assumption that incentives,
whether money or another desirable good, will motivate people to change behavior,
applications of the theory generally make the narrower assumption that money is a strong
incentive, particularly when compensation policies are studied. But the assumption that
money is a strong inducement to change behavior is not supported by the empirical
evidence here. While virtually all of the teachers in this study consider additional pay a
desirable good, few appear to be strongly motivated by the pay incentives.

When 24 non-enrolled teachers were asked why they had decided against
enrollment, most cited non-financial reasons but three of them noted that they did not
believe the rewards were worth the required effort. Peter decided against initial
enrollment because he would be better off financially under steps and lanes: “When this
first came out and I looked at it, I was where I was going to make a fairly substantial
movement in the salary (schedule) and I thought that was a sufficient amount of

movement.” Peter would have been hard pressed to participate in enough activities to

205

 

improve his steps and lanes rewards. Many teachers reported receiving similar advice
from their union representatives. Teachers were told to compare their automatic
increases to the rewards they believed they would receive under Professional Pay when
making their enrollment decision. While Peter cited financial considerations for his
initial decision not to enroll, he failed to enroll in year two as well. When asked about his
decision in year two, Peter explained that he was too busy with his current obligations to
take on rewarded activities so there was still little point in enrolling. Like other teachers,
he was unwilling to substitute rewarded offerings for his own professional development
agenda and felt unable to add new ones to his current schedule.

Two other teachers explained that they decided not to enroll because they were
nearing retirement. With only a few years to reap the rewards, they decided it was not
worth the additional work to earn rewards. While only 12 percent (three out of 24) of
interviewed non-enrollees mentioned that they did not believe enrollment was financially
worthwhile, a larger group — thirty-three percent — of surveyed non-enrollees agree to
some extent that, “The extra money wasn’t worth the effort required to earn it” has
affected their enrollment decision while 16 percent strongly agreed. This is a much
greater percentage than in the interviews and may reflect the fact that teachers were not
presented with specific prompts for reasons in the interviews. When confronted with a
menu of possibilities in the survey, teachers may have tended to endorse any that seemed
reasonable, whether or not they had served as serious enrollment obstacles.

This difference may also be the result of normative pressures in teaching to
appear disinterested in financial rewards. Normative pressures may be depressing the

number of interviewees who cite financial reasons for their lack of active participation,

206

but not having as great an effect on survey responses. The online survey format may
make teachers more comfortable about revealing financial motivation since surveys could
be completed at home, using the relatively anonymous medium of a computer.

When interviewed enrollees were asked why they had not participated in any
rewarded activities, no one cited insufficient financial rewards. This is generally
supported by the survey data as well. When asked whether, “The rewards offered under
Professional Pay are too small for the work required to earn them,” 62.8 percent of
enrolled teachers disagreed with the statement and only 4.7 percent strongly agreed.
Many of the enrolled teachers found the size of rewards reasonable, but they were not
motivated enough by financial rewards to respond. There are three factors that may be
depressing the effectiveness of financial rewards as a motivator. The first has to do with
policy itself, specifically the amount of rewards, the second has to do with the nature of
those who choose to teach, while the third recalls the overwhelming busyness reported by
many teachers.

Small reward amounts.

We begin by considering the amount of the offered rewards. The financial
incentives to participate in rewarded activities are relatively small. During the 2003-4
school year, the minimum base salary for a novice MPS teacher with a bachelor’s degree
was $31,549. Professional pay provided only a few opportunities for teachers on
Professional Pay to earn rewards during this period. If a teacher completed ENVoYTM
and earned the maximum school performance award (which is largely outside of the
individual’s control), he or she would earn an additional $1,500, or 5 percent of novice

pay. Since most MPS teachers earn more than the minimum base salary, the rewards as a

207

percentage of base pay would be even smaller for them. Lawler (1990) notes that
increases of 10 to 20 percent are generally necessary before an employee finds it
worthwhile to change behavior in response to a new pay incentive. The amounts
available during the first few years of Professional Pay are considerably below that 10
percent threshold, particularly for experienced teachers. Thus, teachers have little
financial incentive to participate when doing so will increase their workload. A
perceived lack of credible commitment exacerbates the problem as, with the exception of
critical skill set awards, rewards are added to teachers’ base pay. At the time of the
study, it was not clear if teachers would lose these additions to base pay if Professional
Pay was discontinued. If teachers believe that they will only receive rewards for a few
years, rather than over their careers, the perceived total compensation incentive decreases
and, with it, the financial motivation to do the work needed for rewards.

There is some evidence that simply increasing the amount of rewards may fail to
dramatically increase their motivational effectiveness. In 1998, the Massachusetts
legislature created and funded a program to pay academically gifted prospective teachers
$20,000 signing bonuses to be paid out over four years. The first year bonus was $8,000
and $4,000 was paid in each of the remaining years. These are relatively large incentives
yet researchers found that the money had a limited impact on individuals’ decisions to
enter and continue teaching over the four year rewarded period (Liu, Johnson, & Peske,
2004). Other factors, such as the accelerated certification available to program
participants, appear to have been more attractive to prospective teachers than the signing
bonus. When participants were deciding whether to remain in teaching from year to year,

the financial inducements appeared ineffective as well. Working conditions were of

208

primary importance when most participants were making their stay/leave decisions and
not a single teacher in the study considered the bonus money a decision factor.

Teachers’ relative insensitivity to financial rewards.

Teachers may be relatively insensitive to financial rewards. Teaching is often
viewed as a somewhat poorly paid profession, an impression shared by teachers and the
general public. In a recent study, 75 percent of novice teachers agreed that they are
seriously underpaid and 78 percent of recent college graduates agreed that teachers are
seriously underpaid (Farkas, Johnson, & Foleno, 2000). Given the impression of poor
pay prospects, it would be surprising if those who chose to teach were highly motivated
by pay. As Nieto (2000) notes, “Teachers enter the profession for any number of reasons,
but neither fame nor money nor the promise of lavish working conditions is at the top of
that list” (p. 91).

Research tends to support the hypothesis that teachers are not particularly
motivated by money. When novice teachers were faced with a hypothetical choice
between pairs of schools, one of which offered higher pay and the other of which offered
attractors such as “highly motivated and effective teachers,” over three quarters of
respondents chose to forgo higher pay (Farkas et al., 2000). In other studies, teachers
were asked to rank the attractiveness of several hypothetical teaching opportunities by
reading position advertisements. When economic incentives were emphasized in the
advertisement, teachers found the position less attractive than when organizational
climate (leadership style, etc.) or the work itself (specific responsibilities, etc.) were
emphasized (Young, Rinehart, & Heneman, 1993; Young, Rinehart, & Place, 1989).

Lortie found that only two percent of teachers cite financial reasons as key factors in their

209

own decision to teach and attribute financial motivation to only 37 percent of their
colleagues (1975). School administrators also appear to agree with researchers’ findings
regarding teacher motivation. When school administrators were asked what choice they
believed teachers in their school would make when choosing positions offering higher
pay or better working conditions, almost three-quarters of them believed that teachers
would forgo higher pay (Farkas et al., 2000).

The difference that Lortie (1975) found in teachers’ attribution of their own
motives compared to that of their colleagues does point out one possible interpretive
complication. There are strong normative pressures for teachers to cite altruistic reasons
for teaching (Farkas et al., 2000; Lortie, 1975), so teachers may be reluctant to attribute
financial motivation to their behavior. Lortie (1975) attempted to rrrinimize this problem
when he asked teachers about the influence of money on colleagues’ teaching decisions.
Farkas et a1. asked school administrators what they believed teachers would do for the
same reason. In both cases, money was not considered a strong motivator by the majority
of respondents. The design used in Young and colleagues’ studies (Young et al., 1993;
Young et al., 1989) largely eliminated normative pressure for teachers to nominate
“noble” factors as more important than salary. Teachers were not forced to choose
between higher pay and the other factors; they were asked which form of advertising
would appeal to them if they were searching for employment. They were more
enthusiastic about positions where non-economic position details were emphasized and

relatively disinterested in financial information.

210

Too busy to respond.

The third factor that may be depressing the effectiveness of financial rewards as a
motivator is the overwhelming busyness that many teachers report feeling. With many
demands being made on their time, teachers may require a higher marginal return —
whether measured in financial or professional dividends — before feeling that Professional
Pay is worthwhile. Survey responses to, “I’m already doing as much as I can and
couldn’t take on one more thing” were significantly correlated to responses regarding
preferring to make their own professional development decisions (Spearman’s rho =
-.456, a =.002) and may be somewhat related to the belief that, “The extra money wasn’t
worth the effort required to earn it” (Spearman’s rho = -.214, a =.16). Teachers who did
not believe they were capable of greater effort may be less persuaded by financial
rewards than their less overwhelmed colleagues.

Policy-makers assume that teachers will find the financial rewards embedded in
this policy attractive enticements to change their behavior. This may be overly
optimistic. Both this study and previous empirical work indicate that the amount of
rewards may be insufficient to motivate teachers to change their behavior and that
financial rewards may not be particularly effective teacher motivators.

Some teachers d9 respond.

While the majority of respondents appeared relatively disinterested in rewards,
one group — active participants — did talk about the attractiveness of promised rewards.
Nine out of 13 active participants - 69 percent — said that money played a role in their

decision to enroll in specific rewarded activities. While this is a small group (18 percent

211

of interviewees), it is worthwhile to explore this group since they exhibited the largest
policy responses.

Kathy, a special needs teacher at Pressley, was one of five teachers for whom
additional money was the primary motivation for active participation. Kathy took the
rewarded critical skill set on using data to improve instruction because she was interested
in increasing her salary. She had assumed that the class would not be helpful, noting that,
“It’s really hard because — my special population, it’s [the class] not geared to me.” Had
the money not been attached to the data analysis class, she would have spent the time in
professional development geared towards meeting her students’ special needs instead.

While Kathy enrolled in the data analysis class for financial reasons, she was
being more discriminating about ENVoYTM participation. She has decided not to take
ENVoYTM unless a session of it is offered that is tailored to licensed employees who are
not regular classroom teachers: “Because my population is so different, it’s like it would
be hard for me to go with the regular ed [teachers] and do the ENVoYTM team thing.”
She found adapting the data analysis class to meet her needs a laborious task and one that
she was not anxious to repeat. While motivated by money, Kathy was unwilling to
devote a great deal of energy to adapting another rewarded offering to her situation the
way she did with the data class.

Financial considerations were also Sally’s primary incentive for participating in
the data analysis class: Without the financial incentive, “we flatly would not have had, we
would not have said, ‘Oh, let’s get together and spend a whole afternoon, just looking at
data. Boy, wouldn’t that be fun.”’ The data analysis class would not have appealed to

her if it had not been part of Professional Pay. Sally was planning to take ENVoYTM as

212

well, but money did not play a role in that decision - she was interested in that class
because the course content was consistent with her own plans for professional growth.

Janelle also spoke about the appeal of additional pay. Unwilling to increase her
effort level unless she would be paid to do so, she commented, “We have lives and we’re
very busy and most of us are parents.” With everything else in her life, she rarely finds
time for classes on weekends or in the evenings; they often conflict with activities that
her children are involved in. She has difficulty fitting classes into her schedule and finds
the district offerings, both inside and outside of Professional Pay, limited and
unappealing. Despite this lack of personal or professional motivation, she participated in
what she believes to be a rewarded course so that she can earn additional money.17

In four other cases, money acted as the tipping point, motivating teachers to
participate in offerings that were already somewhat attractive. Consider Andy.

Andy’s story.

Andy is a mid-career elective teacher at Maclaven. Extremely well liked and
respected by his colleagues, at least six of his colleagues told me that I really needed to
talk to him because he was doing all sorts of interesting things in his classroom, and
because he was sharing the skills he was gaining through Professional Pay classes with
colleagues. When we talked, I was not surprised by their recommendations. Andy’s

demeanor combines the enthusiasm of a sixteen year old talking about his new sports car

 

’7 The district course that Janelle had participated in did not qualify for Professional Pay
rewards but she believed that it did at the time of the interview. Three interviewed
teachers mentioned participating in activities outside of Professional Pay that they
believed would be rewarded. These teachers were characterized as active participants in
the analysis.

213

with the thoughtfulness of a monk reflecting on his spiritual journey and it was a pleasure
to listen to him.

Andy is fairly selective about professional development. He tries to attend a few
inservices through district offerings each year; and then focus his energies on integrating
what he learns into his classroom practice. Unlike many teachers I interviewed, he is not
particularly interested in summer workshop opportunities, for he guards this time, using it
to personally recharge and to catch up on professional reading.

Andy did not join Professional Pay until the beginning of its second year. When
the program was first offered, available information was scarce and confusing, so Andy
took a wait and see attitude. By the second year, he had experienced more opportunities
to learn about the policy and the combination of a higher potential salary and expanded
professional development opportunities appealed to him. He had also worried that his
salary could actually decline as a participant but learned that this was not the case.
Another incentive was a particular offering:

I saw the ENVoYTM was coming up and I always wanted to learn some more

nonverbal things so I thought, it can’t hurt me so I’m just going to sign up and do

it because I wanted to take the class. And then I’m sure other good things will
come up in the future that I could do too.

While EN VoYTM was appealing, Andy would not have signed up without the
additional incentive provided by the money: “There’s always that, for me, that
motivation, too — the money. . .It does motivate so I saw that was there.”

Despite his initial hesitation, Andy has become a strong proponent of Professional

Pay. In addition to its financial appeal, he believes that the policy may provide increased

214

opportunities for professional interactions with colleagues. In Andy’s case, this has
already proven true. Several teachers reported doing informal observations in his
classroom to see the ways he incorporates ENVoYTM techniques into his teaching. In
addition, both Andy and his colleagues mentioned ongoing discussions about the
techniques promoted in the workshops.

More than any other teacher I interviewed, Andy appeared to be making changes
to his practice as a result of participation in rewarded offerings. Andy reports that
ENVoYTM training encouraged him to make changes to his classroom teaching. Two
teachers who occasionally observe Andy’s classroom also volunteered to me that Andy is
using the ENVoYTM skills regularly so the self-reported change in practice is apparent to
at least two classroom observers.

Andy is not the only teacher who reports that money was the tipping point in
deciding whether or not to participate in rewarded offerings. When I asked Anastasia
whether she would have taken advantage of rewarded opportunities without the promise
of additional pay, she replied:

I think I would have done them if they had been offered on a professional paid

day. But I don’t know that I would have taken them at night, without the pay,

because I’ve done so many of those [night classes] already for different reading
programs and things. And it just gets to be a lot. 18 hour classes, I mean,
evenings — for free — I don’t know that I would have continued doing that

Chris is more skeptical of the offerings but willing to give them the benefit of the

doubt thanks to the financial incentive. She is skeptical about the content of rewarded

215

 

offerings and believes that “the one shot, canned thing” is unlikely to help her improve

 

her teaching:

There is no teacher that I know alive that goes to a session for four hours and

says, ‘Oh my gosh, I know how to teach much better than I ever taught before.’

Until you really get whatever it is into your routine and you make it a habit, it’s

not going to happen.

Despite her reservations, Chris was predisposed to taking advantage of
professional growth opportunities, and, therefore, she was open to the idea of
Professional Pay’s offerings. Money was the push: “They say you’re going to get a $500
bonus for doing this [data analysis class] so okay, I’ll do it.”

These nine teachers cited financial considerations as significant factors in their
decisions to participate in rewarded activities but they represent a small fraction of
interviewed teachers. The majority of MPS teachers are not responding strongly to the
financial incentives in Professional Pay and small reward amounts, relative insensitivity
to financial incentives, and a perception of overwhelming busyness may be muting their
responses.

Lack of Trust

As discussed in Chapter 4, agency theory assumes that the principal is trustworthy
and that agents are aware of this, but this assumption does not always hold true in this
study. Several teachers expressed reservations about the district’s ability to administer
Professional Pay in an objective, trustworthy manner, believing the district would favor

certain teachers over others or renege on their financial promises. Others erroneously

216

believed rewards were based on teacher performance evaluation and did not believe
evaluations would be done properly.

One of these teachers, Howard, expresses concerns about favoritism when
rewards are distributed or professional development opportunities are made available.
His concern is largely based on prior experiences in the district.

Howard ’s story.

Howard is a seasoned elective teacher with almost 25 years of teaching
experience, the bulk of it at Washington. Initially hesitant about being interviewed,
Howard agreed when he realized he could be interviewed in a private setting (most of the
Washington interviews took place in the teacher’s lounge). He overcame his initial
reluctance because he wants his story, and the stories of other teachers of color, to be
heard and feels that they are often ignored.

Howard is involved in interdisciplinary teaching with several other teachers and
enjoys a good working relationship with much of the staff. As an elective teacher, he
teaches children across grade levels and, based on conversations I observed in the lounge
and in the halls, shares information about students and their needs with a broad cross
section of his colleagues. When he has an occasional problem, he reports being
comfortable asking other teachers to observe his teaching and that he reciprocates as well.
As a seasoned teacher, the younger teachers seem particularly eager to learn from him
and several of them are often drifting around, like shy fans hoping to get noticed by their
favorite celebrity.

Howard finds his principal supportive and accommodating but, given his

experience and the assistance available from colleagues, he rarely needs help from school

217

administrators. Howard’s only real source of frustration within his building is that he

 

finds that he and the other teachers of color are rarely asked to serve on school
committees or placed in leadership roles. He said, “Sometimes you want to be asked
because someone thinks you are capable. Now, you feel sort of special when that
happens. I don’t see that happening fairly and equitably around the building now.”

Howard is quite distrustful of both the district and union, and his first comment
regarding the policy was, “I wonder about their motive.” He feels that it will be difficult
for teachers of color to be assigned additional responsibilities that qualify for rewards or
get into rewarded professional development opportunities. He worries that, “Just like we
[teachers of color] don’t get perks or strokes now, anything having to do with
Professional Pay or paying people more money, I can absolutely see that not going
toward teachers of color.”

As a result of past experiences with the district, Howard is also concerned that the
district will not provide policy support. When asked about his relationship with central
office, Howard notes with a rueful laugh that, “I have really lost my trust and respect over
the years because they promise so many things to us that they haven’t come through
with.” He has little faith in the union either feeling that, “The union and the district are
in bed together. . .we get no representation.” Howard has no interest in participating in a
policy he feels will be used to hurt teachers, believing that the district is dropping a
poorly articulated policy in teachers’ laps with no support. He notes that, “They pass
these things and leave teachers to work it out. . .and when it doesn’t work out, the district

can blame it on teachers, say it was your colleagues that did it.”

218

Howard is not alone in his concerns. Brandon, another teacher of color, said:

I have some serious questions about how people would be rewarded under the

plan. . .Right now, with the steps and lanes, it’s fair. If you have the education,

then you get paid within a certain range. Now it’s been my experience that
there’s a lot of exceptions to the rules and I can see doing that [making sure the
same opportunities to earn additional pay were offered to everyone] would be

very hard to monitor. ’8

Lou worries promised rewards will not be paid. He recalled another policy
where, “I jumped on, we were supposed to get a brand new computer if we did a certain
amount of stuff. And then right at the end, they said, ‘Well, you know, that was an
error.”’ Lou is concerned that participation will involve a great deal of work and he will
never see the prorrrised financial rewards.

Professional Pay is a new concept for MPS and there are few examples in other
districts for teachers to look to for reassurance. Changing to a new and innovative pay
schedule requires a leap of faith on the part of participants. These teachers do not trust
the district enough to make that leap. They all cite negative experiences during their
tenure with the district, making them skeptical about the likelihood the policy will be
implemented in an objective manner and promised rewards paid.

Distrust was a problem for many survey respondents as well. When asked about
factors influencing their non-enrollment decision, 40 percent of teachers agreed to at least
some extent that one factor was, “I don’t believe the district will actually pay us what we

have earned.” As with the interviews, respondents with poor general perceptions of the

 

’8 Given the small sample size in this study, it is unclear whether teachers of color are
generally more likely to be concerned about favoritism than white teachers.

219

 

district and its policies were more likely to doubt that the district would behave in a
trustworthy fashion. The district perception variable was highly correlated with doubts
about receiving earned rewards (Spearman’s rho = -.404, a =.006’9).

Problems with distrust may be very resistant to district intervention. As
previously noted, trust is much easier to destroy than to rebuild (T irole, 1996). When
teachers talk about their lack of trust, they have concrete stories supporting their
concerns. Howard talked about broken district promises during one search for a new
school administrator; Lou told about the promised computer that was never delivered
despite his belief that he had earned it. Both individuals and entire groups of employees
can become distrustful based on one broken promise. This is particularly true in a sector
such as teaching where colleagues have opportunities to share stories with each other.
Repairing this damage is a long term problem and one that, as discussed in Chapter 4,
may be almost impossible for MPS to accomplish.

Summary

Response to Professional Pay has been muted. When this study was conducted,
60 percent of MPS middle school teachers avoided responding altogether, choosing to
remain on steps and lanes. Almost half of enrolled teachers had not participated in any
rewarded activities, leaving fewer than one-fourth of teachers as active participants.

While individual stories are unique, reasons generally fell in one of four
categories: district efforts to build capacity are unpersuasive because teachers do not find
opportunities professionally worthwhile and/or do not have time to take advantage of

them; teachers lack adequate, accurate information; financial rewards are not sufficiently

 

’9 The negative sign reflects the scaling of the variables. More positive perceptions of the
district and higher trust levels were positively related.

220

motivating; and teachers do not trust the district to properly administer the policy. The
most commonly cited of these are insufficient time to take advantage of opportunities and
a lack of adequate, accurate information (Table 5.3). This was true for both non-enrolled
and enrolled teachers. Teachers believe that they are too busy to add more activities to
their busy schedules and are hesitant to actively participate in a policy they only partially

understand.

Table 5.3: Interviewed Teachers’ Reasons For Lack of Active Participation

N on-enrolled Enrolled teachers Total
teachers
Capacity: not professionally 17 % 40 % 24 %
worthwhile
Capacity: insufficient time 46 % 50 % 47 %
Information problems 46 % 50 % 47 %
Not financially worthwhile 13 % 0 % 9 %
Lack of trust 13 % O % 9 %

Note: As many teachers cite more than one reason for a lack of active participation,
column totals will exceed 100 percent.

Teachers, particularly enrolled ones, also cited concerns about whether active
participation would be professionally worthwhile. Assuming that enrolled teachers have
a greater incentive to learn about specific rewarded activities than their non-enrolled
peers, it appears teachers who know more about the rewarded professional development

may be more skeptical of its value than their less knowledgeable peers.

221

Insufficient financial incentives and a lack of trust were also cited as obstacles but
only by non-enrollees. As mentioned earlier, union personnel strongly encouraged
teachers to “run the numbers” before enrolling to be reasonably certain that their salaries
would be higher under Professional Pay than under steps and lanes, making enrolled
teachers less likely to have found insufficient financial rewards a problem. Teachers who
do not trust the district to fairly administer the plan and pay short-term rewards would
also have little incentive to enroll.

Many teachers cited more than one reason for their lack of active participation.
The preceding cases demonstrated some of the ways these reasons interact, making it
particularly difficult for the district to increase the attractiveness of active participation.

Surveyed non-enrollees were also asked about factors affecting their decision
(Table 5.4), and their responses were generally consistent with interview responses. The
most commonly endorsed survey factor was insufficient time but almost as many teachers
agreed that a preference for choosing their own professional development affected their
decision as well. The other common reason in the interviews, insufficient information,
was also endorsed by over half the survey respondents. Concerns that participation
would not be financially worthwhile and that the district would renege on promised
rewards were much more common among survey respondents than among interviewed

teachers.20

 

2” Survey respondents may have agreed to any factors that seemed reasonable, even if
they had not actually been particularly affected by those factors, as they tried to
remember back to their initial enrollment decision. This would tend to increase the
credibility of the interview responses. Another possibility is that the menu of choices
offered in the survey may have jogged respondents’ memories and that survey responses
reflect the complexity of their decision-making more accurately than interview responses.

222

Table 5.4: Surveyed Teachers’ Reasons For Non-enrollment

Reason % Respondents % Respondents
who agree to who agree to a
some extent very great extent

Capacity: not professionally worthwhile 29% 42%

“I prefer to make my own choices about
professional development, rather than
taking the classes the district selects for
MPP ”

Capacity: insufficient time 29% 47%
“I’m already doing as much as I can and
couldn’t take on one more thing”

Information problems 18% 33%
“I don’t know enough about the program
to be comfortable participating in it”

Not financially worthwhile 33% 18%
“The extra money wasn’t worth the effort
required to earn it”

Lack of trust 24% 16%
“I don’t believe the district will actually
pay us what we have earned”

Teachers’ Responses and the Agency Context

Total %
Agreeing

71%

75%

51%

51%

40%

If we measure effectiveness in terms of changing teacher behavior, Professional

Pay does not appear to be a resounding success. The majority of teachers are not enrolled

and almost half of enrollees have not done anything in response to the policy.

These somewhat dismal findings are consistent with the predictions of agency

theory. We expect teachers to avoid additional work and professional development is

hard work. The rewards embedded within this policy appear insufficient to entice the

majority of teachers to even put in the time necessary to earn rewards, never mind the

223

additional time needed to integrate what they have learned into their practice. While the
theory appears to do an adequate job of predicting teacher behavior on the whole, it
leaves policy-makers with little to do but throw up their collective hands in despair.

In this section, I propose a series of theoretical elaborations to agency theory that
provide an alternative to this despair. Current theory does an adequate job of predicting
overall teacher behavior but the assumptions underlying these predictions do not reflect
the richness of teachers’ rationales. Issues of capacity, information problems, a lack of
sufficient incentives, and distrust all appear to be discouraging active enrollment.
Incorporating these issues into agency theory in a more nuanced way — rather than
treating them as easily ignored assumptions - will provide both policy-makers and
theorists with a more useful analytical tool.

Capacity

The first questionable assumption relates to capacity and can be broken into two
components, excess effort capacity and excess technical capacity. Agency theory
assumes that agents are capable of working harder with no deleterious effects on
Performance, in other words, that they can increase their effort. Only a few of the
teachers interviewed in this study seemed to believe that they were capable of doing
more; the vast majority felt that they could not take on another thing. Some teachers
related stories of a personal nature, but many recited an exhausting list of professional
responsibilities. While respondents may have been inflating their actual obligations, it
seems reasonable to assume that at least some honestly believe themselves incapable of
Working harder. For these teachers to participate in rewarded activities, they would have

to 8ch up some of their existing professional activities or the district would have to

224

provide teachers with release time for additional professional development during the
school day.

This brings us to our second component, that of excess technical capacity. The
theory assumes that a lack of will is the only obstacle to improvement and that agents
possess the skills needed to improve their performance. If this were the case, it would
make sense for MPS to pay teachers for outcomes, rather than for participation in
targeted professional development. Teachers could then concentrate their efforts on
working harder in the classroom, rather than devoting time and energy to ongoing
professional development. However, even teachers who are predisposed to reflect
policy-makers’ goals in their own practice may not have the technical capacity to do so
(Ball, 1990b; Cohen, 1990; Cohen & Hill, 2001; Wilson, 1990). As discussed in Chapter
3, MP8 recognizes this problem and took steps to address it in the design of the policy
but, as we have also seen, building technical capacity is a daunting task the district’s
efforts have only begun to address.

The district recognizes that technical capacity must be increased in order for the
policy to be effective and has tried to do so by tying capacity building programs to
rewards. For better or for worse, teachers appear to have their own ideas about
increasing their technical capacity and these ideas are not aligned with district offerings.
Designing effective professional development opportunities for a diverse group of
eligible employees requires an enormous commitment on the part of district
implementers and a leap of faith for teachers. Thus far, the district has not been able to
commit the needed resources and teachers have been unwilling to embrace the district’s

choices.

225

Understanding

Our second questionable assumption is that consistent, complete policy
understanding exists. We have seen that many MPS middle school teachers are confused.
Some believe that rewards are based on student test scores, others believe that they will
be rewarded for activities that do not actually qualify, while still others are unsure about
how to take advantage of professional development opportunities. Some teachers realize
that they know very little about the policy while others believe that they understand the
policy quite well and are passing along their misconceptions to colleagues. None of this
is surprising.

Professional Pay is a complicated, evolving policy. While the district and union
are making good faith efforts to inform teachers about the policy, they have limited
resources to do 50. Time and money spent informing teachers about the policy decreases
the manpower and funds available for developing high quality professional development
offerings and for compensating teachers. In addition, the district’s efforts must compete
with the cacophony of information that teachers are subjected to each day. On a day
when it is entirely possible that a teacher has dealt with an irate parent, two pamphlets
from companies selling teacher resources, an excited student with a rather long story to
tell, yet another IEP form, and what feels like approximately 57 loudspeaker interruptions
during the day, it is understandable if teachers fail to notice the message informing them
about a rewarded professional development opportunity or an informational meeting.
Teachers are not responding to Professional Pay; they are responding to their
understanding of the policy and this is often quite different from the message that policy-

makers hope to disseminate.

226

Financial Motivation

Our third questionable assumption is that money is a strong financial motivator.
While agency theory does not dictate the primacy of financial incentives, policy with
theoretical roots in agency theory often makes this assumption. The hierarchy of
incentives generally assumed in compensation policy places money at or near the top.
While teachers do respond to financial incentives, current reward levels do not appear to
be effective as a primary motivational tool. Only nine interview subjects cite money as a
factor in their decision to participate in rewarded activities and only four of these say that
it was their primary motivator. While this may be partly attributable to professional
norms — teachers may be reluctant to report financial motivation — there is little evidence
that money is a particularly powerful motivator for these teachers since so few enrolled
respondents have taken steps to earn rewards. They have no qualms about expressing
their view of money as a desirable good; it just is not the most important motivator in
their professional lives.

Teachers appear more concerned about their students’ welfare and their own
professional autonomy than about financial rewards, investing time in activities that they
believe help them do a better job of teaching. It is possible that the promise of greater
student achievement is a necessary, but not sufficient, incentive and that financial
rewards must also be part of the package offered to teachers but it is clear that current
financial rewards alone are an insufficient incentive for the majority of MPS middle
school teachers. While teachers are not immune to financial enticements, their response

elasticities may be rather low. It is said that everyone has a price and it is possible that

227

increasing reward amounts would strengthen teachers’ motivation to participate in
rewarded activities but the needed amounts may be quite large.

In trying to overcome this obstacle, the district faces a resource allocation
nightmare. Given that the state grant which funds the program is a fixed amount and the
district has few discretionary funds available to supplement that amount, we can consider
the funding appropriation a zero sum game. In order to make the financial rewards
greater, the district could reallocate some of the available monies from paying for course
development to increasing the amount of rewards. This would, unfortunately, be likely to
decrease the quality and quantity of offerings and worsen problems with unattractive
learning opportunities.

Trust

Our final questionable assumption is trust. We saw in Chapter 4 that trust is an
important component of credible commitment but it appears that trust plays a direct role
in some teachers’ short-term decision-making as well. Short—term concerns about policy
administration and payment of rewards may be more amenable to policy intervention
than longer-term credible commitment concerns, however. More complete policy
specification may lessen teachers’ fears as might more consistent, complete
communication. MPS could also place reward money in a “lockbox” controlled by an
entity outside the district, such as the union, reassuring teachers that promised rewards
will be paid. While measures such as this are unlikely to completely eliminate short-term
trust issues, they may alleviate them somewhat.

These elaborations provide policy-makers with a powerful tool to use when

designing and modifying policy. Many of the teachers interviewed during this study

228

appeared to share the district’s goal of excellent teaching but were, nevertheless, ignoring
this policy. A renewed focus on issues of capacity, understanding, motivation, and trust
may help MPS make Professional Pay more attractive to teachers.

As discussed in Chapter 1, there is no perfect teacher compensation system. This
does not mean that current initiatives cannot be improved upon or that maintaining the
status quo is the only workable choice. It does mean that careful attention must be paid
to teachers’ responses to any compensation policy, including the status quo. It is also
important to look at the rationale underlying teachers’ responses. Teachers have
legitimate concerns about alternative compensation systems and incorporation of these
concerns into our analytical framework will enrich our theoretical and empirical

understanding of compensation policy.

229

Chapter Six

Revisiting Agency Theory

15 Professional Pay “better” than steps and lanes? Will it serve as a catalyst,
encouraging teachers to make positive changes to their behavior? There is theoretical
and empirical evidence that an ideal teacher compensation system does not exist; each
system entails tradeoffs and Professional Pay is no exception.

In this chapter, I focus on the tradeoffs and obstacles that must be faced as MPS
implements Professional Pay. First, I review the theoretical predictions introduced in
Chapter 1 and elaborated on in Chapter 3. Following that, I summarize empirical
evidence supporting or refuting those predictions, unpacking the reasons underlying
principals’ and agents’ policy responses. Finally, I revisit agency theory, proposing
elaborations that will enrich our understanding of teachers’ responses to compensation
systems.

Theoretical Expectations

In Chapter 3, we predicted that teachers would be unwilling to enroll in this
voluntary program unless they believed its benefits would outweigh its costs, primarily
time and effort. When the theory is operationalized, financial benefits are the primary
ones considered, particularly when compensation policy is discussed. We also predicted
that enrolled teachers would be active participants, willing to change their behavior in
order to earn rewards. This changed behavior would not necessarily be consistent with
policy-makers’ goals, but would allow teachers to earn rewards with the least possible

additional effort. Generally, we would predict that teachers would tend to avoid taking

230

on additional work and that they would either avoid responding to the policy altogether or
would do so in ways that maximize rewards while minimizing additional effort. These
predictions are both quite general and would apply to any deviation from a bureaucratic
compensation system.

The remaining predictions were much more closely tied to the details of this
particular alternative. Professional Pay is primarily an individual, piece rate,
performance-based system. This classification has several implications for agent
behavior. Since rewards are based on individual, rather than group, performance, free
riding is not a problem. Each teacher has a financial incentive to perform to the best of
his or her ability. Unfortunately, teachers have no incentive to help colleagues or devote
time to activities that benefit the group rather than themselves. A tournament system
would exacerbate this problem, but Professional Pay is piece rate so teachers have no
incentive to actually undermine colleagues’ work - everyone can be a winner. The
inclusion of one group performance component — the school-wide achievement award -
may also help minimize the tendency of the other rewards to discourage helping
behaviors.” Of course, this group performance component may also encourage free
riding. If teachers are in high achieving schools, they might not maximize their
individual performance, being content to receive the school-wide achievement reward.

The bulk of the rewards are performance-based, giving teachers an incentive to
participate in as many rewarded activities as possible. We would predict that enrolled

teachers would decrease their participation in non-rewarded activities, substituting

 

2’ The district-wide achievement award is also group based but the district is such a large
unit (over 4,500 teachers) that individual teachers are unlikely to see either their direct
contributions or the contribution attributable to their helping behaviors as impacting the
likelihood of earning the reward.

231

rewarded ones whenever possible. In a true performance-based plan, we would expect to
see rewarded behaviors increase as well. For example, teachers might concentrate on
items that appeared on an evaluation checklist — at least during the times that a monitor
was present. In Professional Pay, incentives to change performance at all are minimal.
Performance measures used by the district are largely proxies for actual performance.
For the most part, teachers are rewarded for participation, that is, their successful
completion of courses; the only outcome that comes close to measuring practice is an
action research project. Classroom observations are not included and there is no attempt
to monitor long-term application of what was learned. Under this scenario, we would
expect teachers to move quickly from one rewarded activity to another, maximizing
income while minimizing the effort spent incorporating what was learned into classroom
instruction.

The individual, piece rate, performance-based nature of Professional Pay has
implications for the principal as well. In a piece rate system, the amount of earned
rewards is limited only by the number of rewarded activities offered. It is possible for
every MPS teacher to earn a substantially higher salary by rapidly completing rewarded
activities. If this occurred, the district could quickly find itself with unsustainably high
salary costs. The potential for rising costs provides the district with a strong incentive to
exhibit a lack of credible commitment. We would predict the district would limit the
availability of rewarded opportunities, minimizing total reward payments. As more
teachers participated in rewarded activities and salary costs increased, we would expect
the district to consider eliminating or radically restructuring the policy. It is too early in

the policy’s history for rising costs to encourage elimination or restructuring -— salaries

232

have not increased very much up to this point — but we can test our prediction that
rewarded offerings will be scarce.
Empirical Evidence

This section will test the predictions outlined above, using the empirical evidence
presented in the preceding chapters. Predictions regarding principal and agent behavior
are generally supported by the evidence but the reasons underlying that behavior often
look quite different than those suggested by the theory. This has implications for
appropriate district action. In this section, evidence concerning district behavior will be
presented first, followed by a discussion of teachers’ responses.

District Behavior

Agency theory does a reasonably good job predicting principal behavior, although
the reasons underlying that behavior are not always consistent with theoretical
expectations. We predicted the principal would exhibit a lack of credible commitment.
Since rewards are primarily piece rate, we anticipated the principal might try to control
salary costs by restricting opportunities to earn rewards and, if salaries rose too high, by
restructuring or eliminating the policy altogether. All of these things appear to be
happening in MPS.

We anticipate that the principal might try to restrict opportunities to earn rewards,
particularly critical skill set and MPSU 1-2-3 rewards. The primary ways for teachers to
earn rewards quickly are by participating in critical skill sets and MPSU 1-2-3 courses.
While there are other options, such as attainment of advanced degrees, it is highly
unlikely that a full-time teacher would be able to earn an advanced degree in a short

period of time. The availability of critical skill sets and MPSU 1-2-3 courses is also

233

under the principal’s control. The district cannot prevent teachers from enrolling in
university courses but can restrict the availability of critical skill sets and MPSU 1-2—3
courses by limiting the number of offerings and the number of spaces available within
each offering. There is evidence that the principal is doing just that. i

In year one of the policy, only one rewarded offering, a critical skill set, was
available. In the policy’s second year, there was again only one rewarded offering — an
MPSU 1-2-3 course. The sparse offerings limited the upside potential of teachers’
salaries. It appears that restricted space within each offering may be a problem as well.
One enrolled teacher has been unable to get into the classes, complaining that, “They had
something with a limit so it got filled up right away.” In addition, several participants
who had taken classes commented that classes filled very quickly; nonparticipants had
heard this as well.

It appears that the district is restricting the ability of teachers to earn rewards.
What is not entirely clear is if this is a deliberate attempt to hold down salary costs or if it
is a by-product of the district’s limited capacity to develop and fund a large number of
offerings. The implementation team discussed the difficulty they were experiencing in
finding time and money to develop courses. The district may not be choosing to restrict
offerings; district leaders may desire to offer more but lack the resources to do so. The
net result is the same — teachers have few opportunities to earn rewards - but the policy
implications of these two underlying reasons are quite different. If the district is
deliberately trying to restrict salaries, a movement to a tournament system or a change in
the state funding formula might reduce the incentives to limit offerings. If the problem

stems from the district’s limited capacity to offer professional development, greater

234

attention to the staffing and funding of professional development might be a more
appropriate response.

Our second prediction is that the district might try to eliminate or radically
restructure the policy as salary costs rise. We would not expect to see evidence of this so
early in the policy’s history, yet the future of Professional Pay is already in doubt. The
district has no contractual obligation to offer Professional Pay beyond the 2004-2005
school year. Changes in state and local leadership have eroded support for the policy and
continued funding beyond the initial grant period is far from certain.

Unlike more tractable problems related to limited offerings, changes to design and
funding may have only a limited effect on the district’s ability to make a long—term
credible commitment to this policy. As discussed in Chapter 4, the political and
institutional framework within which public school districts operate may limit their
ability to credibly commit. MPS needs financial support from state policy-makers or
another source to continue offering Professional Pay. In addition to their inability to
credibly commit due to factors largely outside of the district’s control, current district
leadership may also lack the desire to do so. As constant personnel changes reduce the
number of district leaders with a history of supporting Professional Pay, the policy’s
persistence becomes increasingly unlikely. District leaders may choose to let the policy
quietly join the legions of other educational innovations that appear as bright flashes
across the educational landscape, only to die out as quickly as a shooting star (Tyack &
Cuban, 1995).

In summary, it appears that our theoretical predictions are correct. After only two

years, MPS is exhibiting a lack of credible commitment. In the short-term, access to

235

rewarded opportunities is restricted while it appears that the policy itself may disappear
in the slightly longer term. While our predictions are consistent with observed outcomes,
the underlying reasons may not be consistent with the theory of action underlying agency
theory. Districts may lack the capacity, rather than the will, to offer teachers a menu of
rewarded offerings. In addition, their lack of credible commitment may, at least partially,
stem from factors outside of district control, rather than reflecting a deliberate choice.

Let us now compare our predictions of teacher behavior to observed responses.

Teacher Behavior

The ability of agency theory to predict teacher behavior is limited. Our broad
expectation that it is difficult to use incentives to change behavior in ways consistent with
policy-makers’ goals is generally true; however, the assumptions underlying our
predictions are often contradicted by the empirical evidence. This is similar to our
findings regarding principal response.

Generally, we predicted teachers would tend to avoid taking on additional work;
the limited response to this policy is consistent with that. The majority of teachers have
chosen to remain on steps and lanes, and almost no enrolled teachers report making
meaningful changes to their practice. One possible interpretation of this finding is that
teachers will try to avoid the additional work of earning rewards. Some teachers might
find the participation rewards insufficient motivation for changing their behavior,
choosing not to enroll. Others might enroll and then do the minimum work necessary to
earn rewards, rather than embracing the spirit of the policy and using what they have

learned to improve their practice. This interpretation is consistent with agency theory.

236

However, the empirical evidence suggests another possibility. There is evidence
that factors other than distaste for work are causing teachers to avoid active participation,
including a perceived lack of credible commitment from the district, relatively low
interest in financial rewards, valuation of professional autonomy, limited understanding,
and a perceived inability to work harder. These other factors dominate teachers’
responses, so much so that we are unable to tease out whether distaste for work is an
issue as well.

Teachers have little incentive to respond to a policy if they believe the policy will
simply disappear if they wait a while. As discussed above, there is evidence that MPS is
not credibly committed to Professional Pay. On its own, a lack of credible commitment
might not be a problem. Teachers could still be responding to the policy as if credible
commitment existed. This does not appear to be the case in MPS, for the vast majority of
MPS teachers have doubts about the district’s commitment to Professional Pay. Fewer
than 15 percent of interviewed teachers believe the policy will still be in place in several
years and over half believe the policy will not persist much beyond initial funding period.
Surveyed teachers exhibited similar skepticism, with only 27 percent agreeing that the
state or district will continue funding Professional Pay after the initial grant period. Only
43 percent believe that Professional Pay will still be offered in five years; the rest believe
survival is unlikely. Many of their reasons mirror those mentioned above — teachers
believe there is insufficient political will at the state and local levels to sustain the
program and that funding constraints are likely to lead to an early demise. A few

teachers were even more skeptical. These teachers, discussed in Chapter 5, do not even

237

believe that the district is credibly committed in the short-term. They did not trust the
district to fairly distribute promised rewards during the current period.

A lack of credible commitment gives teachers a strong incentive to avoid
responding or to do so in very short term ways since the policy and associated rewards
are expected to disappear quickly. Some teachers avoid the policy altogether. Enrolling
in the policy requires some work on teachers’ part: finding the relevant information,
comparing costs to benefits (including calculating their expected salary change), and
completing enrollment forms. The shorter the amount of time that teachers expect the
policy to persist, the smaller the incentive to devote time to enrollment. The majority of
MPS teachers are not enrolled and the perceived lack of credible commitment may be
discouraging them from doing so. The role that a perceived lack of credible commitment
plays in depressing enrollment may, in turn, increase the likelihood the policy will be
seen as a failure and abandoned. As this cycle of policy churn repeats itself, it will
become increasingly difficult for MPS to convince teachers that credible commitment
exists and that meaningful responses to future policies are worthwhile.

Teachers who choose to enroll may also be skeptical about credible commitment
but enroll in order to be rewarded for activities they had already planned to do. This too
seems to be occurring; several teachers mentioned enrolling in order to receive financial
rewards for participating in professional development they had planned to attend anyway.
These teachers are not changing their behavior; they are simply taking advantage of
opportunities to be paid for preexisting behavior. The district’s long-term commitment is

not particularly important to these teachers; they are doing what they already wanted to

238

do. In this ease, financial resources are being allocated inefficiently as teachers are being
compensated for continuing current behavior, rather than making desired changes.

Some teachers may be trying to take advantage of short-term rewards but
avoiding the work associated with longer term benefits. For example, a teacher might
enroll in a critical skill set rather than begin work on an advanced degree. The critical
skill set entails less total work than earning a degree. In addition, all the required
activities can be completed and the reward earned within a single school year. While the
financial rewards are greater for earning an advanced degree, teachers may be concerned
that the policy will disappear before they are able to complete the degree and qualify for
rewards.22 The evidence supports this — a greater number of teachers were participating
in short-term opportunities than longer term ones — but it would be unwise to place too
much weight on this finding. The policy is so new that teachers may not have had
sufficient opportunity to complete the enrollment process for longer term activities, such
as applying to graduate programs.

A perceived lack of credible commitment may be discouraging some teachers
from enrolling and dissuading others from changing their behavior in substantive ways.
Principals generally assume agents will treat each new idea as a permanent addition to the
landscape, yet education policies often come and go. Seasoned teachers know this,
having lived through various policy cycles, and novices hear the stories from their more
experienced peers (Fuhrman & O'Day, 1996; Hess, 1999; Nieto, 2003). In MPS, many

teachers believe Professional Pay will come and go, having no perceptible impact on their

 

22 While advanced degrees are also rewarded under steps and lanes, the amount of the
reward is reduced. Teachers would weigh the likelihood of receiving the larger amount
paid under Professional Pay against the smaller amount paid under steps and lanes when
making their decisions.

239

daily lives or their long-term compensation. Thus, it makes sense for teachers to wait on
the sidelines as Professional Pay flashes across the sky.

While credible commitment may be a major factor in teachers’ muted responses,
other factors come into play as well. One is the motivational power of the rewards
themselves. Applied agency theory presumes teachers compare expected financial
rewards to the cost of the behavioral changes necessary to earn those rewards. We
predicted that enrolled teachers would believe financial rewards outweighed costs while
non-enrolled teachers would come to the opposite conclusion. The evidence generally
supports this prediction. Sixty-three percent of non-enrolled survey respondents feel
rewards are not worth the effort required to earn them while only 37 percent of enrolled
teachers believe rewards are inadequate (Phi = -.258, 01:.009).23 It appears that a
comparison of expected rewards and required effort plays a role in the enrollment
decision. Enrolling is a relatively low risk decision. Teachers are only committed to
Professional Pay for one year and then they can return to steps and lanes. They are not
required to take on any additional responsibilities as enrollees. They automatically
qualify for school and district performance awards and their only losses are the annual
step increases forgone during the enrolled period. For more experienced teachers who
are frequently frozen on the steps and lanes schedule, this annual loss does not even come
into play. A belief that rewards are a reasonable reflection of increased effort might be

sufficient incentive to enroll, since little is lost by enrolling.

 

23 Among interviewed teachers who responded to the survey, active participants who said
money played a role in their decision to participate in a rewarded activity were more
likely to disagree with the statement, “The rewards offered under Professional Pay are too
small for the work required to earn them,” (83 percent compared to 63 percent) than other
interviewees. This result is also consistent with our theoretical prediction, but the
difference is not statistically significant.

240

Active participation is a different story. We predicted that enrolled teachers
would be active participants, taking steps to earn rewards, but the evidence suggests
otherwise. Only 13 out of 23 enrolled interviewees reported participating in any
rewarded activities and four of them said that they would have completed the activities
whether they were rewarded or not. A favorable reward/effort ratio appears to entice
enrollment, but is insufficient, on its own, to convince teachers to participate in rewarded
activities. Other priorities appear to be taking precedence over active participation,
including teachers’ concerns about their ability to increase effort beyond current levels,
their reluctance to cede authority over their professional development to the district, and
inadequate policy understanding. It does not appear that insufficient reward amounts are
the dominant decision factor suppressing active participation, since 63 percent of
surveyed enrollees believe that rewards are reasonable in light of required effort.

Enrolled teachers report feeling unable to add additional responsibilities into
already crowded schedules. As discussed in Chapter 5, many teachers, both enrolled and
non-enrolled, believe current professional and personal obligations make it virtually
impossible for them to do more. Half of the inactive enrolled teachers say that they do
not have time to participate in any rewarded activities. The agency framework assumes
that agents have excess effort capacity, but this assumption appears inconsistent with
teachers’ perceptions of their ability to work harder.

Given that teachers believe they are already maximizing their effort, we might
expect them to eliminate some current activities, freeing up time to earn rewards. We
predicted teachers would substitute rewarded activities for unrewarded ones in an effort

to minimize effort, even when they were able to increase effort levels; teachers who

241

believe they cannot increase effort should be even more likely to make these
substitutions. The bulk of the empirical evidence does not support this prediction.
Teachers are not generally substituting rewarded activities for unrewarded ones.
Moreover, they often cite time spent on unrewarded activities as a major obstacle to
active participation. When questioned about this, teachers often discuss issues of
professional autonomy. Many believe that they, rather than the district, are the best
judges of what professional development will help them improve their practice. At
current reward levels, teachers are unwilling to surrender their professional autonomy to
participate in district choices. While all of the interviewed teachers clearly found money
a desirable good, many valued their professional autonomy more. Teachers appear to
have a strong preference for activities they feel are professionally worthwhile and
promised financial rewards are insufficient to entice them to abandon this preference.
Unless MPS can convince teachers that district offerings will help them develop as
professionals, many teachers will continue to ignore rewarded offerings. Some teachers
may begin returning to steps and lanes in order to qualify for step (experience) increases,
further eroding the policy’s potential to improve teacher quality.

Limited understanding also appears to be depressing active participation rates.
Half of the inactive, but enrolled, teachers cited information problems as a factor in their
lack of active participation. The agency framework assumes understanding is complete
and consistent but this is not always the casein practice. As former North Carolina
Governor J arnes Hunt Jr. notes, “We haven’t developed many approaches [to alternative
compensation] that are simple enough that people can understand” (Keller, 2004, p. 4).

MPS teachers are confused about Professional Pay. Many recognize that they do not

242

completely understand the policy, but some harbor misconceptions that may be
discouraging active participation.

Agency theory assumes teachers are motivated primarily by financial rewards, yet
many of the teachers in this study appear to value other goods, including their
professional autonomy. While money can be a sufficient inducement for relatively small
changes (e.g., enrollment), it rarely motivated study participants to make larger
behavioral changes, such as participating in rewarded professional development, not to
mention changing their classroom practice. When obstacles such as a lack of time,
reluctance to cede professional autonomy, and limited understanding enter the picture,
current financial rewards are, by and large, ineffective. Even among the small group of
active participants interviewed, only five out of thirteen said money was the primary
motivator in their decision. Of the remainder, four saw money as a tipping point in the
decision to participate in activities that already interested them. The remainder said that
they would have participated regardless of the financial incentives to do so. Unless
teachers are comfortable with their policy understanding and believe that a particular
offering is consistent with their own professional development agenda, they are unlikely
to be swayed by promised financial rewards.

The assumption that money is a primary motivator for changing teacher behavior
has been found suspect by other researchers as well. As Johnson (1986) notes:

Although many change agents presume that teachers are motivated by money and

are intentionally withholding their best efforts, research suggests otherwise.

Discovering what matters to teachers and how best to motivate them for sustained

243

and improved work is apparently a complicated puzzle, one that is yet to be

solved. (p. 74)

While teacher motivation remains a complicated puzzle, I nominate several
factors that may improve the effectiveness of financial motivators. It appears that
teachers must believe: that policy-makers are credibly committed to change, that the
policy is consistent with their own professional frameworks, that they understand the
policy, and that increased effort expectations are reasonable. All of these pieces are not
yet in place in MPS, dampening teachers’ responses.

Agency theory also leads us to predict that teachers would change their behavior
as little as possible in order to earn rewards. We would expect them to do only what is
necessary to earn rewards, minimizing changes to practice. Data about classroom
practice and teacher quality were not collected. Hence, we cannot carefully examine this
prediction here. However, teachers self-reported few significant or sustained changes to
their practice. As with our predictions above, it appears that the prediction is correct but
that the underlying rationale may not hold true in this study. Agency theory predicts
minimal behavioral changes because of a distaste for additional work — change is difficult
and teachers would be expected to avoid it. While this may very well be true, it is
difficult to test in our case because of the policy particulars. Agency theory assumes that
excess technical capacity exists. MPS recognizes that this is often not true for the
district’s teachers, so the district incorporated capacity building activities into the reward
structure. Despite their efforts, these activities may still be insufficient to build technical
capacity. Short-term professional development opportunities may not provide teachers

with sufficient opportunity to learn how to behave in desired ways. Since we are not sure

244

that teachers know how to change in the desired ways, it is impossible to determine
whether the minimal changes they report are due to reluctance to change or the inability
to do so. Indeed, research on the limited effectiveness of traditional professional
development would suggest that teachers did not have sufficient opportunity to learn
(Cohen & Hill, 2001; Fullan, 1991; Hawley & Valli, 1999).

Agency theory also predicted teachers would move quickly from one rewarded
activity to another in order to maximize rewards. Given that only two rewarded
professional development opportunities were offered during the two year period of this
study, it was impossible to test this hypothesis.

We also predicted that the individual nature of the rewards might lead to a
decrease in helping behaviors. Several interviewed teachers expressed concerns about
this, but no one was able to supply examples. The policy is so new and so few teachers
appear to be actively responding to it that the lack of evidence to support this prediction
is not surprising. As the policy matures and more teachers become involved, it would be
worthwhile for future studies to revisit this prediction.

MPS has offered teachers a thoughtfully designed alternative to steps and lanes
yet Professional Pay’s effectiveness has been limited. The district’s commitment appears
weak and teachers know it. In response, many are ignoring the policy. Even those who
express their interest by enrolling are discouraged from actively participating by their
perceived inability to work harder, skepticism regarding the district’s efforts to build
technical capacity, and limited (mis)understanding. Unless steps are taken to address
these problems, Professional Pay may evaporate. While our theoretically-based

predictions are generally supported by empirical evidence, the rationale underlying these

245

predictions is incomplete. Teachers may very well have a distaste for work — a
fundamental assumption of agency theory - but the other factors discussed above are
clearly affecting their behavior as well. We will return to these factors in the next
section, when elaborations to the theory are discussed.
Revisiting Agency Theory:
Towards a More Sophisticated Understanding of Teacher Response

Agency theory generally does an adequate job predicting principal and agent
response but, in its current form, does an inadequate job of explaining the rationale
underlying those responses. This limits the theory’s usefulness both as a conceptual
framework and as a tool policymakers can deploy when designing and implementing
compensation policy. I propose several elaborations to the theory in its current form,
including a greater focus on the impact credible commitment has on principal and agent
behavior, explicit attention to issues of understanding, relaxation of the excess capacity
assumption, and greater attention to the role non-pecuniary factors, such as
professionalism, play in decision-making. These elaborations will enhance our
understanding of the complex interplay of factors affecting principals’ and agents’
decision-making.

Agency theory tends to focus narrowly on reward/effort ratios and agents’ distaste
for work. These are not trivial factors. Agents do appear to consider the work required
to earn rewards when making decisions and there is no evidence that they will seek out
additional work unless there are rewards — intrinsic or extrinsic - attached to that work.

The changes I propose are not inconsistent with these factors but are designed to place

246

them within a broader set of considerations — considerations that appear to affect
behavior as well.

Credible commitment has received some attention in the literature, but is often
couched in terms of the principal’s incentive to behave in ways that are inconsistent with
credible commitment. The ability of public agencies to make a credible commitment,
even if they desire to do so, has not been explicitly considered. In addition, the effect that
a perceived lack of credible commitment has on agent response has been largely
ignored.“ Greater attention to critical commitment’s implications within the agency
framework may improve our ability to predict and explain principal and agent behavior
while deepening our understanding of underlying reasons. In addition to its theoretical
implications, credible commitment has practical ramifications as well. Policy-makers
promoting alternative compensation often base their arguments on neoclassical economic
theory and many of their arguments resonate with a simplified version of agency theory.
It appears that credible commitment may be a fundamental challenge faced by public
agencies as they attempt to design and implement alternative compensation policy, yet
policy-makers often seem unaware that credible commitment is even an issue, never
mind a potentially intractable dilemma. Giving greater theoretical prominence to credible
commitment may encourage them to consider its implications when they design and
implement policy.

The second proposed elaboration is more explicit attention to issues of
understanding. Some compensation policies, such as bureaucratic systems and bonuses

based on absolute student test scores, are relatively straightforward. Others, such as

 

2" Miller’s (1992) and North and Weingast’s (1989) work are notable exceptions.

247

Professional Pay, are complicated. Simple is not always better. It is difficult to use
bureaucratic systems as a tool for improving teacher quality while bonuses based on
absolute student test scores introduce issues of fairness, among other problems. That
said, designing a complex policy in an attempt to minimize the problems of simpler
systems introduces a new problem — that of limited understanding. Rather than assuming
understanding is complete and consistent, agency theorists could explicitly model the
costs associated with the uncertainty introduced by more complex systems, particularly
when those systems are relatively new. This approach might prove more fruitful than
simply ignoring the issue or restricting study to relatively straightforward empirical
examples. As with credible commitment, a greater focus on issues of understanding
could also help policymakers as they seek to minimize undesirable outcomes attached to
compensation policy. Policy-makers would have a tool that explicitly recognizes the
tradeoffs involved as we move from simplicity to complexity. More complex systems
may avoid some of the perverse incentives embedded in simpler ones but are likely to
increase the problem of limited understanding. Including issues of understanding in the
decision matrix will encourage policy-makers to broaden the range of considered factors.
I also propose a relaxation of the assumption that agents possess excess capacity.
It is generally assumed that agents have the ability to work harder, smarter, or both. This
may not be the case. At times, agents may believe that they are incapable of working
harder. They might be right. When this occurs, agents may require a very large financial
incentive to increase their efforts. An alternative would be redirecting existing efforts
rather than asking agents to increase them. For example, MPS could use substitutes to

free up teacher time during the school day, allowing teachers to work on rewarded

248

professional development without increasing their effort levels. Relaxing the assumption
that agents are capable of increasing their effort levels allows both theorists and policy-
makers to explicitly consider tradeoffs between attempts at increasing total effort levels
and encouraging agents to substitute desired activities for other activities they already
participate in.

Agency theorists also assume agents are capable of working smarter - that they
possess the skills needed to do a better job. We have seen that this is not always the case.
For example, teachers may need to learn how to improve student achievement before they
are able to do so. Recognizing this, policy-makers in Minneapolis made professional
development the cornerstone of their new compensation policy. Rather than assuming
excess technical capacity exists in the system, agency theory could recognize technical
capacity as a potential constraint. This would allow theorists to include the cost
associated with increasing technical capacity in their models and would encourage
policy-makers to consider the possibility that technical capacity might be a limiting factor
in their plans.

The final elaboration I propose is greater attention to the role non-pecuniary
factors such as professionalism play in agent behavior. Current theory recognizes that
money is not the only factor in agent behavior. Distaste for work is recognized as a
factor and other considerations such as geographic preference and valuing leisure time
are often included as well. It is assumed that financial rewards can and should be used to
entice agents to ignore these other factors. It is also generally assumed that financial
inducements, rather than other incentives, are the most fruitful way to change behavior.

Financial rewards certainly play a role in decisions but other factors may be at least as

249

important as money for some agents. This appears to be the case with teachers who
enrolled in Professional Pay but then chose not to participate in rewarded activities.
Many of these teachers cited concerns about professional autonomy and the ability of the
district to present them with appropriate professional development opportunities.
Treating money as the most important motivational tool may be discouraging theorists
and policy-makers from giving adequate attention to these other factors, some of which
could conceivably offer a more cost effective way to change behavior. In this study,
teachers’ perceptions of the policy’s impact on their professional autonomy played a
major role in their responses. Broadening the range of potential motivators to include
factors such as professional autonomy would recognize the role of these other motivators,
allowing theorists and policy-makers to consider the interactions among them and
relative costs associated with their manipulation in their work. For example, it might be
more cost effective to focus a larger share of resources on improving the quality and
diversity of rewarded offerings, minimizing teacher’s concerns about professional
autonomy, rather than increasing the amount of rewards.

All of these proposed changes - a greater focus on credible commitment’s impact
on principal and agent behavior, explicit attention to issues of understanding, relaxation
of the excess capacity assumption, and greater attention to the role non-pecuniary factors,
such as professionalism, play in decision-making — are elaborations to the existing
agency framework. They do not question the basic theory of action underlying the
theory, but rather allow for a more nuanced understanding of the likely responses of
principals and agents. Fully incorporating these additional considerations into the

existing framework will improve agency theory’s theoretical and applied utility, assisting

250

researchers and policy-makers as they strive to fit together the pieces of the compensation

puzzle.

251

Appendix I

Instruments

Teacher Interview Protocol
Demographics and background
Q1: I’d like to begin by asking you a little about your background in teaching.
How long have you been a teacher?
How long have you been teaching at this school?
Have you taught at other schools in the district? For how long?
Have you taught in districts other than Minneapolis? For how long?
Q2: Outside of anything that you did because of the plan, what kinds of things do you do

to improve your teaching?
Prompts: How about courses?
How about professional development workshops?
How about professional reading about teaching or about (insert name of subject taught)?
Teacher’s knowledge about the professional pay plan
Q3: Can you please tell me what you know about the MPS professional pay plan?

Prompts: What do you know about receiving additional pay for certain types of
professional development such as critical skill sets (Improving Instruction through

Analysis of the Data) and MPS 1-2-3 ?

What do you know about receiving additional pay for credits towards advanced

degrees ?

252

What do you know about the quality performance awards available if your school

or the district meet improvement goals?

Does this plan or any parts of the plan remind you of any other policies in this

district or in any other districts you’ve worked for?
Q4: Why do you think the district decided to offer this plan?

Prompts: Which groups or individuals have played a critical role in developing

and implementing the plan (ex. union, superintendent, administrators, teachers)?

Why do you think the union decided to work with the district in designing the

plan?

Teacher ’s participation in and assessment of the plan

Q5: Next I’d like to talk to you about participation in the plan. Are you a participant?
Q6 if answer is Yes: Why did you decide to participate?

Prompts: Did you talk to your colleagues or anyone else when you were trying to

decide?
Was there a part of the plan that caught your attention ?

Was there anything specific about this school or your situation that made the plan

attractive ?

Q6a if answer is Yes: Please tell me about what you’ve done so far as a plan participant

and what you plan to do in the next year or so?

Prompts: Can you tell me about one specific activity that you have engaged in

because of the plan and how, if at all, it has affected your classroom teaching?

253

Are there any other things you have done because of the plan?
How is that different from what you would have done without the plan ?
Q6 if answer is No: Why did you decide against participation?

Prompts: Did you talk to your colleagues or anyone else when you were trying to

decide ?
Was there a part of the plan that caught your attention ?

Was there anything specific about this school or your situation that made the plan

unattractive ?
Q7: Do you see any changes in the school or in your colleagues as a result of the plan?

Prompts: Have you noticed teachers paying more or less attention to certain

aspects of their teaching as a result of the plan ?
Have any teachers tried to change their assignment as a result of the plan ?

Have they done anything else that they think might make it easier to qualify for

rewards?
Is there anything else that changed in the school as a result of the plan ?
Q8: What is your professional and personal opinion of the plan?
Prompts: What positive consequences could this plan have for teachers?
How about negative consequences?
What positive consequences could this plan have for students?

How about negative consequences?

254

Do you think the plan rewards things that are reasonable measures of teacher

quality?

Do you think there are any ways that a teacher could make more money under

this plan without actually improving their teaching?

Do you think it’s easier or harder for certain teachers or for the teachers in

certain schools to qualify for rewards?
On balance, will this plan have a negative or positive impact on teaching?

Q9: 15 there anything in the pay plan that would make you reconsider teaching? Are
there any aspects of a non-teaching job, such as sales, that now look more attractive to
you? Unattractive to you?

Q10: What do you think the plan will look like in a few years? How about in a longer

period of time - let’s say 10 years?
Prompts: What about the state funding? Will that still be the same?
Teacher’s relationships with district and school personnel

Q11: Now I’d like to talk about your school and the school community. Would you
please tell me about a typical interaction with a colleague — whether in a formal setting

such as a school committee or informally?
Prompts: Have you been satisfied with your interactions with colleagues?
Are there any other kinds of interactions that you would like to have?
Are there any that you would prefer not to have?

How often does your team get together?

255

Is this similar to or different from other teams in the school?

Q11a: Does the new system give you any reasons to behave differently with colleagues

or have you noticed any changes since Professional Pay started?

Q12: Please tell me about a time that you asked your principal for help with something -

instruction, discipline, etc. — and how that worked out.
Prompts: Is that typical of your interactions with the principal?

Are you satisfied with the kinds of interactions you have with your principal and

how frequently they occur?
Do you think your experience is typical of most teacher/principal interactions?

If the principal wound up doing any sort of evaluation for pay at some point in

the plan, how would you feel about that?

Q12a: Does the new system give you any reasons to behave differently with the principal
or have you noticed your colleagues behaving differently with the principal since

Professional Pay started?

Q13: How often have you interacted with central office? Please tell me about your

experiences.

Prompts: What are the different types (curriculum, special ed. questions, payroll,

health insurance, etc.) of interactions you have had?

Have the people in central ofiice generally been positive and helpfitl or have your

experiences been somewhat frustrating ?

256

Do district policies on things like testing and attendance affect you in the

classroom? Why or why not?

Are policies usually helpful or do they interfere with your teaching? Can you

please give a specific example?

Are these policies pretty stable or do they change frequently? Do they tend to

complement each other or do dijferent policies contradict each other?

If central office wound up doing any sort of evaluation for pay at some point in
the plan, how would you feel about that? Do you think it would be done so that it was

fair to all schools ?

Q13a: Does the new system give you any reasons to change your interactions with central

office?

257

Principal Interview Protocol

Demographics

Q1: I’d like to begin by asking you a little about your background.

How long have you been a principal?

How long have you been principal at this school?

Have you been the principal or assistant principal at any other schools? Where and for
how long?

Have you ever been a teacher? Where and for how long?

Principal’s knowledge about and assessment of the professional pay plan

Q2: Can you please tell me what you know about the MPS professional pay plan?

Prompts: What do you know about receiving additional pay for certain types of
professional development such as critical skill sets (Improving Instruction through

Analysis of the Data) and MPS 1-2-3 ?

What do you know about receiving additional pay for credits towards advanced

degrees ?

What do you know about the quality performance awards available if your school

or the district meet improvement goals?

Does this plan or any parts of the plan remind you of any other policies in this

district or in any other districts you ’ve worked for?
Q3: Why do you think the district decided to offer this plan?

Prompts: Which groups or individuals have played a critical role in developing

and implementing the plan (ex. union, superintendent, administrators, teachers)?

258

Why do you think the union decided to work with the district in designing the

plan ?
Q3a: What is your opinion of the plan?
Prompts: What positive consequences could this plan have for teachers?
How about negative consequences ?
What positive consequences could this plan have for students?
How about negative consequences?

Do you think the plan rewards things that are reasonable measures of teacher
quality?
Do you think there are any ways that a teacher could make more money under

this plan without actually improving their teaching?

Do you think it’s easier or harder for certain teachers or for the teachers in

certain schools to qualify for rewards?

Q4: Next I’d like to talk to you about the ways that the plan has affected your work and
your school. Roughly what proportion of your teachers decided to participate in the
plan?

Did you notice any patterns among participants (e. g., more or less experienced
teachers, teachers who have always been active, your “stars certain departments,

etc. ) ?

Q5: Did any of the teachers talk to you about the plan when they were making their

decision?

259

Prompts: What were some of their comments and questions?
What kinds of information were you able to give them?
If you were unable to help a teacher, who were you likely to refer them to next?

Q6: Please tell me about some of the activities that your participating teachers have been

apart of

Prompts: Have you noticed any changes in their teaching? If so, can you tell me

about those ?

Can you tell me about one specific change that you have heard about or observed

in the school or in a particular teacher’s classroom?
How do you think, if at all, that would have been dijferent without the plan?
Principal’s relationships with district and school personnel

Q7: Now I’d like to talk about your school and the school community. Would you
please tell me about a typical interaction with one of your teachers — whether in a formal

setting such as a school committee or informally?
Prompts: Are you satisfied with your interactions with teachers in this school?
Are there any other kinds of interactions that you would like to have?
Are there any that you would prefer not to have?
Do you think the plan will affect your interactions with teachers?

If the plan was modified so that you did evaluations, how would you feel about

that?

260

Q8: How often do you interact with central office? Please tell me about your

experiences.

Prompts: What are the different types (school issues such as budget, testing
schedules and personal issues such as payroll or health insurance questions) of

interactions you have had?

Have the people in central office generally been positive and helpful or have your

experiences been somewhat frustrating?

Do you ever act as a “go-between ” for one of your teachers and central office

personnel? If so, can you tell me about that?

Do district policies on things like testing and attendance affect your ability to do

your job? If so, how? If not, why is that?

Are policies usually helpful or do they interfere with your ability to be an effective

school leader? Can you please give a specific example?

Are district policies pretty stable or do they change frequently? Do they tend to

complement each other or do different policies contradict each other?

If central oflice ever began to do teacher evaluation under the plan, how would

you feel about that?

Q9: I’m interested in how your teachers are perceived by others, particularly parents.
Would you please provide me with a list of your teachers that characterizes the parental
requests for each teacher in one of three categories: more than can be accommodated,
roughly equal to the number that can be accommodated, and fewer than can be

accommodated.

261

District Personnel Interview Protocol

Q1: I’d like to begin by asking you a little about your involvement with professional pay.
What is your position in the district?
How long have you been involved with the professional pay initiative?
How much of your working time do you spend on matters related to professional pay?
What are your responsibilities related to the professional pay plan?
What other types of responsibilities do you have?
Development of the Plan
Q2: What groups or individuals have been involved in the development of the plan?
Prompts: What did the legislature do?
What did the superintendent do ?
I understand there is a new superintendent. Has this had any effect on the policy?
What did other district administrators do?
Has this changed over time?
What role the union played?
Has this changed over time ?
What about teachers? Who was involved and what have they done?
Has this changed over time?
What about school administrators?
Has this changed over time ?
Who else has been involved and what have they done?

Has this changed over time?

262

The Current Plan
Q3: It seems like the plan has been evolving. From your perspective, what are the plan’s

core ideas?

Q4: Of the various plan components (MPSU, extended professional responsibility

compensation, etc), which do you feel are most critical?

Q5: Which of the plan components are well developed at this time? Which are works in

progress?

Q6: Can you please tell me what you know about planned MPS professional pay plan

components for this year?
Prompts: What are the plans regarding critical skill sets offerings this year?
How about MPS U ?
How are the critical skill set and MPS U oflerings selected? By whom ?

Is the district planning to offer additional pay for credits towards certain

advanced degrees and NBPTS certification?
How about the district and school quality performance awards?

Are there any plans to introduce extended professional responsibility

compensation this year?
Communication of the Plan

Q7: What are some of the opportunities that teachers have had to learn about

professional pay?

Prompts: What materials, if any, have been sent to the schools?

263

To teacher’s homes?

Has the plan been explained/advertised/promoted during any district functions

(workshops, etc.) that teachers attend?
How many presentations has the professional pay team made to schools?

How are these visits generally set up (are you invited and by whom, is the school
administrator involved, are they outside of regular staff meetings or during that time,

etc. ) ?

What other kinds of things has the district done to spread the word about the

professional pay website?

What kinds of things has the union done to inform teachers about professional
pay?
Q8: I have downloaded the literature on professional pay that is available on the website.

Where could I find copies of any other literature on professional pay that has been made

available to teachers and school union representatives?

Q9: When you have personally talked to teachers about professional pay, what’s your

sense of their take on the plan?
Prompts: What parts of the plan seem clear to them ?
Unclear?
Are there any common misconceptions that teachers seem to have about the plan ?
What parts of the plan do they tend to agree with?

What concerns do they commonly express?

264

Early Implementation and Impact of the Plan

Q10: My understanding is that around 1,700 teachers are currently signed up for the
plan. From your perspective, are there certain groups of teachers that are more or less

likely to participate than others?
Prompts: Why do you think these groups are more/less likely to participate?

Q11: Did any of the teachers talk to you about the plan when they were making their

decision?
Prompts: What were some of their comments and questions?
What kinds of things did you talk about?
What kinds of questions did they ask?
What kinds of information did they want?
Q12: From your perspective, how is this plan currently working?
Prompts: What positive consequences could this plan have for teachers?
How about negative consequences?
What positive consequences could this plan have for students?
How about negative consequences?

Do you think the plan rewards things that are reasonable measures of teacher

quality ?

Do you think there are any ways that a teacher could make more money under

this plan without actually improving their teaching?

265

Do you think it ’s easier or harder for certain teachers or for the teachers in

certain schools to qualify for rewards?
The District Policy Environment

Q13: Before I leave, I’d like to ask you a few questions about the district and the larger

context. Are there any other policy initiatives going on in the district right now?
Prompts: Can you tell me about them?
Do district policies usually remain stable, evolve over time, or come and go?
Do these policies generally complement or contradict each other?
Can you please tell me about an example of this?
Do district policies tend to help or hinder teachers in their work?
What set of policies would illustrate this?
Is there anything else?

Q14: Can you tell me a little about the history of district policies related to teacher

quality issues?
Prompts: What about PDP ’s ?
Where did policies such as the PDP come from?
What impact do these policies seem to be having now?

How does professional pay fit into the district 's overall plan for teacher
professional development (PDP ’s, district and other local professional development

offerings, etc.) ?

266

Are there any other district policies that might have implications for professional
pay?
Q15: Can you tell me a little about the union’s involvement in policies and the

relationship between the district and the union in creating and implementing policy?

Q16: I understand that there has been a leadership change and that you have an interim
superintendent. How has this affected the design and implementation of policies? How

has it affected professional pay?

Q17: 15 there anything else I should have asked you to help me understand this plan and

how it is being implemented?

267

Survey Instrument

A Survey for The Minneapolis Professional Pay Plan: The Teacher’s View

As a middle school teacher or other licensed employee, you are invited to participate in a
research project investigating teachers’ reactions to the Minneapolis Professional Pay
Plan and the ways that their reactions are shaped by their interactions with other people in
the system. The results of this study will provide important information to school
districts and states that are trying to develop Professional Pay plans as well as help

Minneapolis public school district and union personnel improve the current plan.

You can help by completing this Web-based survey questionnaire. The questionnaire
should take about 20 minutes to complete and both Professional Pay Plan participants and
non-participants are encouraged to respond. To encourage your participation, surveys
returned by June 12, 2004 will be entered in a drawing for a $50 Barnes and Noble gift

certificate.
The purpose of the survey is to gather information about how teachers are responding to
the plan and what kinds of relationships they have with other people in the system. Your

input about your experiences will provide direction for plan designers and implementers.

Your participation in this survey is entirely voluntary. You are under no obligation to

Participate and, if you choose not to participate, no penalties will occur. If any of the

268

questions make you uncomfortable, you may discontinue participation in the study at any

time without penalty.

Though this survey is coded so that we can identify you (to avoid sending reminders to
those who have responded and identify the gift certificate winner), no one outside of our
research team will ever have access to the codes or your responses. This research is
conducted under stringent university and US. government regulations governing
confidentiality procedures. All data collected will be securely stored. Only researchers
working with the study, all of whom are bound to maintain confidentiality, will have
access to study data. Results of the study will be reported in such a way that individuals
cannot be identified. Your privacy will be protected to the maximum extent allowable by

law.

For questions about this study, you may call Dr. Suzanne M. Wilson, Principal
Investigator at 517-353-9150 or email swilson@msu.edu. If you have any questions
about the protection of human subjects of research, please call Peter Vasilenko, Ph.D.,
Chair of the University Committee on Research Involving Human Subjects, Michigan

State University at 517-355-2180 or email ucrihs@msu.edu.

Sincerely,
Debbi Harris
Research Associate

d;and-z @ iuno.com

269

1. To begin, please enter the five digit id number that was sent to you in your e-mail
notification. If you have any problems logging in, please contact Debbi Harris at d-and-

z@juno.com.

Logon ID Number:

 

Next click the “next” button to go to the first question of the survey.

Thank you for your participation!

270

aerfiv
Klaroldurog

9:;er
£113.13an

93.1388!“
Knurauef)

39.188er
Klaraldruog

among?» 05 E Sam .35 omen—Bos— ofi Co 8283
E288 Stop 0583 EB ham 3:336on .595 moEcntomao
Eofimggoc Ecoimomoa E BEBEE on? Eocene“.

.3532 L8 850 some £5
80988 9 E288 wEfimBEam mowflnooco mam Ecommmomofi

.anwcfi 28
>Eoco=a 3.85805 .2388 FEB 88%an mum 3558.85

.mfiow coEEoo €339 Lonaomoa
x53 9 Eugene wELanEeQ mowfisogo bum .9868on

Sam Ecoamouoi
Lope: E8 53 ES 82: 05 £83 £288 SE“: 2F

.82: Emu 8 @2538 x83
2: c8 =2; 08 2a mam 555835 SES vacate €338 2S.

.5525 5:88 «o 3532:
222538 can :2: mefi €532 :aE bum Ecoumflocm 23.

$2362 c8

woocozcomcoo 338m 02E E3 53 than 3:06.835 2: £830
.332. 2.08863

:98 :33 vogue no» sow—B 3 vogue—o 2: 88:5. 8on .N

mam Esmmsuocm 5;? «design 25> “=2: 3 558m

271

anSV
Kmaldulog

anBV
“18.13an

aafiesga
£113.19an

aaxfissgq
Kmaldwog

in: 3:280:05 80::
8:3B0: F30 0: 3:038 @3020: :wE :0 £0.38: .5: H0330 m: z

43:
3:280:05 50:: €3B0: F30 8 306:8 368 0:: 60:28
.33 0w3:w:3_ .fimfiv 800.38 0:00 :0 80:03: :0: :0330 2 a

.mnuz 05 EB :380: EB
H 35 300—: 0:08 z 003:: ma: :3: 3:280:05 :0 :0:0::0::_ 2F

23: 3:280:05 5:0 0: 0::::00 2:05 “052: 0:9

$5.0: 2.05830
£000 .EB 02w: :0.» :0EB 0. 02%.: 0.: 808:5 800—: .m

03:35.3: 0: :0: 00200: 0::
80:80: :05 t :00: 03: 2:0B >05 :35 to 5:00 00 EB >3:
3:280:05 E 03:3:3: 0:B E0538 :0 850:8 0:: ._3:0>O

00:08: :05 :0 goo—:0: 0: E05 82:50: :3: 0:: 8:38:
E0530: 5:00 0:500: EB >0: 3:280:05 50:: 802::to::0
::0E:0_0>00 3:280:03 E 03:3:3: 0:B E0:030:.

€088.30: :0380: :26: 05 :0 8:300:
8:05:00 .6500 0:500: EB >3: 3:280:05 50:: 80E:::0::0
80:50—05: 35.888: 5 033033: 0:B 305308

272

aslng

an.“

A0900=00000 95000 :06 9.20000 no 02300 0022560008..
:0 00—3 >05 2 0.88 00050000 930 bum $023035

.330
:0 :03 0.800 8:00:00 :05 t 0008 80:88 930 bi 300600005

.5 000903000 .35 005 000800—050 302300000
05 :0 00m .30 3002200 03000.— E0s000~ Sam Ecommmflem 000:3

.m000w00
0000560 E00 >05 3 000.: $050000 930 hi Ecoamfloi

.03.: 0.0 3205033.
w=m§0=0x0SX0 £000 050200 30% 00:00:03 030.003 0803 .v

.000 38 ES»
REE 05 0000.0 >3 30230005 003 00 0:50:00 EB 08:. 0E.

.50 0:3 Efim 008m
REE 05 000% had 300500005 wig—3 0:50:00 EB 0030.8 05.

@000» 0>c E 03¢ 300800.005 mam—0&0 0n :5 EB 00506 05.
.520 xmm 300600005 05 55 005.500 Ea — .2830

.3000 .93 30230005 0 030000
200% 50805200 E0030 0000005 :2: E0050 5 0.002000%

0005000 0050 >00 00 hi 300600.095 880
000030.. wE>_0000 mo 005:0 :02: mm 020; .0053. .50 E 000:000H

273

anBV
Kmaldluog

aolfiv

Knuauag

anBBsgq
Kumauaf)

amfiasgq

.(maldluog

000000000000 mam 9000 00 00:0 00000030000000 000.000.0000
000000000 000 >000 0000000000000 030000 20000 0000000 B

.0000w00 00000>00
000 000000 0w0=00 000 .0000 00000000000 030000 03000 0000000 H.

.0000m 000000000000 30000 00000 000000 00000000
00 000000 000 0003 >00 0000000005 030000 03000 0000000 H

.§0>Zm.. 00 00:0
00000.0 Wm; Dmnfiz 000 >000 000000000000 030000 03000 0000000 0.

...0009 000 00 000300.00 0m=0000 0000000000 w00>000000= 00 00:0
000m :08 003000 000 >00 00000000000 030000 00.000 0000000 0.

.3200 000000080
0000 000.5 000w0 :0» 080B 3 000u00 000 000000.: 00.00..— .m

000000 000 >000 0000 >009:
000_ 00000 >08 >00 000000000000 00 0000000000 003 00000000 0000m

000000003
000 >000 30030000000 030000 00 00000 03 0000000020 00000
E 000002 0>00 >000 000? 000 >000 0000 000000000000 00000 0000000 0.

s00 0000000000000 000 30000
00000300000 00000002030 3000000000 00000030 0000000 .300

274

9313‘;
Kmaldwog

anSv
KneJauaf)

99.13%!(1
“18.13an

anfiesga
Kmaldulog)

0000300 000
30000 000 000 0 00 00000 00000 0000 0000 00 0000 0>00 00003 H 0000
>0m 000000000000 000 0000000 0000 000—0 0 0000 8 >000: 00000 E0 _

.>0m 00000000005 000003 00>0
0000002050 0000000000000 030 >00 00 0000 00 0000. 0003 00003 _

.0m0000000 0000000000000 0000000000000 000 00 0000000000 :00 00003
0 000000000 003 >00 000000000000 00 0000000000 00300 000 00

0000.03 00 0000000000000 000E000 000 00000 005 >0m 00000000005
05 0w00000 000002 0>00 0 0:00 000 m0000 0000000 :03 0

.3200 000000030 0000 003 00000
00> 00003 00 000M000 000 0003000 00005 ”000030.000k 00k .0

|oz l 00>

000000 0000900000

000> 00 w00000000 .>0>000 000 00 0000000 00>0—00

00 0000000000 00 >=0000E0000 :03 00> 0000 0000 0000.5
000E >000 0000000000000 000 00 w0000000000 >0000000 00> 00< .0

0000000000 >000 00003 00 00000000000 000000000050 0000000000000
000 00 :0 000 >000 00000000000 030000 000000 0000000 ,0.

275

@0000 >000000 003 0 0003 000 0000 00w 8 000000

0 m N 0 0 000 0>0w 0000 000 >03>00 00003000 00000300 05 w0000 >00000 003 0
v m N 0 00003008 00.0000 0 003 >058 00000 0000 00 000000 00.0.
m u m. u m u w
m B m S m B 1
... A m. m w W W
H a
m. .u a W W. “...—.32 00 000300000 3 00000000 000> 000.000
W ”00300.00 000 .00 0000 000 000000 0003 0h. 0000000§A~ 00k .0
0>00 0000000000000 000 @0000 000 00 0000
.0 m N 0 00003000 0008000050 0000000000000 00 0000900000 000> 0000000Q
v m m 0 00003000000 30000300 000,0.
00.00w 000000500000 00000 000000.00 000 00
.00000 000> 0 000000>0 000030 000000000000 30000 000 00 0000000
.0 m N 0 w000000_ 0000000 00 0000000 00000 00 00000 $000000 000> 0000.0
0 m N 0 200000000 w0000000 000>v 00000 00> >03 000 0w000U
0>00 000000000000
0 m m 0 000 ©0000 0000 00003000 00000000050 30000000000 00 0w0w0m
D
l
m n H u H o m
19 B u S X B m.
00 mm mm a
W n ... a 1 W W 0000.» 0000 000 0000.5 00 00.: 00> 00 00 00> 0>00 000000

0003 8 . 0 00000000000 00 00.00 000 .0— ..fizummfifium 00% .w

276

m
v—t

(*3
—¢

mama
183-1f)
5.13 A a 01,
lualxfl
smog 01,
lualx'd
news

a 01 N

“V 1V 10M

.05...
0.50. 000 00 000. ..000000 000 000 0 00 0000. 00 w0000 >00000 00.0

000000
0>00 03 .003 00 >00 >000.00 .03 80.0.0 00. .00. 0>00—00 ..0000

... E00 0. 0000000 .0000 00. 0.003 ..0003 >0000. 00.00 00 .0.

0.3—th 0. 0.00.0000.— 0. .00 00.00000 000> .00....0

w0m30=0u 00. ..0 0000 0.0 .0008 .003 0.0. "0.000.005.0002 00k .2
.3200

x00 00. 0. E00. 00.00000 0000.0 0.00.0000.— 0. 030.000 ..00>
00.00....0 .00. 000.00.. 000.0 >00 000 0000. .= 0.00.0000...“ 00.0 .0.

000802050 000000000000
50000 0E0 .00... 00000 0. 3.0000000 00. 0E 002w 03%

0:20 000000000000
>0. flue—0>00 05 E00 .03 .00. 00.00300 0000 0. 80.0.0 00. .00.:

Samoa 00. .0000 >00 0. 0w0_0. 000m 000 000w00=00 >2

000000. 0000000002 00 @0000 00.
0000000. 900 E3 .0 0>00—00 0 0000000 0.00%.... 00. 000000 0. 00.003 0

0.090000 0. 00. 0000000000 3080000000.. 00.00 >2

277

v

anBV
Klmaldwog
anuat)

9913mm

Kmaldluog

33.133ng

anBV Knmauaf)

..00000 0.0. 0. 0.00000. 00. m0000. .00... .0 .000 .00.» 0 0. 0.00 ...

..00E0w.000 .0850 ..00. .0 ...00 00 .00 >00.
.. 0w0o0. 00>0 .00 0.00 o. 0.00.000. 0.0.0 .00.... 00 .0000 000 00>

..00000 >0.

0. 0.00.000. 0.0.0 0000.0 0000 03.0.0080 .0 .000 000.0 0 0. 0.00 b
30.00 30.00.50

0000 003 00.00 00> 00.03 0. 00.000 00. 0.00.00. 0000... .m.

0.0.0...0..0..§<
30000. .000 000000000 003 00000.20me .00... .000... .N 00.00%.

30.00 0.00 00. 0. 0.00. 00.00000 0000... 6.00.0200.—
0. .00 00.0000 .00> 00.00....0 .00. 000.00.. 000.0 >00 0.0 0.00. .— .m—

.m0..00.0...00 0.0.. 00. 00w0.0000.0 0>..0.0000.00. 00.00 >3.
.E0.wo.0 00. .0000 >00 0. 0w0.0. 03.00.00 000 000M00..00 00.00....

... 0. m0..00.0...00
0.0008050 .00. 0. 0.0.3.0 00. .0000 0w0000 3000. ..000 .

>000... 0000 E00 0. 0000.00. 0.0 .00.
0w0.0. 00. w0.00 0003 00 0. .. .0. 0w0o00 w0o. 0000.0 00 ..003 00.0.).

NEE). .0. 0.00.00 .0....0.0 00. 00000.0 00. w00.0. 000. .00.0.
000800.050 3020000000 .0000 000.000 030 >0. 00.0... o. .000... .

278

..00 000.00
0.0 >00. .00. 0000 000 .000... 000.00. 0000000 0.00 000.00.... 00E

>03 00000000>0
...0. 0 0. 0.00000. 00. ..0 .00.. 08.000.000.00 .00000 00%

.0000 ..03 00.. 0 .0. 0000.?00. 0.0 0.00.00... ..0.m
..000. .0 00.0005 0. .003 3000. 0.0080... ..0.0 0.0. 0.0.0000 00..

...0.0
00. >0 0000. 0.00.00... 00. 000.0.0000 08.000.00.000 0.00000 00...

.w0.w0.00000 000
0>...00000 0. ..0.0 00. 0.030. .0.>0000 0..0.0..0.0.0.00 .00000 00..

00.0000. >0. 003 0.0.6.... .0w00 .00. .00000
00. 00.0.00 .00.. 00.0000... 003 >.0>..00..0 0.000 000.00.... 00,0.

.0000. 300 .00 >.. 0. 000w00..00 >0 00000000 .00. .

.0000.
300 w00.000 000 $00.00. >..00.0000 0.0 .00000 0.0. 0. 0.00000 ...

000000...
0.00.0006 000 0.00w w0.0000. 0..00.0 0000 .0 000. 000m 0 0>00 03

279

anSV
KlaJQIdluog

anSV Kumaueg

N

anSBsgq
Kumauag)

9318mm
Kmaldulog

68: 6>o 038m :65»: 86:3 6.52m

.oc 2:2? 96:88
3:3 Son“ 8882: 858 6:8 coco 86:8 6th6 280:5

.3533 HE 6H 8% on 8 88:
x55 H HEB at 2 690 E 86:8 6:66 20st 9 02E coco H

.9668“ m: 8 mans—2 €0sz E8595
co 86:8 £256 £5 5;» 02mm 8 H1656 : H25 H coco

8:52: Bugs—68
£5 a E bi Hmcoawfloi EoEoEEH 9 6558a 65%. as: H

628668 SH x8
H 8;? 6666 ES Ham—o; 2a Hoccofloa 8588 5:5: 6:35

.20: 2 03m
95 w===3 2w m.<mO,H. 05 $55088 5.3 a6: 80: H :25»

.HoEmHHV 25 SH @5103 mue—

9 690 E Eoficwawm 60.6w 6 W86 is 50E? :88“ 2:03 H
.322. 265856

.68 :23 02w“ :9» .653 3 8&2. 2: 8.655 08: .3

86:an HEa
65888 6665 55> 82660an 25> «:25. "m .8398

280

 

 

 

D 062 .: D 0680a .0 89H 0?. .0H

 

060» 3:: :08: no» 0: @68— 3va :80 $300.33 6:3 .3
200:8 0:5 8 Ewan so» 0>m: 800» >88 >25 .5

wmnHHZ 8 Ewsfl so» 0>0: 800% 808 BoHH .oH

$80830 88000: .8 800% :o 6:88 6:8 8% 0: 6:3 .mH

8:88:83 8:.waan ”v 8sz

.08: 80>: 80:80:00 8080: 68 88m .306 8: 680:0 0:,H.

boa 8:600:
>8 :83 80:86:00 03802800 :0 00.88 0 0: 680.8 0:.H.

281

20. If you have any other comments or observations that you would like to make
regarding professional pay, please make them in the space below.

Thank you completing the survey. You are now entered in the drawing for the $50
Barnes and Noble gift certificate. If you would like to make any additional comments,
please e-mail me at d-and-;@iuno.com

282

Appendix H
A Rasch Analysis of Three Measures of Teacher Perception:

Administrators, Colleagues, and District

Three variables are constructed for use in secondary data analysis: Perception of
Administrators, Perception of Colleagues, and Perception of District. Data for each of
the new variables is scaled using a Rasch Rating Scale Model (RRSM). This model
estimates the logit value of a respondent (6n) responding with a category that requires
exceeding k category thresholds (1..) versus k-l to an item (8.) as a linear function of three
parameters that locate the respondent, item, and category threshold onto the same

underlying continuum,

1: 'k

In __m__ = 6 — 8. — ‘r .
1; . n l k

n1k —1
Parameters for this model are estimated using joint maximum likelihood estimation
procedures as implemented in Winsteps (Linacre, 2002b; Wright & Masters, 1982). For
the three variables under consideration, the respondent measure refers to the individual’s
tendency to endorse items as being descriptions of their perceptions of school
administrators, colleagues, and the district. The item calibration refers to the difficulty of
endorsing a particular item, and the threshold calibration refers to the difficulty of
assigning a rating of k versus k — 1 on the rating scale in question. Standard errors for
these estimates are also estimated (e. g., SE9), and the mean of the square of these

standard errors is the mean squared error of the data from the RRSM (MSEe, for

example).

283

When all the ratings for a person are at the extreme ends of the rating scale, raw
scores are adjusted by adding or subtracting 0.3 raw points to the zero or maximum score
to allow for estimation using the maximum likelihood procedure. For the Perception of
Administrators, Perception of Colleagues, and Perception of District variables, these
adjustments are made 10, 4, and 2 times, respectively. In the first two cases, all of the
adjustments are based on perfect positive responses. In the Perception of District
variable, the adjustments are based on perfect negative responses.

For ease of interpretation, the scaled variables are linearly transformed to a Z-
scale. These transformed variables are then used in the survey analysis.

The Rasch Analysis

A Rasch analysis is conducted on the three constructed perception variables:
Perception of Administrators, Perception of Colleagues, and Perception of District. The
quality of the measures for the three scaled variables is examined on the following
dimensions: dimensionality, item quality, effectiveness of rating scale, item hierarchy,
and quality of measures.

Dimensionality

Principal component analysis of the residuals from the RRSM is used to look for
evidence of unidimensionality. Specifically, the eigenvalues for the first factor should
account for a substantial amount of the total variance. The scree plot is also examined.
The plot will asymptote or form an elbow just after the last factor. Factor loadings are
also used to determine if a factor substantially influences the response to each item.

Items with absolute loadings above .4 are considered meaningfully large. When evidence

284

of multidimensionality exists, large factor loadings on the first factor indicate a
dominance of that factor.
Reliability

Reliability of separation, or the consistency of the rank order of variable estimates
with repeated sampling, is examined. There is no agreed upon cutoff for reliability of
separation and the consequences of the rankings for individuals should be considered
when establishing a cutoff score. For low-stakes applications that do not affect
individuals such as this study, a reliability of separation of around .5 is acceptable. The
real RMSE person for non-extreme persons is used in this analysis. The true score
reliability will tend to be higher so the statistic used in this analysis provides a
conservative estimate.

Item Quality and Effectiveness of Rating Scale

Item quality is assessed by looking for evidence that subsamples of items are
responded to by persons with similar levels of satisfaction in similar ways. This is done
using infit mean square statistics with a larger value indicating a mismatch. For items,
infit statistics of greater than 1.4 are cause for concern (Wright & Linacre, 1994).

The effectiveness of the rating scale is assessed using two statistics. Outfit mean
square statistics look at whether hard to endorse items are answered positively by
satisfied persons and easy to endorse items are answered negatively by dissatisfied
persons. A larger value indicates a mismatch. Outfit mean square statistics should be
less than two for rating scale categories (Linacre, 2002a). The coherence of the rating

scale is also assessed by looking at the percentage of measures that we expect to be in a

285

certain category that are actually in that category. Forty percent is the cutoff for a useful
level of coherence between measures and categories (Linacre, 2002a).
Item Hierarchy

We expect some items to be easier to endorse than others. If our scale is
functioning properly, item responses should form a hierarchy that is consistent with those
expectations. Item calibrations within each scale are examined to see whether the items
rank order themselves in the expected manner.
Quality of Measures

Quality of measures is assessed by looking for evidence that subsamples of
persons with similar levels of satisfaction respond to items in similar ways. This is done
using infit mean square statistics with a larger value indicating a mismatch. For persons,
infit statistics of greater than 1.4 are cause for concern (Wright & Linacre, 1994).

Perception of Administrators

Seven items are used to construct the Perception of Administrators variable

(Table AII.1). The items are presented with four Likert-type response options and all

items are polarized to be positive (high scores indicate positive perceptions).

286

Table AII.1: Perception of Administrators Items

Please indicate the degree to which you agree with each statement below

7'? a E a E: T?
2 a. E a. E .3. 2 8
a a 2 3 82° 2::
a e <3 e <3 a
l. The principal deals effectively with pressures l 2 3 4
from outside the school that might interfere with
my teaching.
2. The school administrator’s behavior toward 1 2 3 4
the staff is supportive and encouraging.
3. The school’s administrators understand the 1 2 3 4
problems faced by the staff.
4. The principal lets staff members know what 1 2 3 4
is expected of them.
5. Staff members are recognized for a job well 1 2 3 4
done.
6. The school administrators treat all the 1 2 3 4
teachers in a fair, evenhanded way.
7. The principal sets priorities, makes plans, and 1 2 3 4

sees that they are carried out.

Dimensionality

The scree plot shows strong evidence of unidimensionality (Figure AH. 1).
Eigenvalues indicate that the first component accounts for 73 percent of total variance
while the second component accounts for only 7 percent of variance. For factor 1,
loadings from the exploratory factor analysis range from .82 to .90; substantially above

the .4 value threshold.

287

Figure AII.1: Scree Plot of Eigenvalues for Administrator Items

mmCHm<5tDtQP-m
)ussss )~\\\‘ ).-§\ )-\‘~ )\-§‘ )s\\

)s s V ~ s

§

gffffff’Yfffffffff"ffffffffff‘ffffffffff"ffffffffff"ffffffffff‘ffffffffff"ffffffffff"ffff

fff
0 1 2 3 4 5 6 7

Number

Reliability

The real RMSE person of .88 is high, indicating very good reliability of
separation.
Item Quality and Efiectiveness of Rating Scale

None of the items have infit statistics >1.4. This is excellent and indicates that the

item quality is high.

288

The effectiveness of the rating scale is also high. The four Likert—type categories
had outfit mean square statistics of .81, 1.04, 1.08, and .91. These are all below the
cutoff of two. The coherence between measures and categories is 74, 46, 83, and 80
percent for categories 1, 2, 3, and 4, respectively. These are all above the cutoff level of
40 percent and indicate acceptable coherence of the rating scale.

Item hierarchy

The hierarchy of item endorsability is sensible (Figure AII.2). Items such as being
supportive and encouraging are easier to endorse than items like insulating teachers from
outside pressures and establishing and implementing school priorities. The first is
something that an administrator should have more control over so we would expect more
administrators to be able to do this (and for their teachers to recognize this) than items

over which the principal has less control.

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure Al|.2:
Hlerarchy of Endorsability for Administrator Items
8
'nsulates
7 3_._ e ‘
6 . H e recognizes
priorities
5 4v s
fair
s4 --_,___..__.
g 3 A understands
commnicates
2 1L
1 A supportive
o .r l l I r T
-1.5 -1 -0.5 0 0.5 1 1.5 2
Endorsability

 

 

 

289

Quality of Measures

The mean person measure infit statistic is .90 with a standard deviation of .78
(Table A112). 18 percent of persons have a person measure infit statistic of greater than
1.4. This is somewhat high but an examination of the response patterns for these teachers
indicates a tendency for many of them to find one or sometimes two items either more or
less difficult to endorse than the others. It is possible that these teachers had an
experience related to these the items that make them believe their school administrators
are particularly strong or weak on that aspect of the construct. These teachers are not
responding in illogical ways so the high percentage of flagged responses is not a cause

for concern.

Table A112: Person Measure Summary for

Administrator Items
MS Infit
Mean .90
SD .78
Minimum .07
Maximum 4.44
% Flagged 18% (14/78
persons)

Perception of Colleagues
Six items are used to construct the Perception of Colleagues variable (Table
AII.3). Items are presented with four Likert-type response options and all the items are

polarized to be positive (high scores indicating positive perceptions).

290

Table AII.3: Perception of Colleagues Items

Please indicate the degree to which you agree with each statement below

1. There is a great deal of cooperative effort
among staff members in my school.

2. You can count on most staff members to help
out even though it may not be part of their
official assignment.

3. There is a great deal of trust among the
teachers in this school.

4. We have a good idea of each other’s teaching
goals and classroom practices.

5. Teachers in this school are constantly learning
and seeking new ideas.

6. I feel supported by colleagues to try out new
ideas.

Dimensionality

The scree plot indicates a fairly strong first dimension but that a second

dimension may exist as well (Figure AH.3). The first component accounts for 60 percent

Completely

H

Disagree

Generally

Disagree

N

Generally
Agree

0.)

Completely

Agree

4:.

of total variance, the second component accounts for 14 percent of variance, the third

accounts for 10 percent and the fourth accounts for only 8 percent.

For factor 1, loadings from the exploratory factor analysis range from .70 to .89;

substantially above the .4 value threshold. These loadings indicate that the colleague

291

variable is capturing a common underlying construct. The second factor has loadings
greater than .4 for Items 2 and 4. Both of these items relate to working together with
colleagues either in a helping capacity or by discussions or observations of each others
practices. The second factor appears to be a refinement of the more general Perception of

Colleagues variable that focuses on interactions.

Figure AII.3: Scree Plot of Eigenvalues for Colleague Items

I

4.0 "

mmcwm<smtat+m
)ssss )‘\\‘ )“§\ )ssss )~\§§

5 )s s s §

\

)\ s s s

0.0
éffffffffofffffffff"fffffffffff“fffffffffff‘fffffffffff“fffffffffff”fffffffffffofffff
f

0 1 2 3 4 S 6

Number

292

Reliability

The real RMSE person is .81. As with the administrator variable, this is quite
good.
Item Quality and Effectiveness of Rating Scale

None of the items have infit statistics >1.4, indicating high item quality. The
effectiveness of the rating scale is acceptable. The four Likert—type categories had outfit
mean square statistics of 1.37, .64, 1.00, and .83. These are all below the cutoff of two.
The coherence between measures and categories is 60, 71, 83, and 80 percent for
categories 1, 2, 3, and 4, respectively. These are all above the cutoff level of 40 percent
and indicate acceptable coherence of the rating scale.
Item hierarchy

The hierarchy of item endorsability is sensible (Figure AHA). Most of teaching
takes place behind close doors and teachers are rarely able to observe each others
practices. As a result, they are hesitant to say that they have a good idea of each other’s
classroom practices - they only know about these practices secondhand from the teacher
and her students. The item pertaining to trust is also somewhat difficult to endorse.
Compared to the other items in this section, trust calls for a higher degree of involvement
than the others. You must know someone fairly well before you can claim to trust them
whereas it is fairly easy to feel that a relative stranger can be supportive when you try out

something new since the risk to both parties is much less.

293

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure AI|.4
Hierarchy of Endorsability for Colleague Items
7 supported
6 cooperative v -...
5 A e --
help out
E 4 -- W m— - ~- e --————~-—~——~«
o learners
z 3 a» — o
trust
2 n~———-——-—-—— -- _ 6 ~-—- ~—————~—«—~—
1 ..- -..-.‘SDQYY Hammer- .. - WWW- -W
0 I T T I
-2.5 -1.5 -0.5 0.5 1.5 2.5
Endorsability
Quality of Measures

The mean person measure infit statistic is .90 with a standard deviation of 1.06
(Table AII.4). 20 percent of persons have a person measure infit statistic of greater than
1.4, a fairly high percentage. The most infitting person gave responses that are split
between the two extreme ends of the scale and appear quite inconsistent. Pairs of items
that most persons responded to with similar ratings (e. g., a “4” on one would be found
with a “3” or “4” on the other) would be given disparate responses (e.g., a “4” and a “1”)
by this person. Almost all of the other teachers used a higher or lower rating category for
one item more frequently than we would predict based on their other responses. One
possible explanation for this is that these teachers had recent experiences that related
particularly well to the misfit item and that were dissimilar to their general impressions of
colleagues. The high percentage of flagged responses is a cause for some concern but,

since misfit appears to follow explainable patterns, measure quality is acceptable.

294

Table AII.4: Person Measure Summary for
Colleague Items

MS Infit
Mean .90
SD 1.06
Minimum .12
Maximum 8.47
% Flagged 20% (17/84
persons)

Perception of District

Ten items were initially used to construct the Perception of District variable
(Table AII.5). The items are presented with four Likert-type response options. Seven of
the items are positively polarized with high scores indicating positive perceptions and
three questions are negatively polarized (indicated by an asterisk in Table AII.5). The
scale of negatively polarized items was reversed prior to variable construction (A one
was rescaled as a four, etc.).

In the initial analysis, one of the items, Item 2, did not appear to be measuring the
same underlying construct as the remaining items. It only had a loading of .21,
substantially below the .4 cutoff. This lack of fit was not surprising. During the
interviews, teachers often had very different opinions of TOSA’s and of other district
personnel. Teachers tended to see TOSA’s as supportive of their instructional work
while they were more likely to view other district personnel as either imposing additional
work on them or as bureaucrats who, sometimes ineptly, took care of issues such as

payroll. When discussing “the district,” they did not usually discuss TOSA’s but thought

295

of them as a distinct entity. Some of this may also have flowed from the fact that
TOSA’s are still considered teachers by the district and work under the teacher’s contract.
Since the interview evidence supported the empirical distinction between the district and
TOSA’s, the item referring to TOSA’s was excluded from the latent variable. The

district variable that is used in the analysis includes items 1, 3, 4, 5, 6, 7, 8, 9, and 10.

296

Table AII.5: Perception of District Items

Please indicate the degree to which you agree with each statement below

2' >. >.
3?. "5?. at
a. Q) G) u
E I”: = .g "-' <
c a 5 c: (3
O

1. I would accept almost any class or school 1 2 3

assignment in order to keep working for this

district.

2. When I need help with something, the 1 2 3

TOSA’s are willing and able to help.

3. District human resource personnel are helpful 1 2 3

and cordial when I ask for assistance.

4. I trust district personnel to implement 1 2 3

Professional Pay in a fair, evenhanded manner.

*5. Often I find it difficult to agree with this 1 2 3

district’s policies on important matters relating to

its teachers.

*6. I often have to ignore district policies in l 2 3

order to do what I think needs to be done for my

students.

*7. Different district policies often send mixed 1 2 3

messages about what teachers should do.

8. District policies remain stable over time. 1 2 3

9. The district is a source of considerable 1 2 3

satisfaction with my teaching job.

10. The district has clear goals that remain l 2 3

consistent over time.

297

Completely

Agree

A

Dimensionality

The scree plot shows some evidence of multidimensionality (Figure AII.5). The
first component accounts for 41 percent of total variance, the second component accounts
for 14 percent of variance, the third accounts for 10 percent and the fourth accounts for
only 8 percent. This evidence of multidimensionality is expected given the nature of the
items in this section. Unlike the other two variables, the district variable includes
questions about both personnel and policy. The second factor has loadings greater than .4
for Items 6, 7, and 9. These items all relate to the district’s affect on teaching and the
first two ask directly about policy. These commonalities are resulting in similar
responses to these items.

For factor 1, loadings from the exploratory factor analysis range from .54 to .80;
substantially above the .4 value threshold. Despite the evidence of multidimensionality,
the loadings indicate that the district variable is capturing a common underlying

CODStl'UCt.

298

Figure AII.5: Scree Plot of Eigenvalues for District Items

mmrzwm<bmtowm
)ssss )ssss )sss§ )ssss )ssss )s‘ss )s‘ss
H

)‘ s s s

0.0

gffffffff'fffffff“fffffff‘fffffff‘fffffff'fffffffofffff“fffffff‘fffffff‘fffffff“fffffff

f
0 1 2 3 4 S 6 '7 8 9

Number

Reliability

The real RMSE person is .73. This is above the .5 threshold and indicates
acceptable reliability of separation.
Item Quality and Effectiveness of Rating Scale

None of the items have infit statistics >1.4. All the items appear to be measuring

the underlying construct adequately.

299

The effectiveness of the rating scale is acceptable. The four Likert-type
categories had outfit mean square statistics of 1.00, .91, .88, and 1.65. These are all
acceptable. The coherence between measures and categories is 68, 58, 71, and 87 percent
for categories 1, 2, 3, and 4, respectively.

Item hierarchy

Teachers did not have generally positive opinions of the district’s policies and
personnel. On only one item, the one about policy’s student impact, did the majority of
teachers have positive perceptions.

The hierarchy of item endorsability is sensible (Figure AII.6). The more difficult
to endorse questions generally asked the teachers about the district’s impact on their job
satisfaction while items that were more specific were easier to endorse. They could, for
example, be relatively satisfied with human resource personnel but still not be willing to

accept any assignment to remain with the district.

 

 

 

 

 

 

 

Figure Al|.6
Hierarchy of Endorsability for District Items
10
trust 0 student policies

t h . [hurrah resources _.__- a

eac er ICIeS
S 6 w” M"“';’Eiéér g'o’aE"m "'”-"“'.‘" ' ” "" 29‘"T”W”w— _.,.__.._._ "'“ "' -...___.__.
_ 4 W..._..mw$Q-U§S§§99§M-_-_. ..-m--..-_-._...__-._...r_.-.-__r_-.-..-..wr.-.-_-r.._-..... ”WWW“...

o assignments
2 _ .safsfying
’ stable
-1 -0.5 0 0.5 1 1.5 2
Endorsability

 

 

 

300

Quality of Measures

The mean person measure infit statistic is 1.00 with a standard deviation of .76
(Table AII.6). 17 percent of persons have a person measure infit statistic of greater than
1.4. The most misfitting person felt that policies were quite stable (“4” on item 8) but
gave extreme negative responses to all the other items. This seems reasonable since
stability is not necessarily a desirable attribute if you disagree with the policies that are
remaining in place. The second largest misfit showed a similar pattern. The person did
feel that district policies sent mixed messages but had a generally positive impression of
those policies and the district personnel. This mixed messages item is flagged as
inconsistent more often than any other item and appears to be a lower quality item than
the others. Since most of the persons with a high infit statistic exhibit explainable

response patterns, the high percentage of flagged responses is acceptable.

Table AH.6: Person Measure Summary for

District Items
MS Infit
Mean 1.00
SD .76
Minimum .16
Maximum 4.70
% Flagged 17% (15/86
persons)

301

Discussion

All of these constructed variables, Perception of Administrators, Perception of
Colleagues, and Perception of District, appear to be of adequate quality for inclusion in
analysis. There are some problems with the measures. The Perception of Colleagues and
Perception of District variables exhibit evidence of multidimensionality but a single
dimension does dominate in each case. Given the high reliability of separation, it is
reasonable to treat these as unidimensional variables. Item quality is consistently high
and the rating scales appear to be functioning effectively. The item hierarchy is
reasonable as well. The quality of measures is somewhat low. For all three variables, a
relatively high degree of persons are exhibiting misfit. An examination of these teacher’s
responses shows logical patterns to the misfit so they are not a cause for undue concern.
Overall, the three measures exhibit acceptable quality and are appropriate for use in

further data analysis.

302

Appendix HI

Minnesota 2001 Statute 122a.413-415

The following contains the text of the statutes enabling the state to fund initiatives
such as Professional Pay and outlining the conditions under which this could be done.
122A.413 Educational improvement plan.

Subdivision 1. Qualifying plan. A district may develop an educational improvement
plan for the purpose of qualifying for alternative teacher compensation aid under sections
122A.414 and 122A.415. The plan must include measures for improving school district,
school site, teacher, and individual student performance.

Subd. 2. Plan components. The educational improvement plan must be approved by
the school board and have at least these elements:

(1) assessment and evaluation tools to measure student performance and progress;

(2) performance goals and benchmarks for improvement;

(3) measures of student attendance and completion rates;

(4) a rigorous professional development system that is aligned with educational
improvement, designed to achieve teaching quality improvement, and consistent with
clearly defined research-based standards;

(5) measures of student, family, and community involvement and satisfaction;

(6) a data system about students and their academic progress that provides parents and
the public with understandable information; and

(7) a teacher induction and mentoring program for probationary teachers that provides

continuous learning and sustained teacher support. The process for developing the plan

303

must involve district teachers.

Subd. 3. School site accountability. A district that develops a plan under
subdivisions 1 and 2 must ensure that each school site develops a board-approved
educational improvement plan that is aligned with the district educational improvement

planiunder subdivision 2. While a site plan must be consistent with the district
educational improvement plan, it may establish performance goals and benchmarks that
meet or exceed those of the district. The process for developing the plan must involve

site teachers.

122A.414 Alternative teacher compensation.

Subdivision 1. Restructured pay system. A restructured teacher compensation system is
established under subdivision 2 to provide incentives for teachers to improve their
knowledge and skills and for school districts to recruit and retain highly qualified
teachers, and to support teachers' roles in improving students' educational achievement.

Subd. 2. Alternative teacher Professional Pay system.

(a) To participate in this program, a school district must have an educational
improvement plan as described in section 122A.413 and an alternative teacher
Professional Pay system as described in paragraph (b).

(b) The alternative teacher Professional Pay system must:

(1) describe the conditions necessary for career advancement and additional
compensation;
(2) provide career advancement options for teachers retaining primary roles in student

instruction;

304

(3) use a Professional Pay system that replaces the step and lane salary schedule and is
not based on years of service;

(4) encourage teachers' continuous improvement in content knowledge, pedagogy, and
use of best practices; and

(5) implement an objective evaluation system, including classroom observation, that is
aligned with the district's or the site's educational improvement plan as described in
section 122A.413.

Subd. 3. Report. Participating districts and school sites must report on the
implementation and effectiveness of the alternative teacher compensation plan,
particularly addressing each requirement under subdivision 2 and make biennial

recommendations by January 1 to their school boards. The school boards shall transmit
a summary of the findings and recommendations of their district to the commissioner.
122A.415 Alternative compensation aid.

Subdivision 1. Aid amount. (a) A school district that meets the conditions of section
122A.4l4 and submits an application approved by the commissioner is eligible for
alternative compensation aid. The commissioner must consider only applications
submitted jointly by a school district and the exclusive representative of the teachers for
participation in the program. The application must contain a formally adopted collective
bargaining agreement, memorandum of understanding, or other binding agreement that
implements an alternative teacher Professional Pay system consistent with section
122A.414 and includes all teachers in a district, all teachers at a school site, or at least 25
percent of the teachers in a district. The commissioner, in approving applications, may

give preference to applications involving entire districts or sites or to applications that

305

align measures of teacher performance with student academic achievement and progress
under section 120B.35, subdivision 1.

(b) Alternative compensation aid for a qualifying school district, site, or portion of a
district or school site is as follows:

(1) for a school district in which the school board and the exclusive representative of
the teachers agree to place all teachers in the district or at the site on the alternative
compensation schedule, alternative compensation aid equals $150 times the district's or
the site's number of pupils enrolled on October 1 of the previous fiscal year; or

(2) for a district in which the school board and the exclusive representative of the
teachers agree that at least 25 percent of the district's licensed teachers will be paid on the

alternative compensation schedule, alternative compensation aid equals $150 times the
percentage of participating teachers times the district's number of pupils enrolled as of
October 1 of the previous fiscal year.

Subd. 2. Percentage of teachers. For purposes of this section, the percentage of
teachers participating in the teacher Professional Pay system equals the ratio of the
number of licensed teachers who are working at least 60 percent of a full-time teacher's
hours and agree to participate in the teacher Professional Pay system to the total number
of licensed teachers who are working at least 60 percent of a full-time teacher's hours.

Subd. 3. Aid timing. (a) Districts or sites with approved applications must receive
alternative compensation aid for each school year that the district or site participates in

the program as described in this subdivision. Districts or sites with applications received

by the commissioner before June 1 of the first year of a two-year contract shall receive

306

alternative compensation aid for both years of the contract. Districts or sites with
applications received by the commissioner after June 1 of the first year of a two-year
contract shall receive alternative compensation aid only for the second year of the
contract. A qualifying district or site that received alternative compensation aid for the
previous fiscal year must receive at least an amount equal to the lesser of the amount it
received for the previous fiscal year or its proportionate share of the previous year's
appropriation if the district or site submits a timely application and the commissioner
determines that the district or site continues to implement an alternative teacher
Professional Pay system, consistent with its application under this section. The
commissioner must approve initial applications for school districts qualifying under
subdivision 1, paragraph (b), clause (1), by January 15 of each year. If any money
remains, the commissioner must approve aid amounts for school districts qualifying
under subdivision 1, paragraph (b), clause (2), by February 15 of each year.

(b) The commissioner shall select applicants that qualify for this program, notify
school districts and school sites about the program, develop and disseminate application

materials, and carry out other activities needed to implement this section.

307

Appendix IV

The 2003-2005 Teacher’s Contract for Minneapolis Public Schools

The following text includes all sections of the 2003-2005 contract pertaining
directly to Professional Pay. Roman numerals distinguishing each section and italicized
text are a guide to the reader and are not included in the actual contract language.

Section I

MPS U SYSTEM 1-2-3 PA YMENT

Effective September 15 2003, teachers on active assignments who are members of the
Professional Pay Plan (PPP) shall be eligible to receive an increase of $1,000 to their
annual salary for after completion of a qualified Minneapolis Public Schools

University (hereafter known as MPSU) course.

The teacher must apply in writing to the Human Resources department. The increase
shall be effective the date of certification, but not more than 45 calendar days prior to
the date of application. Teachers may apply for Professional Pay Plan increases up to
two (2) times in one school year, or thirty (30) times in their career in Minneapolis

Public Schools.

The Minneapolis Staff Development Advisory for the Minneapolis Public Schools
Teacher Academy (hereafter known as MPSTA) shall determine the qualifications of

MPSU courses and requirements for certification of completion of such courses.

Completion of an NH’SU course for certification and eligibility to apply for an increase

shall be guided by System 1-2-3, and shall minimally include the following:

308

1. All activities, attendance, and expectations of the MPSU course are met

completely;

2. Practical implementation of the skills and/or strategies with the teacher’s students
or within other appropriate, assigned, non-classroom duties is clearly demonstrated

for observation and feedback by an instructor or qualified designee; and,

3. Thorough and meaningful assessment of the impact of those skills and/or strategies
and reasonable efforts to share their conclusions with colleagues and others in the

district can be demonstrated.

Teachers who are not members of the Professional Pay Plan, upon being certified as
completing a qualified MPSU course are eligible to use such course completion as

professional growth credits equivalent to three (3) quarter credits toward a lane change.

This section of the contract formal izes the requirements that teachers must
complete three activities in order to qualify for rewards under MPS U 1-2-3. They must
successfidly complete all required coursework, they must demonstrate that they are using
their acquired skills in the classroom or other appropriate setting, and they must conduct
action research about the impact of their new skills and share the results of this research
with others. This also guarantees that teachers will receive a $1 ,000 increase to their

base salaries for successfid completion of each course.
Section II

Section A. Salary: Schedule: Teachers shall be paid in accordance with one of two
salary schedule plans, the standard salary schedules or the Professional Pay Plan option,

adopted by the Board of Education for the 2003-2005 contract period.

309

Subd. 1. Standard Salary Schedules: The salaries for teachers employed under regular
contract in the Minneapolis Public Schools are reflected in standard salary Schedule A
and shall be a part of this Agreement for the period July 1, 20013, through June 30,
20024, and standard salary Schedule B shall be part of this Agreement for the period July

1, 20024, through June 30, 20035.

Subd. 2. Professional Pay Plan Option: Effective upon adoption of this agreement,
teachers shall continue to have the option to participate in the Professional Pay Plan.
Teachers currently enrolled in the Professional Pay Plan shall have a period of forty-
five (45) duty days following publication of the terms and conditions of the
Professional Pay Plan in which to review the plan and exercise the option to
withdraw from the Professional Pay Plan without loss of pay, change of rightful
placement on the standard salary schedule, or other penalty. The Professional Pay
Plan shall continue in effect and remain an option for teachers and shall be part of

this Agreement for the period July 1, 2003, through June 30, 2005. '

Curtailment: In the event that the Professional Pay Plan option should cease to be
available at any time, salaries of current members of the Professional Pay Plan shall
remain unchanged until or unless their current experience and education (step and
lane) placement on the Standard salary schedule places them at a higher amount than

their salary under the Professional Pay Plan at the time of its cessation.

Section B. Relationship of Continuing Contract: The standard salary schedules and
Professional Pay Plan are a part of a teacher's continuing contract as outlined in this

Agreement (Article I, Section C.1.). Pending continued, adequate Minnesota State

310

funding for the plan and beginning prior to June 30, 2005, the Union and the District shall
renegotiate continuation of the Professional Pay Plan.

This section makes it clear that plan participation is optional and that teachers
can withdraw without penalty during the enrollment/disenrollment period. It also
guarantees that, should the plan be eliminated, participating teachers will not have their
salaries reduced upon returning to steps and lanes. Plan continuation beyond the 2004-
2005 school year is contingent upon continued state funding.

Section III
SALARY IN THE PROFESSIONAL PAY PLAN
The following shall be applicable in determining salary of a MPS teacher under the
Professional Pay Plan option:
Subd. 1. Base Salary. The salary of a teacher under the Professional Pay Plan shall be
determined upon the establishment of a base salary upon which the teacher shall
adjust salary through post-secondary education and professional development through

the Minneapolis Public Schools University (MPS-U).

Subd. 2. Conversion of current teachers to the Professional Pay Plan: A teacher’s base
salary will be determined by their placement on the 2004-2005 standard salary

schedule upon converting to the Professional Pay Plan.
Subd. 3. Salary Adjustments Based on Education and Professional Development

Teachers are encouraged to continuously improve their professional skills and
knowledge through coursework and other professional development activities
which enhance their instructional skills and help them meet the needs of the

students they teach. Requests for reclassification on the salary schedule may be

311

made at any time the teacher completes the required work. All requests must be

accompanied by appropriate certification:

° A certified transcript from an accredited training institution showing credits

earned and credit hours completed and degree granted, if any,

° A certificate of professional growth credits for Minneapolis Public School
University (MPS-U) professional development for which completion of training,
observation of implementation and documentation of use of student data have

been duly certified.

° A copy of the National Board Certificate (NBC) or a recognized equivalent

national certificate as agreed upon by the Union and the District.

Certification must be submitted with the salary adjustment request form to the Human
Resources department for evaluation. When determining salary adjustments, credits

earned shall be converted, when necessary, and calculated as quarter credits.

Any increase in salary to which the individual is entitled by reason of salary
adjustment will be made effective at the beginning of the payroll period following
submission of all required documents. However, retroactive salary adjustments shall
not exceed forty-five (45) days nor extend prior to the date of the confirmation of the

degree or credits.

A. Professional Development Advancement: All certified, fully completed MPS-
U courses shall qualify such individual for salary adjustment according to the
specifications prescribed by the agreed upon Professional Pay Plan. Note,

however, that MPS—U credits submitted as professional development

312

advancement, as described herein, may not subsequently be applied as

additional educational credits, as described below.

B. Additional Educational Credits: All accredited college and university credits,
and professional growth credits, including MPS-U credits, earned after an
individual's initial Bachelor's Degree shall qualify such individual for salary
adjustment according to the specifications prescribed by the agreed upon
Professional Pay Plan. Note, however, that MPS-U credits submitted as
additional educational credits, as described herein, may not subsequently be

applied as professional development advancement, as described above.

C. Montessori Credits: For purposes of this Article only, a teacher who
successfully completes a training program in a school accredited by the
American Montessori Society or International Montessori Association shall
receive (1) one quarter credit, applicable as additional educational credits, for
each thirteen (13) classroom contact hours of instruction received without

regard to college or university credits assigned by the training institution.

D. PDP Credits: Teachers shall earn one (1) professional growth credit for PDP

work per year upon demonstration of ten (10) or more hours of PDP.
E. AYP Service Credits: Teachers may earn up to three (3) professional growth

credits (PGC’s) for each year of teaching in a designated Tier 1 or Tier 2 or 3rd
year AYP school setting. Credit shall be given for teaching experience of

more than a semester in a year.

The Human Resources department will not be responsible for making

interpretations of transcripts or certificates without a written specific request.

313

When teachers join Professional Pay, their base salary will be the salary they
would have earned under steps and lanes in the 2004-2005 school year. A teacher can
increase this base salary through the successful completion of MPSU 1-2-3 courses,
through eaming of academic credits through an accredited college or university, and by

earning a National Board Certificate.
Section I V

Subd. 4. Curtailment of the Professional Pay Plan: In the event that the Professional
Pay Plan option should cease to be available at any time, salaries of current members
of the Professional Pay Plan shall remain unchanged until or unless their current
experience and education (step and lane) placement on the Standard salary schedule
places them at a higher amount than their salary under the Professional Pay Plan at

the time of its cessation.

This section reiterates the last part of section II.

314

References

Alchian, A. A., & Demsetz, H. (1972). Production, information costs, and economic
organization. The American Economic Review, 62(5), 777-795.

Al-Naj jar, N. I., & Casadesus-Masanell, R. (2001). Trust and discretion in agency
contracts (Working Paper No. 02-06). Cambridge, MA: Strategy Unit Harvard

University.

Arrow, K. J. (1985). The econorrrics of agency. In J. W. Pratt & R. J. Zeckhauser (Eds),
Principals and agents: The structure of business. Boston, MA: Harvard Business
School Press.

Asch, B. J. (1990). Do incentives matter? The case of navy recruiters. Industrial and
Labor Relations Review, I 990(43), 3.

Bailey, S. K., & Mosher, E. K. (1968). Chapter 2 (The enactment of public law 89-10). In
ESEA: The ofi'ice of education administers a law. Syracuse, NY: Syracuse
University Press.

Ball, D. L. (1990a). The mathematical understandings that prospective teachers bring to
teacher education. The Elementary School Journal, 90(4), 449-466.

Ball, D. L. (1990b). Reflections and deflections of policy: The case of Carol Turner.
Educational Evaluation and Policy Analysis, 12(3), 247-259.

Ball, D. L., Lubienski, S. T., & Mewbom, D. S. (2001). Research on teaching
mathematics: the unsolved problem of teachers’ mathematical knowledge. In V.
Richardson (Ed.), Handbook of Research on Teaching (4th ed., pp. 433-456).
Washington, DC: American Educational Research Association.

Bi g-city districts scrap reward-based systems of evaluating teachers. (2003, October 22).
Education Week.

Blom, D., Kelley, K., Bonaiuto, J ., Skinner, C., Tucker, J ., Stover, D., et al. (2002).
Superintendent tenure (CUBE Survey Report). Alexandria, VA: Council of Urban
Boards of Education.

Borman, G. D., & Kimball, S. M. (2004). Teacher quality and educational equality: Do
teachers with higher standards-based evaluation ratings close student
achievement gaps? (No. TC-04-12). Madison, WI: University of Wisconsin -
Madison.

Brown, C. (1990). Firms' choice of method of pay. Industrial and Labor Relations
Review, 43(3), 165-181.

315

Brown, K. C., Harlow, W. V., & Starks, L. T. (1996). Of tournaments and temptations:
An analysis of managerial incentives in the mutual fund industry. The Journal Of
Finance, 51(1), 85-110.

Buchmann, M. (1990). Beyond the lonely, choosing will: Professional development in
teacher thinking. Teachers College Record, 91(4), 481-508.

Calhoun, F. S., & Protheroe, N. J. (1983). Merit pay plans for teachers: Status and
descriptions. Arlington, VA: Educational Research Service.

Campbell, D. T. (1975). "Degrees of freedom" and the case study. Comparative Political
Studies, 8, 178-193.

Campbell, D. T., & Stanley, J. C. (1966). Experimental and quasi-experimental designs
for research. Boston: Houghton Mifflin Company.

Ceroni, K. M., & Garman, N. B. (1994). The empowerment movement: Genuine
collegiality or yet another hierarchy? In P. P. Grimmett & J. Neufeld (Eds.),
Teacher development and the struggle for authenticity: Professional growth and
restructuring in the context of change (pp. 141-161). New York, NY: Teachers
College Press.

Clowse, B. B. (1981). Brainpower for the Cold War: The Sputnik Crisis and National
Defense Act of 1958. Westport, CT: Greenwood Press.

Coase, R. H. (1937). The nature of the firm. Economica, 4(16), 386-405.

Cochran-Smith. (2004). Walking the road: Race, diversity, and social justice in teacher
education. New York, NY: Teachers College Press.

Cohen, D. K. (1988). Teaching Practice: Plus Ca Change. East Lansing, MI: National
Center for Research on Teacher Education.

Cohen, D. K. (1990). A Revolution in One Classroom: The Case of Mrs. Oublier.
Educational Evaluation and Policy Analysis, 12(3), 311-329.

Cohen, D. K., & Hill, H. C. (2001). Learning policy: When state education reform
works. New Haven, CT: Yale University Press.

Coltham, J. B. (1972). Educational accountability: An English experiment and its
outcome. School Review, 81(1), 15-34.

Conley, S. C., Muncey, D. E., & Gould, J. (2002). Negotiating teacher compensation:
Three views of comprehensive reform. Educational Policy, 16(5), 675-706.

316

Courty, P., & Marschke, G. (1997). Measuring government performance: Lessons from a
federal job-training program. The American Economic Review, 87(2), 383-3 88.

Cuban, L. (1993). How Teachers Taught: Constancy and Change in American
Classrooms 1890-1990. New York: Teachers College Press.

Cusick, P. (1983). The Egalitarian Ideal and the American High School. New York:
Longman.

Danielson, C. (Ed.). (1996). Enhancing professional practice: A framework for teaching.
Alexandria, VA: Association for Supervision & Curriculum Development.

Delpit, L. (1988). The silenced dialogue: Power and pedagogy in educating other people's
children. Harvard Educational Review, 58, 280-298.

Dillman, D. A. (2000). Mail and Internet Surveys: The Tailored Design Method (2nd
ed.). New York: John Wiley and Sons.

Dixit, A. (1997). Power of incentives in private versus public organizations. American
Economic Review, 87(2), 378-382.

Dow, P. B. (1991). Schoolhouse Politics: Lessons from the Sputnik Era. Cambridge, MA:
Harvard University Press.

Drago, R., & Garvey, G. T. (1998). Incentives for helping on the job: Theory and
evidence. Journal of Labor Economics, 16(1), 1-25.

Draper, N. (2004, September 14). State will test teacher merit pay. Minneapolis Star
Tribune.

Eberts, R. W., Hollenbeck, K. M., & Stone, J. (2002). Teacher perfomrance incentives,
collective bargaining, and student outcomes. Kalamazoo, MI: W.E. Upjohn
Institute for Employment Research.

Ehrenberg, R. G., & Bognanno, M. L. (1990). The incentive effects of tournaments
revisited: Evidence from the European PGA tour. Industrial and Labor Relations
Review, 43(3), 748-883.

Eisenhardt, K. M. (1988). Agency- and institutional-theory explanations: the case of
retail sales compensation. Academy of Management Journal, 31(3), 488-511.

Elsbree, W. S. (1939). The American teacher: Evolution of a profession in a democracy.
New York: American Book Company.

317

Encinosa, W. E., Gaynor, M., & Rebitzer, J. B. (1997). The sociology of groups and the
economics of incentives: Theory and evidence on compensation systems (No.
NBER Working Paper No. 5953): National Bureau of Economic Research.

Facts you should know. (2003). Retrieved June 1, 2003, from
http://www.mpls.k12.mn.us/about/facts.shtml

Farkas, S., Johnson, J ., & Foleno, T. (2000). A sense of calling: Who teaches and why.
New York: Public Agenda.

Faul, F., & Erdfelder, E. (1992). GPOWER: A priori, post-hoe, and compromise power
analyses (Version 2.0). Bonn, FRG: Bonn University, Department of Psychology.

Firestone, W. A. (1994). Redesigning teacher salary systems for educational reform.
American Educational Research Journal, 31(3), 549-574.

Frase, L. E. (Ed). (1992). Teacher compensation and motivation. Lancaster, PA:
Technomic Publishing.

Fuhrman, S. H., & O'Day, J. A. (1996). Rewards and reform: Creating educational
incentives that work.'San Francisco: Jossey-Bass Publishers.

Fullan, M. (1991). The New Meaning of Educational Change (2nd ed.). New York:
Teachers College Press.

Gallagher, H. A. (2004). Vaughn Elementary's Innovative Teacher Evaluation System:
Are Teacher Evaluation Scores Related to Growth in Student Achievement?
Peabody Journal of Education, 79(4), 79-107.

Goldhaber, D. D. (2001). How has teacher compensation changed? In Selected Papers in
School Finance, 2000-2001. Washington, DC: National Center for Educational
Statistics.

Goldhaber, D. D. (2002). The mystery of good teaching. Education Next, 2(1), 50-55.

Grimmett, P. P., & Neufeld, J. (Eds). (1994). Teacher development and the struggle for
authenticity. New York: Teachers College Press.

Guernsey, M. A. (1986). Review of related literature and research: History of merit pay,
differentiated stafiing, and incentive pay programs (N o. ED267513). Washington,
DC: ERIC.

Hansen, D. G. (1997). Worker performance and group incentives: A case study.
Industrial and Labor Relations Review, 51(1), 37-49.

318

Hatry, H. P., Greiner, J. M., & Ashford, B. G. (1994). Issues and case studies in teacher
incentive plans (2nd ed.). Washington, DC: The Urban Institute Press.

Hawley, W. D., & Valli, L. (1999). The essentials of effective professional development:
A new consensus. In L. Darling-Hammond & G. Sykes (Eds), Teaching as the
Learning Profession: Handbook of Policy and Practice. San Francisco, CA:
Jossey-Bass.

Heneman, H. I. (1998). Assessment of the motivational reactions of teachers to a school-
based performance award program. Journal of Personnel Evaluation in
Education, 12(1), 43-59.

Heneman, H. I., & Milanowski, A. (1999). Teachers attitudes about teacher bonuses
under school-based performance award programs. Journal of Personnel
Evaluation in Education, 12(4), 327-341.

Hess, F. M. (1999). Spinning Wheels: The Politics of Urban School Reform. Washington,
DC: Brookings Institution Press.

In gersoll, R. M. (2001). The problem of underqualified teachers in American secondary
schools. Educational Researcher, 28(2), 26-37.

Jensen, M. C. (1994). Self-interest, altruism, incentives, and agency theory. Journal of
Applied Corporate Finance, 7(2).

Johnson, R. B., & Onwuegbuzie, A. J. (2004). Mixed methods research: A research
paradigm whose time has come. Educational Researcher, 33(7), 14-26.

Johnson, S. M. (1984). Merit pay for teachers: A poor prescription for reform. Harvard
Educational Review, 54(2), 175-185.

Johnson, 8. M. (1986). Incentives for teachers: What motivates, what matters.
Educational Administration Quarterly, 22(3), 54-79.

Kane, T. J ., & Staiger, D. O. (2002). Volatility in school test scores: Implications for test-
based accountability systems. In D. Ravitch (Ed.), Brookings Papers on
Education Policy, 2002 (pp. 235-283). Washington, DC: Brookings Institution.

Keller, B. (2004, December 1, 2004). Tight budgets prompt more talk than action on
teacher-pay plans. Education Week.

Kelley, C. (1999). The motivational impact of school-based performance awards. Journal
of Personnel Evaluation in Education, 12(4), 309-326.

Kelley, C., & Finnigan, K. (2003). The effects of organizational context on teacher
expectancy. Educational Administration Quarterly, 39(5), 603-634.

319

Kelley, C., Heneman, H. 1., & Milanowski, A. (2002). Teacher motivation and school-
based performance awards. Educational Administration Quarterly, 38(3), 372-
401.

Kelley, C., & Protsik, J. (1997). Risk and reward: Perspectives on the implementation of
Kentucky's school-based performance award program. Educational
Administration Quarterly, 33(4), 474-505.

Kreps, D. M. (1990). Corporate culture and economic theory. In J. E. Alt & K. A.
Shepsle (Eds.), Perpectives on Positive Political Economy (pp. 90-143).
Cambridge: Cambridge University Press.

Ladd, H. F. (1999). The Dallas school accountability and incentive program: An
evaluation of its impacts on student outcomes. Economics of Education Review,
18, 1-16.

Ladson-Billings, G. J. (1999). Preparing teachers for diverse student populations: A
critical race theory perspective. In A. Iran-Nejad & P. D. Pearson (Eds.), Review
of Research in Education (Vol. 24, pp. 211-247). Washington, DC: American
Educational Research Association.

Lawler, E. E. I. (1990). Strategic pay: Aligning organizational strategies and pay
systems. San Francisco, CA: Jossey-Bass Publishers.

Lazear, E. P. (1996). Performance pay and productivity. Cambridge, MA: National
Bureau of Economic Research.

Lee, K., & Weimer, D. (2002). Building value-added assessment into Michigan 's
accountability system: Lessons from other states (Research Report N o. 1). East
Lansing, MI: Michigan State University.

Linacre, J. M. (2002a). Understanding Rasch measurement: Optimal rating scale category
effectiveness. Journal of Applied Measurement, 3(1), 85-106.

Linacre, J. M. (2002b). A user's guide to WINSTEPS/MINISTEP Rasch-model computer
programs (Version 3.36). Chicago, IL: MESA Press.

Lipsky, M. (1983). Street-level bureaucracy: Dilemmas of the individual in public
services. New York: Russell Sage Foundation.

Liu, E., Johnson, S. M., & Peske, H. G. (2004). New teachers and the Massachusetts
signing bonus: The limits of inducements. Educational Evaluation and Policy
Analysis, 26(3), 217—236.

320

Lortie, D. C. (1975). Schoolteacher: A sociological study. Chicago: University of
Chicago Press.

Ma, L. (1999). Knowing and teaching elementary mathematics: Teacher's understanding
of fimdamental mathematics in China and the United States. Hillsdale, NJ:
Lawrence Erlbaum Associates, Inc.

Malen, B. (1999). On rewards, punishments, and possibilities: Teacher compensation as
an instrument for education reform. Journal of Personnel Evaluation in
Education, 12(4), 387-394.

Martineau, J. A., & Plank, D. N. (2004). Fairness in accountability policy: Is value-
added assessment the answer? East Lansing, MI: Education Policy Center at
Michigan State University.

Maxwell, J. A. (1992). Understanding and Validity in Qualitative Research. Harvard
Educational Review, 62(3), 279-301.

McLaughlin, M. W., & Talbert, J. E. (2001). Professional communities and the work of
high school teaching. Chicago, IL: University of Chicago Press.

Milanowski, A. (1999). Measurement error or meaningful change? The consistency of
school achievement in two school-based performance award programs. Journal of
Personnel Evaluation in Education, 12(4), 343-363.

Milanowski, A. (2004). Relationships among dimension scores of standards-based
teacher evaluation systems, and the stability of evaluation score - student
achievement relationships over time (CPRE-UW Working Paper No. TC-04-11).
Madison, WI: Consortium for Policy Research in Education.

Milanowski, A., Kimball, S. M., & White, B. (2004). The relationship between
standards-based teacher evaluation scores and student achievement: Replication
and extensions at three sites (CPRE-UW Working Paper No. TC-04-10).
Madison, WI: Consortium for Policy Research in Education.

Miles, J. (2004). Getting the sample size right: A brief introduction to power analysis,
from www.ieremymiles.co.uk/misc/power/

Miles, M. B., & Huberman, A. M. (1984). Qualitative Data Analysis: A Sourcebook of
New Methods. Beverly Hills, CA: Sage Publications.

Miles, M. B., & Huberman, A. M. (1994). Qualitative data analysis (Second ed.).
Thousand Oaks, CA: Sage Publications.

321

Milgrom, P., & Roberts, J. (1988). An economic approach to influence activities in
organizations. American Journal of Sociology, 94(Issue Supplement:

Organizations and Institutions: Sociological and Economic Approaches to the
Analysis of Social Structure), S154-Al79.

Milgrom, P., & Roberts, J. (1990). Bargaining costs, influence costs, and the organization

of economic activity. In J. E. Alt & K. A. Shepsle (Eds.), Perspectives on positive
political economy (pp. 57-89). Cambridge: Cambridge University Press.

Milgrom, P., & Roberts, J. (1992). Economics, Organization, and Management.
Englewood Cliffs, NJ: Prentice Hall.

Miller, G. J. (1992). Managerial dilemmas: The political economy of hierarchy.
Cambridge, UK: Cambridge University Press.

Miller, G. J ., & Hammond, T. H. (1994). Why politics is more fundamental than

econorrrics: Incentive-compatible mechanisms are not credible. Journal of
Theoretical Politics, 6(1), 5-26.

Miller, L. M., & Say, E. (1982). Incentive pay for teachers: Impacts in an urban district
(No. ED216438). Washington, DC: ERIC.

Mumane, R. J ., & Cohen, D. K. (1986). Merit pay and the evaluation problem: Why most
merit plans fail and a few survive. Harvard Educational Review, 56(1), 1-17.

Murphy, J. T. (1971). Title I of ESEA: The politics of implementing federal education
reform. Harvard Educational Review, 41(1), 35-63.

Natkin, G. (2002, May). Myth of the revolving-door superintendency. School
Administrator.

Nieto, S. (2003). What keeps teachers going? New York: Teacher's College Press.

North, D. C., & Weingast, B. R. (1989). Constitutions and commitment: The evolution of

institutions governing public choice in seventeenth-century England. Journal of
Economic History, 49(4), 803-832.

Odden, A., & Kelley, C. (1997). Paying teachers for what they know and do: New and
smarter compensation strategies to improve schools. Thousand Oaks, CA:
Corwin Press.

Peterson, P., Rabe, B. G., & Wong, K. (1986). When Federalism Works. Washington,
D.D.: Brookings.

Popper, K. R. (1963). Conjectures and refutations: The growth of scientific knowledge.
London: Routledge & K. Paul.

322

Porter, A. C., Youngs, R, & Odden, A. (2001). Advances in teacher assessments and
their uses. In V. Richardson (Ed.), Handbook of research on teaching (4th ed., pp.
259-297). Washington, DC: American Educational Research Association.

Porwoll, P. J. (1979). Merit pay for teachers. Arlington, VA: Educational Research
Service.

Posner, E. A. (2000). Agency models in law and economics (John M. Olin Law and
Economics Working Paper (2nd Series) No. 92). Chicago, IL: The University of
Chicago Law School.

Prendergast, C. (1999). The provision of incentives in firms. Journal of Economic
Literature, 37, 7-63.

Public school district finance peer search. (2004). Retrieved October 18, 2004, from
http://nces.edgov/edfin/search/search intro.asp

Pugmire, T. (2004). Seagren brings different style but same philosophy. Retrieved
September 15, 2004, from
http://news.minnesota.publicradio.org/features/2004/07/l9 scheckt edcommissio
m

Rapple, B. A. (1992). A Victorian experiment in economic efficiency in education.
Economics of Education Review, 11(4), 301-316.

Real incentives for professionals. (Special Report)(2001). Washington, DC: American
Federation of Teachers.

Reid, K. S. (2003, September 10). Iowa's move toward pay for performance on verge of
collapse. Education Week.

Results of the first NAEP trial urban district assessment in mathematics. (2004).
Retrieved October 18, 2004, from
http:llnces.ed.gov/nationsreportcard/pdf/dst2003/2004458.pdf

Results of the first NAEP trial urban district assessment in reading. (2004). Retrieved
October 18, 2004, from
http://nces.edgov/nationsreportcard/pdf/dst2003/2004459.pdf

Richards, C. E., & Sheu, T. M. (1992). The South Carolina school incentive reward
program: A policy analysis. Economics of Education Review, 11(1), 71-86.

Ritterson, J. (2003). Minneapolis Public Schools Teacher Academy Professional Pay

Plan Report for the Year Ending December 2002. Minneapolis, MN: Minneapolis
Public Schools.

323

Rivkin, S. G., Hanushek, E. A., & Kain, J. F. (2000). Teachers, schools and academic
achievement.Unpublished manuscript, Cambridge, MA.

Rogosa, D. (1999). Accuracy of Individual Scores Expressed in Percentile Ranks:
Classical Test Theory Calculations (No. R305B60002). Washington, DC:
National Center for Research on Evaluation, Standards, and Student Testing.

Salisbury, B. (2004, September 14). Teachers are trying out performance pay system.
Pioneer Press.

Salzer, J. (2003, September 21). Teacher bonuses strain budget Atlanta Journal
Constitution.

Sanders, W. L., & Horn, S. P. (1998). Research findings from the Tennessee value-added
assessment system (TVAAS) database: Implications for educational evaluation
and research. Journal of Personnel Evaluation in Education, 12(3), 247-256.

Sappington, D. E. M. (1991). Incentives in principal-agent relationships. The Journal of
Economic Perspectives, 5(2), 45-66.

Seiler, G. (2001). Reversing the 'standard' direction: Science emerging from the lives of
African American students. Journal of Research in Science Teaching, 38(9),
1000-1014.

Shah, A. (2002, June 18). Minneapolis teachers approve new pay system. Star-Tribune.

Silberman, T. (2003, September 30). Teachers to get ABCs bonuses. Raleigh News and
Observer.

St. John, E. (2003). Revenues and expenditures for public elementary and secondary
education: School year 2000-2001 (Statistics in Brief). Washington, DC: National
Center for Education Statistics.

Stevens, D. E., & Thevaranjan, A. (2003, November 25). Ethics and agency theory:
Incorporating a standard for eflort and an ethically sensitive agent, from
httpz/lssm.com/abstract=34222 1

Tashakkori, A., & Teddlie, C. (1998). Mixed Methodology: Combining Qualitative and
Quantitative Approaches (Vol. 46). Thousand Oaks, CA: Sage.

Teacher quality bulletin. (2004). Retrieved January 23, 2004, from
http://www.nctcLorg/bulletin/v5n3.html

Teacher Union Reform Network. Retrieved November 8, 2004, from
http://www.gseis.ucla.edu/hosted/tum/proposal.html

324

Thomas, W. B., & Moran, K. J. (1992). Reconsidering the power of the superintendent in
the progressive period. American Educational Research Journal, 29(1), 22-50.

Tirole, J. (1996). A theory of collective reputations (with applications to the persistence
of corruption and to firm quality). The Review of Economic Studies, 63(1), 1-22.

Tosi, H. L. J ., & Gomez-Mejia, L. R. (1989). The decoupling of CEO pay and
performance: An agency theory perspective. Administrative Science Quarterly,
34(2), 169-189.

Tschannen-Moran, M., & Hoy, W. K. (2000). A multidisciplinary analysis of the nature,
meaning and measurement of trust: Ohio State University.

Tyack, D., & Cuban, L. (1995). Tinkering toward utopia: Reflections on a century of
public school reform. Cambridge: Harvard University Press.

Walsh, J. (2003, October 2). Pawlenty pitches paying "super teachers" up to $100,000.
Minneapolis Star Tribune.

Webster, W. J ., & Menro, R. L. (1997). The Dallas value-added accountability system. In
J. Millman (Ed.), Grading teachers, grading schools: Is student achievement a
valid evaluation measure? (pp. 81-99). Thousand Oaks, CA: Corwin Press.

Welbourne, T. M., & Gomez-Mejia, L. R. (1995). Gainsharing: A critical review and a
future research agenda. Journal of Management, 21(3), 559-609.

White, B. (2003). Performance-based teacher compensation in Minneapolis (Working
Paper No. CPRE-UW-TC-03-3). Madison, WI: Consortium for Policy Research
in Education.

White, B. (2004). The relationship between teacher evaluation scores and student
achievement: Evidence form Coventry, RI (CPRE-UW Working Paper No. TC-
04-13). Madison, WI: Consortium for Policy Research in Education.

Whyte, W. F. (1955). Money and motivation: An analysis of incentives in industry. New
York: Harper and Row.

Williamson, 0. E. (1985). The economic institutions of capitalism: Firms, markets,
relational contracting. New York: Macmillan.

Wilson, S. M. (1990). A conflict of interests: The case of Mark Black. Educational
Evaluation and Policy Analysis, 12(3), 309-326.

325

Wilson, 8. M., Floden, R. E., & Ferrini-Mundy, J. (2002). Teacher preparation research:
An insider's view from the outside. Journal of Teacher Education, 53(3), 190-
204.

Wragg, C. M., Haynes, G. S., Wragg, E. C., & Chamberlin, R. P. (2004, April 16). Merit
pay for teachers in the UK. Paper presented at the Annual Meeting of the
American Educational Research Association, San Diego, CA.

Wright, B. D., & Linacre, J. M. (1994). Reasonable mean-square fit values. Rasch
Measurement Transactions, 8, 370.

Wright, B. D., & Masters, G. N. (1982). Rating scale analysis: Rasch measurement.
Chicago, IL: MESA Press.

Yee, G., & Cuban, L. (1996). When is tenure long enough? A historical analysis of
superintendent turnover and tenure in urban school districts. Educational
Administration Quarterly, 32(supplemental), 615-641.

Young, 1. P., Rinehart, J. S., & Heneman, H. I. (1993). Effects of job attribute categories,
applicant job experience, and recruiter sex on applicant job attractiveness ratings.
Journal of Personnel Evaluation in Education, 7, 55-66.

Young, 1. P., Rinehart, J. S., & Place, W. A. (1989). Theories for teacher selection:

Objective , subjective, and critical contact. Teaching and Teacher Education,
5(4), 329-336.

326

M

nl\jllijljljiljjifiljll13111111