PLACE ll RETURN BOX to roman this checkout tram your record.
TO AVOID FINES Mum on or before data duo.

DATE DUE DATE DUE DATE DUE

A‘L

      

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU lsAn Afﬁrmative ActionIEqud Opportunity Inﬂation
Wanna-m

THE INNOVATION DECISION DESIGN
IN THE DEVELOPMENT OF
STATE AGING POLICY
By

Jeanette M. Hercik

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Political Science

1997

ABSTRACT

THE INNOVATION DECISION DESIGN
IN THE DEVELOPMENT OF STATE AGING POLICY

By

Jeanette M. Hercik

This study examines the incentives and constraints on state-based innovative decision-
makinginagingpolicy. Utilizingdatagathered froma 1995 national survey ofstatehealthand
human service agencies and the state units on aging regarding plans and efforts underway in
regards to the aging Of the baby boom population in their state, this study explores the
demographic, economic, political and internal organizational inﬂuences on the innovative
decision-making process. In addition to this aggregate data analysis, four case studies of states
deﬁned as ”outliers" in the aggregate analysis are highlighted.

This dissertation builds on the eaﬂier work OfLammers and Klingman (1984) and tests
hypotheses regarding the determinants of innovation advanced by Mohr (1969) and Downs
and Mohr (1976; 1979). Innovation is deﬁned as a process of decision-making which is
consistent with the Downs/Mohr deﬁnition (1976), and focuses on the internal determinants of
innovation involving issues of organizational and leadership capacity within the state and links
the innovation literature soundly with the agenda-setting literature. This study proposes that a
state's ability to plan for and irmovatively respond to the forthcoming demographic challenges
of the 21st century is directly associated with the ability of the state to provide a collaborative
enviromnent among state agencies where current issues and problems are addressed and

strategies for ﬁrture policies are developed.

The ﬁndings indicate that the most consistent signiﬁcant indicator of innovative state
decision-making is the governance structure of state agencies. Simply put, when state agencies
are encouraged to collaborate, there seems to be more innovation. Interagency collaboration is
a signiﬁcant and reliable determinant of innovative decision-making.

Four comparative studies of the states of California, Indiana, South Carolina and
Vermont provide a more in-depth analysis of the level and type of "collaboration" necessary to
make a difference in stimulating policy change. Contrary to the conﬂict-resolution hypotheses
advanced by Baumgartner and Jones (1993), it is in this spirit of collaboration and within this

cooperative environment where most policy change is evident and innovative policies found.

Copyright by
JEANETTE M. HERCIK
l 997

TO

Christie, Katie and Scott,
And in memory of my Father

ACKNOWLEDGMENTS

This journey has been a long one. It's completion would not have been possrble
without the love and support of my family. SO, this is ﬁrst for my daughters, Katie and
Christie, who gave up many a weekend with Mom SO that I could write. They will always be
the greatest accomplishment of my life. I look at them with wonder and awe, for they are SO
very special. I love them desperately. This is also dedicated to my husband, Scott, whose kind
words, cheerﬁrl smiles, and constant encouragement were the only things that Often made me
stay on task. I will never be able to say thank you for his hours and hours of taking care of
“ ' e” SO that I could work on this dissertation. At times, when I didn't think I would ﬁnish, he
was there with gentle words and loving support. I would be lost without him.

Cleo Cherryholrnes was not only the chair of my doctoral committee, but he was my
mentor in many ways, as well as a guardian angel who guided me through the many pitfalls of
this manuscript. He constantly improved on what I thought was my best work He made me
tell a story that I wanted to tell. In spite of his own health concerns, he was a constant source
Of support for me. I wish him the best, and can only say “thank you.”

Also, this is for the other members of my doctoral committee, Jim Granato, Carol
Weissert, and Ric Hula Jim Granato forced me to wrestle with methods. His patience with

my seemingly unending questions and his overall kindness to me did not go unnoticed, nor

unappreciated. Carol Weissert and Ric Hula each brought to this work their unique talents and
provided a richness in perspective which is reﬂected here. Carol Weissert has served as a
model Of the kind of professor I would want to be like, and she is responsrble for peaking my
interest and desire to bridge the distance between the academic and practitioner. Each member
of my committee leﬁ their mark on this dissertation, and I thank them for the time and energy
they dedicated to making my work better. Lastly, this is also for Ada F iniﬂer. Although not a
member of my dissertation committee, She was my guidance chair. I believe she was the ﬁrst
one who taught me what research was all about.

There are so many family members and ﬁiends who gave me encouragement along the
way. Wrthout them, I know I would not have ﬁnished. So this is especially for my Mom who
has taken constant pride in my accomplishments, and has been a source of loving support. To
my sister, Charlotte, who patiently listened as I complained over all the work, over all the
years. And to my brother, Joey, who is always there for me. To Inn and Luan, whose
acceptance ofme and constant loving support of my pursuits made me always feel like a
daughter. And, in memory of Wabash, and for our new little Maggie. They were my constant
companions as I wrote late into the night.

Andﬁnally,tomyDad,whopassedawaylastyearduringmywritingofthis
dissertation. A man with little formal education, he instilled in me the desire to be well
educated and Showed me the importance of education. He was the bravest of hearts and the
kindest of souls. It saddens me to think he will not be here to see me graduate, but I know he

is proud of this accomplishment, and that he’s smiling right now. I do miss him terribly.

vii

TABLE OF CONTENTS

LIST OF TABLES .......................................................................................................... xi
LIST OF FIGURES ....................................................................................................... xiii
PROLOGUE: STATE POLICY AND ITS CENTRALITY TO THE
NATIONAL SOCIAL POLICY DEBATE ...................................... 1
Introduction .................................................................................... 1
Laboratories of Democracy ............................................................ 2
The Devolution Revolution ............................................................. 4
Summary ......................................................................................... 8
CHAPTER I: ESTABLISHING THE IMPORTANCE OF THE AGING
ISSUE ON STATES’ PUBLIC POLICY AGENDAS .................... 12
Introduction .................................................................................... 12
The Changing Social and Economic Context of
Growing Old in America .............................................................. 16
The Demographic Imperative ......................................................... 19
A Global Issue .................................................................... 19
America's Baby Boom ........................................................ 23
Implications for States ........................................................ 38
Summary—Convergence of the Policy and Political Streams-
States at the Center ...................................................................... 46
CHAPTER 2: BUILDING A THEORY OF INNOVATION .................................... 48
Introduction .................................................................................... 48
A Historical Review of the Innovation Research ............................ 50
Toward a Theory of Innovation—A Focus on Process .................... 60
Linking the Literatures—Innovation and Agenda Setting ................ 69
Summary—A Paradigm Shift or a Theory Building Loop? .............. 77

CHAPTER 3: RESEARCH DESIGN AND NIETHODOLOGY:
CREATING A MODEL OF STATE LEVEL INNOVATION
DECISION DESIGN FOR THE DEVELOPNIENT OF
STATE AGING POLICY .................................................................. 82

viii

Introduction .................................................................................... 82
The Research Design ...................................................................... 83
Data Collection ............................................................................... 87
Methodology Employed ................................................................. 90
Creating the Model ......................................................................... 91
The Dependent Variable—A Measure of Innovation .......... 92
The Inﬂuence of Independent Factors ................................ 97
Demographic Factors ............................................. 97
Socioeconomic Factors .......................................... 99

Political Factors ..................................................... 102

Organizational Factors ........................................... 105

Summary ............................................................................ 107

CHAPTER 4: THE FINDINGS: UNDERSTANDING THE
INCENTIVES AND CONSTRAINTS ON
STATE-BASED INNOVATIVE DECISION-MAKING

THROUGH AGGREGATE ANALYSIS ........................................ 109
Introduction .................................................................................... 109
The Hypotheses .............................................................................. 1 12
Testing the Model ........................................................................... 114
Validating the Model ...................................................................... 130
Summary ......................................................................................... 144

CHAPTER 5: A COMPARATIVE REVIEW OF FOUR STATES:

WHAT MAKES A MAVERICK INNOVATOR? ......................... 148
Introduction .................................................................................... 148
Selecting the States ......................................................................... 150
State Proﬁles ................................................................................... 153

California ............................................................................ 153
Indiana ................................................................................ 158
South Carolina .................................................................... 165
Vermont ............................................................................. 169
Summary—What Makes a Maverick Innovator? ............................. 172

CHAPTER 6: CREATING A NEW STRUCTURE FOR INNOVATION:
THE IMPORTANCE OF COLLABORATION TO THE

INNOVATIVE DECISION-MAKIN G PROCESSES ...................... 175
Introduction .................................................................................... 175
Explaining State Variation in Innovation ........................................ 176
The Challenge of Collaboration ...................................................... 170
The Challenge to States:

Change the Culture and Create Innovation ......................... 180

Building a More Complete Theory of Innovation ........................... 183

Conclusion ..................................................................................... 185
LIST OF REFERENCES .................................................................................................. 187
APPENDIX A: List of Agencies Surveyed ..................................................................... 201
APPENDIX B: Survey ..................................................................................................... 202
APPENDIX C: Interview Protocol .................................................................................. 209
APPENDIX D: White House Letter ............................................................................... 212

LIST OF TABLES

Table 1 State by State Analysis of Population Statistics ................................................... 37
Table 2 Correlates: Capacity/Policy Community/Irmovation ............................................ 96
Table 3 Validation Tests ................................................................................................. 114
Table 4 Initial Model ...................................................................................................... 1 15
Table 5 Model #1 ........................................................................................................... 117
Table 6 Model #2 ........................................................................................................... 1 19
Table 7 Model #3 ........................................................................................................... 120
Table 8 Model #4 ........................................................................................................... 123
Table 9 Model #5 ........................................................................................................... 124
Table 10 Model #6 ......................................................................................................... 125
Table 11 Correlation Matrix ........................................................................................... 136

Table 12 Validity Test for Multicollinearity .................................................................... 137

Table 13 Comparison OLS and Robust Regression ....................................................... 143
Table 14 State by State Comparisons of Selected Variables .......................................... 152

LIST OF FIGURES

Figure l—Federal Grants-In-Aid as Percent of Total State-Local Outlays ......................... 5
Figure 2—Age and Race Comparisons between 1995-2050 ............................................. 14
Figure 3-Projections ofElderiy in Selected Countries in the Year 2025 ......................... 22
Figure 4—Total Fertility Rate and Live Birth Rate: U.S. 1920-1990 ................................ 24
Figure 5—U.S. Population Pyramids: 1960-2040 ............................................................. 29
Figure 6—Change in the Composition of the Federal Budget, 1950-2002 ........................ 31
Figure 7-Percentage of Population Over the Age of 65 in 1995 ..................................... 41
Figure 8—Projected Percentage of Population Over the Age of 65 in 2020 ..................... 42

Figure 9—Projected Percentage Change in Population
Over the Age of 65 from 1995-2020 ................................................................... 45

Figure 10—Adoption of an Innovation by an Individual
within a Social System ......................................................................................... 52

Figure Ila—Plot of Relationship between Independent Variable
“pctc9520” and the Innovation Index ................................................................ 130

Figure llb—Plot of Relationship between Independent Variable
“percapin” and the Innovation Index .................................................................. 131

xiii

Figure llc—Plot of Relationship between Independent Variable
“derngov” and the Innovation Index .................................................................. 131

Figure lld—Plot of Relationship between Independent Variable
‘ﬁmiﬁed” and the Innovation Index .................................................................... 132

Figure lle—Plot of Relationship between Independent Variable
“orgstruc” and the Innovation Index .................................................................. 132

Figure llf—Plot of Relationship between Independent Variable

“va120_1” and the Innovation Index .................................................................. 133
Figure 12—Histograrn of Residuals ................................................................................. 133
Figure l3-Quantile Norm of Residuals .......................................................................... 134
Figure 14-Boxplot of Residuals .................................................................................... 134
Figure 15-Interagency Collaboration and the Innovation Index .................................... 151
Figure 16-Indiana Policy Council and Working Group ................................................. 161
Figure 17—The Indiana Collaboration Project ................................................................ 164

xiv

Prologue

STATE POLICY AND ITS CENTRALITY TO

THE NATIONAL SOCIAL POLICY DEBATE
Introduction

The focus ofthis dissertation is state policy, and speciﬁcally, state aging policy. State

policy, as a base of comparative analysis, has long been a focus of political research. Thomas
Dye argued that states provide an ideal opportunity for comparative analysis, given that all
states operate under written constitutions which divide authority among executive, legislative
and judicial branches, and that the structures of state governments are similar from state to
state, making it easier to isolate causal factors in analysis of public policy outcomes (Dye,
1966). Utilizing the “innovative decision-making design” advanced by Downs and Mohr
(1976) as a theoretical framework this dissertation explores the enablers and constraints for the
development of innovative state based aging policy. This dissertation is primarily an aggregate
analysis of data gathered from a written survey, but is enriched with in-depth information
reﬂecting four state experiences.

There are numerous conﬂicting opinions about the value of state studies. Some
political scientists over the years have determined that there is so much variance among states
that it is challenging to eﬂ‘ectively and eﬂiciently conduct multi-state analysis (Jewell, 1982).
Malcom E. Jewel] in his 1982 article criticized political scientists of not paying enough

attention to what was happening on the state leve. We have given too little thought and

2
devoted too little of our research resources to the ﬁeld of state government and politics”
(Jewel, 1982: 638). Fourteen years later at a speech at the 1996 annual policy conference of a
national organization dealing with state issues, Peter Harkness, editor of Governing magazine,
suggested that the same thing about political journalists.
So much power is being devolved to the states, yet the media has yet to catch
up. It continues to focus on Washington, and not on the state capitals where

real domestic policy is being made, and signiﬁcant changes are happening
(Harkness, 1996).

Laboratories of Democracy

Historically, states have ﬁrnctioned as “incubators” for new ideas and as “test runs” of
national policy. In 1932, Supreme Court Justice Louis Brandeis hailed the states as
“laboratories capable of launching novel social and economic experiments.” Although
consistent data gathering for multi-state analysis is difﬁcult, there have been signiﬁcant state
based studies conducted, particularly limited comparative studies on a variety of issues
including political culture, state legislatures, state parties and elections, governors, state
administrators and interest groups (Elazar, 1972; Lowery and Sigelrnan, 1982, Erikson,
McIver and Wright, 1987; Chubb, 1988; Patterson, in Gray, Jacob and Albritton, 1990; Harnm,
1980; Ray, 1986; Squire, 1988; Songer et a1., 1986; Weissert, 1991; Beyle, 1989, 1992;
Sigelman and Dometrius, 1988; Bnrdney and Herbert, 1987; Gormley, 1982; Berry, 1982;
Lammers and Klingrnan, 1984; Berry and Berry, 1990; and Osborne 1988). Especially over
the past decade and half; it seems that state policy has been extremely relevant for explaining
and predicting ﬁxture national policy. “..the record of innovation in the states in the past
suggests that not infrequently the ultimate consequence of state innovativeness is to provide a

basis for subsequent action on the part of the federal government” (Lammers, 1989: 64).

3

David Osborne in his boolg Matories of Democragy: A New Breed of Governor
Cgeates Models for National Growth, advances this notion of policy generation at the state
level which ultimately moves into the national mainstream. He points to the progressive
movement, which originated at the state and local level and grew up in response to the many
problems created by rapid industrialization, urban growth and corrupt urban political machines.
He suggests that many of the progressive reforms introduced at the city or state level were
gradually institutionalized at the federal level as a part of the New Deal. He proposes that
governors and their work in the states are foreshadong national politics and policies. He
advances this notion of a “new political paradigm” involving new assumptions about the proper
roles of federal, state and local governments and predicts that the party that embraces the new
paradigm will win a realigning election and donrinate the following decades (Osborne, 1988:
330)

The 19905 ﬁnd the states ﬁrll-front-and-center at making domestic policy. All states
are being called upon to rise to the challenge of devolution “The devolution-revolution will
shift states’ roles from that of being laboratories—testing and reﬁning, piloting and
experimenting—to that of ultimate deﬁner and provider of virtually all essential public services
to our citizens” (Gross, 1996). Many state policymakers and administrators are wrestling with
some critical social issues for the ﬁrst time. With a new emphasis on block grants and states’
rights, and an apparent continuing devolvement in social domestic policy by the federal
government, it is likely that states will play a pivotal role in developing policy innovations to

address the greatest demographic phenomenon of the next century—the aging of America.

The Devolution Revolution

The division of responsrbility for domestic aﬂ'airs between the national government and
state governments has been an important political theme throughout much of our history. “In
ﬂaming the Constitution, Madison sought a ‘middle ground’ that would provide ‘due
supremacy’ to a national government while leaving the states intact in order that they might be
‘subordinately useﬁrl’” (Derthick, 1987). From Madison to Gingrich—the equilrbrium of
powers between the federal and state governments has been unsteady. Often, the Sharing of
federal power is linked with the sharing of the power of the purse and the allocation of
responsrbility for policies and programs.

In the late nineteenth and early twentieth centuries grants were given by the federal
government to states and localities for agricultural experiment stations, state forestry
promotion, merchant marine schools and highways. Although there was a slow and steady rise
inspeciﬁccashassistancegrantsmadeto statesduringtheearlypart ofthetwentiethcentury,
even after Franklin Roosevelt’s New DeaL federal aid still accounted for less than ten percent
of total state and local spending (Nathan, 1996). Categorical grants grew dramatically during
the 19608 as the Kennedy and Johnson administrations created programs to address poverty,
social inequalities, and pressing urban problems. In the 19705, the growth in federal aid as a
percentage of state-local outlays continued with the initiation of block grants. To give some
perspective on the scope and proliferation of the federal grants-in-aid, in ﬁscal year 1993, 578
federal categorical programs with $182 billion in ﬁrnding provided assistance to states and
localities (GAO, 1995). From the 19603 to the present, federal grants-in-aid have represented

on average about 20 percent of total state and local spending.

 

 

 

a /‘

 

 

 

Percentage of State-Local Outlays
G

 

 

 

 

Vi o V) O in o in c in

in O ‘0 l‘ l‘ 00 so 0 Ox

as O as O as a a o ax

— H F- F! —- .— — on- we
F'scal Year

Figure l—Federal Grants-in-Aid as Percent of Total State-Local Outlays
Source: Advisory Commission on Intergovernmental Relations
President Nixon’s New Federalism program in 1969 developed a rationale for
centralizing some governmental ﬁmctions, while decentralizing others (Bonnett et al., 1995,
Nathan, 1996). Nixon sought to shift power, funds and responsibility from Washington to the
states and cities. Block grants were seen as the best means of enhancing state authority. Nixon
focused on capital and operating types of grants and did not include entitlement programs such
as Medicaid and AFDC in his block grant proposals. Nixon did propose a Family Assistance
Plan (F AP) for national welfare reform and a Farme Health Insurance Plan (FHIP), but both

proposals were never enacted and were lost in the “Watergate” scandal (Nathan, 1996).

6

Block grants again took center stage during the 19803 as part of President Reagan’s
strategy to reduce federal spending and decentralize federal programs by giving states program
oversight responsrbility. In contrast with previous block grant legislation, President Reagan’s
proposals gave states greater ﬂexrbility in managing the programs, but with decreased federal
ﬁnancial support. F iﬂy—seven federal categorical programs were consolidated into nine block
grants with the passage of the Ommbus Reconciliation Act of 1981 (Hayes and Danegger,
1995).

The scale and complexity of many of the proposals for block grants to states before the
104th Congress was much greater than the Reagan precedent (Hayes and Danegger, 1995).
During much of the past two years Congress has been debating proposals that would shift
programmatic responsrhility from the federal government to the states for welfare, Medicaid
and employment and training programs. The magnitude of these programs is huge: federal
payments of $15-17 billion for welfare, $92 billion for Medicaid, and $5-7 billion for
employment and training programs (Stanﬁeld, 1995: 2206). AS a result of the 1990s federal
devolution eﬂ’orts, states are currently playing a more central role in the development of
domestic social policy. During the 1995-1996 National Governors’ Association summer and
winter meetings in Boston and Washington, DC, the nation’s governors played a critical role
in drafting national welfare reform and Medicaid block grant legislation, which subsequently
formed a basis of the negotiations between President Clinton and the Congress (National
Governors’ Association Resolutions, winter meeting, 1996). There seemed to be agreement
on both sides of the aisle, as well as on the federal and state level, that devolvement of power

and purse to state govermnent was a good thing and inevitable.

7

Our next goal must be to dramatically restructure the relationship between the

federal government and the states. . .We can meet national obligations and

pumre our national interest with dramatic devolution of power (President Bill

Clinton, 1996).

This is an exciting time to serve in Congress. And it’s an exciting time to serve

as a governor....The debate today is not whether power should be shifted out

of Washington, it’s how fast we should do it (Senator Bob Dole, 1996).

The recent passage of PLlO4-193, the Personal Responsibility and Work Opportunity
Reconciliation Act of 1996, redeﬁned the relationship between the federal government and
states and shifted the responsibility and costs for poor individuals from the federal government
to state government. The change from a national welfare system rooted in entitlements to a
state-based welfare-to-work initiative with capitated federal ﬁmding puts the states in the
center of domestic policymaking, with power and responsibility which they did not before have.

Although block granting to states is not a new practice, PL103-193 is breaking new
ground and giving governors, in particular, new control and responsrbility for the nation’s
welfare system. The scale of the proposed shift of domestic responsrbilities qualiﬁes it to be
called a “Devolution Revolution” (Nathan, 1995).

In the 20th century the clear momentum has been national. There have been

ﬂurries of state activism but the great movement has been to the center. Now

we are in the midst of a shift to the states that could well involve basic changes

in our governmental system, not just at the edges, but at its core (Nathan,

1996: 7).

Given the overall political and social environment of America in the mid-19903, states
will continue to play a critical and central role in creating domestic social policy as part of the

answer to out-of-control federal expenditures and too much federal involvement in state and

local concerns (Cox, 1995). Because states will have to be so “out ﬂoat” on social policy

8
issues and thus, subject to much media and public scrutiny, the most innovative states will not
only have to wrestle with the crisis of the day, but also need to be prepared for what is coming

intheﬁrture.

Summary

In 1920, only 4.6% of the US. population was over the age of sixty-ﬁve. “Taking care
of the elderly” was viewed primarily as a family and church responsrbility. The numbers of
elderly were fairiy insigniﬁcant, given that life expectancy in the early 1900s was forty-nine.
Historically, if there was any governmental role for “taking care of the elderly,” it was played
on the state and local level, and typically focused on health and housing provisions for indigent
elderiy. Wrth the creation of Social Security as one of the New Deal efforts to address poverty,
a federal role in aging policy emerged. However, in 19605, with the passage of the Older
Americans Act, hberal increases in Social Security beneﬁts, and the creation of Medicare and
Medicaid, the Federal government became the focus of policy development for the elderly, and
moved center stage for “taking care of the elderiy.”

During the 19805, in the wake of federal budget cuts aﬁ‘ecting programs for the elderiy,
many advocates for the elderly turned for assistance to state governments. States responded
haphazardly with a variety of diﬂ‘erent programs, but again somewhat limited their assistance
efforts to the health and housing needs of the elderiy. States established prescription drug
assistance programs, Medicaid and Medicare clearinghouses, construction programs for
building low and moderate income housing for elderly, as well as assisted living facilities, and

provided tax credits or tax breaks for housing for moderate and middle income elderiy.

9

By the year 2020, it is anticipated that 20% of the nation will be over the age of Sixty-
ﬁve. The demographic phenomenon of the 215t century poses signiﬁcant societal and
economic problems beginning in the year 2010 (Chrystal, 1982). This impending demographic
crisis is accompanied by the political realities of “devolution.” It appears likely that states will
be the focal point for developing and implementing aging policy for the next century. In John
Kingdon’s terms, it seems that what we have before state policymakers is an “open policy
window.” “Policy windows are opened either by the appearance of compelling problems or by
happenings in the political stream” (Kingdon, 1984: 204).

Kingdon emphasized that most policy change grows out of the coupling of problems,
policy proposals and politics (Kingdon, 1984: 20). Kingdon was weak in articulating how
“solutions” come about. He presented the concept of “Policy Primeval Soup” involving
numerous players, particularly policy entrepreneurs, interacting in a variety of “policy
communities” (Kingdon, 1984). Kingdon did not Speciﬁcally address the issue of innovation or
how the “new idea” comes about. Michael Hayes’ presentation of a prototype of “concentric
circles of policymaking” with “technical experts at the core of policymaking and moving out to
a wider policy arena” relied heavily on the Kingdon model. However, like Kingdon, he also
did not satisfactorily address the issue of the development of solutions or the issue of
innovation (Hayes, 1992).

The mq'or emphasis of my research is exploring the linkages and connections between
“solutions” as identiﬁed in the agenda-setting literature and “innovation decision design”
deﬁned in the irmovation literature. By linking these literatures, factors are identiﬁed which

inﬂuence the capacity of state policymakers and administrators to develop “solutions” or

10
“policy alternatives” and allow them to take advantage of this “open policy window.” How
that policy change will happen and the process of developing “solutions” or “policy proposals”
in states is the focus ofthis dissertation.
Speciﬁcally, this dissertation will examine the following two questions:
0 What internal determinants within a state—demographic,
socioeconomic and political-are plausibly causal in state planning for
the aging of the baby boom population, and in the subsequent
development of aging policy for the 21 st century?
0 What governance structures end practices within state government are
associated with policy innovation and provide for an irmovative

environment within which to respond to the demographic realities of
the 215t century?

In the next chapters these questions are pursued. In chapter one, the anticipated demographic
challenges of the 21 st century, the aging of the baby boom population and the implications of
these changes for state policy is reviewed. In chapter two, a theory of innovation is built which
is centered on the process of irmovative decision-making. This chapter focuses on the planning
for change and the process that ignites new ideas and allows them to ﬂourish and be
implemented and not simply a discussion around diffusion of interesting ideas. A new
dimension of innovation reﬂecting the role of collaboration and cooperative work environments
is presented. In chapter three an econometric model for state level innovation decision design
for the development of state aging policy is created. Chapter four tests the model through an
aggregate analysis. Chapter ﬁve highlights four state experiences and the role of collaboration
in the development of irmovative decision-making processes in these states. And ﬁnally, the
conclusion focuses on the creation of a new structure for innovation which appropriately takes

into account this new element of collaboration.

11

Given “devolution,” individual state efforts at addressing the demographic challenges
of the next century might well be the basis for subsequent national aging policy and the essence
of innovation for other states’ aging policy efforts. It is timely to reﬂect on our ability as a
nation to plan adequately for the economic, social and political changes caused by these shifting
demographics, and critical that we look at the capacity of state policymakers and administrators
to meet the challenges of these new responsibilities. Gaining insights and perspectives
regarding the determinants of state level irmovative decision-making processes, and the
potential role of collaborative governance structures and practices Should be helpﬁrl to states
because state policymakers and administrators will be primarily responsible for developing the

solutions and policy alternatives to meet the demographic challenges of the 215t century.

Chapter 1

ESTABLISHING THE INIPORTANCE OF THE
AGING ISSUE ON STATES’ PUBLIC POLICY AGENDAS

“....the economic implications of America ’s aging population over the next several
decades will aivarfwry other big issue one might name. ”

Peter G. Peterson,
Atlantic Monthly, May 1996

Introduction

America is aging. Due to advancements in medical technology and healthier lifestyles,
Americans are living longer. The fertility rate has decreased over the last several decades and
plummeted to its lowest point about a decade ago, staying there ever since (Dychtwald and
Flower, 1990). Also, most Signiﬁcantly, “baby boomers”—a ﬁill one-third of our nation-are
aging. All of these factors are profoundly shiﬂing the demographic balance of our society.

We are currently experiencing a lull in the growth of the elderly population because of
the low level of live births irmnediately following the Depression. This lull will continue
throughthe 19905, butgrowthinthenumberofelderlywillbeginincreasing aﬁerthe year
2000 and is projected to continue increasing through the year 2050. The changing
demographics are well documented (De Vita, 1989, 1996; Bouvier and De Vita, 1991;

Kingson, 1992; Hodgldnson, 1992; Baerand Cohen, 1996; AARP Proﬁle, 1995; U. S. Bureau

12

13
Of the Census Special Study, 1993; Atkins et al, 1994; Wakins, 1994; Treas, 1995; and US.
Bureau of the Census, 1994). The fastest growing segment of the population is that over the
age of 80. The average age of the population is increasing and it is estimated by 2010 that
therewillalreadybetwiceasmanyAmericansages 85 andolderastherewerein 1990(De
Vita, 1989). By 2030, more than one in ﬁve Americans will be over the age of sixty-ﬁve and
there will be more people over sixty-ﬁve in the country than under eighteen. Not only will the
numbers of elderly be changing, but they will also be much more diverse. The baby boom
generation is racially and ethnically diverse, and there is signiﬁcant variation within these racial
and ethnic groups. Approximately 9.5 million African-Americans, 6.1 million Hispanics, and
2.8 million of other minority groups are currently part of these baby boom cohorts (Kingson,

1992). Figure 2 shows age and race comparisons between the years 1995 and 2050.

Pueauof

'8°BNWP

100

80 ‘-

40 "

20 "

 

  

.‘II‘III’III’III’IIIIII'I‘I‘IJIZIIIII.
:5xxxxxxxxxxxxx5
..,............................
.AA AAAAAAAAAAAA
..... ........55555 55..........
.... .....A ..........
..............555555.........,.
5555555555555555555555555555555
.IIIuIIII :IIHII IIIIIIIIII‘
.......... ,......,.. ...I...
.......................I...I...
. ........ ....................
...............................
..............I..I..........II.
........... ...................
. ......... .I.......I..I...II
. ......... I... . ............
555555555555555555I555' 5555555
”i"’£115$1$$$$’5$“ 151535,
x 255 55

'5 555555

5555551

nun?

 

 

.IIIII/Iur ”r/IIIIt/II'
.5555555555 5555555555
55555555555 55555555 ‘
55555555555? 5555555555
:555555555555555555555555555555
vuln/ununu/ IIIIIIIIII/I
...................5...........
5.55....55.5...55...5..55..5..5
:555555555555555555555555555555
.....I..5.....................I
...............................
............................5..
5555555555555555555555555555 55
............................555:
...............................
..................I............
.........I................... .
”IIIIICIZIIISI147715571352
............................ .
........... ... .I. .........I
........... .,........I..I..
........... ...........,...
.,,,,.,,,, I......I..........
..... .... ..I .........I.
:5'5555555555'555555555555..
..,.......................
32555555555555555555555555'5
................5.........
............... A.......
.............,.5 ...,.....
.............. .5.........
CIISIII’IISIIII’ IZI’ZiiiiII
..,.......... ... ............
iSIISIIIZSSIIIIZ ”I55?
555555 5555555555.. 55555

ZAIxIA5A55 I45x5x5x5x5
.5.II.I.I.II.IIIIII, IIIIIIIII.
5555.55 .I I..5I55 IIIII..I
III . .IIIIII

 

      

. ....... .............
.II .......5555 A...........
II. ...............5......I
..I .I..........
.I. ...III..IIII
II. .II IIIIIII
III III II.IIII
I . I... .I.....
. .... .....I.
. .III ..IA..
. I... I I...
5 555 5 5555
. ......
. . ....
. I...

5 5555

III...

. I.

I. .

.

 

5............
..55........5.
........
555....I.....
A. A.
555.. . I
......, I
5555555 5

I

:§\\\\~s»\“
\

n

\\“

LIJ
II”
IuII
IHI

I III

IIIIIIIIIIIIIIIIIIII

III$5IIIII IIII III
In

I. . I................
I. ....5......... .......
I......55.................

.I.........I..

........... ..
....5 I
555 5
.I. .
... I I
.... .........
................II
5.................

\s\tt\\\
\»\~t\\\
\\\\\\\\\

I
IIIIII I
IIIIII I
IIIIII I
IIIIII I
III II I
II

tastt

63
II

,\\...,...\...
.t.......\...
it

«‘\\\\\«\~\\\\

IIIIIIIIIIIIIIIIIIIIIIIIIIIII
IIIIIIIIIIIIIIIIIIIIIIIIIIIII
IIIIIIIIIII

2050

 

    
 

I.....I........ .. I
5555555555555555552555555555555
...... ...................
55 ..,....5....................
5........ ..I ..................
...... .. ...................

............... ...... .. .5.
.......................... ....
..I....................... ....
......I. .. . .. ...........
............... .....I....
...... ...... ..........
555555555..""'5 55555555555
II... I.. I.........
I I... ..... ....I..II.
..II.II..II. .II.......
5... ....... ...I.I...I
............ ..........
......... ....I.....
...I..I...I.. ......I.....I.
......I......., ........I.....
I............I..I.............
.I.II.I.... ..................
555555555555555555555555'555555
...............................
....................... .......
.........II...... ......I.5...I
.......I.................I.....
I..............I.....I.II......
I.....I.I.........I......I.....
.III.I................I........
...............................
.I.....I............I..........
I................ ...........I.
5555555555555555555555555555555
5555555555555555555555555555555
............I..................
55555555555555555555555555555. 5
I....I.......................5
... ..........I....I.......I.I:
... I....... ................
55555 5555555555555555555555555
.... .....I.I..... ..,........
............... ....... ...
55555555555555555. 5555555.:5
55555.555555555555..55555555555
5555555 5555555555555555555555
5.... .......................
.I. 5 ...............I...I..
... .I......................
...5555.I....................I.
55555555IIIIIIIIII. I...I..I.II
.........I..II.........
5555........................
555.....I......I...............
.........I...I.I...............
I........I.....II..............
.I.........................I...
............. ......... .......
555555555555 555555555555555:
.......... .............I...I
.....I.......I..............
... ......I......I..........
I... . ...........I .I.....
555 '75555555555...’5555555
5. III... I. .II..... .. . .
I....I.. 5... .I... ........
.5.......... ..........I.......
.... .....I.............. I..55
55.5... .5.5555 555555555555555 5
...............................
5.5.5.........................I
.555. Ht 555555555555555555 55555
.............. ... ..... ....
.............5...5..I.. .I.
........................ II.

1995

Under-go‘s

.Ama'nmrnai-t

 

    

 

55555555555555555555555555555?5
.I..................5..........
.AAA AAAAA. .AAAAA
I...... .......... ...........
..I. I I........I,..........
......I ..................II.
5555.55555555555555555555555555
555555555. :2 5555555555555555
55555555555 ' 5555555555
55555555555 55555555555
........... ...........
‘55555555555 5555555555
55555555555 55555555555
5555555555.55555555555555555555
...............................
AA AAAAAAAAAAAA.
.... .................. .....
........................ ...I.
...................I...........
5555555555555555555555I55555555
......I.............. ......,..
.................III .,,.......
.I.............. .I............
.......................5.......
I..............................
I.............................
.I...I.I....5I........ .I...I..
..........II. II... ......I.
.......,............ ........,.
.AAAAAAAAA..AAAA
................. ........
.... .. ....... ......
...I.5555... ... . ...,.....I..
5.... A..I..I.I.5..5555555.
.....5555...........I..
....I.555..5. I.........5555. 55
III... ............I......5
.....A 5..............5.
I.......5. 5............ ,55;.5.
................. .I.....I..A
................. .,......... .
.........................,.....
...............I...... ........
.......... .... ... ....
5555555555555555555555555555555
I.........I....................
.......... .. I..............
............ .. ............
.I................ .. ....I....
.............. .I. .......I
I.I............................
...............................
.I..I....................I.....
.. .. ....... .... ....I..
1151” ' 42175175575122:
5555.. 5555'55'55'555'5'555'5
....I A I

2050

Figure 2—Age and Race Comparison Between 1995 and 2050
Source: US. Bureau of Census, CurrentPopulation Reports P25-1 104 (1993

15

The cohort of 76 million “baby boomers,” a diverse group of people born between
1946 and 1964, has substantially altered many aspects ofAmerican society. Gerber foods and
diaper services expanded in the late 19405 and early 19503. Then these “baby boomers”
entered the school system. Schools became overcrowded—causing a rush of catch-up
construction and often half-day programs to accommodate the large number of school children.
The social fabric of the country was shaken by the student demonstrations and the counter-
cultures of the sixties and early seventies—as these “boomers” moved through their turbulent
teen and young adult years. The real estate market exploded as “boomers” entered the housing
market to buy their homes, driving prices upward. At the stroke of midnight, January 1, 1996,
the ﬁrst of the “baby boom” population entered its ﬁfth decade. These “baby boomers” are
turning ﬁfty at an average rate of one every eight seconds for the next eighteen years.

As these “boomers” age, they will continue to inﬂuence dramatically the social,
economic and political systems of the country. The expectations and needs of the “baby
boomers” in retirement will be signiﬁcantly different than the generations before them
(Dychtwald and Flower, 1990). The aging population of the future will impose changes in the
types of services demanded by citizens from their federal, state and local governments, and
these delivery systems will have to change the way they are organized and ﬁrnded. Stephen
Chrystal hypothesizes that the problems and difﬁculties of the aging “baby boomers” will create
a crisis situation in American society (Chrystal, 1982).

Some policymakers and researchers suggest that because of the post-depression birth

dearth, we have a window of opportunity in the 19903 to ready the nation for this demographic

l 6
metamorphosis (T orres-Gil, 1995, Dychtwald and Flower, 1990). We now stand poised at the
edge ofthe 21st century, and policymakers must begin to plan for the social and economic
implications of the aging of America The challenge to them is to identify comprehensively the
effects that this large aging population will have on the sociaL political, and economic systems
in the nation, and then begin to plan systematically for these changes and develop a strategies

which will alter negative trends and encourage positive options and alternatives.

The Changing Social and Economic Context of Growing Old in America

From the New Deal through the mid 19703 popular stereotypes of older Americans
were that they were poor, frail, dependent, and above all else, deserving of ﬁnancial assistance.
The federal government responded with a variety of compassionate programs, starting with the
New Deal’s Social Security, the Great Society’s Medicare and the Older Americans Act, and
special tax exemptions and credits for being age 65 or older initiated during Nixon’s New
Federalism (Atkins et al., 1994).

Since the late 1970s, however, the long-standing compassionate stereotypes of older
persons have been undergoing a reversal. Through the 1980s and into the 19903 new
stereotypes have emerged in popular culture depicting older persons as prosperous, hedonistic,
selﬁsh and politically powerful—“greedy geezers” (Atkins et al., 1994). Although there are a
multiplicity of factors that contribute to this reversal of stereotypes, the major factor is
ultimately “money”-theirs (older persons) and ours (taxpayers).

There has been a dramatic improvement in the aggregate economic status of older

people in this country. Social Security and Medicare has helped to reduce the proportion of

1 7
elderly persons in poverty from about 35% in 1965 to 12.2% today. However, this economic
enhancement of living standard of the elderly is reﬂected in the graying of the federal budget.
During the past 15 years, as the proportion of the budget devoted to beneﬁts for older people
became increasingly recognized, programs for the elderly have become important tradeoff
elements in any attempt to deal with American economic and social problems.

The overall size of the federal government budget has been stable over the last several
decades, though there have been signiﬁcant shifts in priorities of spending (Quinn: 1996).
Health, retirement and disability, and interest on the federal debt currently exceed two-thirds of
all federal spending leaving less than one-third for defense and all other expenditures. Persons
aged 65 years and older already account for one-third of the nation’s annual health care
expenditures, or $300 billion of an estimated total of $900 billion in 1993. Per capita
expenditures for Americans age 65 and older are four times as much as for those under the age
of 65 (Atkins et al., 1994). Social insurance expenditures, speciﬁcally Social Security and the
hospital insurance component of Medicare, are responsible for most of the increase in total
public expenditures over the last 40 years. The past rate of growth of these two programs,
coupled with the approaching retirement of the baby boom cohorts, place these programs in
the forefront of policy concern into the next century (Rand Research Brief, 1995).

When members of the baby boom population become aged—into their 70s and 805-
some project that they will be healthier, on average, than that of preceding cohorts. This
general expectation is based on numerous factors that have been unique to the baby boom’s life

course experiences including: better prenatal care, optimum childhood preventive practices

18
such as imrmrnization, better nutrition, more healthful work enviromnents with lower work-
related injury rates, reduced exposure to known carcinogens, better health practices throughout
adult life such as lower rates of smoking, and more participation in exercise programs.
However, by their sheer numbers, the baby boomers will place an enormous drain on the
nation’s health and economic support systems as they age, and most likely as they join the
ranks of the oldest-old—over 85.

There are many challenges that the aging of the baby boom pose to America.
However, the potential enormous increase in health care expenditures appears to be the focus
of the current policy debate. Health delivery systems and structures, public and private
insurance mechanisms, long-term care insurance, and terms of beneﬁts and eligibility appear to
be shifting daily. Although President Clinton’s health care reform package did not pass
Congress, health care reform is occurring all over America Some suggest for the better,
others for the worse (Moon; Washington Post, September 24, 1996). This health care debate
will continue, and a part of this public policy discourse will be the issue of aﬁ’ordability for
Medicaid and Medicare as the boomers age.

However, the aging of the baby boom generation is not solely a health care issue. The
aging of the boomers will have enormous impact on the economic, ﬁnancial and social
constructsofoursociety. Itinﬂuencesallfacetsofourtaxandﬁnance structures, our
transportation systems, our economic development and workforce policies, and even the way

we design and develop communities. The aging of America is an inter-generational issue-

l9
impacting the young, the working adult and the old—causing the existing compacts between

generations to be rewritten.

The Demographic Imperative

A Global Issue:

America does not stand alone in meeting the challenges and understanding the
opportunities of an aging society. “Aging” of society is a global issue. The age structure of a
population is determined primarily by fertility and mortality. Most societies historically have
had high levels of both birth rates and death rates. Whole populations begin to “age” when
fertility falls and mortality rates continue to improve or remain at low levels.

Low birth rates during and following World War I resulted in the global elderly
population to plunge in the early 1980s Similarly, it is anticipated that the low birth rates
duringthedepressionandduringWorldWarIIwillresultinadipinnmnberofelderlyinthe
late 1990s and into the beginning of the twenty-ﬁrst century. The global increase in fertility
following Worid War H and through the 19505 coupled with medical and technological
advancements of the late twentieth century will result in rapid acceleration of the world’s
elderly population starting in the year 2010 (Torrey et al., 1987). The current growth rate of
elderly is 2.4% per year, which is much faster than the global population as a whole. This
growth rate will result in a worldwide population of more than 410 million elderly by the year
2000.

Immigration patterns and policies also affect population structures. Outmigration of

young working-age adults from “poorer” countries to “richer” countries can initially raise the

20
proportion of younger adults in receiving countries as well as the proportion of older persons in
sending countries. The long range effects of such migration depend on a variety of factors,
most important of which is if the immigrant remains in the country of destination Age-
selective international migration appears to have decreased over the last two decades (Torrey,
1987)

The working deﬁnition of elderly has been historically speciﬁed as a chronological age
of 65. Using this as a benchmark, comparisons between countries can easily be made. Sweden
has a total fertility rate well below the natural replacement level and thus, currently has the
highest proportion of elderly among the major countries of the world with 17% of its
population being over the age of 65. However, Japan which currently has the highest life
expectancy rate in the world-78 years—should “age” most rapidly during the next ten years.
Japan has more than doubled its percentage of elderly from 1970 at 7% to 1996 at 14%.
However, throughout much of Europe, the proportion 65 and over will increase modestly
through the year 2005, with only relatively large gains seen in the Soviet Union (from 9 to
13%) and in Germany (ﬁom 14 to 19%). This demographic pattern will be mirrored in the
United States. The proportion of persons 65 years and over will change very little by the year
2005.

However, this slow growth in elderly will change alter the year 2005. The post-World
War 11 baby boom coming into retirement should have societal and economic implications for
most developed nations, and it is predicted that the number of elderly in all developed countries

will expand noticeably. It is anticipated the annual growth rate for persons 65 years and over

21
will reach 4 % by the year 2010 and remain at that level through the ﬁrst halfofthe next
century (Torrey et al., 1987). Japan will see a continued expansion of the proportion of elderly
in their country throughout the ﬁrst quarter of the next century, with a projected 20% of their
population being elderly by the year 2025. This phenomena will be reﬂected in the United
States, which will also experience a rapid growth in the percentage of elderly during the latter
part of the ﬁrst quarter of the next century—from 14% in 2010 to nearly 20% in the year 2025.
However, it is predicted that by the middle of the next century, China will be the “oldest”
country primarily because of its oﬂicial policy of one child per married couple. Demographers
project that this policy will result in 40 percent of China’s total population being 65 years or

older by the middle of the next century.

22

3

$5.5
{99¢
$m.m

 

 

 

 

$02
$6:
$0.2

 

 

 

 

.65 new 38» no
.85 93 E8» we
~96 93 E8» ow

[Ill]

 

 

 

mg— 03on

@533
525.
.5936
, moon
_ 32 .m.D

 

 

Figure 3—Projections of Elderly Population in Selected Countries in the Year 2025

Source: U. S. Bureau of the Census,SpeciaI Population Reports P-95, No.78

23
America’s Baby Boom:

Changes in population size and composition can greatly inﬂuence a nation’s policies
and programs. Some have even argued that “demographics is destiny” (Easterlin, 1968). The
baby boom generation is America’s largest generation in history resulting from an
unprecedented decade-and-a-half long fertility splurge. In 1943, demographers Warren
Thompson and Pascal Whelpton projected that the nation’s population would peak at 161
million in 1985 and then begin to decline. Instead, because of the baby boom and increased
immigration to the United States, the US. population in 1985 was close to 240 million and
growing by approximately 2 million people per year (Bouvier and De Vita, 1991). It has been
suggested that this generation has reshaped US. society in many ways. Its size alone—over 75
million—a ﬁrll one-third of our nation’s population—has required adjustments in our schools,
labor markets, housing markets, consumer markets and government programs.

The baby boom is often referred to as the “post World War H baby boom.” However,
this generation spans 19 years-from 1946 to 1964. The end of World War 11—1945-1947—
wasmetwithincreasedmmﬁageandferﬁlitymtes,butthebirthratefeﬂinﬂre subsequentthree
years. In 1951 these rates rose again and stayed high for another 13 years. Over this 19 year
period, there were 3.8 million births per year in the early years ofthe boom, 4.6 million births
peryearinthepeakyears, and4.3 millionbirthsperyearintheﬁnalyears. Inmarked contrast
to this baby boom generation is the baby bust generation born primarily in the 1970s. There

were 7 million fewer births during the 19703 than during the 1950s.

24

urn-roar Fertil'nykde

 

 

 

 

 

 

 

 

r ofliﬂh (million) (W number of children par woman)
5 - 5
13m _ 4
r- 3
rm
P 2
l r "1 1 I l
1920 1930 1940 1950 1960 1910 1980 1990

Figure 4—Total Fertility Rate and Live Birth Rate: Us. 1920-1990
Source: U. S. Bureau ofthe Census

Tlﬁsmcreasemferﬁlﬁyrate,althoughrdatedtotheendofWWH,redlymﬂeaeda
positive mood of the country and an upbeat public opinion about the ﬁxture of the world that
lasted from the end of WWII to the assassination of President Kennedy. Landon Jones offers
the “Procreation Ethic” as explanation of this 1950s and early 1960s phenomenon (Jones,
1980). He suggests tlmt the rrrilitary victory of WWII and the subsequent economic prosperity
of the country renewed people’s faith in the future and encouraged early marriages and a boom
inbirths. PaulLightalsoemphasizestheimportanceofthe subsequentculturalchangesin
society following WWII, and the “social conformity” of the “ideal” family with the male as

breadwinner and the female as full-time homemaker, which resulted in the conventional

25
family having two-to-four children. (Light, 1988). All of these factors contnbuted to the
creation of the baby boom generation.

Much of the interest in the baby boom generation and its effects on society stems from
the fact that it is a large generation—numerically—sandwiched between two substantially smaller
generations. Also, the baby boom generation is often referred to as a single entity because
there are things about this generation which distinguishes it ﬁom previous generations. The
baby boom generation is the most highly educated generation in American history. Baby
boomers have also contributed to redeﬁning the “traditional American family” in many ways.
Baby boomers have typically delayed entry into marriage and they have been more likely to
dissolve a marriage than previous generations. In addition to delaying marriage, baby boomers
have postponed having children. Nearly 30% of all births in 1988 were to women age 30 and
older (Bouvier and Devita, 1991). Baby boomers also weakened the tie between wedlock and
childbearing.

The US. labor force added 2 million new workers per year between 1968 and 1980
(Bouvier and De Vita, 1988). Besides aﬂ’ecting the size of the labor force, the baby boom
generation also affected its composition. In 1995, 75% of baby boom women were in the labor
forcewiththevast majority ofthese women in ﬁrll time career positions. In 1990, there were
93 million households in the United States, 30 million more than in 1970, representing a 50%
increase in 20 years (Allen, 1993; Bouvier and De Vita, 1988). The sheer numbers of the baby
boom generation accounts for much of the increase, however, it also reﬂects the changing

family patterns and lifestyles of this generation.

26

Although baby boomers share similar life experiences, there are also signiﬁcant and
important differences among baby boomers. Paul Light identiﬁes several of these cleavages, of
which the largest one being that of the early versus late boomers. Baby boomers are typically
broken down into two age groups or birth cohorts—those born between 1946 and 1954 and
between 1955-1964. It has been suggested that the ﬁrst cohort of baby boomers are better oﬂ’
thanthe second wave, inthatthe early baby boomers entered the laborforce during aperiod of
strong economic performance receiving the “lion’s share” of economic and social beneﬁts
(Light, 1988: 77).

There are numerous other diﬂ’erences among the baby boomers, including gender, race,
amount of education, level of income, marital status, employment status, geographic region and
the role played in the Vretnam War (Light, 1988). The baby boom generation is racially and
ethnically diverse including approximately 9.5 million Aﬁican-Americans, 6.1 million Hispanics,
and 2.8 million persons of other minority groups. Approximately 18 million baby boomers are
members ofwhat has been called “minorities at risk”-that is, groups who, by virtue ofrace
and/or ethnic status, experience barriers that signiﬁcantly restrict their opportunities for social
and economic well-being (Kingson, 1992). Some baby boomers are very well off and many
live comfortable middle-class lives. However, there are 7.8 million baby boomers oﬂicially
deﬁned as poor or near poor (Kingson, 1992).

This generation continues to inﬂuence future generations in many ways. It is
important to understand the social, economic and demographic dimensions that created the

' baby boom, however it is also very important to understand the ﬁrrther impact it will have on

27

the ﬁrture of American society. We are now witnessing the baby boom echo or baby boomlet.
Althoughtheydonotmatchthepeakyears ofthepost-WWIIbabyboom, thecurrentbirth
rates are edging upward according to the Census Bureau. Part of the explanation for this
increase could be that the baby boom women who postponed their childbearing are now having
children Also, younger women, who fear infertility problems with delayed childbearing are
having babies. Even though the economic situation of most families require two incomes,
given the greater availability of childcare options and increased employment ﬂexrbility
regarding family-ﬁiendly policies, more women are not inhrbited from having at least two
children.

These children are creating crisis-level crowding in schools, similarly to the way their
parents did two-to-three decades earlier. From Washington State to Florida, school systems
are tying to accommodate the large increase in school enrollments by converting gymnasiums
and cafeterias into classrooms and increasing class size from 30 students to 50 students
(Russakoﬂ’; Washington Post, September 14, 1996). It is projected that they will continue to
overstuﬁ‘ elementary and secondary schools and colleges for the next two decades.

There are mrmerous societal and economic implications associated with the aging of
the baby boom generation. Dealing with the realities of the baby boom echo is just one of
them If the population projections are reasonably accurate, policymakers and planners should
be concerned with the demographic shifts underway in our society, in which the ratio of elderly

to younger persons has permanently changed. The ratio of elderly persons to those of working

28
age will nearly double ﬁom 1990 to 2050.1 Although the dependency ratio has undergone
some criticism over the last few years as being too crude a measure, it is an important
demographic trend that cannot be ignored.

In 1900, one in 25 Americans was elderly. This proportion had risen to 1 in 8 by the
year 1990, and is projected to be 1 in 5 by the year 2025. The demographic importance of the
baby boom can be best illustrated through a series of population pyramids tracing the age
structure of the United States ﬁom 1960 to 2040. The baby boom cohort stands out in all the
pyramids. In 1960 it forms the base of the pyramid, by 1990 it extends across the middle
section, and by 2040 it represents the protruding bands of people age 75 and older.

In the past, declines in the number of births have been the most important contributor
to the long-term aging trend. However, the improved chance of survival to the oldest ages, is
now the most important factor in the growth of the very old population. The oldest old are a
small but rapidly growing group. In 1900, 122,000 people were 85 years or older. Their
mrmbers reached 3 million in 1990. By 2030, it is projected that there will be 8.6 million
people over the age of 85 in the United States. This is prior to the entrance of the baby

boomers into the ranks of the oldest old.

 

1' The societal support ratio (SR) or what is more commonly referred to as the

dependency ratio is computed as the number of youth under age 20 and the elderly over
age 65 per one hundred person aged 20-64. It is criticized as not being a relevant measure
of economic and societal well being because it uses age as its only criteria and ignores the
fact that there are many economically independent older persons, economically dependent
unemployed adults, and that the costs of young people, other than public education, are
primarily borne by families.

29

85+
80-84
75-79
70-74
65-69
60-64
55-59
50-54
45-49
40-44
35-39
30-34
25-29
20-24
15-19
10-14

5-9

0-4

 

4 2 0 2 4 6 2

Percent of Population

2020

85+ Male

80-84
75-79

0

 

 

2

2040

 

70-74

4

Percent of Powlation

Female

 

65-69 [1

 

60-64

 

 

55-59

 

 

50-54
45-49
40-44

 

35-39

 

30-34

Depension Cohort.

 

25-29

' ' born 1930-39

 

20-24

 

BabyboomCohotL
b0ml946-64

 

 

 

15-19

 

10-14

 

Baby bust Cohort,
born 197079

 

5-9

 

g BabyboomEcho

 

0-4

 

 

 

 

 

Colon born 1985-95

 

 

 

I
0 2 4 2

4

4 2

Percart of Population

0

r
2

Figure 5--U.S. Population pyramids, 1960-2040

Source: US. Bureau of the Census

1
4 6 8

Percart of Powlation

30

As the individual members of the baby boom age, so does the nation as a whole. The
median age of the population is gradually increasing. As baby boomers become senior
boomers, the number of older people in the population will double, and by 203 0, there will be
more people over the age of 65 than under the age of 18. By the year 2030, there will be 65
million people age 65 and older, as compared to 30 million senior citizens today. In 2030, the
oldest boomer will be 84 and the youngest turning 65. In 2030, the elderly will be comprised
of a larger proportion of minorities, as well as a larger share of the oldest old. The aging of the
baby boom implies more than a simple increase in the size of the older population. There are
overall implications for our society and economy as our overall society ages (Foster and
Brizius, 1993).

Changes in age composition can have dramatic political, economic, and social effects
on a society. We need not have to wait for the boomers to enter the rank and ﬁle ofthe
“senior” generation to begin to observe and experience the implications of the aging of baby
boomers. By the year 2000, baby boomers will account for more than halfof all workers, and
this will raise the median age of the work force to 39 from 36 today. A declining number of
younger workers in the population may create a new demand to retain, recruit and retrain older
workers. Age-based policies in the workplace which ignore individual differences, may well
need to be adjusted. As boomers become “older workers” they might well reshape work
retirement policies and deﬁnitions of age discrimination.

Given the desire of Congress and President Clinton to balance the budget by the year

2002, there are signiﬁcant political pressures to change the ﬁmding of the Social Security and

31
Medicare systems today. A larger and larger share of the federal budget has been comprised of
support for the nation’s retirement and health systems. In 1950, only 10% of the federal
budget outlays were dedicated to health and retirement spending, as compared to the projected
budget for 2002, which includes 50% of the budget dedicated to health and retirement
spending (Steurele and Merrnin, 1996). This spending increase does not reﬂect signiﬁcant
growth in tax revenue, but a “peace dividend” as funding was shifted from military spending to
health and retirement spending. This shift in federal spending also reﬂects a societal value to

spend our nation’s wealth on the elderly versus other domestic programs, i.e. education.

 

 

 

 

 

 

100% All Other
90% -
Military
80% -
70% -« Net Interest
3 60% " OASDI
8
m
g 50% -
'0
0
Lu
‘5'
40% 4
ﬁ
“-4
O
23, 30% . Other Retirement
g and Drsabrhty
G4
20% - Medicare
10% _‘ Other Health
\/
0% I l I I
1950 1960 I970 1980 1990 2000

Figure 6—Change in the Composition of the Federal Budget, 1950-2002
Source: C. Eugene Steuerle and Gordon Merrnin, The Urban Institute. Calculations based on
data ﬁ'om the Presidents’ budget proposal in the Budget of the United States Government
Fiscal Year 1997, OMB (1996)

32

When Social Security was designed, life expectancy in the country was 68 years of age-
tlrree years beyond eligrbility. Today life expectancy is 75 years of age. Men turning 65 in
2040 can expect to live another 17.6 years, and women turning 65 in 2040 can expect to live
another 22 years (Steuerle and Merrnin, 1996). Increased longevity should be good news, but
it appears that society has not ﬁgured out how to “manage its’ miracles” (Fahey, 1996).
Health care costs have grown rapidly and have become increasingly expensive. Medicare is not
adequately ﬁnanced. Currently, most elders are not protected against the risk of long-term
care. Theluxruyoflonglifemeansthatthereisarealpotentialthatthebabyboomerswillbe
consumers of services long after they stop being producers of economic and societal goods.

Most people see 65 as the retirement age, but nearly 60% of men and 70% of women
collect Social Security beneﬁts beginning at age 62. In 1950, 87% of the men age 55—64
worked. Through the 1970s and early 1980s the average retirement age in the country was
falling. Althoughtheretirementage ratebeganto stabilizeinthe late 19808, andinfact
recently rise, only 68% of 55-64 year-old men currently participate in the workforce.
Currently, only 12% of persons over the age of 65 worlg even on a part-time basis. It has been
suggested that the “early out” policies in private companies and government, and the structure
of the Social Security system encourage early retirement. Wrth the potential need for older
workers,andtheinabilitytoﬁnancetheexistingretirement systemstherewillbelikely
adjustments in the nation’s tax structure and systems to induce baby boom workers to stay in

the workforce longer.

33

It is anticipated that the aging of the baby boom generation, particularly when they
begin to hit the age of retirement in 2010, will place unbearable pressure on the existing social
security and health care systems. These pressures will only increase with the age of the
boomers, particularly after 2030, as large numbers of baby boomers begin to swell the ranks of
the very old. Financial security in retirement is primarily based on three components: social
security, pensions, and savings/assets. This is often referred to as the three-legged stool of
retirement planning. However, to continue this analogy, the stool is deﬁnitely tilted. Today,
approximately 40% of all retirement income comes ﬁ'om Social Security payments, and a ﬁll]
60% of today’s 65+ population relies of Social Security for 80% of their income (Salisbury,
1994; Easterlin, 1990).

The viability and solvency of the Social Security system is one of the most persistent
public policy questions debated today. Social Security saw real increases in beneﬁts in the late
1960s, reﬂecting the public desire to lift the elderly out of poverty. However, there was a
dehberate change in Social Security policy with the passage of the Social Security Reform Act
of 1983, which required that 50% of the Social Security beneﬁts be subject to taxation for
individuals with income over $25,000 and for couples with income over $32,000. This policy
was extended in the Omnibus Budget Reconciliation Act of 1993, which subjected 85% of
beneﬁts to taxation if individuals nrade over $34,000 and couples nrade over $44,000.

Even with these adjustments, the current Social Security and Medicare systems are not
sustainable. With the anticipated number of retirees and the proportion of retirees to workers

rising, there is concern that society will not be able to meet ﬁiture Social Security obligations.

34
It is anticipated that ﬁrture changes to Social Security will be a decrease in beneﬁts and a delay
in eligibility age. These changes have already been enacted for the latter cohort of baby
boomers, in that they are not eligible for ﬁrll Social Security beneﬁts until the age of 67, with
signiﬁcant penalties levied for early withdrawals at age 62. Additional changes being debated
include “means testing” Medicare; “means testing” Social Security; sliding scale Social
Security; moving eligibility age to 72; and many other options to contain public investment in
our health and retirement systems.

Pensions, the second leg of the stool, are also undergoing changes in structure.
Approximately 50% of the current retirees have pension income. For the most part, these
pensions are deﬁned beneﬁt plans and not deﬁned contribution plans.2 The number of private
pension plans has more than doubled in the past 15 years from 340,000 to 870,000, however
the nrajority of these plans have been deﬁned contribution plans (Salisbury, 1994). Three out
of every ﬁve baby boomers in 1988 worked for an employer who sponsored a pension plan,
and nearly halfofthese individuals reported that they were fully vested in their plan (Salisbury,
1994). However, private pension plans were more common among men, and persons who
worked for large companies. Although employment tenure rates seem to be stable, with the

19903 even higher than the 19703 decade, employers do not appear to be offering the security

 

2 Deﬁned beneﬁt pension plan means that there will be a deﬁned beneﬁt at
retirement usually reﬂecting the salary of the employee at termination of employment or
retirement. Deﬁned contribution means that the employee opts to participate in a
retirement plan which he! she ﬁnancially participates in as well as the employer, and that
the amount of ﬁrnding available at retirement reﬂects the amount the employer/employee
contributed. The deﬁned contribution pension plan shiﬁs much of the risk of investment of
retirement ﬁrnds from the employer to the employee.

35
of paternalistic beneﬁt packages as in the past (Salisbury, 1994). These practices will inﬂuence
ﬂreabilityofthebabyboomerstobeﬁnanciaﬂysearreintheirretiremem years.

How well oﬂ’ baby boomers will be during their retirement years will depend on their
sources of income, their lifestyles, and their health The baby boomers are likely to experience
astandard oflivinginold agethat isatleastequaltothatofcurrent retirees. It is less clearthat
theywillbeabletomaintainastandardoflivinginretirernerrtcomparabletothatoftheir
working years. Those with additional sources, such as private pensions, savings, assets’
income, and income from employment will probably be substantially better oﬂ‘.

Savings is a critical piece to the ﬁrture ability of the boomers being able to escape
poverty in old age. To date, baby boomers are known more for their spending than their
savings. Savings rates peaked in 1973 at a rate of 9.4% of an individual’s after tax income,
bottoming out at 2.9% in 1987. By 1990 the savings rate had risen to 4.6% and is staying at
about the 5% mark. It has been argued that baby boomers are saving at one-third the amount
that they should be doing—saving too little, too late (Lavery, 1996). Some propose that the
lack of saving practices among boomers is overstated because the boomers have invested so
heavilyinhousing,andiftheytapirnotheirhousingwealﬂttheywilldowellinreﬁrement
(Salisbury, 1994). Home ownership has signiﬁcant implications for baby boomers in
retirement. Homeequityhasalwaysbeenthesinglemostirnportarrt sourceofbothindividual
and household wealth in the United States. However, some have suggested that as the

boomersretireinforce,theywillliquidatetheirassetsdrivingtherealestaterrrarketintoa

36
downward plunge (Quinn, 1996). This “asset meltdown” will cause signiﬁcant ﬁnancial
hardship for boomers in retirement, particularly the youngest cohort of boomers.

Given changes in pension structure, lack of adequate savings and depletion of “housing
wealth,” as with today’s retirees, Social Security seems likely to be the most important source
of income for most baby boomers, especially at the lower income levels. Given the
heterogeneity within the baby boom generation, the anticipated changes in the federal
retirement and health care systems for older Americans pose risk of poverty or near-poverty for
many baby boomers in their later years. Particularly at risk are non-homeowners, less educated
boomers, the single and the youngest baby boomers.

Family caregiving will accelerate in importance and signiﬁcance in the ﬁxture. More
and more women of the baby-boom generation will face the demands and trade-oﬂ’s of caring
for an older relative in the decades ahead. Currently, 44% of adult daughters or daughters-in-
law who care for an impaired parent are employed, and another 12% report that they quit their
job to provide the care (Stone et al., 1987: 622). Also, the current work-family policies and
support services may need profound adjustments as the baby boom generation enters old age
and needs caregiving assistance. The baby boom generation has had a reduced number of
children and an increased incidence of divorce and rernarriage. These factors might well create
a sense of ambiguity over family roles and responsibilities for caring for the vulnerable or nwdy
parent.

The aging ofthe baby boom generation and of society as a whole will raise important

questions concerning generational equity and fairness. Some have suggested that there will be

3 7
a struggle over scarce resources and inevitably generational conﬂict and tradeoﬂ‘s will need to
occur. However, others challenge the baby boom generation to seize the opportunity to play a
pivotal role in this equity debate, and create policies which provide for their children, assist
their aging parents, and allow them to adequately plan for their own ﬁrtures.

The baby boomers’ entrance into and movement through their retirement years will
span a halfa centruy. The policy debate on the national level has already begun. Clearly, there
is a desire to avoid having the aging of baby boomers strain the retirement, health, and other
social institutions to the breaking point. The emphasis of the debate is that of ﬁnancial
contaimnent. This implies Social Security and Medicare cost containment through federal tax
restructuring, pension regulation (or deregulation) and ﬁnancial incentives to keep the boomers
working longer. It also suggests that states will play a large role in the development of these
policies, and also possrbly be most at risk of supporting “at-risk baby boomers” as the federal
govemment continues to devolve authority and responsrbility for domestic programs to the
states. Understanding and appreciating the implications of the demographic metamorphosis
underway in this county for each speciﬁc state is critical. Thoughtful and deliberate
construction of policies that recognize the broad societal and economic implications of the
“coming of retirement age” of the baby boomers, and the diversity within this generation, is
needed, if the boomers are to have the necessary ﬁnancial support for retirement, and be able to

continue to be active producers of societal and economic goods.

38

Implications for States:
These international and national demographics play out in a variety of diﬂ‘erent ways

when doing a state-by—state analysis. The implications of the aging of the baby boomers vary
signiﬁcantly by state and region. The shitting demographic balance of the country is not
replicated for each individual state.

Over the decade of the 19803 the largest percent increases in elderly population were
mostly in the West, particularly the Mountain States, and in the South, especially the South
Atlantic States of Florida, South Carolina, and Delaware. These areas also reﬂected the
highest growth in the oldest old. The percent change in the elderly population from 1980 to
1990 ranged from a low of 9% in Nebraska, to a high of 94% in Nevada The most populous
states tend to also have the most elderly. In 1995, nine states had more than one million
elderly: California, Florida, Illinois, Nﬁchigan, New Jersey, New Yorlg Ohio, Pennsylvania,
and Texas.

Some states age because of in-migration of the elderly, some because of out-migration
of the young, and some because of sustained low fertility or a combination of these
factors. The states with the greatest proportion of elderly are generally diﬂ’erent ﬁ'om those
withthegreatestnumber. While Califomiahasbyfarthelargestnumber ofpersons aged 65
and over, only 10.2% of its population is elderly, and it ranks 46th among the states. For the
most part, the farm belt states have a higher proportion of elderly than for the total United
States, primarily because of out-migration of the young. Although Iowa has only 438,000
people over the age of 65, it is the state with the fourth highest proportion of elderly in the with

nation at 15.3%. Florida, on the other hand, has the highest proportion of elderly in the nation,

39
almost 19% of its population over the age of 65, and ranks second for having the highest
mrmber of elderly, with over 2.7 million seniors.

While in 1995, Florida is the only state with more than 16% of its population aged 65
and over, by 2020 a projected 32 states will fall into this category. (See Figure 7 and 8). In
2020, it is forecasted that about one out of every ﬁve persons in the United States will be
elderly, compared to about one out of eight persons today. In 2020, it is projected that one-

fourth of the population of Florida will be elderly.

Table l--State by State Analysis of Population Statistics
Source: US. Bureau of Census

- % change in 65+ .
65+ 1995 65+ 2020
m from 1995 to 2020 m

 

40

Table 1 (cont’d)

 

41

 

 

.82 8 ans -
4.4.2 8 sad -

s3. 2 so... -
+3 092580.—

 

 

     
 

Figure 7-Percentage of Population Over the Age of 65 in 1995
Source: US. Bureau ofthe Census

42

 

 

scan a see. I
at: 8 sad I

s2. axed -
+8 nuance».—

  

Figure 8—Projected Percentage of Population Over the Age of 65 in 2020
Source: US. Bureau of the Census

43

Irmnigration plays a signiﬁcant role in explaining increased fertility rates, particularly on
thestatelevel, andpotentiallycanirrﬂuencetheagingdynamicsinastate. Fertilityofminority
populations usually tends to be higher than that of the majority, and therefore the increases in
the proportion (or share) of minorities in the population may result in an overall increase in
fertility. This results in a phenomenon known as “shitting shares” (Bouvier and De Vita,
1991). In example, in California, the fertility rate rose from 1.9 in 1982 to 2.5 in 1989, of
which almost 40% of the gain is attributable to the expanding proportion of minorities in the
state population (Bouvier and De Vrta, 1991). Although on a national level, the “shilling
shares” phenomenon does not necessarily inﬂuence the demographic balance in society, on a
state basis, speciﬁcally states like California, Texas and Florida, which experience signiﬁcant
levels of new immigrants, it can undoubtedly impact their demographic balance.

Census Bureau projections indicate that the West and South will increase their elderly
population by 99% and 81% respectively. Over halfof the United States’ elderly will likely live
in just ten states in 2020, California, Florida, New York, Pennsylvania, Texas, Illinois, Ohio,
Michigan, New Jersey and North Carolina. The elderly population is projected to double in
eight states ﬁ‘om 1990-2020. Seven of these states—Alaska, Arizona, California, Colorado,
Nevada, Utah and Washington-are in the West (Georgia is the only exception). However, the
percent of the oldest old population will be highest in the nridwestern states. The ﬁve states
with the highest proportion of persons aged 85 years and over of their total population are all
farm states, Iowa, North Dakota, South Dakota, Nebraska and Kansas. In 1990, only Iowa

had more than 2% of its population over 85. By 2020, 34 states fall into this category.

44

In 2020, the states with the highest proportion of elderly will be East of the Mississippi
River, and the states that will experience the most rapid growth in the number of elderly will be
primarily in the West. (See Figure 9). It is arguable that the Northeastern, Midwestem and
South Atlantic states already have an established inﬁastructure for older citizens, since these
states currently have a signiﬁcantly large proportion of elderly. Given the 15-25 year time span
available to adjust these systems, these states might be able to slowly assimilate their policies
for the aging of the baby boomers. The Western states, which will be experiencing rapid
growth in elderly—85%-105% growth—will be especially challenged by these rapid changes.
Since they do not currently have a large proportion of elderly, these states most likely do not
presently have adequate systems in place. However, all states will challenged by these new
demographic realities and either have to substantially shift their existing approaches to
addressingtheneeds oftheir oldercitizens, ortheywillneed to designnewapproachestomeet
the challenges and opportunities of an aging society.

Overall international and national demographic changes, the health of the nation’s
economy, and the capacity of economic, political, and social institutions to adapt to the aging
of the baby boomers will substantially determine the circumstances under which individual
states will have to wrestle with the issue. However, since devolution appears to be reality,
state government will most likely be at the center of developing domestic social policies to

meet the challenges and opportunities of the let century.

45

 

 

«$8. 2 $3.” I
:3» 9 $2.“ I
.53 2 sham I
seen 23%.: a

cassava—520m

 

 

     
 

Figure 9—Projected Percentage Change in Population from 1995-2020

Source: US. Bureau of the Census

46

Summary-Convergence of the Policy and Political Streams—States at the Center

Demographics can be a powerful inﬂuence in shaping the life changes of individuals
and societies, but they do not operate in a vacuum. Economic, technological, cultural and
political changes affect our daily lives. They interact with demographic trends, and inﬂuence
the well-being of individuals, and society at large.

The aging of America will present one of the toughest public policy challenges ever
faced by American society. “The US. society has not yet agreed on the public values that
should drive the public policies in an aging society” (Comman and Kingson, 1995: 25). Also
not taken place, is an informed public and political dialogue regarding the broader context of
agingandtheimplications ofanagingsociety. Noristheresocietalagreernentonthe
individual, family and community responsrbilities across life course (Comman and Kingson,
1995:27)

Devolution puts states at core of the challenge of designing and developing ﬁrture
aging policy in this country. The public policy agenda developed to meet the challenges and
opportunities of an aging America should not take a “business as usual” approach toward
ﬁnding solutions (Zedlewski, et al., 1990). It should examine and seriously consider irmovative
proposals that take into accormt the changing characteristics of the elderly, changes in family
constructs, changes in societal and cultural norms, political changes, and the impact of the
forever changing global economy.

States need to begin to understand the projected demographic changes, and the

implications of these changes for their state. State policymakers and administrators need to

47
sort through how these demographic shifts might well impact the state’s tax and ﬁnance
structure, its tax base for K-12 education, its existing health care capacity, its available
workforce, its transportation systems, its community development policies, and its econonric
development strategies. Although “demographics are not destiny,” the projected demographic
changes are likely.

The old saying, “timing is everything,” is particularly relevant as state leaders and
policymakers face the challenges of the 2 1 st century. There are approximately 15 years before
baby boomers begin the mass exodus into retirement. Innovative states will take advantage of
the “time factor,” and ﬁrlly engage in long-range planning in the mid-to-late 19903, and they
will plan to begin phased-in implementation of broad policy changes beginning with the new

Millennium.

Chapter 2

BUILDING A THEORY OF INNOVATION

“A paradigm sets the stwrdards for legitimate work within the science
it governs. ”

A.F. Chalmers
What Is This Thing Called Science? 1982

Introduction

What nukes a good theory? Chalmers suggests that theories must be seen as
organized, open-ended structures involving concepts with precise meanings, and contain within
them prescriptions as to how they should be developed and extended (Chalmers, 1982: 79).
Eckstein (in Greenstein and Polsby, 1975) discusses the two polar positions on what
constitutes theory in political science. He outlines the constructs of “formal theory,” which are
modeled after contemporary theoretical physics involving elements of formal and elaborate
deduction. He also discusses the “soft line” of theory which is simply regarded as any mental
construct which orders phenomena or inquiry. Eckstein argues that regardless of the
theoretical construct, the goals of theoretical inquiry remain consistent regarding the need for
regularity, reliability, validity, foreknowledge and parsimony (Eckstein, in Greenstein and
Polsby, 197 5). And thus, theories can be more or less “good” depending on the “nrlefulness of

regularity statements, the amount of reliability and validity they possess, the amount and kinds

48

49
of foreknowledge they provide, and how parsimonious they are” (Eckstein, in Greenstein and
Polsby, 1975:90).

The process of theory building begins with the development of good questions for
which answers are wanted, as well as a strong hypothesis regarding the “candidate-solution” to
the question (Eckstein, in Greenstein and Polsby, 1975). Testing of hypotheses is viewed by
some as the end of the theory building process, but in reality can feed back into the process,
and allow for further theory building to take place. This cumulative progress of theory building
is characteristic of the inductivist accounts of science, in that scientiﬁc knowledge is growing
continuously as more numerous and more various observations are made, enabling new
concepts to be formed, and old ones to be reﬁned.

This dissertation draws upon established theories within the public policymaking,
agenda setting, and policy imovation literatures. Some of these theories are well formulated
and tested, others are less so. Regardless, it has been argued that good theory is validated
when the discourse is persuasive (Chenyholrnes, 1988). Given the broad questions that I am
pursuing—factors associated with and plausrbly causal in state-based planning and innovative
policy development—the theoretical framework I require is found in part, in each of these
literatures. My theoretical ﬁarnework is grounded in the state-based policy innovation research
of Rogers, Walker, Downs, Mohr, Gray, Thompson, Frendreis, Klingman and Lammers,

Polsby, Glick and Berry and Berry. It relies heavily on the Kingdon/Hayes and March and

Olson model of agenda-setting; and builds off of the work done by Dye and Lindblom in

 

50
policymaking theory. State policy innovation theory and agenda-setting theory are directly
linked.
A Historical Review of the Innovation Research

The study of “diffusion of ideas” is long and has its roots in agricultural research with
work done in the 19303 regarding hybrid corn seeds (Ryan and Gross, 1943). However, it was
within social science, in particular the work done by Everett Rogers in 1962, in which much of
the theory on diffusion of innovations was developed. Rogers reviewed 506 different studies
that were completed ﬁom the late 19303 through the 19503 in a variety of ﬁelds from
sociology, education, medicine and agriculture. He wanted to identify the common threads
running through all of the various research traditions on diﬂ’usion of innovations.

In this work, Rogers clearly deﬁned innovation and diffusion. Rogers argued that a
deﬁnition of innovation needed to be universally relevant and mutually exclusive if there was
any hope of approaching a theory of innovation. He outlined ﬁve characteristics of
innovations: (1) relative advantage or the degree to which the innovation is superior to the idea
it supersedes; (2) compatibility or the degree to which an irmovation is consistent with existing
values and past experiences of the adopters; (3) complexity or the degree to which an
innovation is relatively diﬂicult to understand and use; (4) divisrbility or the degree to which an
innovation may be tried on a limited basis; and (5) comrnunicability or the degree to which the

results of an innovation may be diﬂirsed to others (Rogers, 1962: 124).

51

He identiﬁed four elements crucial to an analysis of the diffusion of innovation: (1)
“the irmovation” was deﬁned as an idea perceived as new by an individual; (2) “its
communication”, how the idea spreads; (3) “the social system” or population of concerned
individuals in a collective-problem solving mode; and (4) “the adoption” of innovation which
was deﬁned as a ﬁve-step decision-making process involving the stages of awareness, interest,
evaluation, trial use and adoption. Rogers deﬁned diﬂirsion as the process by which an
innovation spreads and “irmovativeness” as the degree to which an individual adopts an
innovation earlier relative to other members of his social system. Rogers suggested that the
adoption of an innovation is a process of decision-malang and that... “decisionmaking is the
process by which an evaluation of the meaning and consequences of alternative lines of
conduct are made” (Rogers, 1962: 77-78). (Figure 10 outlines the Rogers’ paradigm for
innovation diﬂusion.)

Rogers also introduced the notion of the role of the change agent and the consequences
of innovation. He deﬁned a change agent as a professional who attempts to inﬂuence adoption
decisions in a direction that he feels is desirable (Rogers, 1962: 254). He suggested that most
change agents are local-level bureaucrats whose purpose is to inject a cosmopolitan inﬂuence
to innovate into a client social system (Rogers, 1962:255). The change agent serves as a
communication link between a professional system and his client system. (Rogers, 1962:283)

Although innovation theory has its roots in Rogers research, it was Jack Walker’s

seminal work in 1969 on state policy diffusion that structured much of the debate and

52

discussion in the literature around policy innovation diﬂirsion among the states. Walker’s

“diﬂirsion” of ideas and not “invention.”

deﬁnition of innovation is conﬁned to

183.8: Base—co

i gang—cone

11 gas. :3

'83.? 3380'

mPADmmm

gauged

Easing”. .n
£5.35 .4
€29.50 .n

Eﬁiaso .u
omega->34. Pawn—um A

mmmoomm ZOCLOQ<

E E =

2 germ— ; gut—'3‘.
I

A8055; .338 .533

EB 2:0 835650 .n
Btu—02:

‘5 5588.50 uiseeom .N
38355:

and 5:8: :83? 38m .—

 

wmmOOMm

 

/ :3. 188:8
an E5. .88: .n
B=_a> .N
bare $.36 ._

gun—9143‘

thQOOm—E

Figure 10 — Adoption of an Innovation by an Individual Within a Social System

Source: Diffusion of Innovations (1962)

53

“An innovation will be deﬁned simply as a program or policy which is new to the states
adopting it, no matter how old the program may be or how many other states may have
adopted it” (Walker, 1969: 881). Walker investigated two major questions. (1) Why do some
states act as pioneers by adopting new programs more readily than others? And, (2) How do
new forms of service or regulation spread among the American states? Walker constructed an
innovation score for the American states based upon elapsed time between the ﬁrst state
adoption of a program and its later adoption by other states. Walker monitored eighty-eight
different programs adopted by twenty or more states, and he averaged each state’s score on
each program adoption to produce an index of innovation of each state. The larger the
innovation score, the faster the state responded to new ideas or policies. Walker explored
relationships between the irmovation scores of the 50 states and socioeconomic, political, and
regional variables. He found that innovation was more readily accepted in the urban,
industrialized, wealthy states (Walker, 1969: 884). He also found that state decision makers,
although seemingly inﬂuenced by states in their region, seemed to be adopting a broader,
national focus based on new lines of communication extending beyond regional boundaries
(Walker, 19692896).

In a subsequent study of policy innovation, Vuginia Gray criticized Walker’s ﬁndings
and argued that no general tendency toward “innovativeness” really existed—states that are
innovative in one policy area are not necessarily the same states that are innovative in other

areas. Gray deﬁned innovation as “an idea perceived as new by an individual; the perception

54
takes place after invention of the idea and prior to the decision to adopt or reject the new idea”
(Gray, 1973:1174). She limited her discussion of innovation to speciﬁc laws adopted by states.
She examined the adoption of twelve speciﬁc innovations in civil rights, welfare, and education
(including the adoption of state public accommodations, fair housing and fair employment
laws, and merit systems and compulsory school attendance). States that were innovative in
education were not necessarily innovative in civil rights or welfare. Diffusion patterns differed
by issue area and by degree of federal involvement (Gray, 1973: 1185). However, similar to
Walker’s analysis, she also discovered that “ﬁrst adopters” of most innovations tended to be
wealthier states, leading one to conclude that ﬁnancial capacity of a state inﬂuenced the ability

of the state to be innovative (Gray, 1973: 1182).

Robert Eyestone explained the instability of ﬁndings in the Walker and Gray studies as
caused by the complexity of the “policy content” of the “ideas” being diﬂirsed. He suggested
that “policy content” inﬂuences the diﬂirsion of ideas from one state to another. He also
proposed that there are several distinct diﬂirsion models, and that the model operational in a
state will be dependent upon the “policy content” of the idea (Eyestone, 1977: 447). “States
may want to review the experiences of other states before taking the plunge themselves,
depending on the strength of the incentives put forward by federal agencies, and perhaps on the
level of residual state resistance to federal innovations” (Eyestone, 1977 : 447). He also
suggested that the wealthier, industrial states might be ﬁrst to “innovate”, not because of

availableﬁrrancesbutbecausetheyarealsotheﬁrsttosuﬂ’ertheundesirablesideeﬂ’ectsof

55
urban and industrial growth which create demands for state policy responses (Eyestone, 1977 :
446-447).

Savage (1978) suggested that the differences in ﬁndings from Walker and Gray
reﬂected a “sampling problem.” He proposed that the data bases used by Walker and Gray
were not sufﬁciently large and representative enough to create a statistic or index to measure
policy innovativeness or to discount such a measure (Savage, 1978: 213). Using a total of 181
policy measures from diverse areas such as agriculture, business regulation, conservation,
crime, education electoral regulation, governmental structure and operation, taxation,
transportation and welfare, Savage created a measure of innovation based on the relative speed
of adoption of the given policy innovation. He found that there seemed to be relative stability
across time regarding the rate of diﬂirsion in states. He disagreed with Gray’s criticism of
Walker’s ﬁndings, citing her “too hasty in discounting a general irmovativeness trait as a
variable characteristic of the American states” (Savage, 1978: 218). However, he found that
nationalizing forces played a strong role in inﬂuencing the speed of policy adoption across state
lines inthe latterpart ofthe twentieth century.

Eyestone pointed out that a state adopts or rejects a policy due to a complex web of
factors, of which a federal incentive is one. Savage’s ﬁndings regarding the nationalizing forces
of the twentieth century seemed to conﬁrm the role of the federal government in state policy
innovation. In 1980, Susan Welch and Kay Thompson speciﬁcally explored the impacts of

federal incentives on state policy innovation and the diﬂirsion rates of public policies

56

throughout the American states. They deﬁned “innovation” as Walker did—”a program or
policythatisnewtothestateadoptingit, nomatterhow old the program maybeorhow many
states have adopted it” (Walker, 1969). They felt that Walker’s analysis did not include much
consideration of the federal government and its potential for inﬂuencing irmovation on the state
level (Welch and Thompson, 1980: 716). They argued that there were two ways in which the
federal government places pressure on the states to conform: (1) the federal government gives
states monies to implement or improve a program—the carrot; or (2) it threatens to deprive
states of existing ﬁrnds if certain requirements are not met—the stick (Welch and Thompson,
1980:719)

Welch and Thompson found that the initial rate of adoption was most inﬂuenced by
whether the policy had positive incentives, while adoption by the laggard states was more
inﬂuenced by whether there is a federal incentive of any sort (Welch and Thompson, 1980:
724). They also conﬁrmed Walker’s conclusions regarding the linkage between diffusion rates
and the ability of state policymakers to communicate across state lines, sharing information and
ideas. In the Welch/Thompson research, they looked at policies enacted pre and post New
Deal. The myriad of national organizations of policy specialists and governmental oﬁcials did
not exist in the early 20th Century. There was less communication among states pre-New
Deal, and thus, less opportunity for them to exchange policy ideas. They attributed the faster
diﬂ‘usion rate of post New Deal policies to increased state networking (Welch and Thompson,

1980:723)

57

Canon and Baum in 1981 expanded on the work of state level diﬂirsion studies ﬁom
dealing solely with legislative or administrative innovations to judicial doctrines. They studied
the diﬂirsion of 23 innovative tort doctrines among state court systems between 1876 and
1975. They examined the correlates of innovativeness and the pattern of diﬂ’usion. Basing
much of their efforts on Walker’s research and ﬁndings, they tested the relationships between
population, urbanization, wealth, industrialization, political culture, party atliliatiorr, and
prevailing ideology with judicial innovativeness. Contrary to Walker’s ﬁndings, Canon and
Baum found very weak evidence that regionalism played a role in diﬂirsion of judicial
innovations. They detennined that developing social and technological avenues of information
exchange overrides geographical barriers, and they suggested that this would be even more
true in the future (Canon and Baum, 1981: 985). Also, they found that neither urbanization,
wealth, industrialization, political party or ideology played a signiﬁcant role in determining
judicial irmovations. The most signiﬁcant predictor of judicial innovation was population.
CanonandBaumattributedtheirﬁndingstothefactthatthecomtsystemisreactive,inthe
sense that the courts must depend on litigants to provide opportunities for innovation. Their
ﬁndings suggest that the diﬂirsion ofjudicial doctrines is a diﬂ'erent process from the diffusion
of legislation.

Lammers and Klingman in 1984 explored the determinants of state based aging policy
innovation Essentially, they were asking many of the same questions that Canon and Baum

did in their work in the judicial ﬁeld. They wanted to gain a better understanding of what

58
prompted innovative state responses in the past, so to predict ﬁrture state actions and the
appropriateness of different state strategies. Lammers and Klingman conducted a ﬁfty-state
aggregate analysis examining variations in state based aging policies over a twenty year period
(1955-1975).

Primarily grounding their work in Walker’s research, Lammers and Klingman created
an “index for irmovation” utilizing single indicators and composite factors measuring aging
policy eﬂ’ort. These dependent variables reﬂected four issue areas/categories: (1) the state’s
efforts at income maintenance, (2) the state’s social services programs, (3) the state’s health
and long-temr care delivery systems, and (4) the state’s efforts at regulatory protection for the
elderly. These dependent variables were then analyzed with the use of socioeconomic and
political variables as potential predictors of diﬂ’erent state responses.

Lammers and Klingman considered a plethora of independent factors including aging
advocacy in the state, general policy lrberalism of the state,3 ﬁscal capacity of the state, political
capacity within the state, political openness, socioeconomic development, and status of the
aging population within the state. Based on this regression analysis, Larmners and Klingman
classiﬁed the states into a four quadrant matrix: strong achieving states, underachieving states,
low achieving states, and maverick innovators. In addition to this aggregate analysis, Lammers
and Klingman conducted a comparative case study involving eight states to assess more fully

some of the relationships which emerged in the aggregate-level analysis. The selection of case

 

3 Policy liberalism deﬁned in accordance to Walker (1969) and the work of Klingman
and Lammers (1984).

59
study states was based on the quadrant matrix and was undertaken to provide maximum
variation on three basic dimensions: (1) percent of aging population; (2) level of policy effort
for the older population; and (3) general liberalism in overall policy responses (Lammers and
Klingman, 1984). The states selected included Minnesota, Iowa, California, North Carolina,
Maine, Florida, Ohio and Washington.

In general, they found that policy liberalism and ﬁscal capacity were positively related
to states’ policy innovations for their aging populations. The states with a more “active”
government, were more likely to seek additional sources of revenue to fund projects/programs.
They deﬁned “active” government as primarily a state with “strong institutions” and “political
openness” (Lammers and Klingman, 1984: 14). Innovation in aging policy on the state level
was dependent on the state’s political capacity—that is the level of development of its
“Policymaking Institutions,” primarily the governor and the legislature. Thus, states with a
power'ﬁrl and successful governor, and a professional legislature were more likely to make
substantial policy eﬂ’orts in a variety of areas, and to seek the sources of revenue to underwrite
those eﬂ’orts. Through the aggregate analysis and comparative case studies, Lammers and
Klingman conducted a comprehensive review of the development of state aging policy over
this twenty year period. They evaluated the role of demographic, socioeconomic and political
variables in determining variation in state aging policy. However, they did not address the issue
of organizational capacity to innovate or the role of governance structure in creating capacity

to implement an innovation.

60
Toward a Theory of Innovation—A Focus on Process

The root word of “innovation,” a noun, is “innovate,” a verb, and as deﬁned by the
OED, innovate means “to change a thing into something new; to alter; to renew” (girth—rd
Engh'sh Dictiom, p. 997). For the most part in the literature, “innovation,” 8 “product” has
been studied and not the “process” of change. The research of Walker, 1969; Gray, 1973;
Eyestone, 1977; Bingharn, 1977; Savage, 1978; Welch/Thompson, 1980; and Canon/Baum,
1981 centered on the study of diﬂ’usion of innovations, with “innovations” being a ﬁrnction of a
product—either a law, a policy, an administrative rule or a tort doctrine. Although Lammers’
and Klingman’s “aging policy eﬂ’ort” included several different indicators and arguably
provided a broader and deeper measure of “policy,” they still were dealing primarily with
passage of legislation or implementation of a certain policy or program.

Lawrence B. Mohr sought an explanation for the varied ﬁndings in the innovation
research. He argued that the research had not yielded a theory to permit scholars to predict the
extent to which a given organization will adopt a given innovation (Mohr, 1969; Downs/Mohr,
1976, 1979). In his 1969 article, Mohr attempted to shift the focus ﬁom a discussion around
the diffusion of a speciﬁc innovation or type of policy innovation to the process of innovation
within public organizations. He wanted to identify the determinants of innovation in public
agencies. He brought back into the debate many of the earlier concepts of the “innovation

process” or “adoption decision-making process” highlighted in the founding work of Rogers.

61

Using a 1965 survey of 93 public health agencies in the states of Illinois, Michigan,
New York, and Ohio and the province of Ontario, Mohr explored his hypotheses regarding the
process of innovation in organizations. He suggested that innovation was a ﬁrnction of the
interaction among the motivation to innovate, the strength of obstacles against innovation, and
the availability of resources for overcoming such obstacles (Mohr, 1969: 111).

The variable emerging by far as the most powerﬁil predictor of innovation was size of
the organization However, Mohr concluded that “size” was a theoretically incomplete ﬁnding.
Size of the organimtion should predict innovativeness only insofar as it implies (1) the
presence of motivation to innovate; (2) constraints on obstacles to innovation; and (3)
provision of adequate resources for innovation (Mohr, 1969: 126). Mohr proposed that
innovation theory needed to be further developed, but that his ﬁndings suggested that the
theoretical construct of innovation should be viewed as a multiplicative function of the
motivation to innovate and the balance between the obstacles and resources bearing upon
innovation (Mohr, 1969: 126).

In a 1976 article, Mohr with his co-author George Downs continued to explore the
complexity of issues regarding innovation in organizations and conﬁrmed to attempt to address
the instability in ﬁndings of the previous research efforts. Downs and Mohr suggested that
there was not a single, unitary theory of innovation, but rather diﬂ’erent theories to explain
different aspects of innovation (Downs/Mohr, 1976: 713). Relying heavily on the early work

of Rogers, they intimated that innovations and organizations had primary and secondary

62
attributes which require innovations to be interpreted differently by organizations and
individuals”. an innovation might be seen as minor or routine by some organizations but as
major or radical by others” (Downs and Mohr, 1976: 704).

They proposed that an innovation does not exist as a separate unit of analysis, nor do
organizations have consistent and constant properties. They recommended that irmovations be
viewed within the context of the organizational unit. “The unit of analysis is no longer the
organization but the organization with respect to a particular innovation, no longer the
irmovation, but the innovation with respect to a particular organization” (Downs and Mohr,
1976: 706). This “innovation decision-design” should focus the attention of research on the
shifting incentives and constraints that are relevant to the decision to innovate in complex
organizations and away ﬁom single policy initiatives or passage of certain pieces of legislation.

In another joint article in 1979, Downs and Mohr built upon the concepts of innovation
formulated in their 1976 article, and stressed the importance of focusing on the process of
decision-making. They argued that for innovation theory to advance, it needed to be centered
on the “innovation decision” rather than on the innovations themselves. They moved away
from Walker’s and Gray’s use of “innovation” and “innovativeness” as a product-a law or a
program-and considered it a process-the decision to do something new. In building a theory
arormd the “irmovation decision,” they considered organizational capacity and leadership

factors, as well as the contextual issues surrounding a particular decision to innovate.

63

In 1983, Frendreis utilized many of Downs’ and Mohr’s suggestions regarding
innovation research and explored the process of innovation adoption in an environment of
linrited resources in 45 American cities. The basic model Frendreis tested was that decisions
by cities to innovate are due to organizational characteristics of the municipality, attributes of
the innovations themselves, and a combination or interaction of these two classes of variables
(Frendreis, 1983: 111). He followed Downs/Mohr’s suggestion to focus on the “decision to
innovate” versus the “innovation” itself; and narrowed his study to what he determined to be a
“similar policy issue.”

F rendreis deﬁned innovation as: “a delrberate, novel, and speciﬁc change which is
thought to be more eﬂ’ective in accomplishing the goals of city government” (Frendreis, 1983:
112). He examined the “movement towards” or the process of adoption of two innovations by
city government—program budgeting and zero-based budgeting. He proposed that these two
innovations were similar and that it was reasonable to expect that they should show similar
patterns of adoption. However, he found that “none of the best ﬁve predictors of adoption
movement for program budgeting show similar power for zero-based budgeting” (F rendreis,
1983: 118).

Frendreis concluded that his “research efforts reveal a disappointing tendency toward
idiosyncratic results,” and in essence the search for a general theory of innovation was “a
ﬁuitless enterprise” (Frendreis, 1983: 109, 120). Unfortunately, it is clear that Frendreis was

not very familiar with budget development theory. Program budgeting allows ﬂexrbility on the

64
part of those administering the budget, whereas zero-based budgeting requires the
administrator to justify his/her budget (Brizius, 1994; erdavsky, 1975, 1992). It can be
understood why “zero-based budgeting” was unpopular with ﬁnance oﬂicers and city budget
staffs. It could easily be argued that these “innovations” are counter to one another, and thus it
is reasonable to suggest that they would not follow the same innovation pattern.

Although my inferences from Frendreis’s study differ greatly ﬁom the author, his
research focused on the dynamic decision-making processes within organizations. His main
determinants of “movement towards innovation” included issues of public saliency,
demographic needs, organizational knowledge, political leadership and consensus that the issue
was a problem or something which needed to be addressed (Frendreis, 1983).

Glick and Hayes in 1991 focused their study of state policy irmovation on this
“continuous, dynamic policymaking process.” Glick and Hayes criticized much of the previous
innovation research as being too narrowly focused on the decision to adopt a speciﬁc policy or
law and not enough on the “extent of policy innovation” or what they called the “policy
reinvention” (Glick and Hayes, 1991: 836-837). They examined the extent to which 38 states
adopted and implemented living will laws between the years of 1976 and 1988. They stressed
the need for political scientists to examine more than the chronology of the adoption of a
presumed uniform policy.

Glick and Hayes discovered that policy innovation was an evolutionary process. Early

innovation and reinvention through later adoption and amendment are important and distinctive

65
parts of the contimring innovation process. They found that though the earliest adopting state
provided important policy leadership, the later adopting states were often more innovative,
because they learned from the earlier adopting states about what worked and what didn’t.
Glick and Hayes suggested that this “reinvention process” was an important part of the
innovation process (Glick and Hayes, 1991: 848).

In 1990, Berry and Berry in response to the varied ﬁndings from the innovation
research and the suggestion from Mohr (1969) regarding an interactive model or theoretical
construct for innovation, determined that signiﬁcant work needed to be done on the
methodology used for innovation research. Berry and Berry suggested that the two types of
explanations of state government innovation: (1) internal deterrrrinarrts; and (2) diffusion
models—are not inconsistent, but that they must be considered in a uniﬁed model. They deﬁned
internal determinants as the political, economic and social characteristics of a state; and
diﬂ’usion as the inﬂuence of neighboring states in prompting a state to adopt a certain policy or
law. Berry and Beny proposed that it is the interaction of these factors which can best predict
state innovation

Using Event History Analysis (EHA), through a probit model, Berry and Berry
explored the probability that a state will adopt a lottery in a given year based on a variety of
independent factors: the ﬁscal health of state government in the previous year, per capita
income from the previous year, the proportion of a state’s population adhering to

ﬁrndarnentalist religion; the degree to which there is uniﬁed party in control of the executive

66

and legislative units in a state; a dummy variable indicates if it is an election year", and the
number of states sharing a border with the state which have already adopted a lottery. They
found that neighboring states are found to have a stronger impact on the likelihood of a lottery
adoption when the internal characteristics of a state are themselves favorable for innovation
(Berry and Berry, 1990: 411). Berry and Berry conﬁrmed the essential elements of Mohr’s
theory that the probability of state innovation is directly related to the motivation to innovate,
inversely related to the strength of obstacles to innovation, and directly related to the
availability of resources for overcoming these obstacles.

In 1992, Berry and Beny repeated this methodology using “state taxes” as a data base,
and had similar ﬁndings regarding the inﬂuences ofboth internal determinants and regional
diﬂ’usion (Berry and Berry, 1992). They suggested that Event History Analysis is a suﬁciently
promising research mode to use in the development of innovation theory. They proposed that
scholars can subject theories of state government innovation to a powerful test by assessing
whetherthesetheoriescanpredict the probabilitythataparticulartypeofstatewilladopta
particular policy in a particular year.

In a 1994 article, Frances Stokes Beny, building on her previous research, reviewed
the three dominant explanations of policy innovation—internal determinants, regional diﬂirsion
and national interaction models. She argued that the single-explanation methodologies used
with each of these models-cross-sectional analysis, factor analysis, and time series analysis,

respectively—did not recognize the inﬂuence of other models, and thus could not be a complete

67
explanation for innovation (Berry, 1994: 443). Using simulated data, she tested the “single-
explanation mode ” and concluded that the conventional single-explanation methodology in
state innovation literature is inadequate and oﬁen produces wrong results (Berry, 1994: 452).

Berry and Berry in their research do not rely on (nor do they cite) the later work of
Mohr, in which he and Downs call for a greater focus on the organization and its role in the
innovative decision-making process. The Beny and Berry model does not include a variable
reﬂecting organizational capacity, or the role of the policy expert/entrepreneur/change agent.
Their uniﬁed explanation methodology of innovation may appropriately account for diﬂ’erences
in previous innovation studies, however, it reverts the dialogue back to a focus on product-a
passage of a law—and not on the decision-making process.

Francis Beny in her 1994 article somewhat addresses this concern with her suggestion
that the Event History Analysis methodology must be modiﬁed if it is to be appropriate for use
in state innovation research. Given that policy innovation is somewhat a “continuous process”
the “discrete time method” inherent in EHA must be adjusted to more appropriately reﬂect the
contimrum of change associated with innovation (Berry, 1994: 453). Berry recommended that
any future model for state policy innovation must allow for the simultaneous speciﬁcation of
inﬂuences by both internal state characteristics and the behavior of other states.

In 1994, David C. Nice, in Poligg Innovation in State Govemm_em, examined the
diﬂ’erent state characteristics that impact adoption of various state policies. Nice utilized a

cross-sectional design and a quantitative 50-state comparative approach. Nrce used a

68

comparative method to examine the causal processes underlying the adoption of eight distinct
policy innovations. These issues included teacher competency testing, ratiﬁcation of the
Balanced Budget Amendment to the United States Constitution, sunset laws, public ﬁnancing
of election campaigns, rail passenger service, property tax relief, deregulation of intimate
behavior and state ownership of freight railroads. Nice looked to the adoption of a policy
innovation as a function of the problem environment, slack economic resources, and a state’s
general orientation toward innovativeness. Nice found that a state’s problem environment
created a signiﬁcant stimulus for policy innovation, but discovered little support for the role of
slack economic resources for any of his eight policies. Also, he could not make many broad
generalizations about the factors inﬂuencing innovation, and found that there were many
different causal processes that underlie the adoption of diﬂ‘erent policies.

A limitation to the Nice study was the dichotomous nature of the policy adoption
variable. He simply reverted to an “adopt” or “not adopt” measure and did not heed the
recommendations of Glick and Hayes to recognize the “extent of adoption” of the legislation.

Nice also failed to integrate other factors into his analysis, including the role of pressure
groups, neighboring states, and policy entrepreneurs in stimulating policy innovation.

Gray suggested that process studies and variance studies could learn from one another
(Gray, in Dodd and Jilison, 1994). She agreed with Mohr’s criticism that innovation research
was too focused on variance studies, and proposed that the agenda formation literature,

speciﬁcally the “policy process model” could provide a basis for broadening innovation theory.

69
Gray suggested that more national diﬂirsion was occurring because of emerging policy
networks, active professional associations and federal government incentives. She linked the
irmovation diﬂirsion literature with the agenda formation literature and called for the inclusion
of the concepts of policy windows, policy communities and policy entrepreneurs as critical
components when explaining the process of innovation. Once irmovation is centered on the

process of decision-making, this linkage to agenda-setting can be quickly made.

Linking the Literatures—Innovation and Agenda-Setting

Downs and Mohr cited the complexity of public policymaking as undermining the
ability of the sub-ﬁeld of innovation research to be truly scientiﬁc (Downs/Mohr, 1979: 379).
They argued that “social scientists have allowed innovation to take on too many different
meanings and have allowed its meaning to be ambiguous in too many signiﬁcant respects”
(Downs/Mohr, 1979: 385). Downs and Mohr suggested that incompatible deﬁnitions and
inconsistent ﬁndings have resulted in research which is not cumulative, and that this level of
instability has hindered the development of a core theory of innovation The work of Berry and
Berry regarding models and methods might well speak to these early concerns which were
expressed in the literature. However, Downs and Mohr also argued that innovation theory can
be advanced only if the focus remains on the decision-making process.

In building a theory around the “innovation decision,” Downs and Mohr suggested that
issues such as organizational capacity, leadership factors, and contextual issues surrounding a

particular decision to innovate must be considered. Many of these factors are mirrored in the

7O

policymaking and agenda-setting literature as determinants of policy initiation or agenda-
setting. The agenda-building or agenda-setting literature points out the dynamic ﬂuidity of the
decision-making process involving people, problems, solutions and opportunities (Cohen,
March and Olson, 1972; Kingdon, 1984; Elder and Cobb, 1984; Hayes, 1992; and
Baumgartner and Jones, 1993). By narrowing the focus of innovation theory to the
Downs/Mohr “innovation decision design,” these sub-ﬁelds can easily be crossed and
innovation theory linked with agenda-setting theory.

Nelson Polsby purposely crossed the literatures in his 1984 Political Innovation in
A_mg_ric_a. Although Polsby’s review of innovations reﬂected national policy changes,4 he
questioned the basics of how public policies come to be. He highlighted eight case studies of
new policies in recent American political history. Polsby used the terms “policy initiation” and
“innovation” interchangeably. He argued that the American political system is too complex
and too diﬂicult to settle deﬁnitely on the exact point in time at which any particular innovation
emerges from the “primordial ooze” (Polsby, 1984: 13). He also suggested that those who are
more curious about the shape of the real world must be willing to accept some necessary

deﬁnitional compromises (Polsby, 1984).

 

‘ The innovations reviewed in this publication include: (1) the creation of the Civilian
Control of Atomic Energy; (2) the Creation of the National Science Foundation; (3) the
Nuclear Test Ban Treaty; (4) the Truman Doctrine: Aid to Greece and Turkey; (5) the
Formation of the Peace Corps; (6) the Creation of the Council of Economic Advisors; (7)
the Creation of National Health Insurance for the Aged; and (8) the Local Participation in,
Community Action Programs.

71

Polsby proposed three criteria to delineate a policy innovation or initiation: (1) large
scale and visrble; (2) a purposeﬁrl break with preceding habit; and (3) lasting consequences
(Polsby, 1984: 13). Polsby cited limitations in the study of policy innovation as being too
narrowly focused on the “event” of the initiation, which artiﬁcially restricts an understanding of
the process of change. Also, he suggested that such narrow focus did not satisfactorily address
the issues of incentives and constraints on successﬁrl strategies of innovation implementation

In his case studies, Polsby modeled his review of policy innovation or initiation alter the
Cohen, March and Olson’s “garbage can model”5 (1972) of agenda-setting and policy

development. He addressed issues of policy alternatives and the process of policy elimination.

He highlighted the importance of policy entrepreneurs and organizational culture to spur
innovation. He described political innovation as a combination of two processes.

The ﬁrst, the process of invention, causes policy options to come into
existence. This is the domain of interest groups and their interests, or persons
who specialize in acquiring and deploying knowledge about policies and their
intellectual convictions, of persons who are aware of contextually applicable
experiences of foreign nations, and of policy entrepreneurs, whose careers and
ambitions are focused on the employment of their expertise and on the
elaboration and adaptation of knowledge to problems. The second process is a
process of systemic search, a process that senses and responds to problems,
that harvests policy options and turns them to the purposes, both public and
career—related, of politician and public ofﬁcials. As we have seen, in the
American political system, search processes can be activated by exogenously
. genaated crises and by constitutional routines, by bureaucratic needs and by
political necessities. Describing political innovation in any particular instance
thus entails describing how these two processes interact (Polsby, 1984: 173).

 

5 “A Garbage Can Model of Organizational Choice,” Administrative Science Quarterly
17 (March 1972).

72

The linking of the literatures on the state level was speciﬁcally done in 1992 by Henry
R Glick in his book, Polig Innovation and Its Consequences. Glick connected agenda-setting
theory with state irmovation theory in his review of Right to Die legislation in states.

If it were possible to neatly separate parts of the political process, innovation

would begin where agenda setting ends, but the two clearly are connected.

Most research on policy innovation assumes earlier agenda setting, but it rarely

inquires into that process. Since studies of irmovation take place after

substantial diffusion and adoption has occurred, there is no need to discover

how issues were placed before the government in the ﬁrst place. Tracing the

pattern of policy diﬂ’usion can stand on its own as a separate research

enterprise. But to understand how particular issues emerge, evolve and are
adopted, it is essential to begin to forge links between agenda setting and

irmovation (Glick, 1992: 41).

He proposed that agenda-setting is a process that focuses on how problems transform
from a general social concern into speciﬁc items for oﬂicial governmental action. Glick
suggested that agenda-setting theories lean toward description and comparison of the strategies
of political processes. In contrast, he argued that innovation deals with goverrnnent adoption
of new programs and policies, and the extent of how these programs are implemented. He
suggested that the process of innovation was much more complex than the date a speciﬁc law
is enacted (Gliclg 1992: 42). In his comparative case studies of Massachusetts, Florida and
California, he explored in detail how these three states reﬂect diﬂ’erent patterns in agenda-
setting and subsequent policy and program irmovation. As suggested in his earlier work with
Hayes (1991), Glick stated that a signiﬁcant component of irmovation was “reinvention,” and
that few studies have considered both dimensions of “earliness of adoption” and “extent of

adoption” for the same policy.

73

Mooney and Lee speciﬁcally focus on this concept of “policy reinvention.” in their
review of pre-Roe abortion regulation reform in states. They suggested that policy adoption
was similar to a social learning process. “The experiences of the states that have adopted the
policy previously provide information that allows a later adopting state to predict these effects
better’ ’(Mooney and Lee, 1995 2608). They found that the later adopters of abortion reform
(Pre-Roe) took into account the experiences, both policy and political of the earlier adopting
states, and that abortion reform efforts were both unidirectional and incremental and thus,
consistent with “social learning theory.”

Mintrom and Vergari (1996) also continue building an explicit link between irmovation
and agenda setting in their research about the diffusion of school choice ideas across the United
States. They focus on the role of “policy networks” or “plarnring comnnmities” in ensuring
approval for policy innovation. They emphasize the importance of the policy entrepreneur in
manipulating the resources of the policy commrurity and directing its support for the policy
irmovation being pursued.

By linking the literatures and focusing innovation on the process of policy change
suggests that the “organization,” “institution,” or the “administration” are the critical elements
to investigate when looking at determinants of state based policy innovation. In building a
theory of state irmovation, this dissertation proposes that it is critical that the focus be on
organizational structures and governance practices. George W. Downs Jr., in Bureaucracy,

Innovation, and Public Policy. argued that the policy determinants literature provided no

74

guidance in selecting those characteristics of the bureaucracy and its environment that might be
responsible for the differential irmovativeness of states. He suggested that previous irmovation
research was unstable and that too much variance was left unexplained, because researchers
were paying attention only to a state's socioeconomic development or political attrrbutes
(Downs, 1976:41). Downs proposed that knowledge of key bureaucratic and task-
envirornnent characteristics would substantially increase the ability to understand and predict
how states react to policy irmovations. Downs argued that variables such as complexity of the
organization and hierarchy are important dimensions of policy irmovation, and that they
deserve attention when analyzing the process of irmovative decision-making.

Bureaucracies are made up of hierarchical structures with clear lines of authority and
speciﬁc expertise. They ﬁrnction within established rules and regulations. These institutions of
goverrnnent are typically not collaborative nor irmovative. "Nobody can be at the same time a
correct bureaucrat and an innovator. Progress is precisely that which the rules and regulations
did not foresee; it is necessarily outside the ﬁeld of bureaucratic activities" (von Misses,
1944:67). The conventional criticism of bureaucracy is that it is inﬂexible and uncreative; it
stiﬂes spontaneity of the employees; and it is primarily urnesponsive to the public (Downs,
1967).

James Q. erson, building on the earlier work of Weber (Weber, 1946) suggested that
bureaucracy, given the regiment of the hierarchical structure, would be an ideal organization

for the adoption/unplementation of irmovations, but not the conception of innovations (Wilson,

75
1966). Vrctor Thompson also argued that the conditions within bureaucracy are inappropriate
for creativity, and are determined by a drive for productivity and control (Thompson, 1965).

Thompson examined the obstacles to innovation, and suggested that alterations in
bureaucratic structures be made to increase innovativeness. He reconnnended that
bureaucracies increase professionalization of employees; decentralize; create a looser and more
tmtidy structure, including ﬁeer communications systems; engage teams of employees to
organize around projects; rotate assigrnnents; expand reliance on group processes; modify the
incentive system; and change management structure. Thompson suggested that "conﬂict or
coalition structures" encouraged innovation.

Unlike the bureaucracy literature, the literature on collaboration is short and primarily
anecdotal. The root of the deﬁnition of collaboration can be found in the seminal work of
Axelrod, The Evolution of Cooperatr_'<_)_r_t Using concepts advanced within game theory,
Axelrod investigated, in the absence of a central authority, how to inspire an individual to
cooperate rather than pursue his own self-interest. Axelrod suggested that this mutual
cooperation can be promoted in three diﬂ’erent ways: making the ﬁrture more important
relative to the present; changing the payoffs to the players of the four possrble outcomes of a
move; and teaching the players values, facts, and skills that promote cooperation (Axelrod,
1984:126).

Collaboration involves individuals getting to know each other, the sharing of

information and/or ideas, and also "making decisions together" (Chynoweth, 1993).

76

Collaborative strategic plarming is long-range planning. Collaboration is about consensus
planning. It deﬁnes the desired outcome, comprehensively assesses problems and
opportrmities, and most effectively utilizes limited resources to achieve results. The systematic
use of collaboration as an irmovative decision-making process tool extends beyond the
interaction of participants and the sharing of information and resources, to include fundamental
change in organizations and institutions as they are altered to support these collaborative
endeavors. "Institutions matter in that they contribute to or impede particular policy
capabilities" (Rockman,1994:149). Collaboration might well be the explicit organizational
procedure for linking irmovation to other aspects of agency operations (Yin, 1979).

Collaboration is a process tool to create change. It is reasonable to suggest that
process causes policy, however, this is just one-half of a complete theory, because it is also
necessary to understand what causes certain processes. The necessary ingredients for a
successﬁrl collaboration includes leadership, trust, commitment to reﬂective action, sharing of
information and resources, clearly stated outcomes and ﬁnally action‘5 Leaders of collaborative
efforts often face similar challenges—addressing turf issues, working through conﬂict,
developing ﬁnancial strategies, recognizing issues of diversity, agreeing on desired results, and

developing strategies to achieve those results through consensus. The people who lead,

 

6 For a detailed discussion of collaborative processes, see Judith Chynoweth's boolg A
Guide to Comm__umty' -based Collaborative Stratgic leg,’ CGPA, 1993; and the Melaville

and Blank booklet, V_V_hat It Takes: Structurrng’ Intmgency Partnerships to Connect Children
and Families with Comprehensive Services, 1991.

 

77
participate in, and eventually implement the activities of interagency collaboratives need to have
a common vision and joint commitment to or ownership of the process and/or product to make
the collaborative successﬁrl.

The factors inﬂuencing policy innovation explored by Polsby (1984), Glick (1992),
Mooney and Lee (1995), and Mirtrom and Vergari (1996) link directly with agenda-setting
theory. According to the Kingdon model of agenda-setting, it is a multitude of factors and
players that determine issues which get on the public agenda He suggested that it is not simply
a crisis or disaster which will launch something onto a policy agenda and deﬁne it as a
“problem” to be resolved, but that “something else” needs to accompany this crisis (Kingdon,
1984:103). He also presented the need for “knowledge” to be resident among the policy
specialists so that solutions can be developed for these problems (Kingdon, 1984:134-138).
He also highlighted the role of the “political processes” which affect agenda setting (Kingdon,
1984:153-172). As Polsby, Glick, and MirtromNergari maintain, there is a deﬁnite comrection
between agenda-setting and innovation theory.

Summary—A Paradigm Shift or a Theory Building Loop?

It seems apparent that efforts at building an “theory of policy innovation” have been
stymied by a variety of problems, one of which is the desire for the discipline to distinguish
between the sub-ﬁelds related to the development of public policy. Lindblom argued that the

policy-making process can explain partially how governments pursue their various policy

78
targets, but not why the targets are chosen (Lindblom, 1980). Thomas Dye and Virginia Gray
argued that: “The explanation of public policy can be aided by the construction of a model
which portrays the relationships between policy outcomes and the forces which shape them”
(Dye and Gray, 1980: 2). As Baumgartner and Jones suggested, broad research questions are
sometimes not pursued because of the narrow focus of researchers who center their attention
on issues of agenda-setting, policy implementation or policy evaluation and never make the
connections between these various elements of the same policy cycle.
Those who have studied policy implementation typically have not emphasized

thedrarnaticchangesthatoftenoccurinthepublicagenda, andthosewho
focus on the agenda often discount the strong elements of stability or

incrementalism present in other parts of the policy cycle” (Baumgartner and

Jones, 1993: 10).

Polsby (1984), Glick (1992). Mooney and Lee (1995), and Mintrom and Vergrari
(1996) bridged sub-ﬁelds in their work Also, Baumgartner and Jones (1993) explored the
consequences of issue deﬁnition and how it is related to agenda processes and subsequent
policy development. In their work on “comparative issue dynamics,” they link the policy
development literature with that of the agenda-setting literature. I suggest that when exploring
the factors that inﬂuence irmovative decision-making by state policymakers and administrators
as they develop long-range plans for the aging of the baby boom population, we must look
beyond the irmovation liteature. We should consider the early work of Dye, Gray and

Lindblom in isolating the determinants of public policy. Also, we should review the agenda-

79
setting literature—the work of Kingdon, March and Olson, Hayes, and Heclo—and the roles of
policies, politics, and individuals in the decision-making processes. These theories are
interconnected and should be considered jointly if researchers are to wrestle with broad
questions in a complex political environment.

It seems that with the genesis of the agenda-setting literature in 1984, there is a
signiﬁcant fall-oﬂ’ in the development of innovation theory. Irmovation theory has been plagued
by inconsistent deﬁnitions, unreliable ﬁndings, and contrasting methodologies. However, a
conﬁrmed focus on “irmovation” is critical if we are to isolate factors contributing to the
processes of innovation. Although connected, agenda-setting theory does not replace the basic
elements of irmovation theory. Kingdon emphasized the importance of linkages between
problem identiﬁcation and solutions. However, he was weak in articulating how “solutions”
come about. He presented the concept of “Policy Primeval Soup” involving mrmerous players,
particularly policy entrepreneurs, interacting in a variety of policy communities (Kingdon
1984). However, Kingdon did not speciﬁcally address the issue of irmovation—how the “new
idea” comes about. This is the work of irmovation theory-to discover how solutions are
developed. It reﬂects the ﬁrst of the “processes” which Polsby deﬁned as irmovation (Polsby,
1984)

Inthisdissertation, Ibuildonthetheoryintlre agenda-settingliteratmeandthepolicy
development literature. I expand on some of the hypotheses presented by Baumgartner and
Jones regarding conﬂict resolution, and submit that irmovation theory, as developed by Downs

80
and Mohr, signiﬁcantly assists in answering the question posed by Lindblom—why targets are
chosen It also addresses Kingdon’s weakness in identifying “how solutions come about.” I
maintain that these literatures are congruent. Theories within each sub-ﬁeld need to be
merged, if we are to address the complex political, social, economic and organizational factors
which explain and predict state-based aging policy, and if we are to come to a better and ﬁrller
appreciation of the factors which can inﬂuence state-based policy innovation.

Using the “irmovation design theory” of Downs/Mohr as my theoretical framework, I
focus on the “process” of innovation-the act of changing, altering or renewing. By limiting
my work to the “process of irmovation” I have the opportunity to respond to some of the
challenges put forth by Downs and Mohr almost 20 years ago regarding the conceptualization
of dependent variable and the interaction among independent variables (Downs and Mohr,
1976)

BybringinginthetheoriesadvancedbyDyeandGrayandothersregarding
determinants of public policy, and the work of Kingdon, Hayes and Batnngartner and Jones on
agenda-setting, I advance the discussion of “policy irmovation” on the state level. By linking
the distinct theories in these literatures, I explore the conelates of policy innovation on the state
level, and discuss the conditions under which state public administrators and policymakers
respond irmovatively to the demographic, social, economic and political challenges of the aging
of America. As Eckstein suggested, the work of this dissertation should “feed back” to the

existing theoretical constructs, and assist in the theory building process, allowing for ﬁuther

81
reﬁnement and adjustment to innovation theory, agenda-setting theory and policy development

theory (Eckstein in Greenstein and Polsby, 1975).

Chapter 3

RESEARCH DESIGN AND METHODOLOGY
CREATING A MODEL OF STATE LEVEL INNOVATION DECISION DESIGN
FOR THE DEVELOPMENT OF STATE AGING POLICY
“Intellectual progress proceeds by ﬁts and starts, and cannot be
sustained solely by methodology. ”

Christopher H. Achen
Interpreting and Using Regression, 1982

Introduction

There are numerous policy analysis models and methods developed to assist political
scientists in their research pursuits. These methods are easily grouped into two general
categories of quantiﬁable and qualitative research. Diﬂ’erent research approaches are
appropriate for different types of problems, and insights from quantitative and qualitative
approaches can be viewed as complementary rather than conﬂicting (King et al, 1991).

The more methodologically rigorous quantiﬁable method used by political scientists is
regression analysis. Regression analysis is based in microeconomic theory and allows
researchers to determine relationships between variables (Lewis-Beck, 1980). Regression

analysis is hailed by social scientists as a strong predicting tool, permitting them to advance the

82

83
discussion about causal inference. The other side of quantiﬁable research is qualitative
research, which is thick in description and often criticized for being methodologically weak.
Qualitative research tools include case studies, ethnography, historiography, participant

observation, and comparative studies.

The Research Design

A research design is basically a blueprint of research to be conducted. The research
design is “the logical sequence that connects the empirical data to a study’s initial research
questions, and ultimately, to its conclusions” (Y 1n, 1984: 28). In developing the research
design for this dissertation, I heeded the advice of Lowi (1964) to be both relevant and
rigorous. I followed the advice from Campbell that methodologists must achieve an applied
epistemology which integrates both qualitative and quantitative research (Campbell, 1975:
191). This dissertation combines the rigor of aggregate regression analysis with the richness of
a comparative case study.

The research questions are the core of the research design. As submitted by Yin
(1984), it is ﬁom these questions, that a researcher develops his/her hypothesis, test this
hypothesis through both quantitative and qualitative methods, draws conclusions, and hopeﬁrlly
advances the political science discourse. Speciﬁcally, at the core of this research design are the
following questions:

(1) What internal determinants within a state—demographic, socioeconomic

and political—are plausibly causal in state planning for the aging of the baby

boom population, and the subsequent development of aging policy for the 21 st
century?

84

(2) What governance structures and practices within state government are
associated with policy irmovation and provide for an “irmovative” environment
within which to respond to the demographic realities of the 21 st century?

In an attempt to answer these questions, two corresponding hypotheses are developed,
which reﬂect the theoretical framework used in this dissertation.

H1: States that have a signiﬁcant mrmber of older citizens currently, or

anticipate notable growth in the mrmber of elderly, larger, wealthier states;

politically liberal; and have a uniﬁed political base between the executive and
legislative branches are more likely to engage in long-range plamring for the

aging of the baby boom population and be better prepared to develop

irmovative strategies regarding the aging of America

H2: States in which an aging agenda is visrble, or with governance structrnes

that provide for and encourage irrteragency collaboration on the state level, or

both are more likely to engage in long-range planning for the aging of the baby

boom population and be better prepared to develop irmovative strategies

regarding the aging of America

In testing these hypotheses, this dissertation utilizes two distinct research tools. First, a
written survey is conducted sampling the perceptions of state policymakers and administrators
about aging in America and its implications for each state and their state agencies. These data
were collected in a 1995 50-state survey of state policymakers and administrators in the health,
social services and aging ﬁelds. Second, a comparative case study involving four states is
convened, in which contextual information is gathered to provide more in-depth analysis of the
assumptions conﬁrmed in the written survey. This comparative study complements the
aggregate analysis because the states selected, California, Indiana, South Carolina, and

Vermont, are “outlier states” based upon the regression analysis. (See Interview Protocol used

85

withthecasestudystatesinAppendixC.) Thiscomparativestudyisinterpretedasavalidity
test for the aggregate analysis. In exploring the enablers of and constraints on the development
of state aging policy for the let century, this dissertation relies upon Downs/Mohr’s
“Innovation Decision Design” as a ﬁarnework for analysis. Its focus is on the “ ” of
imnovation—the act of changing, altering or renewing. Consistent with the Downs/Mohr’s
design, it considers the motivation for and obstacles to innovation. This dissertation examines
the “irnternal detemirnants” that might inﬂuence innovative decision-making, as deﬁned by
Berry and Beny (1990), as well as the processes or governance practices which ignite new
ideas or allow new ideas to be adopted and/or implemented. This review of the inﬂuence of
collaborative and cooperative work errviromnents in state goverrnnernt on the innovative
decision-making process signiﬁcantly adds to the existing literanne on policy determinants and
also continues to build an innovation theory.

I propose that a state’s ability to plan for and irmovatively respond to the anticipated
demogaphic challenges of the let century is a direct ﬁnnction of four inter-related state
characteristics reﬂecting: 1) demogaphic aspects of the state; 2) a composite of
socioeconomic characteristics of the state; 3) political factors within the state; and 4) the
govennarnce structure and practices within state government. Exploring these “internal
determinants” within a state-the dernogaplnic, socioeconomic and political factors-has along
and rich lnistory within state policy research generally, and speciﬁcally within the innovation

literature. This examination of the deterrnirnarnts of state policy development can be fournd in

86
the work of Rogers (1962); Easton (1965); Dye (1966, 1978); Walker (1969); Baybrooke and
Lindblom (1970); Lindblom (1959,1979); Gray (1973); Savage (1978); Dye and Gray (1980);
Carmon and Baum (1981); Lammers and Klingnan (1984); and Berry and Beny (1990, 1992).
Although the ﬁndings from this body of research have been conﬂicting at times, this
dissertation re-examines the potential inﬂuence of selected dernogaplnic, socioeconomic and
political factors on innovative decision-making processes.

This exploration irnto the fourth component of these “state characteristics”-the
governance structure and practices within state goverrnnent—reﬂects the work of Frendreis
(1983), Polsby (1984) and Glick (1992). It focuses on the complex organizational elements in
innovative decision-making, and examines the role of collaborative and cooperative processes
withirn states and their irnﬂuence on the development of aging policy. Although Lammers and
Klingnan (1984) looked at the capacity of the State Unit on Aging they did so to deterrrnine if
the state ﬁilﬁlled the federal requirerrnerrt to have such an oﬂice in order to get funds ﬁom the
Older Americans Act of 1965. They did not consider governance issues or practices within the
oﬂice of aging nor did they review cross state goverrnnernt or irrteragency collaboration efforts.

It is arguable that “governance issues and practices” witlnin state goverrnnent inﬂuence
the ability of states to respond to the “aging crisis” and to plan innovatively for the year 2010
and beyond. In examining the role of governance structures and collaborative work
environments, issues of entreprenenial leadership and irnrnovative orgarnizational design, as well

as the level of cross agency collaboration are considered. This builds upon the early work of

87
Polsby (1984), in which he considered the role of policy entrepreneurs and organizational
culture in policy irnitiation and innovation. Also, it addresses the appraisal of Glick and Hayes
(1991) and Glick (1992) that irnnovation is a “complex process” involving “reinverntion”, and
their criticism of previous innovation research focused narrowly on the adoption or non-
adoption of a law or progarn This dissertation also challernges the conﬂict-resolution
hypothesis advanced by Baumgartner and Jones (1993), and suggests an alternative policy

development process built upon irnteragency collaboration and cooperation.

Data Collection

Theﬁrststepinfalsilyingorvahdaﬁngtheassnnnpﬁonshnthehypotheses,isthe
collection of empirical data. There are a variety of diﬂ’erent data collection efforts used irn this
dissertation, some of which are primary sources and some secondary sources. The two
primary sources of data involve a written survey, with a follow-up phone interview with
representatives from selected states. The secondary sources consist of 1990 census
information and factual details included in The Booﬁk of Sta_te_s.

The ﬁrst step in implementing the research design was to gather information
systematically from states about on-going, long-range planning for the shifting demographics,
andtheprocessbywhichstatesarepreparingfortlre213tcenturyandtheagingofthebaby
boomers. This effort was accomplished through a written survey to state level policymakers
and public administrators. The second primary data collection effort was through a phone

irnterview with represerntatives from four selected states, which enhances information gathered

88
ﬁom the writtern survey, and speciﬁcally focuses on the level of policy innovation and
orgarnizational entrepreneurslnip in their state.

Four-to-ﬁve different agencies were identiﬁed within each state that poterntially could
be inﬂuenced by the aging dernogaphics arnd that had programs or policies currently affecting
the elderly population I focused only on those agencies which would be typically engaged in
aging policy such as the departments of social services, mental health, public health, oﬂice of
disabilities, oﬂice of veterans aﬂ’airs, insurance connnissioners, and ofﬁces/departments of
aging. (See attached listing of the agencies surveyed in each state in Appendix A).

The names, titles, addresses and phone numbers of these agency heads/policy directors
were veriﬁed during the months oprril and May 1995. Four different survey instruments
werecreatedinordertolirrnitsome ofthequestionstotlneareaofexpertiseoftheagency
srnveyed. (See Appendix B for actual survey instrument used.) However, all survey
instruments included the same questions about the awareness of demogaphic shifts; the
plarnning capacity within the state to meet these demogaphic challenges; the estimation of
inrnovativeness irn plarnning for these changes; and the level of state collaboration in plarnning for
this demographic shift.

In total, 324 state policymakers/public administrators were surveyed nationwide.
These surveys were distributed arnd collected over a seven-week period during the latter part of
the Surrnner of 1995 through a national orgarnization krnown as the Council of Governors’

Policy Advisors (CGPA). CGPA is a Washington based membership organization, which is an

89
aﬂiliate of the National Governors’ Association. CGPA had undertaken an irnitiative to educate
and inform state policymakers about the shifting dernogaplnics of the 213t century and the
potential implications for state policy. This subset of surveys was a part of the information
gathered by CGPA. Also, as a part of this project, the two Lead Governors, Governor Chiles
of Florida and Governor Branstad of Iowa, sernt a letter to all 50 governors erncouraging
participation in this national survey.

The last of the surveys were returned by the ernd of September, 1995. Response rates
varied depending upon agency and state. Of the 324 surveys distributed, 122 (38%) were
connpleted and returned. This response rate is deceiving in that several states submitted only
one response. For example, the departments of health and mental health might well refer their
surveys to the department of social services for completion Or, irn some cases, oﬂices of
disabilities or aging were subsumed under another agency such as mental health or social
services. Overall, 48 of the 50 states (96%) responded, and data are lacking only ﬁ‘om the
states of Permsylvania arnd Delaware.

Information gleamed from these surveys provided the basis for the analysis conducted
inn this dissertation regarding the developmernt of innovative state aging policy and plarnning
strategies for the 21 st century. However, it is also necessary to obtain factual information
regarding demogaphic data, socioeconomic information and political factors for each state.
Thus, in addition to the irnfornrnation gathered from these surveys, data from the 1990 census,

made available tlnrough the Census Bureau, was used to get a variety of state dernogaplnic and

90
socioeconomic details. Also, irnfonnation from the BooJk of the Stat_es regarding partisanship
control of the executive and legislative branches in each of the 50 states over the last 15 years
was utilized to allow for the construction of the political variables.

Firmlly, based on the ﬁndings ﬁom the aggregate analysis, four states are selected for
in-depth exploration irnto the dyrnarnics of long-range plannirng and imnovative decision-making
processes. This comparative case study was conducted as a validity check for the aggegate
analysis. Data collection for this connparative case study of Califorrnia, Irndiana, South Carolina,
and Vermont involves three sources. First, the written snnveys which were completed in the
surrnner of 1995 by their state policymakers or public administrators. Second, the plethora of
demographic and socioeconomic data available on the state level from the Census Bureau.
And lastly, a phone interview protocol was used with the representatives from these states on
planning underway and the level of state agency collaboration irn developing irnnovative
strategies to deal with the demogaphic challenges of the 213t century. These interviews

provide additional contextual information to complement the aggregate analysis.

Methodology Employed

As previously stated, this dissertation ennploys both regression analysis and a
comparative case study involving four states which are identiﬁed as “outliers” in the OLS
regression. The data analysis progarn used was STATA, which provides easy access to

diagnostic tests which were run on the regession model. The comparative case study was a

91
validity check for the ﬁndings from the aggegate analysis and was done to complement the
ﬁndings in the aggregate arnalysis.

Although the four states selected, Vermont, Indiana, Califonnia and South Carolina,
have different socioeconomic, political and organizational traits, the aging of the baby boom
population should be of equal political and public policy concern among the state leaders,
policymakers and public administrators because of the impending dernograplnic imbalance.
However, two states are plarnning irnnovatively for the aging of the baby boom population,
while the other two states are not. This comparative case study should ernlnance the
information gathered through the survey and provide contextual details regarding the
relationship between interagency collaboration and innovative decision-making and strategic
long-range planning. Dissertations are often done as single case studies or comparative case
studies. This comparative case study simply allows for a more complete and interesting
response to the questions posed in the research design, and should not be viewed as a thorough

comparative review of the development of aging policies in these four states.

Creating the Model

Regession analysis assumes that the structure of a relationship is systematic, and
places on this relationship a series of conditions which must be met if the model is to have
predictive power. “Wrthout correct speciﬁcations, conventional statistical theory gives no
assurance that the impact of a variable will be estimated accurately” (Achern, 1982: 11). The

hypotheses examined irn this dissertation, previously stated, argues that a positive relationslnip

92
exists between certain demographic, socioeconomic, political and governance factors and the
on-going, long-range planning and innovative policy development in states regarding aging
policy for the 215t century. Speciﬁcally suggested is a set of conditions, which when existing,
will result in states being better prepared for the 213t century. These conditions include:

o Ifa state currently has a signiﬁcant number ofolder citizens, or anticipates
notable grth in the number of elderly;

0 Is a larger and wealthier state;
0 Is politically liberal;

0 Beneﬁts from a uniﬁed political base between the executive and legislative
branches;

0 Has a visrble aging agenda; and

0 Has governance structures that provide for and encourage irrteragency
collaboration on the state level;

0 Then this state will be more likely to actively engage in long-range planning

for the aging of the baby boom population and be better prepared to
develop innovative strategies regarding the aging of America

The Dependent Variable—A Measure of Innovation:

In chapter two of this dissertation, I presented the concept of “innovation” as a
“process of decision-making.” This concept is consistent with the early innovation theory
building eﬁ‘orts of Rogers (1962), and speciﬁcally compatible with the deﬁnition of irmovation
proposed by Mohr (1969) and Downs and Mohr (1976) in their theoretical construct—the

innovation decision design. Also, by deﬁning a measure of irmovation as a “process” by which

93

new policy or program ideas are generated within a state, the caution expressed by Polsby
(1984), and Glick (1992) about “innovation” reﬂecting the complexity of policy change is
heeded. Criticisms have been levied against previous innovation research because of its over-
reliance on the creation of a dichotomous dependent variable - the adoption or non-adoption
of a speciﬁc piece of legislation (Glick and Hays, 1991; Glick, 1992; Berry, 1994; and Hays,
1996). Creating a measure of innovation which considers an on-going process of change
addnnwesﬂﬁscnﬁcbnr

Also, in chapter two, I highlighted how Polsby (1984) and Glick (1992) precisely
ﬁngedtheﬁnklxﬁumenthezmmndasennu;wmihuxnmﬁonlhenuunx.Ikrﬂxzdewﬂopmwntof
‘ﬂm2depemhnn vmmﬂﬂe ﬁn ﬂﬁsrﬁaurunknt dﬁslhumgw:u)the qgmuhkaﬂﬁng hunauneib
critical. The dependent variable in this dissertation is a measure of “innovation” deﬁned as a
process of decision-making. This measure of innovation is grounded in the agenda-setting
literature in the concepts advanced by Kingdon (1984), Hayes (1992) and Baumgartner and
Jones (1993).

Kingdon (1984) proposed that issues become a part of the public agenda when they are
ckﬁhuxlaspuobknnsandlhﬂaxivdﬂrposnbknyﬂuﬁons

“We conceive of three process streams ﬂowing through the system—streams of

problems, policies, and politics. They are largely independent of one another,

and.ewdrrhnnﬂopsruxxndhugto hsrywnldynanﬁcsruulruks. Ihnzatrxnne

cﬁﬁcaljunctmestheﬂneestremnsarejoinedandﬂregreatestpoﬁcychanges

grow out of the coupling of problems, policy proposals and politics” (Kingdon,
1984: 20).

94

He articulated the importance of “capacity” to develop policy alternatives for the identiﬁed
problems as critical to the agenda-setting process. Kingdon also discussed the cnrcial role of
the “policy communities” and “policy entrepreneurs” in this agenda-setting process.

Michael Hayes (1992) did not limit himself solely to agenda-setting, but considered the
more extensive processes surrounding public policy development. He suggested that there are
four stages of policy development-problem identiﬁcation, agenda setting, policy adoption and
policy implementation. He portrayed these stages as concentric circles involving diﬁ‘erent
players who take on different roles. Hayes suggested that at the center of the circle sits the
“change agent,” who can be either a political player or a policy entrepreneur. The next circle
involves more players from the “policy connmrnity”, in which the “suggested policy change” is
reﬁned and ﬁrrther developed. In the third circle, the proposed policy change moves into the
larger political arena and becomes part of a more generalized public and political debate, and is
subjected to both political and bureaucratic scrutiny. Finally, the fourth circle entails the
implementation of the policy change and the need to effectively engage the bureaucracy and
successfully engross the public support for the speciﬁc policy change.

Baumgartner and Jones (1993) proposed a conﬂict-resolution model as a basis for
explaining how issues become part of a public/government agenda They, like Kingdon and
Hayes, emphasized the importance of problem or issue identiﬁcation and deﬁnition, and they
stressed the critical role of the policy expert or political leader in shaping the “policy image”

and determining the “policy venue” in which the debate and discussion about policy change is

95
to take place. Baumgartner and Jones highlighted the function of “institutional structures” in
encouraging or suppressing policy changes.

There are three common threads which run through these examples from the agenda-
setting literature: (1) the need for “policy capacity” to deﬁne issues/problems and to develop
sohrtions to these identiﬁed problems; (2) a speciﬁc role for “policy communities” or plarming
groups; and (3) the importance of “policy or political entrepreneurship and/or innovative
strategy.” These “threads” can be arguably linked easily with the concepts of “enablers and
constraints to innovation” advanced by Downs and Mohr (1969). These common “threads”
form the basis of the dependent variable in this dissertation.

In developing a measure of the “innovative decision-making process” within states, I
focus on: (1) policy capacity; (2) policy development and planning functions; and (3) the
“innovation” of the strategy or the organization. The dependent variable is constructed out of
survey responses, which is consistent with the research design of both Mohr (1969) and
Frendreis (1983). Three questions on the survey are utilized to test these components of the

“innovation decision design.” The question used to test policy capacity was:

(1) “Please evaluate the current capacity of the state to effectively meet the
challenges and opportunities of these shitting aging demographics in enhancing
and assuring the quality oflife of older Americans including health care and
social support services?”

minimal suﬂicient superior

The question used to test “policy communities” or planning underway was:

96
(2) “How would you describe your state’s planning for a coordinated support

system to detect gaps in services and develop new resources to meet the needs
of a changing older American cohort?”

poor fair good

The ﬁnal question, testing “innovative strategy” was:

(3) “How innovative is your state’s strategy for providing human/social
services to the aging baby boomer cohort?”

not at all somewhat very
There is minimal correlation between these three questions, and it is arguable that they

are measuring different aspects of the “innovation decision design” The correlations are:

Table 2—Correlates: Capacity/Policy Community/Innovation

 

 

 

 

Capacity Policy Community Innovation
Capacity 1.0000
Policy Community .1336 1.0000
Innovation .3055 .4451 1.0000

 

 

 

 

 

 

The responses to these questions are then collapsed into a single dependent variable-
an “innovative decision-making index.” The dependent variable is ordinal, and based on a scale
of 0-3, with “0” representing “poor,” “minimal” or “not at all,” and “3” representing “good,”
“superior,” or “very.” This variable is titled “index” and state scores are continuous ranging

ﬁom .83 to 2.50.

97

The Inﬂuence of Independent Factors:
Berry and Berry, in their 1990 article, outlined “internal determinants” of innovation.
They deﬁned these internal determinants as the political, social and economic characteristics of
a state and distinguished them from regional diffusion factors or the inﬂuence from
“nationalization.” Use of such “determinants” has a long history within the discipline generally,
and speciﬁcally within the innovation research Walker, 1969; Gray, 1973; Savage, 1978;
Cannon and Baum, 1981; Frendreis, 1983; Lammers and Klingman, 1984a; Lammers, 1989;
Glick and Hayes, 1991; Beny and Berry, 1990, 1992; and Glick, 1992 utilized a variety of
demographic, economic, social and political variables in their analysis of the independent
inﬂuences on irmovation. The explanatory factors used in this model include many of the same
independent variables as employed in these previous studies, and are grouped into four general

categories: demographic, socioeconomic, political and organizational.

Demographic Factors:

Population statistics are readily used by social scientists in nmny different areas of
study. The use of state population and population density as a independent factor in innovation
diﬂ’usion studies dates back to some of the earliest work in the ﬁeld by Ryan and Gross (1943).
State population has been found to be a signiﬁcant predictor of innovation (Walker, 1969;
Savage 1978; Cannon and Baum 1981). Lammers and Klingman (1984) in their analysis of
aging policy innovation ﬁom 1955-1975 considered percentage growth in states’ elderiy

population during the 20-year time ﬂame of the study, and its potential inﬂuence on the

98
development of aging policy. They, however, did not ﬁnd that the percentage of elderly in a
state’s population was a strong predictor of the development of aging policy in that state.
Also, Lammers (1989) found that “there is no consistent tendency for states with higher
concentrations of elderly to have either earlier or more substantial policy responses” (Larmners;
1989:52)

Although the ﬁndings about the inﬂuence of “population” were inconsistent in past
studies, there is enough established research correlating demographics and subsequent policy
development, that this regression model uses demographics as an independent factor. There
are four separate demographic statistics initially considered for the model. The demographic
variables considered for each state included:

(1) Percent change in persons over the age of 65 over the last 15 years—1980-
1995;

(2) Percent of population of persons over the age of 65 in the state in 1995;

(3) Projected change in the percent of persons over the age of 65 ﬁom 1995-
2020;

(4) Projected percent of population of persons over the age of 65 in the state in
the year 2020.

The year 2020 is selected as the “out year” because it reﬂects the middle of the agewave of
retiringbaby boomers. The birth span for baby boomers covered an 18 year time span, as will
also their retirement. The ﬁrst cohort of baby boomers will begin turning age 65 in the year

2010, with the last of the boomers entering the traditional “retirement age” in the year 2029.

99

Itisreasonableto suggestthatifastateexperiencedrecerrtgrowthinthenumberof
elderly, or if such state has a signiﬁcant number of elderly anrently, then there is a potential
that this population will inﬂuence the development of aging policy. Since this dissertation
focuses on the long range planning for the baby boom population, it is also hypothesized that
the anticipated growth in elderiy in the state would also potentially inﬂuence the planning
underway for these demographic changes.

There is signiﬁcant correlation-.89—between past growth, current number of elderly,
and the projected percentage of persons age 65+ in 2020. Therefore, the only demographic
statistic selectedforthemodeloutofthesethreeisthepopulationofpersons 65+irrthestatein
1995. The percentage growth in elderly projected from 1995-2020 is also used as an
independent variable in the model. This statistic is not signiﬁcantly correlated with the current
level of elderly, and can be used as an explanatory variable for planning for anticipated
demographic changes. Thus, these two demographic statistics are used in the regression model

as explanatory factors for the innovation index.

Socioeconomic Factors:

Throughout the political science literature, a variety of socioeconomic variables are
considered. Thomas Dye (1966), in creating his model for the analysis of policy outcomes in
states, looked speciﬁcally at urbanization, per capita income in a state, poverty level, and
education level. He saw these as critical inputs in the policymaking process. Walker, 1969;

Gray, 1973; Savage, 1978; Cannon and Baum, 1981; Larmners and Klingman, 1984; Larmners,

100
1989; and Berry and Berry, 1990 and 1992 considered many of these same independent
variables when analyzing the inﬂuence of socioeconomic factors on policy innovation. In the
work of Berry and Berry (1990 and 1992), and particularly in Nice’s research (1994), not only
was individual ﬁnancial well being considered, but they also emplmsized state resources—ability
to pay—and the role they play in the initiation and implementation of new programs.

There are numerous socioeconomic variables that can be included in the model. The
challenge is determining the most relevant and signiﬁcant variables to incorporate into the
model to assure that it is speciﬁed correctly. Given the signiﬁcant ﬁndings of Walker (1969)
and Savage (1978) and Berry and Berry (1990 and 1992) regarding the potential correlation
between large, wealthy urbanized states and innovation, it is reasonable that these factors be
considered as a part of my model. Although Nice (1994) found little support for his hypothesis
regarding “slack resources” and innovation, given the importance of state ﬁscal capacity
emphasized in the Larmners and Klingman study (1984) on aging policy irmovation, a state’s
general fund budget is also incorporated as a potential explanatory variable in the model. The
initial socioeconomic variables considered included:

(1) Per capita income ofall persons in the state (1990 census);

(2) Percent of population living in a metropolitan area (1990 census);

(3) Poverty level of all persons in the state (1990 census);

(4) Poverty level of elderly (65+) in the state (1990 census);

(5) Population size (1990 census)-states are categorized into three categories
-the largest 10 states, the middle 30 states, and the smallest 10 states; and

101

(6) State General Fund (in millions of dollars) appropriated in the 1995
state budget.

Given the high correlation rate of .79 between the elderly poverty rate and the overall
poverty rate, and the relatively high negative correlation of -.60 between the poverty rate and
per capita income, any model including all three variables would suffer multicollinearity
problems. I determined that the closest measure of ‘eralth” as deﬁned by Walker (1969) is the
“per capita income” variable. Therefore, I dropped the poverty variables from consideration
Also, in that the correlation between urbanization and per capita income is moderately high at
.58, it is arguable that one of these variables should be excluded.7 Since the emphasis in the
ﬁndings from Walker (1969) and Savage (1978) and Berry and Berry (1990 and 1992) reﬂect

the importance of ‘Xvealth”, and per capita income is a stronger predictor of the irmovation

 

7 A model with only these two explanatory variables show that the "Per capita
income" variable is stronger, and thus should be selected over the level of urbanization in a
state.

 

regress index percapin urbanpct

 

 

 

 

Source | SS df MS Number of obs = 45
F( 2, 42)= 2.20
Modell .613762336 2 .306881168 Prob>F= 0.1236
Residuall 5.86194898 42 .139570214 R-squared= 0.0948
Adj R-squared= 0.0517
Total | 6.47571132 44 .147175257 Root MSE = .37359
index | Coef. Std. Err. t P>|t| [95% Conf. Interval]
percapin | .0000509 .0000243 2.092 0.043 1.79e-06 .0001
urbanpct |-.0054896 .0047758 -l.l49 0.257 -.0151275 .0041483

_cons |l.066826 .3901532 2.734 0.009 .2794646 1.854187

 

102
index, the regression model includes only per capita income, population size and the state’s

general ﬁmd budget, as the socioeconomic explanatory factors.

Political Factors:

There has been signiﬁcant research regarding the relevance and importance of political
systems and their potential inﬂuence on public policy outcomes in states. The role of politics,
party systems and power structures have been examined by many political scientists over the
years with mixed ﬁndings (Easton, 1965; Dye, 1966, 1978; Fry and Winters, 1970; Uslaner,
1978; Lewis-Beck, 1977; Stonecash, 1980; Lammers and Klingman, 1984; Berry and Berry,
1990, and Lowry, 1996). Oﬂen, the debate in the literature is the relative importance of
socioeconomic factors over political factors. In Dye’s model (1966) of the policymaking
process, he presents a structure in which both socioeconomic and political factors are relevant
explanatory variables. He suggests that socioeconomic factors are ﬁltered through the political
system—party systems and power structures—to develop state policies. Stonecash (1980) built
on this concept, and suggests that “politics” play a facilitative role in the creation of policy.

Speciﬁcally within the innovation literature, there also has been mixed ﬁndings
regarding the relevant signiﬁcance of politics, partisanship and ideology. Cannon and Baum
(1981) looked at political party, political ideology and political culture in their study. They
found no correlation between politics and innovation. However, in Lammers and Klingman
(1984), Klingman and Lammers (1984) and Lammers (1989), political liberalism seemed to be

a signiﬁcant explanatory variable for innovativeness in aging policy.

103

Given the Lammers and Klingman (1984); Klingman and Lammers (1984) and
Lammers (1989) ﬁndings regarding political liberalism, I determined that some measure of
political liberalism should be incorporated into the initial model. Historically, the Democratic
party is known for a more socially hberal agenda and a pro-government activist political culture
or ideology. Therefore, it is arguable that a Dernocratically controlled state, either the
executive or legislative branches, should be more innovative. As a measure of political
liberalism, this dissertation examined the control of both the executive ofﬁce and legislature
over a ﬁfteen year period—19804995. This time frame is selected for a variety of reasons.
Generally, on the national level, the decade of the 1980s is considered ﬁscally conservative,
given the two Reagan administrations. There were also signiﬁcant policy changes implemented
in 1981, divesting more domestic policy responsibilities to the states through block grants.
Also, the aging issue was on the national agenda, in that a Social Security Review Commission
was established to examine the solvency of Social Security. In addition, important tax changes
took place in 1986 regarding the taxability of Social Security income. In the early 19903, with
the election Bill Clinton, aging is once again on the forefront of national policymaking with the
health care reform initiatives and the discussion of Medigrants.

Not only is political hberalism or partisanship examined as relevant political explanatory
variables in the literature, there is also a fair amount ofresearch regarding the importance of
political uniﬁcation of the executive and legislative branches. Jacobson (1990) in his work on

Congress advances the theoretical ﬁarnework of “divided government’ ’ and suggests that the

104
electorate purposely selects different parties to control different branches of government
because they want a system of political checks and balances. Lowry, et a1 (1996) explored the
role of uniﬁed party government in explaining the connection between state spending and
election outcomes. They found that when there was uniﬁed government, the electorate was
more speciﬁc on blaming the party in control for economic conditions and state spending.

In addition to political lrberalism, and uniﬁed party government, the role of interest
groups camrot be discounted. It has been argued that interest groups are signiﬁcant motivators
for agenda setting and policy development (Dye, 1966; Kingdon, 1984; Elder and Cobb, 1984;
Hayes, 1992; Baumgartner and Jones, 1993). Speciﬁcally, when considering the aging issue,
interest groups have played an important part in developing policy and ﬁuthering an aging
agenda (Binstock, 1972, 1991; and Cutler, 1977). Lammers and Klingman (1984) did not ﬁnd
interest groups per se as a signiﬁcant predictor of innovation in their study. However, it is still
feasrble to examine the role of aging interest groups in state policy innovation, given these
other studies.

The American Association of Retired Persons (AARP) is the largest representative
group for the elderiy. It currently has thirty-three million members. AARP has historically
beenactiveinlobbyingonbehalfofitsmembersinCongress. MostrecentlyAARP hasbeen
credited (or blamed, depending on your perspective) with the stalling of the recent eﬂ‘orts by
Congress to balance the budget, and with blocking their efforts to block grant Medicaid to the

states.

105
The political variables initially examined included:

(1) Governor partisanship over the last ﬁfteen years—this is constructed as a

numeric value representing the number of years the state was under
Democratic executive control through the years of 1980-1995. The scale
can run from 0-15.

(2) Legislative partisanship over the last ﬁﬁeen years—this is constructed as a
mnnericvaluerepresentingthenmnberofyearsthestatewasunder
Democratic legislative control through the years of 1980-1995. The scale
can run from 0-15.

(3) Uniﬁed government—deﬁned as the number of years from 1980-1995,
which the executive and state legislature was ofthe same party. This
variable is constructed as a durmny variable—”+1” if the governor’s oﬂice,
thestate senateandstatehousewerecontrolledbythesameparty, a“0”if
the control of the state legislature was split (or if an Independent
controlled the governor’s ofﬁce), and a “-1” if the governor’s oﬂice was
controlled by one party and the state legislature was controlled by the
otherparty. (Nebraskabecauseithasanon-partisanandmricameral
legislature, it was not computed.) There were eight elections over the
courseofﬁﬁeenyears, andthereforetherangeofthisvariableisﬁem-S
to +8;

(4) Percent of persons over the age of ﬁfty (deﬁned eligibility) in the state who
are members of the American Association of Retired Persons (AARP).

Organ'national Factors:
The last group of independent variables considered in the model were organizational.

To examine organizational capacity is diﬂicult because of the numerous ways in which to
measure it, along with the diﬂiculty of measuring it. From the eariy work on the bureaucracy
done by Downs (1967) to the most recent efforts by Osborne and Plastrick (1997), the

importance of individual leadership and personal relationships in work enviromnents is stressed.

106
As already pointed out, the exploration of organizational capacity is extensive in the agenda-
setting literature (Kingdon, 1984; Hayes, 1992; and Baumgartner and Jones, 1993), and again
the elements of personal leadership, entrepreneurship and technical capacity are highlighted.
Speciﬁcally within the irmovation literature, Polsby (1984); Glick and Hayes (1991); and Glick
(1992) emphasize the importance of the entrepreneurial organization and its’ linkage to policy
innovation.

Peters and Austin (1985); Peters, (1988) suggested that one of the most critical
elements in successful organizations is the ability of the organization to form informal working
networks. Speciﬁcally, Peters tied the capacity to innovate with the creation of “skunkwo ”
within an organization—infonnal teams of people working together on a single problem or on a
single project. Osborne and Gaebler (1993) transferred many of these concepts to the public
sector, and they evaluate the effect of “teams” and “informal participatory management
practices” in public administration Osborne and Gaebler showed that there are many examples
in the public sector where entrepreneurship and informal networks yield more eﬂicient and
eﬂ‘ective services. They, too, make the tie between these informal networks and innovation.

The hypothesis in this dissertation regarding collaboration reﬂects the emphasis in the
literature on the beneﬁts of entrepreneurship, informal networks, “skunkworks,” and
interpersonal relationships. Also, it is proposed that if there is a cabinet level department of

elderaﬂ’airswhichisinchargeofagingissuesandadvancingtheagingagenda,therrtherewill

107
be more aging policy innovation Thus, the model includes two explanatory organizational
variables for state policy irmovation. These organizational variables are:

(l) Bureaucratic structure of the state function for providing services to
seniors/elderly in the state—this is constructed as an ordinal variable
ranging from 1-4, with 1=within another department; 2=autonomous
organization; 3=an ofﬁce within governor’s oﬂice; and ha stand alone
cabinet level department;

(2) Level of Interagency state collaboration—this explanatory variable is

developed from a question on the survey regarding policy development
for older citizens and the level of interagency collaboration. This is a

continuous variable with a scale of 0-3. The question used is:

“Within your state, how would you descnbe the level of collaboration
among state agencies and departments in developing a strategy for
meetingthechangingneedsoftheolderAmericansinthenextfew
decades?”

poor good excellent

Summary

Primarily, this dissertation is about theory building. It builds upon the work of Mohr
and Downs, and ﬁrrther develops a theory of innovation that is focused on the process of
innovative decision-making. The “innovation index” is constructed from three questions from
the survey dealing with state capacity, planning underway and level of innovativeness. The
regression model tests two separate hypothesis about the independent eﬂ‘ects of demographics,
socioeconomic, political and organizational factors on innovative decision-making. Measures

are developed to test the signiﬁcance of these explanatory variables. The model attempts to

1 08

explain and predict state level policy innovation in aging. The initial model constructed to test
these hypotheses is:
Y=a + Bxl + Bx; + Bx3 +Bx4 + Bx; + Bxs +Bx7 + Bxs + Bx9 + me + Bx” +e

Y=innovative decision-making—the innovation index

Bx, = Percent of population of persons 65+ in the state in 1995 (1990 census);

Bx; = Projected percent change of persons 65+ from 1995-2020 (1990 census);

Bx3 = Per capita income ofall persons in the state (1990 census);

Bx4 = Population size of state (1990 census);

Bxs = State general fund appropriation in 1995 (NASBO Report);

Bx6 = Years of Democratic control of executive branch (Book of States)

Bx7 = Years of Democratic control of state legislature (Book of States)

Bx; = uniﬁed government—executive and legislative branches (MM)

Bx9 = Percentage of eligrble 50+ persons belonging to AARP (AARP publication)

Bxlo = Bureaucratic structure (survey data);

Bx" = Level of interagency state collaboration (survey question)
In the ensuing chapters, the validity of this model is explored through both regression analysis

and a comparative case study.

Chapter 4
THE FINDINGS
UNDERSTANDING THE INCENTIVES AND CONSTRAINTS

ON STATE-BASED INNOVATIVE DECISION-MAKIN G
THROUGH AGGREGATE ANALYSIS

“The unraveling of the determintmts of public policy
is one that has preoccupied social scientists since the
advent of the behavioral revolution of the 19603. ”
George Downs
Bureaucracy, Innovation and Public Policy, 19 7 6
Introduction
Social science research, whether qualitative or quantitative, seeks to explain and/or
predict some or political phenomena. Researchers collect information, make observations if
you will, about this phenomena and then attempt to process this information or these
observations into coherent surmnaries—to tell a story. They try to discover what causes the
patterns they observe. Causal mechanisms are impossible to determine with certainty, given
the complexity of our social systems and the problem of inductive inference. Social theories
rarely can say more than that certain variables are related to each other. However, one of the
fundamental goals of inference is to distinguish the systematic ﬁ'om the random component of
the phenomena studied. There are many statistical guideposts established within the social

science discipline, as well as methodological diagnostics, which assist researchers in

109

110

distinguishing patterns of relationships from random acts. By applying these statistical and
methodological standards to the observations or data gathered researchers tell the most
plausible story. Regression analysis, discussed at length in chapter three, is based in
microeconomic theory. It is a statistical tool that helps political scientists advance their
theoretical arguments and tell their story. The most critical component of using this statistical
tool is to be assured that the regression model is accurately speciﬁed. “Wrthout correct
speciﬁcations, conventioml statistical theory gives no assurance that the impact of the variable
will be estimated accurately” (Achen, 1982: 11). Ordinary Least Squares (OLS) assumes that
themodelis speciﬁedcorrectly. Itassumesthattherelationshipbetweeeraninslineanthat
there are no relevant independent variables excluded from the model (omitted variable bias),
and that no irrelevant independent variables have been included (Lewis-Beck, 1980).

A regression model should not be viewed as ﬁnal or complete (Achen, 1982: 52). The
task of the researcher is to formulate a manageable description of the data that allows him/her
to exclude competing theories. Given a set of dependable and meaningﬁrl independent
variables with a linear relationship to the dependent variable, then the task for the researcher
becomes one of variable selection. Variables are incorporated into the regression equation
based on the theory being advanced. Variables are excluded or inchrded in order to check
speciﬁc hypotheses or counter-hypotheses. In determining which independent variables to
inchide in the equation, the goal of the researcher is to decide if the model is a “good ﬁt.”
There are a variety of diagnostic tests which assist the researcher in making this determination,

and assists him/her in upgrading his/her theoretical and/or empirical models.

1 l l

The R2 gives the percentage of the variance explained by the regression model. R2 is
oﬁen reported as a measure 0 “goodness of ﬁt”, but it has been argued that the standard error
of the regression is a far better measure (Achen, 1982). Some researchers attempt to maximize
the R2 by including irrelevant variables in the model. Although having more variables in a
modelmayincreasetheRz,itisnotareasonableprocedmeinthatthemodelwillnotbe
speciﬁed correctly, nor be theoretically relevant. The F-statistic is also used to test the
speciﬁcation of the entire model. This statistic is the “explained variance divided by K - 1, 1
divided by the unexplained variance divided by T - K” (Hanushek and Jackson, 1977).

In regression analysis OLS is used ﬁequently as the estimator. It is critical that these
estimators are BLUE, signifying that the model is the Best Linear Unbiased Estimator. If the
estimators are not BLUE than the model may be biased or ineﬂicierrt and the explanatory
power of the model is jeopardized. There are ﬁve assumptions around OLS, if maintained,
OLS is proven to be BLUE.

(1) That the independent variable (X) and the error term (U) are
independent of one another.

(2) That the estimator is unbiased; E[U]=0

(3) That U is not correlated with any other U term; E[UU’]=oZI. Violation
of this assumption is known as autocorrelation.

(4) That the variance of U is constant and ﬁnite; E[UU’]=021. Violation of
this assumption is known as heteroskedascity.

(5) That U is normally distributed.
Thereareavarietyofstaﬁsﬁcaltestsﬂratcanbeperfonnedtovaﬁdatethese

assumptions. Some of these tests require simple visual examination of the residual plots, while

112
other tests involve more sophisticated diagnostics. All of the methodological work must be
considered part of the theory building process. “Wisdom must be guided by theory, and some
of the necessary theory is statistical.” (Achen, 1982:78). This dissertation attempts to explain
and predict the level of innovative decision-making in states regarding the aging of the baby
boom population. The ﬁrst step in this eﬂ’ort is the development of a theory, which is
grounded in the literature, and is testable, given the information or observations gathered.
Again, the emphasis is on developing a parsimonious, reliable, and valid theory and to test this
theory through speciﬁed hypotheses. It is about telling the most interesting and plausible story

possible.

The Hypotheses
As highlighted in chapter three, at the core of the research design for this dissertation
are two speciﬁc questions:

(1) What internal determinants within a state—dernographic, socioeconomic
and political factors—are plausrbly causal in state planning for the aging of
the baby boom population, and the subsequent development of aging
policy for the 21st century?

(2) What governance structures and practices within state government are
associated with policy innovation and provide for an “irmovative”
environment within which to respond to the demographic realities of the
21 st century?

In an attempt to answer these questions, two corresponding hypotheses are developed,
that reﬂect the theoretical ﬁarnework used in this dissertation

H1: States that have a signiﬁcant number of older citizens currently, or
anticipate notable growth in the number of elderly, are larger, wealthier states;
are politically liberal; and have a uniﬁed political base between the executive
and legislative branches will be more likely to actively engage in long-range

113

planning for the aging of the baby boom population and be better prepared to
develop innovative strategies regarding the aging of America.

H2: States in which an aging agenda is visible, and/or with governance
structures that provide for and encourage interagency collaboration on the state
level will be more likely to actively engage in long-range planning for the aging
of the baby boom population and be better prepared to develop innovative
strategies regarding the aging of America
The initial model constructed in chapter three to test these hypotheses is:

Y=a + BX] + BX2 + BX3 +BX4 + BX5 + 8X5 +BX7 + BXs + BX9 + BXlo + BX” +8
Y=innovative decision-making—the innovation index
Bxl = Percent of population of persons 65+ in the state in 1995 (1990 census);

Bx; = Projected percent change of persons 65+ from 1995-2020 (1990
census);

Bx3 == Per capita income ofall persons in the state (1990 census);

Bx4 = Population size of state (1990 census);

Bxs = State general fund appropriation in 1995 (NASBO Report);

Bxé = Years of Democratic control of executive branchtBook of States)

Bx; = Years of Democratic control of state legislature( Book of States)

Bx; = uniﬁed goverrnnent—executive and legislative branches( Book of States)

Bx; = Percentage of eligible 50+ persons belonging to AARP (AARP
publication)

Bxlo = Bureaucratic structure (survey data);

Bx” = Level of interagency state collaboration (survey question).

114
This model attempts to explain and predict state level policy innovation in aging. Through the
use of the STATA statistical soﬁware package, a regression analysis was conducted to test the

validity and reliability of the model, as well as attempt to make the model most parsimonious.

Testing the Model
The following chart8 outlines a variety of tests to run to detemrine if the regression

model meets the central considerations of econometric inference.

 

 

 

 

 

Table 3—Validation Tests
Econometric Issue Diagnostic Test Source
Speciﬁeation Reset Ramsey, 1969
Collinearity RzlLinear Transformation Johnston 1984; Hendry and
Morgan 1989
Hetereskedascity White Test White, 1980
Residual autocorrelation Durbin-Watson Durbin-Waltggrlr, 1950 and

 

 

 

 

 

This initial model9 includes all of the variables reﬂected in my hypotheses. A reduction process
is followed which eliminates extraneous variables from the equation, thus reﬂecting a better ﬁt.

The following are the results:

 

8 The following chart is adapted from article written by James Granato (1991).

9 For purposes of all models detailed in Chapter 4, the following deﬁnitions apply:

pct65_95 = Percent of population of persons 65+ in the state in 1995 (1990 census);
pctc9520 = Projected percent change of persons 65+ from 1995-2020 (1990 census);
percapin=Percapitaincomeofallpersonsinthestate (1990 census);
popsizeg = Population size of state (1990 census);
ﬁscalre = State general fund appropriation in 1995 (NASBO Report);

(Continued Next Page)

115

Table 4-Initial Model

 

regesixbrpdﬁS_%p®9520WMepqﬁmgwmdchrgwbnhgmﬁﬁdagsnmm20_l

 

 

 

 

 

 

 

 

Source | SS df MS Number of obs = 3510
F( 11, 23) = 4.22

Model | 3.94819414 11 .35892674 Prob > F = 0.0017

Residual | 1.95536026 23 .085015664 R-squared = 0.6688
Adj R-squared = 0.5104

Total | 5.9035544 34 .173633953 Root MSE = .29157

index | Coef. Std. Err. t P>|t| [95% Conf. Interval]
pct65_95 | -.0051062 .0357007 -0. 143 0.888 -.0789587 .0687463
pctc9520 | -.0039427 .0025663 -1 .536 0.138 -.0092515 .001366
percapin | .0000466 .0000258 1.804 0.084 -6.82e-06 .0001
ﬁscalre | -.0000156 .000011 -1.418 0.170 -.0000383 7.15e-06
popsizeg | -.029658 .039092 -0.759 0.456 -.1 105258 .0512099
aarppct | .0133107 .0104879 1.269 0.217 -.0083853 .0350066
derngov | -.0337577 .0136543 -2.472 0.021 -.0620039 -.0055116
dernleg | .0050758 .0120298 0.422 0.677 -.0198098 .0299614
uniﬁed | .0314877 .0151939 2.072 0.050 .0000566 .0629188
orgstruc | -. 1040456 .0422042 -2.465 0.022 -.1913516 -.0167396
var20_l | .2942613 .1025482 2.869 0.009 .0821243 .5063984
_cons | .5379964 .8953606 0.601 0.554 -1 .3 14198 2.390191

(Continued from Previous Page)

derngov = Years of Democratic control of executive branch( Book of States)

demleg = Years of Democratic control of state legislature( Book of States)

uniﬁed = uniﬁed goverrnnent—executive and legislative branchest Book of States)
aarppct = Percentage of eligible 50+ persons belonging to AARP(AARP publication)
orgstruc = Bureaucratic structure (survey data);

var20_1 = Level of interagency state collaboration (survey question).

1° Observations fall from 48 states to 35 because of absence of complete
information. The states eliminated include: Nebraska, Louisiana, North Dakota,
Colorado, New Hampshire, Alaska, Utah, Montana, Rhode Island, Arkansas, Arizona, and
North C arolina.

1 16

According to the F test, the overall model is sigrniﬁcarrt at >.99 level, and approximately
51% of the overall variance in state based innovation in aging policy, as measured by the
adjusted R2 is being explained. The MSE equals .29. The next step in the analysis was to look
more closely at the individual variables, examining them for their relative signiﬁcance in
explaining and predicting aging policy innovation is states, as well as exploring the direction of
the relationship between the independent and dependent variables.

Several variables in this initial model are not signiﬁcant. These variables appear to be
extraneous and are not needed to explain or predict state aging policy irmovation. Although
theRzandMSEmightappearhighirnthismodelthesevariablescausethemodeltobe
nnisspeciﬁed thereby jeopardizing its signiﬁcance in explaining and predicting innovative
decision making regarding state aging policy. Therefore, it is necessary to recast this model to
see if it can be strengthened.

The ﬁrst set of variables examined are the demographic variables. It is hypothesized
that states with a sigrniﬁcarnt number of elderly currently or anticipated growth in elderly
between the year 1995 and 2020, will be more likely to be actively engaged in innovatively
planning and preparing for this demographic challenge of the 21 st century. However, both
variables are insigniﬁcant. Larmners and Klingman (1984) found that there was very little
relationship betweern dernograplnics and aging policy innovation and policy development. Their
ﬁndings seem to be validated irn this original model.

Both demographic variables are not in the anticipated direction. This model shows that

the number of current elderly residing in a state or the anticipated growth irn the number of

117
elderly in a state are inversely related to state aging policy innovation and development. Given
that the individual variables are insigniﬁcant, it is diﬂicult to draw a conclusion.

In this initial regression model the weaker of the demographic statistics is measured by
the percentage of .current elderly residing in a state. Since this dissertation is about the plannirng
and preparation for the aging of the baby boom population, it is arguable that the number of
current elderly nniglnt not be a relevant variable in explaining or predicting ﬁxture aging policy.
Thus, this variable is dropped from the model. The model is recast using only one
dernograplnic explanatory factor—the percentage change irn elderly population between the

years 1995 and 2020.

 

 

 

 

Table S—Model #1
regssshﬂexpac9520pampmﬁsmhepopdmgamppachngovdanhgummdmgsumm20j
Source | SS df MS Number of obs = 35
F( 10, 24) = 4.84
Model 3.94645496 10 .394645496 Prob > F = 0.0007
Residual | 1.95709944 24 .08154581 R-squared = 0.6685
Adj R-squared = 0.5304
Total | 5.9035544 34 .173633953 Root MSE = .28556
index | Coef. Std. Err. t P>|t| [95% Conf. Interval

pctc9520 | -.0037784 .0022472 -1 .681 0.106 -.0084l64 .0008597
percapin | .0000463 .0000252 1.837 0.079 -5.73e-06 .0000983
ﬁscalre | -.0000152 .0000104 -1 .462 0.157 -.0000365 6.24e-06
popsizeg | -.0280889 .0367475 —0.764 0.452 -. 103932 .0477542
aarppct | .0134078 .0102501 1.308 0.203 -.0077474 .034563
demgov | -.033401 .0131478 -2.540 0.018 -.0605369 -.0062652
dernleg | .005136 .0117746 0.436 0.667 -.0191655 .0294375
uniﬁed | .0311651 .0147157 2.118 0.045 .0007933 .0615369
orgstruc | .1038468 .0413115 -2.514 0.019 .1891095 -.018584
var20_1 | .296127 .0996178 2.973 0.007 .0905259 .5017281
cons l .4462768 .6119719 0.729 0.473 .8167711 1.709325

 

 

 

118
This appears to be a slightly stronger model. The overall signiﬁcance of the model
remains at > .99; and the adjusted R2 rises from .51 to .53 with one less variable, but the
MSE decreases to .28556. However, the individual demographic statistic is only
signiﬁcant at the .90 level. This measure is unacceptable, if I am attempting to make any
causal inferences regarding the sigrniﬁcance of this explanatory variable for state aging
policy innovation.

However, there are additional weak variables in the model that might cause it to be
nnisspeciﬁed. Possibly, by excluding other irrelevant variables, the explanatory power of
this demographic measure rrnight increase. The second set of variables examined are
socioeconomic variables. In Model #1, the weakest socioeconomic variable is population
size. This variable is considered primarily because of the ﬁndings from Fabricarnt (1952);
Walker (1969); Savage (1978); and Gray, in Dodd and Jilson (1994). Both of these
previous studies dealt with the diffusion of innovation and not with internal innovative
decision-making eﬂ'orts. Perhaps big states have a greater number of staﬂ‘ to send to
national or regional conferences to gather new information, making size of state a

‘6 ' ,3

relatively sigrniﬁcant predictor of innovation diﬂ‘usion. However, srze appears
insigrniﬁcant when considering innovative decision-making and long range planning
underway within a state.

The model, excluding the measure of population size, is estimated again with the

following results:

119

Table 6—Model #2

 

regress index pctc9520 percapin ﬁscalre aarppct demgov dennleg uniﬁed orgstruc var20_1

 

 

 

 

 

Source | SS df MS Number of obs = 35
F( 9, 25) = 5.40
Model | 3.89881009 9 .433201121 Prob > F = 0.0004
Residual | 2.00474432 25 .080189773 R-squared = 0.6604
Adj R-squared = 0.5382
Total | 5.9035544 34 .173633953 Root MSE = .28318
index | Coef. Std. Err. t P>|t| [95% Conf. Interval]
pctc9520 | -.0039635 .0022155 -1.789 0.086 -.0085264 .0005994
percapin | .0000443 .0000249 1.783 0.087 -6.88e—06 .0000955
ﬁscalre | -9.50e-06 7.20e-06 -l .320 0.199 -.0000243 5 .33e-06
aarppct | .014451 .010074 1.434 0.164 -.0062968 .0351989
demgov | -.0340088 .0130142 -2.613 0.015 -.0608121 -.0072056
demleg | .0087125 .0107148 0.813 0.424 -.013355 .03078
uniﬁed | .031213 .0145927 2.139 0.042 .0011587 .0612673
orgstruc | -.0963829 .0398057 -2.421 0.023 -.1783643 -.0144014
var20_1 | .3102205 .0970793 3.196 0.004 .1102819 .510159
cons I .19906 .5151882 0.386 0.702 -.86199 1.26011

 

 

Once again the overall model is sigrniﬁcant at > .99 level, and the adjusted R2 of .53 does
not decrease when excluding this variable, nor is there sigrniﬁcant decrease of MSE at
.28318. Thus, it is reasonable to suggest that this variable was extraneous that adds
nothing to the predictive power of the model, therefore it is dropped.

Also, the level of state general ﬁmd appears to be an insigrniﬁcant factor. This
ﬁnding seems to validate Nice’s (1994) ﬁnding that there is little relationship between
ﬁscal capacity and innovative policy development. Given that per capita income is an

acceptable measure of “wealth” utilized liberally in the political science literature, as well

120
as in the innovation literature, the general fund variable is deleted, and per capita income
retained.
The regression model is once again estimated with only one demographic factor

and one socioeconomic factor. The following are the results:

Table 7—Model #3

 

regress index pctc9520 percapin aarppct denngov dennleg uniﬁed orgstruc var20_1

 

 

 

 

Source | SS df MS Number of obs = 35
F( 8, 26) = 5.70
Model | 3.75915523 8 .469894404 Prob > F = 0.0003
Residual | 2.14439917 26 .082476891 R-squared = 0.6368
Adj R—squared = 0.5250
Total | 5.9035544 34 .173633953 Root MSE = .28719
index | Coef. Std. Err. t P>|t| [95% Conf. Interval]
pctc9520 | -.003963 .0022469 -1.764 0.090 -.0085815 .0006555
percapin | .0000255 .0000207 1.235 0.228 -.0000169 .000068
aarppct | .0192946 .0095145 2.028 0.053 -.0002627 .0388519
demgov | -.O325 808 .0131528 -2.477 0.020 -.0596l67 -.0055449
dennleg | .0073414 .0108153 0.679 0.503 -.0148897 .0295726
uniﬁed | .0362409 .0142861 2.537 0.018 .0068754 .0656064
orgstruc | -.0825194 .0389382 -2.119 0.044 -. 162558 -.0024808
var20_1 | .2979915 .0980044 3.041 0.005 .0965405 .4994425
cons I .2577451 .5205335 0.495 0.625 -.8122268 1.327717

 

 

 

The per capita irncome variable remains insigniﬁcant in this newly deﬁned model. My
hypotheses state that aging policy irmovation is inﬂuenced by socioeconomic factors. The per
capita income variable is the strongest and seemingly most reliable of the socioeconomic
factors. Before eliminating all socioeconomic factors from the model, it is critical to examine

the other variables iii the model. It is plausrble that there nniglrt be a multicollinearity problem

121
between this socioeconomic factor and some other variable in the model thus irnﬂuencing the
explanatory strength of this factor.

In examining the political variables, it is quickly apparent that the weakest political
variable in the model is the measurement 0 “political liberalism” constructed as the number of
years the state legislature is controlled by Democrats. Since this dissertation is about long
range plarnning and inrnovative decision-making, it is arguable that the state legislature is more
reactive and responsive to executive irnitiated progarns and policies. Therefore, it is likely that
it would not be a signiﬁcant factor in explairning irnnovative decision-making, and the variable is
eliminated from the model.

It is feasible to suggest tlnat the executive brarnch plays a much more proactive role in
plarnning and policy development. It is not surprising to ﬁnd that the measurement 0 “political
hheralisrn” measured by number of years the governor’s oﬂice is controlled by a Democrat
proves signiﬁcant at >.95. However, it is surprising to discover that this relationship is not in
the anticipated direction. There is an irnverse relationship between Dernocratically controlled
governors’ oﬂices and long-range planning and innovative decision-making. This ﬁnding

causes me to question if this variable is a fair masure of “political liberalism.”ll

 

" In Wright, Erikson and McIver (1985) they argued that “partisanship and ideology
irn the states are not measures of the same thing” (Wright et al, 1985: 475). The correlation
between the state ideology measure developed by Wright, Erikson and McIver and the
“political liberalism” variable designed in this dissertation is .06. This suggests that this variable
is not a good measure of state political liberalism. However, Wright, Erikson and McIver were
exannirnirng electorate party identiﬁcation and irndividual ideology. Given that my measure of
political liberalism is constructed as consistent partisanship of elected elites, it is arguable that
ﬂnesetwonmuesneednﬂwnelﬂegvenﬂmﬂneyarelmldngmdiﬂ’eremaspecmﬁ
politics arnd ideology.

l 22

As stated in chapter three, the time period utilized to construct this factor is 1980-
1995. For the most part, twelve of these ﬁfteen years reﬂect a substantial level of conservatism
on the national level, and generally, a more conservative public sentiment. It is reasonable to
suggest that a state with a relatively consistently Dernocratically controlled governor’s oﬂice is
indicative of a politically hberal state. I ardently examined the data to determine if the most
consistently Dernocratically controlled governor’s ofﬁces from around the nation are only in
the South. If this measure is simply reﬂecting Southern Democrats, this factor would be
rendered useless in measuring political liberalism irn states. ‘2

All of the soutlnern states are in the highest tier of the scale (11+), which suggests that
this would account for the negative relationship between this “political hhemlisrn” variable and
the irnrnovation index However, there are also twelve non-southern states in the highest level
ofthe scale (11+). Thaefore,1suggestthatit isinconclusive ifthisisafairmeasmeofpolitical
hberalisrn, but, given these ﬁndings it does somewhat challenge tlnose of Larrnrners and
Klingnan (1984), which stated that political hberalism and political openrness were the nnq'or
determinants of aging policy innovation.

Thestatelegislatmevariableiseliminated, andthemodelisestimatedonceagairn, with

the following results:

 

‘2 See V.O. Key, Politics Parties and Presgue Groups, ﬁfth editiorn, arnd Southern
Politics, in which Key speaks to the conservative nature of the one party Democrat system in
the South and the level 0 “conservatism”—both from the standpoint of ideology and readiness
to innovate. .

 

123

Table 8—Model #4

 

regress index pctc9520 percapin aarppct demgov uniﬁed orgstruc var20_1

 

 

 

Source | SS df MS Number of obs = 35
F( 7, 27) = 6.58
Model | 3.72115229 7 .531593184 Prob > F = 0.0001
Residual | 2.1824021] 27 .080829708 R-squared = 0.6303
Adj R-squared = 0.5345
Total | 5.9035544 34 .173633953 Root MSE = .28431
index | Coef. Std. Err. t P>|t| [95% Conf. Interval]

 

pctc9520| -.0035675 .0021482 -l.66l 0.108 -.0079753 .0008403
percapin | .0000265 .0000204 1.299 0.205 -.0000153 .0000684

aarppct | .018115 .0092606 1.956 0.061 -.0008861 .0371162
demgov | -.0344449 .0127338 -2.705 0.012 -.0605725 -.0083 173
uniﬁed | .0393893 .0133766 2.945 0.007 .0119429 .0668358

orgstruc | -.0772658 .0377783 -2.045 0.051 -. 1547804 .0002488
var20_l | .2668735 .0857524 3.1 12 0.004 .0909241 .4428229
_cons | .4096665 .4652461 0.881 0.386 -.5449397 1.364273

 

 

 

In elinninating this variable the F-test is still signiﬁcant > .99, and the adjusted R2 for the
model increases with dropping this variable, and there is no real difference in the MSE. The t-
scores for two of the political variables—political hheralisrn and urniﬁed govemment—are
signiﬁcarnt. The relationship between uniﬁed government and innovative decision-making is in
the anticipated positive direction. This suggests tlnat a uniﬁed political base between the
executive and legislative branches assists in the irmovative decision-making in states and their

long range planning irn preparing for the aging of the baby boomers.

124

The last political variable in the model is a measure of interest group inﬂuence on the
innovation index. In Model #4, the percentage of eligble members in the state who belong to
AARP is not signiﬁcant at > .95 (p > .94). According to normal standards (.95 signiﬁcance
level) this variable could be dropped. However, given my hypotheses, it is necessary to test the
model further to determine if this variable should be maintained.

There is relatively moderate correlation (. 53) between the per capita income variable
and the percentage of AARP membership. Thus, in testing the model, per capita income is

dropped ﬁom the model, with the following results:

Table 9—Model #5

 

regress index pctc9520 aarppct demgov uniﬁed orgstruc var20_1

 

 

 

 

Source | SS df MS Number of obs = 35
F( 6, 28)= 7.21
Model | 3.58465849 6 .597443082 Prob > F = 0.0001
Residual | 2.3188959] 28 .082817711 R-squared = 0.6072
Adj R-squared = 0.5230
Total | 5.9035544 34 .173633953 Root MSE = .28778
index | Coef. Std. Err. t P>|t [95% Conf. Interval]
pctc9520 | -.0034132 .0021712 -1.572 0.127 -.0078607 .0010342
aarppct I .0241143 .0081258 2.968 0.006 .0074693 .0407594
demgov | -.034l602 .0128875 -2.651 0.013 -.0605591 -.0077612
uniﬁed | .0416656 .0134235 3.104 0.004 .0141689 .0691623
orgstruc | -.0708409 .03791 11 -l .869 0.072 -. 1484983 .0068165
var20 1 | .2459242 .0852528 2.885 0.007 .0712917 .4205567
I

.6619107 .4279857 1.547 0.133 -.2147784 1.5386

 

 

 

125
This shows that the overall model is still signiﬁcarnt > .99, and the AARP variable becomes
signiﬁcant at > .99. The adjusted R2 is slightly reduced, but the MSE increases to .28778.
However, the demographic variable is still not signiﬁcant, and this model does not account for
any inﬂuence from socioeconomic factors, which is a part of my hypotheses. The model is
estimated one more time, elinninating the AARP variable and maintaining the per capita income

variable, with the following results:

Table lO—Model #6

 

regress index pctc95 20 percapin demgov uniﬁed orgstruc var20_1

 

 

 

Source I SS df MS Number of obs = 35
F( 6, 28) = 6.39
Model | 3.41185674 6 .56864279 Prob > F = 0.0003
Residual | 2.49169766 28 .088989202 R-squared = 0.5779
Adj R-squared = 0.4875
Total I 5.9035544 34 .173633953 Root MSE = .2983]
index I Coef. Std. Err. t P>It| [95% Conf. Interval]

 

pctc9520| -.004497 .0021982 -2.046 0.050 -.0089999 5.8le-06

percapin | .0000464 .0000186 2.501 0.019 8.39e-06 .0000844
demgov | -.027422 .012819 -2. 139 0.041 -.0536806 -.0011634
uniﬁed .0289331 .0128662 2.249 0.033 .0025779 .0552883

|
orgstruc | -.0922425 .0388167 -2.376 0.025 -.1717548 -.0127301
var20_l I .318701 .0855743 3.724 0.001 .1434101 .4939919
cons I .8216884 .4352712 1.888 0.069 -.0699243 1.713301

 

 

 

126

By elinninating the AARP variable, all of the other variables in the model become
signiﬁcant, and all of the factors anticipated as relevarnt in my hypotheses are irncluded. The F-
test is > .99, the adjusted R2 falls to .48, but the MSE irncreases to .2983]. Although the R2
decreases, it is arguable that this is a much stronger model and a better “goodness of ﬁt,” given
that each of the explanatory factors are signiﬁcant, the F-test > .99, and the MSE increases.

It is now possrble to draw some conclusions regarding the impact of demogaphics on
aging policy innovation and plarmirng. There is an irnverse relationship between the anticipated
percerntage increase change in elderly population irn a state and innovative decision-making and
planning underway to prepare for the aging of the baby boomers. In essence, those states
which will undergo the most drastic increases in elderly populatiorn, are being least irnrnovative in
plannirng and aging policy development. This is connpletely opposite the arnticipated
relationship stated in the hypotheses.

Finally, irn reviewing the orgarnizational factors, both are signiﬁcant, with collaboration
> .99, and in the anticipated direction. However, the bureaucratic structure variable appws to
be the reverse of the anticipated relationship stated in the hypotheses. The hypothesis suggests
that in the states where there is a cabinet department of elder aﬂ’airs, then the aging agenda is
higher on the governor’s agenda. Also, it is proposed that if there is a separate department of
elder affairs, then the “aging issue” has more visibility. It is possrble that planning for the 21 st
century would be more irnnovative if there is a departrnernt of elder affairs than if aging issues
are handled by an oﬂice witlnin the governor’s oﬂice; is a separate omce, but without cabinet

status; or is a division or bureau under a departrnernt of social services or human services.

127

Thisﬁndingindicatestlnattlnishypothesisisirncorrect,andthatthereisaninverse
relationship between bureaucratic structure and innovative decision-making processes. Two
possible explanations for this inverse relationship is that if a state has a cabinet level department
of elder aﬁ’airs13 then it more than likely has a substantial number of elderly currently. It is
possible that such departrnernts are invested in serving their current elderly constituents arnd,
thus, have not been concerned with the future elderly. Also, perhaps, when there is a single
department of elder aﬂ’airs, then all of the aging policy development function is delegated to
that department, with a very professional bureaucracy serving the needs of elders. This kind of
ﬁnding based on this second rationale, is sinrilar to that of Peterson et a1. (1986), irn wlnich they
found that a professionalized bureaucracy was less likely to change and/or adopt reforms.
Regardless, this ﬁnding supports the hypothesis that collaboration and cooperative work
enviromnents irnstill more innovative decision-making and long range plarnning.

Model #6 appears to be the best estimator for OLS.l4 However, it is now critical to

validate this model to determine if it is BLUE. As stated previously, there are a variety of

 

’3 There are ten states which have department level status for aging/elder affairs at the
time of the survey: Alabama, Florida, Illinois, Kansas, Maryland, Massachusetts, Nebraska,
Olnio, Pennsylvania, and Rhode Island.

1" The construction of the dependent variable was also tested in this model. As
discussed in length inn chapter tlnee, this deperndent variable is constructed as an irnrnovation
irndex utilizing three separate questions from the survey. It rs suggested that this "innovation
index" is measuring the three components of rrmovative decision-making processes—capacity,
policy communities/planning, and innovative strategies. The correlation between the
”innovation index" arnd the three questions is obviously signiﬁcarnt, given that the irndex was
constructed from the questions.

(Continued Next Page)

128

 

 

 

 

 

 

(Continued from Previous Page)
Correlations for Index
INDEX Innovative Planning Capacity
INDEX 1.0000
Innovative 0.8055 1.0000
Planning 0.7265 0.4451 1.0000
Capacity 0.6470 0.3055 0.1336 1.0000

 

 

 

 

 

 

 

Each question is estimated separately as the dependent variable to deterrrrine if arny single
question would be a stronger measure of innovation and reﬂect a better "goodness of ﬁt" given
the independent factors. The "index" appeals to have the best "ﬁt".

Question 1: irmovative strategies:

regess var3 1_1 pctc9520 percapin demgov urniﬁed orgstruc var20_1

 

 

 

 

Source I SS df MS Number of obs = 35
F( 6, 28) = 6.70
Model I 6.32714979 6 1.05452497 Prob > F = 0.0002
Residual | 4.40685023 28 .157387508 R-squared = 0.5894
Adj R—squared = 0.5015
Total I 10.734 34 .315705883 Root MSE = .39672
var31_1 I Coef. Std. Err. t P>ItI [95% Conf. Interval]
pctc9520 | -.0093406 .0029234 -3.195 0.003 -.0153289 -.0033523
percapin I .0000317 .0000247 1.286 0.209 -.0000188 .0000823
demgov I -.0242271 .0170479 -1.421 0.166 -.0591482 .0106941
uniﬁed I .0499954 .0171106 2.922 0.007 .0149458 .0850449
orgstruc | -. 1477307 .051622 -2.862 0.008 -.2534735 -.0419878
var20_l I .352047 .1138045 3.093 0.004 .118929 .585165
_cons I 1.37827 .5788638 2.381 0.024 .1925211 2.564019

 

(Continued Next Page)

129

 

(Continued from Previous Page)
Question 2: planning underway:

regress var34_1 pctc9520 percapin demgov uniﬁed orgstruc var20_1

 

 

 

 

Source I SS df MS Number of obs = 35
F( 6, 28) = 4.00
Model I 4.75685322 6 .79280887 Prob > F = 0.0052
Residual I 5.55267064 28 .198309666 R—squared = 0.4614
Adj R-squared = 0.3460
Total I 10.3095239 34 .30322129 Root MSE = .44532
var34_1 | Coef. Std. Err. t P>It| [95% Conf. Interval]
pctc9520 I .0015653 .0032815 0.477 0.637 -.0051566 .0082872
percapin I .0000479 .0000277 1.730 0.095 -8.81e—06 .0001047
demgov I -.0556743 .0191363 -2.909 0.007 -.0948733 -.0164753
uniﬁed I .0362248 .0192067 1.886 0.070 -.0031183 .075568
orgstruc | .0427677 .0579458 0.738 0.467 -.0759288 .1614642
var20_1 I .4113886 .1277457 3.220 0.003 .1497134 .6730639
_cons I .2862711 .6497753 0.441 0.663 -1.044733 1.617276

 

Question 3: policy capacity:

regress gen3a_1 pctc9520 percapin demgov uniﬁed orgstruc var20_1

 

Source I SS df MS Number of obs = 35
+ F( 6, 28) = 2.25
Model I 3.40605894 6 .56767649 Prob > F = 0.0674

Residual I 7.06050854

 

Total I 10.4665675

28 .252161019

34 .30784022

R-squared = 0.3254
Adj R-squared = 0.1809
Root MSE = .50216

 

 

gen3a_l I Coef. Std. Err. t P>It| [95% Conf. Interval]
pctc9520| -.005739 .0037003 -l.551 0.132 -.0133l87 .0018408
percapin I .0000604 .0000312 1.934 0.063 -3.57e-06 .0001244
demgov | -.0022609 .0215787 -0.105 0.917 -.0464629 .0419411
uniﬁed I .0009501 .0216581 0.044 0.965 -.0434144 .0453147
orgstruc I -.1717628 .0653414 -2.629 0.014 -.3056087 -.037917
var20_1 I .1931874 .1440501 1.341 0.191 -.1018858 .4882606
_cons I .7830445 .7327069 1.069 0.294 -.7178374 2.283926

130
diagnostic tests to perform which will assist in making this determination, and possibly result in

a better theoretical and empirical model.

Validating The Model

The ﬁrst basic assumption of regession analysis is that the model is speciﬁed correctly.
Speciﬁcation relates to three conditions: (1) the model is of tlne correct functional form; (2) the
error terms in the model are normally distributed; and (3) that the model does not suﬂ‘er ﬁem
omitted variable bias. With a visual inspection of scatterplots of the relationslnip between the
independent and dependent variables, X and Y, the researcher can quickly determine if the

relationship is linear. See Figures 1 la-f below:

coef- -.CIJ449703, se - 00219821, 1 = -2.05

 

 

 

595783 ‘ °
0
Q
X
(D —l
U
E
75 o
— o
o
O
0
-.721162 °
I I l I
-43.‘i485 44.1654

e( pctc9520 IX)

Figure 1 1a- Plot of Relationslnip Between Independernt Variable “pctc9520” and Index

BI, Index I X)

-.632077 ‘i
1
6464.51

“Indexln

564898 "

131

coef- .ClIIMBll, se = .00001855,1= 2.5

 

 

7333.7'4
e( percapin IX)

Figure 1 1b- Plot of Relationslnip Between Irndependernt Variable “percapin” arnd Irndex

chef = —.02742199, 52 = 111231903, t = -2.:l«4

.716432 "I

 

-.7004 ‘

 

 

~asiaza

seeds
at dunnov I x )

Figure llc- Plot of Relationslnip Betweern Independent Variable “demgov” and Index

BI Index I X)

132

coef= 0289331, so = .01288618,1= 2.25
.784101 " °

 

-.722801 ‘ °

8( Index I X)

 

 

I I l
e( uniﬁed IX)

-1 1 .1236 8.527i8

Figure 11d- Plot of Relationslnip Between Independent Variable ‘uniﬁ ” arnd Index

coef= -.09224248, se 2 03.1868, 1 = -2.38
.664255 "‘ o

 

 

-.780281 ‘

0
2215251 ' ' ' 225515

a( orgstruc | X)

Figure lle— Plot of Relationslnip Between Independent Variable “orgstruc” arnd Index

133

coef- 31870393, so - 08557426, t - 3.72

at Index I X)

 

 

-.795501 " °
1

-1 78688 ' j

e( var20_1 I X)

I
1.“

Figure llf- Plot of Relationship Betweern Independent Variable ‘Var20_1” and Index

A visual inspection of a lnistogarrn, boxplot or quantile-normal plot will indicate if the

residuals are normally distributed. See Figure 12, 13 and 14.

Fraction

 

 

 

I ' ' .7179dr

miduﬂ

Figure lZ—Histogam of residuals

134

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.717981 0
E
3
u _
5
an
I.
-.605845 ‘ °
1 I
-513231 I 518281
Inverse Normal
Figure l3—Quantile Norm of Residuals
9 residual
.717981 1
..805845 ‘

 

Figure l4—Boxplot of Residuals

 

13 5
One of the most common speciﬁcation errors is a model with onnitted variables.

Onnitted variables particularly damage causal interpretations and can result in the relationship
between X and Y to be substantially overstated or understated. “When a relevant variable is
omitted and it is correlated with one of the variables in the model, the residual of the
nnisspeciﬁed model picks up the omitted variables inﬂuence” (Granato, 1991:131). One way to
test for this speciﬁcation error is the RESET test (Ramsey, 1969).” The I~Io = model has no
omitted variables. Using “ovtest” fnmction on STATA, the Ramsey RESET test is run with the
following results:

using ﬁtted values of index

F(3, 25) = 0.57

Prob > F = 0.6431
Giventhesemnutslcanacceptﬂnenuﬂhypothesisthattherearenoonuued variablesinthe
model. Also, due to the visual review of the graphs and plots, I can assume that the regession

model #6 is speciﬁed correctly.

 

1’ Ramsey has proposed a general test of speciﬁcation error called RESET
(regression speciﬁcation error test). RESET tests are used to test whether unknown
variables have been onnitted from a regression speciﬁcation, and are not to be confused
with OV tests that test for zero coefﬁcients on krnown variables. They can also be used to
detect a nnisspeciﬁed functional form. Although the RESET test was designed to be used
to test for nnissing regressors, it turns out to be powerful for detecting nonlinearities. This
weakens its overall attractiveness, since rejection of a model could be due to either a
nornlinearity or an omitted explanatory variable. (No test can discrinninate between
unknown omitted variables and unknown ﬁlnctional fornn; a strong case can be made that
the RESET test can only test for ﬁnnctional form.) (Kennedy, 1992).

The next diagnostic test to nm is to check the explarnatory variables for evidence of

136

multicollinearity. The most commonly used procedure to detect collinearity is an examination

of the correlation matrix (Granato, 19912132).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table ll-Correlation Matrix
Pctc9520 percapin demgov uniﬁed orgstruc var20_1
pctc9520 1.000
pereapin 00235 1.000
demgov 0.2664 0.1628 1.000
uniﬁed 0.1014 —0.0327 0.3410 1.000
orgstruc -0. 1218 0.0467 -0.0120 -0.0963 1.000
var20_1 -0.0230 -0.0592 -0.0730 0.0756 0.0645 1.000

 

In reviewirng this table, in does not appear that there is a multicollinearity problem However, a

more rigorous test for multicollinearity is to regess each of the indeperndent variables on the

remaining independent variables. If the R2 is lnigher irn any of these “restricted models” as

compared with the original model", then there is evidence of nullticollinearity and the model

should be adjusted.

 

‘6 For the purposes of this dissertation, the original model being referred to in this
statement is model #6.

 

137

Table ll—Validity Test for Multicollinearity

 

regress pctc9520 percapin demgov uniﬁed orgstruc var20_1

 

 

 

 

 

 

 

 

 

 

Source I SS df MS Number of obs = 35
F ( 5, 29) = 0.57
Model I 1797.24104 5 359.448209 Prob > F = 0.7252
Residual I 18416.0904 29 635.037599 R-squared = 0.0889
Adj R-squared = -0.0682
Total | 20213.3314 34 594.509747 Root MSE = 25.20
pctc9520 | Coef. Std. Err. t P>ItI [95% Conf. Interval]
percapin I -.0005491 .0015641 0351 0.728 -.003748 .0026497
demgov | 1.506604 1.046131 1.440 0.161 -.6329745 3.646183
uniﬁed | -.0378712 1.086855 —0.035 0.972 -2.260739 2.184997
orgstruc I -2. 1 17134 3.255407 -0.650 0.521 -8.77519 4.540921
var20_1 I .0633217 7.228923 0.009 0.993 -14.72149 14.84813
_cons I 59.22424 35.08658 1.688 0.102 -12.53586 130.9844
regress percapin pctc9520 demgov uniﬁed orgstruc varZO_l
Source I SS df MS Number of obs = 35
F( 5, 29) = 0.26
Model I 115077999 5 230155997 Prob > F = 0.9321
Residual I 258497074 29 891369220 R-squared = 0.0426
Adj R-squared = -0. 1224
Total | 270004874 34 794131981 Root MSE = 2985.6
percapin | Coef. Std. Err. t P>ItI [95% Conf. Interval]
pctc9520 I -7.707957 21.95377 0351 0.728 -52.60845 37.19253
demgov I 131.0565 125.9674 1.040 0.307 -126.5756 388.6887
uniﬁed -59.56966 128.2925 -0.464 0.646 -321.9574 202.818
orgstruc I 73.54359 388.2487 0.189 0.851 -720.5142 867.6013
var20_lI -190.0193 855.7253 -0.222 0.826 -1940. 174 1560.136
_cons I 19513.18 2418.212 8.069 0.000 14567.39 24458.98

 

 

138

Table 12 (cont’d)

 

regress demgov percapin pctc9520 uniﬁed orgstruc var20_1

 

 

 

 

 

 

 

 

 

 

Source I SS df MS Number of obs = 35
F( 5, 29) = 1.55
Model | 145.150547 5 29.0301095 Prob > F = 0.2042
Residual I 541.535167 29 18.6736264 R-squared = 0.2114
Adj R-squared = 0.0754
Total | 686.685714 34 20.1966387 Root MSE = 4.3213
demgov I Coef. Std. Err. P>ItI [95% Conf. Interval]
percapin | .0002746 .0002639 1.040 0.307 -.0002652 .0008143
pctc9520 I .0443025 .0307621 1.440 0.161 -.018613 .107218
uniﬁed I .3487433 .1747656 1.995 0.055 -.0086924 .706179
orgstruc I .1572337 .5615362 0.280 0.781 -.9912368 1.305704
var20_1 I -.6354828 1.233991 -0.515 0.610 -3. 159278 1.888312
_cons I 1.7376 6.297037 0.276 0.785 -11.14129 14.61649
regress uniﬁed demgov percapin pctc9520 orgstruc var20_1
Source I SS df MS Number of obs = 35
F( 5, 29) = 0.96
Model I 893976625 5 17.8795325 Prob > F = 0.4556
Residual | 537.573766 29 18.5370264 R-squared = 0.1426
Adj R-squared = 00052
Total | 626.971429 34 18.440336] Root MSE = 4.3055
uniﬁed I Coef. Std. Err. t P>ItI [95% Conf. Interval]
demgov | .3461922 .1734871 1.995 0.055 -.0086288 .7010132
percapin I -.0001239 .0002668 -0.464 0.646 -.0006695 .0004218
pctc9520 I -.0011055 .0317258 0035 0.972 -.065992 .063781
orgstruc I -.3065213 .5573354 0550 0.587 -1.4464 .8333574
var20_1 I .7331545 1.227552 0.597 0.555 -1.777471 3.24378
_cons I -.46l9694 6.281608 -0.074 0.942 43.3093 12.38536

 

 

139

Table 12 (cont’d)

 

regress orgstruc uniﬁed demgov percapin pctc9520 var20_1

 

 

 

 

Source I SS df MS Number of obs = 35
F( 5, 29)= 0.19
Model I 1.91045696 5 .382091391 Prob > F = 0.9650
Residual I 590609716 29 2.03658523 R-squared = 0.0313
Adj R-squared = 01357
Total I 60.9714286 34 1.79327731 Root MSE = 1.4271
orgstruc I Coef. Std. Err. t P>ItI [95% Conf. Interval]
uniﬁed I -.0336762 .0612321 0550 0.587 -.1589099 .0915575
demgov | .0171482 .0612423 0.280 0.781 -. 1081064 .1424029
percapin I .0000168 .0000887 0.189 0.851 -.0001646 .0001982
pctc9520 I -.0067897 .0104402 -0.650 0.521 -.0281423 .0145629
var20_1 I .168512 .4081815 0.413 0.683 -.6663129 1.003337
cons I 1.446268 2.064905 0.700 0.489 -2.776937 5.669472

 

 

regress var20_1 orgstruc uniﬁed demgov percapin pctc9520

 

 

 

 

Source I SS df MS Number of obs = 35
F( 5, 29) = 0.14
Model I .298872511 5 .059774502 Prob > F = 0.9807
Residual I 12.1520917 29 .419037645 R-squared = 0.0240
Adj R-squared = 01443
Total I 12.4509642 34 .36620483 Root MSE = .64733
var20_1 I Coef. Std. Err. t P>ItI [95% Conf. Interval]
orgstruc I .0346722 .0839854 0.413 0.683 -. 1370972 .2064416
uniﬁed I .0165733 .0277494 0.597 0.555 -.0401805 .0733271
demgov I -.0142603 .0276909 -0.515 0.610 -.0708944 .0423739
percapin I -8.93e-06 .0000402 -0.222 0.826 -.0000912 .0000733
pctc9520 I .0000418 .0047701 0.009 0.993 -.0097142 .0097977
_cons I 2.119833 .858598 2.469 0.020 .3638025 3.875863

 

 

140
In reviewing the st of these “restricted models”, it is clear that there is no evidence of
multicollinearity. The model need not be adjusted for this reason.

The third test to be performed on this regression model is for heteroskedascity.
Heteroskedascity is a violation of one of the basic assumptions of OLS, and indicates that
the variance of the disturbance terms is not constant. The points in a regression are
suppose to “snuggle in a band of equal width above and below the regession line” (Lewis-
Beck, 1980228). Evaluating if the model suffers from heteroskedascity, the researchér can
do a visual inspection of the plot to determine if the points tend to fan in or out, thus
indicating heteroskedascity. In reviewing Figure 12, it does not appear that this model
suffers from heteroskedascity. However, an additional diagnostic test can be run to
determine if the problem exists.

Using “hettest” function of STATA, the White test” is performed on the
model. The White test uses a chi-square distribution. The H0 = Constant variance. The
following are the results:

Cook-Weisberg test for heteroskedascity using ﬁtted values of index:

chi2(1) = 3.00
Prob>chi2 = 0.0832

 

‘7 Unlike the Goldfeld-Quandt test, which requires reordering the observations
with respect to the X variable that supposedly caused heteroskedasticity, or the BGP test,
which is sensitive to the normality assumption, the general test of heteroskedasticity
proposed by White does not rely on the normality assumption and is easy to implement.
This test examines whether the error variance is affected by any of the regressors, their
squares or their cross-products. The strength of this test is that it tests speciﬁcally for
whether or not any heteroskedasticity present causes the variance-covariance matrix of the
OLS estimator to differ ﬁ'om its usual formula. (Gujarati, 1995; Kennedy, 1992).

141

Given that the p > .05, I can accept the null hypothesis that the error terms have constant
variance, and therefore determine that the model does not suffer from heteroskedascity.

Residual autocorrelation oﬁen plagues time series data. The causes of serial
correlation can be attributed to the result of a random shock, which has continuing
inﬂuence, or inertia, reﬂecting a slow response time to policy changes. Ifthere is evidence
of serial correlation, then the model is inefﬁcient. The data used for this dissertation is not
time series data, but cross sectional data. Therefore the check of the data is not one for
serial correlation, which reﬂects this time factor, but of spacial correlation. Spacial
correlation means that one grouping of data points are affected by another. The Durbin-
Watson statistic18 can be used to determine ﬁrst order serial correlation or spacial
correlation.

Prior to getting a Durbin-Watson statistic for the regression model, it is necessary
to regroup the data according to region. The data is currently sorted in alphabetical order,

and therefore if it were not sorted appropriately by region, then the Durbin-Watson

 

‘8 TheDurbin-Watsonstaﬁsﬁcisastaﬁsﬁcaltestofﬂnermﬂhypoﬂnesisﬂnatﬂnesuccessive
error terms are uncorrelated, that r=o. Iftlne serial correlation parameter r=o then d=2, and we
canbeassuredtlnatthereisno serial or spacial correlation. Theﬁlrtherawaydisﬁ'omZ, then
the less conﬁdernt we can be that our model corntains no ﬁrst order correlation. Ifd>2 then you
have negative correlation. If (1 < 2 then you have positive correlation There is an upper and
lower barnd for the d statistic, with the critical values calculated and irnterpreted depending on
sample size. “Tlms, for a given data set and model to be estimated, ifthe value oftlne Durbin-
Watson statistic is greater than this upper bound for a speciﬁed conﬁdence level, we shall not
reject the null hypothesis of no serial correlation. Likewise, if the value of the d-statistic is less
thentlnelowerbound, weshallrejectthishypothesisandproceedtousegeneralizedleast
squares. Ifthevaluefallsbetweenthelowerandupperbounds, weareuncertairnwhetherto
accept or reject the null hypothesis” (I-Ianushek and Jackson, 1977: 165).

142
statistic would simply be measuring the relationship between states that begin with the
letter “A” as compared to those that begin with “B”, etc. The data is sorted in accordance
with the eight census bureau regions, and then the model is tested for spacial correlation.
The Durbin-Watson statistic is 2.31640, which is greater than the upper bound of the d-
statistic (1.77), and close to “2.”19 Thus, I can accept the null hypothesis that the data
does not suffer from spacial correlation.

Seemingly, Model #6 has been validated as “BEST. ” However, it is important to
do one ﬁnal evaluation on the model to review the implications of outliers. Outliers aﬂ’ect
OLS slopes, standard errors, hypothesis tests, R2, and other statistics. OLS is not robust
in that a single case can have an arbitrarily large impact on sample estimates. Robust
regression is designed to perform well under a broader range of conditions than OLS.
Robust and OLS regressions complement each other in that discrepancies between OLS
and robust results reveal the effects of outliers and warn that OLS may be untrustworthy
(Hamilton, 1992: 200). OLS is simpler and preferable to robust regression, if both models
produce the same results. Coefﬁcients of OLS and robust regression are evaluated to
check whether any of the OLS coefﬁcients are more than one (robust) standard error from
the corresponding robust coefﬁcient (Hamilton, 1992: 200). Robust ﬁndings can be used
to conﬁrm the validity of OLS. The following table shows the comparison between OLS

and Robust regression

 

’9 In accordance to the critical value table for the Durbin-Watson Test for
Autocorrelmion P: .05, (Harvey, 1991:362) the upper bound d statistic is 1.80, given sarrnple
size of35 and six degrees offreedorn

143

Table l3--Comparison of OLS and Robust Regression

 

 

 

 

 

 

 

 

OLS Regression
ﬁt index pctc9520 percapin demgov uniﬁed orgstruc var20_1
Source I SS df MS Nmnber of obs = 35
F( 6, 28) = 6.39
Model I 3.41185674 6 .56864279 Prob > F = 0.0003
Residual I 2.49169766 28 .088989202 R—squared = 0.5779
Adj R—squared = 0.4875
Total I 5.9035544 34 .173633953 Root MSE = .29831
index I Coef. Std. Err. t P>ItI [95% Conf. Interval]
pctc9520 I -.004497 .0021982 —2.046 0.050 -.0089999 5.81e—06
percapin | .0000464 .0000186 2.501 0.019 8.39e-06 .0000844
demgov I -.027422 .012819 -2.139 0.041 -.0536806 -.0011634
uniﬁed I , .0289331 .0128662 2.249 0.033 .0025779 .0552883
orgstruc | -.0922425 .0388167 -2.376 0.025 -. 1717548 -.0127301
var20_1 I .318701 .0855743 3.724 0.001 .1434101 .4939919
_cons I .8216884 .4352712 1.888 0.069 -.0699243 1.713301
Robust Regression
rreg index pctc9520 percapin demgov uniﬁed orgstruc var20_1
Huber iteration 1: maximum difference in weights = .45270417
Huber iteration 2: maximum difference in weights = .12480343
Huber iteration 3: maximum difference in weights = .11172285
Huber iteration 4: maximum difference in weights = .01895246
Biweight iteration 5: maximum difference in weights = .16872751
Biweight iteration 6: maximum difference in weights = .02826756
Biweight iteration 7: maximum difference in weights = .02363256
Biweight iteration 8: maximum diﬁ’erence in weights = .008426
Robust regression estimates Number of obs = 35
F( 6, 28) = 6.69
Prob > F = 0.0002
index I Coef. Std. Err. t P>ItI [95% Conf. Interval]
pctc9520 I -.0049412 .0022466 -2. 199 0.036 -.0095432 -.0003392
percapin | .000058 .000019 3.056 0.005 .0000191 .0000968
demgov I -.0285632 .0131013 -2. 180 0.038 -.0554 -.0017264
uniﬁed I .0229314 .0131495 1.744 0.092 -.0040041 .0498669
orgstruc I -.09l3406 .0396714 -2.302 0.029 -. 1726039 -.0100774
varZO_1 I .3297427 .0874586 3.770 0.001 .1505919 .5088936
cons I .6088476 .4448559 1.369 0.182 -.3023985 1.520094

 

 

144
There is little difference in the standard errors ofthese two regession equations, suggestirng
that the robust estimates validate the OLS regession model. OLS passes this diagnostic check.
Thus, our conﬁdence in the conclusions we can draw from this model is enlnanced, as is the
story we can tell regarding innovative decision making processes in states as they prepare for
the aging of the baby boomers.
Summary

In essence, some of the irnitially hypothesized relationslnips have been shown to be
relevarnt while others have been found inconsequential or sinnply wrong. The ﬁrst of the two
hypotlneses states that there is a positive relationship between a variety of demographic,
socioeconomic and political factors and inrnovative decision-making and long-range plarnning
for the aging of America

H1: States which have a signiﬁcant number of older citizens currently, or

anticipate notable growth in the mlmber of elderly; are larger, wealthier states;

are politically liberal; arnd have a urniﬁed political base between the executive

and legislative branches will be more likely to actively engage irn long-range
planningfortheagingofthebabyboompopulationandbebetterpreparedto

develop innovative strategies regarding the aging of America

Through this aggegate analysis, I can conclude the following regarding this
hypothesis. Similar to the ﬁndings articulated inn the Lammers and Klirngnan study (1984), the
cmrerunumberofelderlyhnastateisnotsiguﬁcaminprojecﬁngthelevel ofagingpolicy
innovation irn that state. However, my ﬁndings indicate that there is a strong, statistically

signiﬁcarnt (p>.05) negative relationship between states that will undergo the largest gewth in

145
the percent of elderly in their state and innovative decision-making and long-range planning.
This suggests that these states will be least prepared for the aging ofAmerica.

Secondly, there is a statistically signiﬁcant (p>.01) positive relationship between
wealth, as deﬁned by per capita income, and innovative decision-making and long-range
plarnning underway in states for the aging of baby boomers. This ﬁnding conﬁrms the study of
Walker (1969) and Savage (1978) regarding the positive relationship between wealthy states
(sophistication and education) and innovation However, size of state had no bearirng on the
irnrnovative decision-making processes within a state, as it did on the diﬁ‘usion of innovation

Lastly, on the political front, this model suggests that the hypothesis regarding political
hberalism and irnrnovation is not correct. There is a statistically signiﬁcant (p>.05) negative
relationship betweern states which consisterntly have Democratic executive leadership and
innovative plarnning and decision-making processes. It is arguable that this variable is not a
measureofpoliticalﬁber‘alisrn,20 butatminimalwecansaythatthereappearstobeaninverse
relationship between Democratically controlled executive oﬂices and innovative decision-
maldng.1wggestﬂmﬂﬁsvaﬁableisbeingmdMymﬂuancedbythesouthemstates
phenomenon, and that ﬁnrther research is necessary to determine the relationship between
ideology21 and partisanship on state policy irnnovation. As anticipated, there is a statistically
relevarnt (p>.05) positive relationship between urniﬁed party control of the executive and the

legislature arnd innovative decision-making and long-range plarnning.

 

2° See Wright, Erikson and McIver (1985).

2‘ The Wright, Erikson and McIver ideology scale was incorporated into the model and
run as a test to an alternative measure of ideology, and it proved as an insigniﬁcant variable.

146

Findings regarding the second hypothesis are very interesting, and possibly most
signiﬁcarnt,22 when considering the potential for adding to the political science discourse about
policy development and innovation.

H2: States in which an aging agenda is visnble, and/or with governance

structures that provide for and encourage irnteragency collaboration on the state

level will be more likely to actively engage in long-range planrning for the aging

of the baby boom population and be better prepared to develop irnnovative

strategies regarding the aging of America

The strongest individual independent variable irn the OLS model in explainirng and
predicting innovative decision-making processes in states regarding the preparation for the
aging of America is the level of collaboration A more cooperative work ernviromnernt and
more opportunity for collaboration among state agencies is related to more innovative
decision-making and long range planrning. Although a positive relationship was initially
hypothesized betweern bureaucratic structure (cabinet level department status) and innovation,
tlneﬁndingtlnatthisrelationslnipisanegativeoneactuallybuttressestheﬁndingregardirng
collaboration. Ifaging issues are delegated to a single department, then it appears that there is
less long range planning and policy innovation Possrbly, other agencies do not feel responsible
nor a need to involve themselves in the exploring aging issues, because tlnere is a “place” in
charge of tlnose matters.

This ﬁnding regarding collaboration signiﬁcantly adds to the story which can be told

about irnrnovative plarnrnirng arnd aging policy developmernt irn states. This ﬁnding challenges the

 

22 See Edward E. Leamer article “Sensitivity Analyses Would Help” (1985) for a
discussion about “important” and “doubtful” categorization of variables.

147
conﬂict-resolution model advanced by Baumgartner and Jones (1993 ), and thus, merits a closer
look. The following chapters highlight the responses from a follow-up irnterview with four
states and discusses the elemernts of collaboration. The states of California, Indiana, South
Carolina, and Vermont have been identiﬁed as outlier states in this aggegate arnalysis. (See
Figure 15.) The states of Vermont and Indiarna have been both collaborative and irnrnovative in
their plans for the changing demographics of the 213t Century, whereas both California and

South Carolina have been neither collaborative nor innovative in their eﬂ‘orts.

Chapter 5

A COMPARATIVE REVIEW OF FOUR STATES:
WHAT MAKES A MAVERICK IN NOVATOR?

"Case studies are ideal in assisting political scientists in understantﬁng complex
social and political phenomenon, and are a preferred research method when
examining contemporary events in which behaviors of individuals cannot be
manipulate "
Robert Yin,
Case Study Research Design and Methods, 1984
Introduction

This quantitative analysis of innovative processes which stimulate policy development
and policy change is different than most oftlne previous irnrnovation studies. As already stated,
most of the irmovation research has been variance studies, using regression analysis to
statistically explain the rate of adoption of a certain law or policy. Virginia Gray (in Dodd and
Jilson, 1994), speciﬁcally called for the focus of irmovation studies to become more process
oriernted, and she suggestedushngthecasesmdyresearchintheagendafonnaﬁonﬁteranneas
a beginning poirnt for studying innovative processes. Gray proposed exploring the same factors
in the agenda literature as done in this dissertation—institutional capacity, policy entrepreneurs
and policy networks or policy communities (Gray, 1994). She argued that the capacity of

govermnerntal institutions account for diﬁ'erences in imnovativeness, and that understanding the

148

l 49
stateprocessesthatleadtoacertainpolicywould assistresearchersinexplaining diﬁ‘erencesin
state level innovation. (Gray, in Dodd and Jilson, 1994223 4).

The aggregate analysis identiﬁed collaboration arnd cooperative work environmernts as
critical variables in explaining and predicting the level of irmovative decision-making processes
underway in states as they plan for the shifting demographics of the 213t Century. In depth
research was conducted in order to more fully understand these processes and the impact of
collaboration on irmovative decision-making and long-range plarnning.

Case studies can be used to test individual, orgarnizational or social theories. Yin
suggested that case studies contribute urniquely to our knowledge of individual, organimtional,
social and political phenomena Case studies, in particular, are used regularly in public policy
analysistogaininsightsirnto speciﬁceventsorhappernings. Lowi arguedthat casestudies of
the policy-making process constitute one of the more irnportarnt methods of political science
arnalysis (Lowi, 1964) and Ecksteirn suggested that case studies are particulariy valuable in the
theory-building process (Eckstein, in Greenstein and Polsby, 197 5).

Often, dissertations involve an intensive single case study or a comparative case study
comprised of two or more cases. It is arguable, that only tlnrough a case study approach can
researchersbecomefamiliarwiththe actors, processes, and issuescentralto policymakingin
the ﬁeld (Downs, 1976). Case studies designed as "comparative studies" arguably have certain
intrinsic advantages when compared to the single case study or to "large-N" statistical analysis
(Lijphart, 19752165). Relying on a comparative case study framework the inﬂuence of

collaboration on state irmovative decision-making processes in four states is explored. The

l 50
same types offactors examined in the agenda formation case study literature is pursued in this
analysis—institutional capacity; the role of policy/political entrepreneurs; and, the importance of
policy networks or policy communities.
Selecting the States

The aggegate analysis showed that collaboration was the critical variable in
determining innovative decision-making. In order to explore this dimension of innovation
deeper, it was irnportarnt that states be selected which varied irn their level of collaboration and
innovation. Some states are historically noted as “irnnovative.” To accurately evaluate the
inﬂuence of collaboratiorn, it was irnportarnt to select states that were not typical “innovators.”
Based on the regession analysis, four "outlier" states were chosen for further review. These
states fell outside of the "norm" of the regression line, in that they were extraordinarily
innovative and collaborative, or they were not irnrnovative and they reportedly did not have a
high level of interagency collaboration.

There is a cluster of states at the top end of the gaph reﬂecting the most irnrnovative
and most collaborative states (See Figure 15.) These are the states of Indiana(14),
Verrnont(45), Minnesota(23), Michigan(22) and New Jersey(30). It is not surprising to ﬁnd
the states of Michigan, Minnesota and New Jersey at the top end of the innovation index.
These states have lnistorically been viewed as innovative. However, it is surprising to see both

Indiana and Vermont as leaders in innovation. 23

 

23 See Virginia Gray chapter in Dodd and Jilson, where she reviews the innovation
diﬁinsion literature and outlines the historical ﬁndings regarding state innovation.

151

 

 

2.5 d “
23
u a
31 I
t!
7 a as
5 3r ,9 ns «5
o e 9
E as 1"
u a a 2 a
1
11 c3
_ as a
ll 12
2t 5
.83 ~ 16
I I I I I
0 3

Interagency Collaboration

Figure 15 — Interagency Collaboration and the Innovation Index

At the bottom end of the gaph, there is also a cluster of states reﬂecting a lack of innovation
and long range plarnning. These states are Mississippi (24), California (5) and South Carolina
(40). It is expected that Mnssissippi and South Carolina would be less innovative,24 but it is
surprising to ﬁnd California, historically a leader in irnnovation, at the bottom of the scale.

The states of California, Indiana, South Carolina, and Vermont were chosen as a part
of the interview protocol. (See Appendix C.) The states of Vermont and Indiana were chosen
because they are typically not viewed as innovative states, yet irn this survey, they were at the

high end of the scale, reﬂecting both high levels of innovation and collaboration The state of

 

24' Virginia Gray argues that moralistic states engage irn the most irmovatiorn, and
traditionalistic states the least (Gray, irn Dodd and Irlson, 1994). The states in the deep South,
in particular, have been slow to innovate. This is also discussed in length by V.O. Key in
Politics, Parties and Pressure Groups and Southern Politics.

152
California, a historically innovative state, scored low on both innovation and collaboration.
The states of South Carolina and Mississippi, southern states, are typically low irnnovative
states. However, for the purposes of this comparative analysis, South Carolina was chosen to
be part of the study because of the potential demogaphic implications for the state given the
agingofthebabyboomers. South Carolinaisviewed asaretiremerntandresort area.

Table 14
State by State Comparisons of Selected Variables

 

 

 

 

 

 

 

 

 

CALIFORNIA INDIANA SOUTH VERMONT
CAROLINA
1995 Population 32.4 5.8 3.7 0.6
(irn rrnillions)
% 65+ in 1995 10.6% 12.8% 11.9% 12.1%
% 65+ irn 2020 13.8% 16.2% 16.8% 16.7%
% Change in 65+ 93.5% 40.5% 77.3% 57.1%
1995-2020
Per Capita Income $ 21,821 $ 19,203 $ 16,923 $ 19,467
Urbanization 92.6% 64.9% 54.6% 32.2%
% AARP Mernberslnip 40.2% 47.9% 44.8% 65.0%
General Fund Budget $ 39.0 $ 6.9 $ 4.2 $ 0.7
(in billions)

 

 

 

 

 

 

 

Source for population ﬁgures, per capita income and urbanization; U.S. Buearu of the Census;
Source for % AARP membership, The State Economic, Demogrcphic and Fiscal ”Wk,
1995: published by AARP Public Policy Institute; Source for General Fund Budget, The Fiscal
Survey of States, National Association of State Budget Oﬂicers

153

The diversity of these states provided for a rich comparative review. (See Table 14).
The states varied demogaphically, socio-economically, politically and orgarnizationally. They
are regionally balanced. They allow for an analysis of the importance of big state/small state
and urban/rural differences. Their resource capacity, both ﬁrnancially and orgarnizationally, was
signiﬁcantly different. Their state political culture and ideological history varied. There was
nmchdiﬂ‘erencebetweenthesestates,butatthe sarnetimethere are sinnilaritiesintlnelevel of
irnnovative decision-making processes and the importance of interagency-collaboration.
According to the survey information, Vermont and Indiana, have active collaborative processes
in place and are planning for the aging oftlne baby boomers; while California and South
Carolina, do not have these cooperative work systems in place and are not doirng long-range
planrning for these shifting demogaphics. However, givern the poterntial irnpernding crisis of the
shitting demographics in this courntry, the aging ofthe baby boomers should be oftlne same
political and public policy concern among state leaders, policymakers and public administrators

inallfourstates.

State Proﬁles

California:

California is the largest state in the nation with a population of 32.4 million people.25
California, being a large coastal/border state, is a main attraction poirnt for irnnnigratiorn,

particularly ﬁom Asian countries, as well as from Mexico. California is expected to experience

 

2’ Based on the 1995 population estimates ﬁ'om the Census Bureau.

1 54
signiﬁcant growth in population over the next twenty-ﬁve years. With a projected total
population increase of approximately 48%, California will be home to a total of 48 million
people by the year 2020. The ethnic and cultural diversity within the state is also expected to
increase substantially.

In 1995, approximately 10.6% of California's population was over the age of sixty-ﬁve.
In comparison to Florida and several other eastern or nnidwestern states, this percentage of
elderly does not seem excessive. However, Califonnia has the largest mrmber of older
Americans living within its borders. There are 3 .4 million persons over the age of sixty-ﬁve
living irn California Between 1995 and 2020, it is anticipated that the number of older
Americans living in the state will increase by 93.5%, resulting in 13.8% ofits' population over
the age of sixty-ﬁve.

Approximately 92.6% of the state's population lives in a metropolitan area The
percapita income is $21,821. The overall poverty rate in the state is 15.8%. The elderiy fare
much better, with only 7.6% of those over age sixty-ﬁve living in poverty. The state's general
ﬁnnd budget was $ 39 billion for ﬁscal year 1995. The appropriation ofstate dollars to the
aging oﬂice irn 1995 was 3 4.9 million, with most ﬁnancial support for older Californians
coming ﬁom the federal government—Medicaid, the Social Services Block Grants and Older
Americans Act funding. The Department of Aging, which has approximately 145 ﬁnll-tirne
equivalent staﬁ‘members,isoneofthirteenentitieswithintheAgencyofHealthandWelfare.

Although 40.2% of the eligble population (age ﬁﬁy and above) are members of the

American Association of Retired Persons, there is not a strong active gassroots sernior lobby in

155

the state. The most effective lobbying for serniors is done by the local professional delivery
system-the local units on aging. It is not surprising that power rests with these local agencies,
given tlnat the cournty system is strong in California. On the county level, there is considerable
evidence of collaboration in a variety of diﬂ‘erent issue areas. For example, in 1991, Governor
Wilson launched the Healthy Start Initiative that focused on school-based health care services
to families. In this delivery system model all health and human services are coordinated on the
local (neighborhood) level and concentrated on serving the needs of the entire family.

Over the last ﬁfteen years, the partisanship of the Governor’s oﬂice has resided
primarily with the Republican party which has had control of the executive branch for ten
years. However, during this same timefrarne, the legislature was controlled by the Democrats.

Governor Pete Wilson (R) was elected into office in 1990, and is now serving his second term
as Governor.

The early 1990s found California irn serious ﬁnancial trouble. The overall economy of
the state was suﬂ‘eiing a recession, as well as the state had to deal with a major budget deﬁcit.
This difﬁcult ﬁnancial situation put the state in a position to focus almost wholly on the presernt
and forego long-range plarnning. Also required were signiﬁcant cutbacks in state services. In
1992, the Ueberroth Council for California Competitiveness connrnissioned a report which
cited a lack of coordination between goverrnnent agencies. This report suggested that if
Califomia was going to be economically competitive in the 21st century, the State needed to

launch a collaborative long-range planrning process.

156

In the 1995 survey, the respondents rated themselves "poor" in the level of interagency
collaboration in relationship to developing policies for the currernt older Americans living iii
their state, as well as in developing strategies for the aging baby boom population. In
interviewing California public administrators in February 1997 as part of the follow-up to the
survey they attributed this rating to a variety of issues not the least of which was the hard
economic times the state had been facing over the early 19905. California, in the ﬁrst halfof
the decade, also survived earthquakes, ﬂoods, and riots.

The irnterviewees stated that state level collaboration in a big state like California is
difﬁcult. Several of these agencies have in excess of 25,000 employees. Each agerncy works
on its own mission and is connected to other agencies only through the Governor's ofﬁce. One
staﬂ‘personrefeiredtothegovernment structureasawheel, witheachagencyasaseparate
spoke, and the Governor’s ofﬁce as the middle hub. In fact, iii the Governor‘s oﬂice, there is an
OﬁceofCabinetAﬂ‘airs, staﬂ‘edwithﬁvepeoplewhosekeyresponsibilityisto interactwith
the agencies and get them to "talk" to one another.

Evidence of cross-agency collaborative plarnning is visrble on the state level, when it is
speciﬁcally focused on a single priority issue of the Governor’s. For example, most recently a
Construction Summit was held, in which the Housing Agency, the Transportation Departrnernt,
the Health and Welfare Agency and Governor’s oﬂice were brought together. These agencies
had to jointly examine the inﬁastructure needs of Califonnia, and to explore diﬂ‘erent
alternatives to encourage the construction of a suﬂicient number of housing urnits to

accommodate the anticipated population patterns in the State.

157

As one interviewee stated, "By Summer of 1995, at the time of the survey, we were
just emerging from the darkness of physical, social and ﬁnancial disaster. We were ornly
startingto seethelight." Muchhas occurred ontlne state level overtlnelastyearthat speaksto
the issue of state innovation in aging policy development. In 1996, the Hoover Commission on
Eﬂiciency and Economy issued its report callirng for the integation of all state agerncies which
deal with Long Term Care. There is a proposal under consideration that would consolidate
progarns from seven different departments irnto a single Department of Long Term Care and
Connrnurnity Services. Secondly, late in 1996 the Older Californians Act was reauthoiized for
the ﬁrst time in 15 years. This Reauthorization Act speciﬁcally addressed the issue of
collaboration on the state level and between state and local partners. This Act, which wernt irnto
eﬂ‘ect January 1997, divested all contract authority and funding—$2.5 billion—to the 33 local
units on aging. It also assigned responsibilities to the Agency Director of the Department of
Health and Welfare to coordinate all the agencies irnvolved in long term care, so to "give voice"
to the redeﬁned LTC agenda in the state.

The interviewees stated that they did see a need to encourage and entice state agencies
to collaborate more in the future. However, givein how "big" California state government is,
they felt that it was unlikely that tlnere would be a cross-agency collaborative process
established as an on-going ﬂmction. In California, policy developmernt and funding control is
pushedtothecourntylevel, andtheinterviewees suggestedthatitwasonthecountyleveltlnat

collaboration and irnrnovation is taking place.

l 5 8

On the state leveL in the aging policy development area, there has been some new
energies dedicated to innovative long-range plarnning. Late in 1995, Governor Wilson brought
in a new Aging Department Director to reform the agency and prepare it for the aging ofthe
boomers. Governor Wilson is cited as having the foresight to initiate these changes. Since this
new director has started, and with the mandates irncorporated iii the reauthorization ofthe
Older Californians Act, agencies have begun to "talk" to one anotlner and to think through
some of these long-range issues, most speciﬁcally iii the health care area.

Although there is rmrch more that could be done in the area of irnteragency
collaboration, such as bringing irnto the collaborative process departments outside of the health
care arena, California is moving along the path to more innovative policies apparently because
of the opportunities presented through cooperative work enviromnents. In a state as large and
diverse as California, irnteragency collaboration is not a "natural act." Each agency sees to its
ownnnissionandhasitsownpriorities. Wrthouttheimpetusaﬂ‘ordedbythepersonal
involvemernt and leadership of the Governor, collaboration will not occur, and policy
developmernt and policy inrnovation will suﬁ'er.

Indiana:

Indiana is iii the heartland ofAmerica A nnidwestem state with a population of 5.8
million people.26 Based on population, it is one of the larger states, ranking 14th irn the nation.

In 1995, approxinnately 12.8% of its population was over the age of sixty-ﬁve. It is anticipated

that the number of older Americans living in Indiana will g'ow over the next twenty-ﬁve years,

 

26 Based on population estimates of 1995.

1 59
particularly the "old old'm. Between 1995 arnd 2020, it is anticipated that the irncrease in
Indiana's older population will be 40.5%, resulting in 16.2% of its' population over the age of
sixty-ﬁve.

Approximately 64.9% of the state's population lives in a metropolitan area. The
percapita income is $19,203. The overall poverty rate iii the state is 11.7%, with the elderly
faring slightly better than the overall population, with a 10.8% poverty rate. 47.9% of the
eligible population (age ﬁﬂy and above) are members of the American Association of Retired
Persons. The senior lobby iii the state is a strong one, and many seniors are active parliciparnts
in the local community groups called " Step Ahead Councils."

The state's general hind budget was 3 6.9 billion for ﬁscal year 1995. The
appropriation of $37.1 million in state dollars to the aging oﬂice in 1995 was signiﬁcant. This
ftmding supports the Oﬁce for Aging with a staﬂ‘of39 full-time-equivalents, as well as many
state-funded community initiatives for seniors. The Omce of Aging is a separate division
within the Indiana Family and Social Services Administration.

Over the last 15 years, the partisarnslnip of the Governor’s oﬂice has resided with the
Republican party for 10 years, as has the legislature. During this entire period, there have been
5 years in which there was uniﬁed political control of the executive and legislative branches
under Republicans. Evan Bayh (D) was elected in 1988 and served his two terms as permitted

bytheteimlimitstatutesinthestate. It wasundertheBayh adnnirnistrationthatsigniﬁcant

 

27 This refers to the population over the age of 85. The midwestern states, particularly
the rural states, which will experience an out-migration of young, are projected to share in a
larger number of the 85+ population than the rest of the nation.

160

reorganization of the state took place with a speciﬁc emphasis on collaboration and a focus on
family. Governor Bayh's vision was to make goverrnnent work for families. He mandated his
agencies to work together collaboratively and rethink the delivery systems to families in the
state.

In 1991, the name of the Social Services Department was changed to the Indiana
Family and Social Services Administration. Also, a new network of community entities were
established called "Step Ahead Councils." These local councils, eventually created in all 92
counties, were empowered by the state as "local voices" and were a mecharnisrn to eﬂ'ectively
do community-based planning and service delivery. An Indiana Policy Conmcil and Working
Group, involving ten different agencies and the governor’s oﬂice, was created as a response
mechanism for the Step Ahead Councils. (See Figure 16.) This policy council was comprised
of the agency heads from each of the ten departments. Governor Bayh was Chairman of the
Courncil and Cheryl Sullivarn, Secretary of the Family and Social Services Administration, was

Vice Chair.

161

 

 

 

 

 

 

 

 

 

Figure 16-Indiana Policy Council and Working Group

The Policy Council meets montlnly to irnitiate policy priorities and to establish
collaborative agendas and responses to meet the needs expressed by the local Step Ahead
Councils. The Policy Council also created a working group of agency progarn administrators
to implement the policy directions of the Council. This Working Group's mandate is to work
together collaboratively—share resources-both people and ﬁnancial—to meet the needs of the
Step Ahead Courncils. This Working Group actually makes "ﬁeld trips" to the communities to
meet with their local partners to better imderstand issues and know how to be most responsive.

All of these efforts were undertaken without new ﬁinding or new staﬁ‘.

162

The agency directors were held accourntable by the Governor for their agency's
collaboration irn solving problems articulated by these Step Ahead Councils. The Governor
emphasized that the Step Ahead Councils were partners with the state irn meeting his vision of
serving families. However, by the second year of the plarn, it was clear that new partners on the
federal level needed to be leveraged to truly make a diﬂ‘erence iii the service delivery system

In 1994, Indiana, along with the State ofWest Virginia, was selected by the White
House to be a part of a national pilot effort to irntegate services at the federal level. As a part
of the Community Enterprise Board and nmder the leaderslnip of Carol Rasco, Director of the
Domestic Policy Council at the White House, seven federal agencies joined together
collaboratively to respond to the needs as articulated by the Indiana Policy Council. (See
appendix D for the notiﬁcation letter from President Clinton.) These agencies included the
departments of health and human services, labor, education, agiculture, housing and urban
developmernt, ofﬁce of management and budget and the attorney gerneral. (See Figure 17 for
the irnterrelationslnips between the comnnurnity, state and federal partrners.)

Given the emphasis on collaboration iii the state, it is not surprising that in the 1995
survey, the respondents rated themselves "excellernt" in the level of irnteragency collaboration in
relationship to developing policies for the currernt older Americarns living irn their state, as well
as in developing strategies for the aging baby boom population. Aging issues and the plarmirng
for the shifting demogaphics of tire 21 st Century is a part of the ongoing agenda of the Indiana
Policy Council. Many of the model progams in long-term health care is a result of the level of

collaboration underway in the state.

163

Indiana completed a long-term care analysis out to the year 2020, and has developed a
long-term care strategy to address the issue of rising heath care costs. With a gant ﬁom the
Robert Wood Johnson Foundation, they have instituted a "Choice" program that focuses on
home-based health care tlnat keeps aging seniors in their homes. Also, as a part of this venture
is a healtln and wellness progarn, that reaches out not only to seniors but also to the middle-
aged population. Indiana has a Medicare/Medicaid Clearinghouse which provides direct
assistance and education to their seniors, and the Step Ahead Councils provide the single poirnt
of erntry for services. They have also developed a private long-term care insurance market in
the state.

In interviewing the Indiarna public administrators in February 1997, they stated that
collaboration was now a "standard" of operation in the state. The new governor, Governor
O'Bannon (D), elected in November of 1996, strongly supports the administrative structure and
the local partnerships with the Step Ahead Councils. The irnterviewees stated that they saw the

focus on collaboration continuing with this new administration.

164

 

 

 

 

 

 

 

 

 

 

 

Access to Qualiy
Impoved coordimtion
of children, youth and
family savices. FAMILY
‘ COUNTY
F ACILITATOR
Routine Problems SIEP CO
Implementation of AHEAD MULTI AGENCY
Local Plans of Action COUNCIL TEAM
that reﬂect commnnnity
goals andpriorities. ‘
Policy and Procerhres
WORKING GROUP WORKING GROUP
Statewlrb Polcy . . State
CHAIRPERSGJ
Devebpment Employees
Affecting
Maui-agencies INDIANA POLICY COUNCIL AGENCY ms
Coordination of .
services for children VICE CHAIRPERSON
and families.
State Plan
mm and REGION V TEAM REGION V DIRECTORS
Federal Waivers
‘ CHAIRPERSON
WHITE HOUSE
Conﬂict Resolution EMPOWERMENT BOARD CABINET
WORKING GROUP
WHITE
HOUSE
COORDINATOR

Figure l7—The Indiana Collaboration Project

Governor Bayh was credited with being the "political entrepreneur" whose insistence
on "continuous quality improvement" pushed the process of change in the way government-

federal, state and local—responded to the needs of Indiana's families. It is clear from the survey

l 65
ﬁndings and the irnterviews that collaboration was a major factor in promoting innovative
decision-making processes and providing a forum in which long-range plarnning took place.
Indiana is a model of "post-bureaucratic goverrnnent" in which collaboration and cooperative
work environmernts create opportunity for innovation and strategic policy developmennt and
policy change.

South Carolina:

South Carolina is a southern state with a population of 3.7 million people.28 Based on
population estimates, it is a mid-sized state, ranking 25th. irn the nation. In 1995,
approximately 11.9% of its population was over the age of sixty-ﬁve. Between 1995 and
2020, it is arnticipated that the number of older Americarns living irn South Carolina will increase
by 77.3%, resulting irn 16.8% of its' population over the age of sixty-ﬁve.

Soutln Carolina is considered a vacation spot. Most recently, the state has developed as
a part of its economic development strategy, a plan to attract more retirees to the state. South
Carohmhasbeenmccessﬁﬂinatﬁacﬁngﬂnemoreﬁnandallyaﬂluernreﬁreesto itscoast areas.

Currently, the state of South Carolina ranks ﬁﬂh in the nation in retirement income. In fact, in
this 1997 legislative sessiorn, a bill exempting the ﬁrst $50,000 of retirement income from the
personal income tax is being considered.

Approximately 54.6% of the state's population lives in a metropolitan area. The
percapita income is $16,923. The overall poverty rate iii the state is 18.9%. Although the state

hasbeensuccessﬁﬂmgetﬁngﬂnemoreatﬂueMreﬁreestomovehnohsborders, one-in-ﬁve

 

28 Based on population estimates of 1995.

166

elderly South Carolirnians live in poverty. Although 44.8% of the eligble population (age 50
and above) are members of the American Association of Retired Persons, there is not a strong
active senior lobby in the state. The most effective lobbying for serniors is done by the local
professional delivery system—the local units on aging. In fact, the 1991 model legislation in the
state that restricted bingo revenue into an infrastructure ﬁnd for senior citizen centers and to
be a support structure for community based homecare was spearheaded by these local councils
on aging.

The state's general fund budget was $ 4.2 billion for ﬁscal year 1995. The
appropriation of state dollars to the aging ofﬁce in 1995 was $2.3 million, with most ﬁnancial
support for older South Carolinians coming from the federal government-Medicaid, the Social
Services Block Grants and Older Americarns Act funding. At the time of the survey, Sunnrner
1995, the Oﬂice of Agirng was a separate ofﬁce within the Governor‘s Oﬂice. However, as a
part of Governor Beasly's 1997 State of the State address, he proposed that the Oﬂice on
AgingbetransferredtotheDepartmentofHealthandHuman Services,thusgivingaging
issues cabirnet level status. The department will be renamed to the Departmernt of the Health
and Senior Services.

The partisanship of the Governor's oﬂice has been fairly evenly divided between the
two parties over the last 15 years. However, Democrats have controlled the legislature during
this entire period. Until 1993, South Carolina was a legislatively controlled state, in that each
of the departments was not a part of the Governor's cabinet, but answerable to the legislative

connrrnittees and/or separate connrrnissions. Durirng the Campbell (R) adnnirnistration, legislation

167
was passed late in 1993 converting the state into an executive controlled state, thus, vesting
more control for the administration of the state to the Governor. In November of 1994,
Governor Beasly (R) was elected29 into oﬂice, with an agenda for "reegineeiing and
consolidating state govermnernt."

In the 1995 survey, the respondents rated themselves "poor" iii the level of interagency
collaboration iii relationship to developing policies for the currernt older Americarns livirng irn
their state, as well as in developing strategies for the aging baby boom population. In
interviewirng the public administrators in February 1997 as a part of the follow-up to the
survey, they attributed this earlier rating to the changes state government was undergoing at
the time. Governor Beasly took over stewardslnip of the state in January of 1995 and quickly
attempted to move the departments under his purview. The Sunnrner of 1995 was a critical
time for this reorganization, and it is very likely that the level of irnteragency collaboration was
very low.

The irnterviewees commented that the level of collaboration has irncreased signiﬁcantly
since the state has come under executive control. Prior to this administrative shift, the agencies
hadnorealopportunityfororreasontocollaborate. Infacthwasstatedthattheonly
opporturnity to inﬂuence policy on behalfoftlne aging constituents in the state was through
legislative channels and the only real inﬂuence on the legislature was the local councils on

aging-

 

29' Governor Campbell did not seek re-election.

168

Forthelast 15 years, therehasbeen aLong Term Care Connrnittee furnctioningthat
involved several state agencies. This Connrnittee was successful in coordinating heath care
policy as it related to the Medicaid waiver the state had received. However, since the changes
in adnnirnistratiorn, this Corrnrnnittee has become more empowered to do long range plarnning iii
relationship to health care needs. In 1996, this Committee issued its ﬁrst Long Term Care
(LTC) strategic plan with nine reconnnendations for action. This plan speciﬁcally addressed
the issue of the aging baby boomers. Also, an Adult Protection Coordirnating Council was
created in late 1995, which included the private sector, law enforcement, comrinunity leaders,
local aging council members, state aging ofﬁce, and social services on the state level. This
coordinating council meets on a regular basis to wrestle with issues of elder abuse and
determine statewide policy options.

ALong Term CareProvisopassedasapart ofthe1996 appropriationsbill required
thatthelocalconmcilsonagingﬂmctionasasinglepoint oferntryforservicesto elders. This
single application poirnt of entry proviso referred to the need for collaboration with state and
local transportation entities. Within the health care arena, there seemed to be much energy
directed at collaboration in the state, between agencies, as well as between the state and their
local partrners. However, outside of tlne health care and elder abuse area, there is little evidernce
of cross-agency collaboration in South Carolina, particularly in preparation for the aging of the
baby boomers.

The irnterviewees stated that they see the focus on collaboration increasing, given the

administrative changes underway in state government. Governor Beasly was credited with

169

being the "political errtreprerneur" who pushed the process of change. It is clear from the
survey ﬁndings and the interviews that the lack of irnteragency collaboration prior to 1996
stiﬂed the ability for innovative decision-making processes and for any long-range planning to
take place. Although there is much more that could be done in the area of interagency
collaboration-bringing in other departments, such as Labor, Commerce, and the Ofﬁce of the
Budget-South Carolina is moving along the path to more irmovative policies apparently
because of tlne opporturnities presernted tlnrough cooperative work enviromnents.

Vermont:

Vermont is a small New England state. In fact, with a population of 579,000,30 it is
the second smallest state in the natiorn, with only the state of Wyoming having fewer people
livirngwitlninitsborders. Vermontisprimarilyaruralstate(ithasbeensaidthatithasmore
cows than people living there), with only 32.2% of the state living in a metropolitan area In
1995, approximately 12.1% of Vermont's population was over the age of 65. It is anticipated
that the number of older Americans living in Vermont will increase by 57.1% betweern 1995
and 2020, resulting in 16.7% of its' population over the age of 65. The per capita irncorne in
Vermont is $19,467. The overall poverty rate in the state is 10.4%, with the elderly being a bit
worse oﬁ‘than the overall population, with a 12.4% poverty rate.

65% of the eligble population (age 50 and above) are members of the American

Association of Retired Persons. The senior lobby iii the state is a strong one, arnd many seniors

areactiveparticipantsincommurnityaﬂ‘airs. Infactitwastheagingadvocacygnoupsinthe

 

30' Based on population estimates of 1995.

170

state that brought the issue of the shifting demogaphics to the public's attention and pushed the
state aging agency to survey the state's middle-aged population regarding their expectations
about the quality oflife they hope to have in the future. Also, at the urging the ofthe Vermont
chapter of AARP and the Council of Vermont Elders (COVE) the Department of Aging and
Disabilities was created in 1990 as a separate division within the Agency of Human Services.
Out of the state's general fund budget of $ 657 million for ﬁscal year 1995, the state
appropriation to the Aging Oﬂice was sliglntly over one million dollars. It is staffed with six
ﬁill-time equivalents.

Overthelast15 years, thepartisanship oftlne Governor’s oﬂicehasbeen split, withnirne
years ofexecutive control by the Democrats and six years by a Republican. During this same
timeframe, the Republicarns had control of both houses of the legislature for six years, but only
three of these years represent a urniﬁed party control of both brarnches of government by the
Republicans. Governor Dean (D) ﬁrst elected irn 1990,31 was chair of the National Governors’
Association in 1994 and championed the issue of early childhood development and school
readirness. His emphasis irn his administration on the importance of investment in young
children and families also included a focus on aging issues, particularly family caregiving.

The respondernts to the 1995 survey rated themselves as "excellernt" irn collaboration on
several fronts. Vermont does not have a county system and authority is vested irn local
goveimnent. Local comrnurnities primarily govern schools and roads, however, there are local

collaboratives that work to create and envision the desired human service outcomes for their

 

31' Vermont is the only state in the nation which has elections for governor every
two years. Governor Dean is now in his fourth term.

171
community. The promirnernt state role in Vermont is the coordirnation and funding of these
locally deﬁned outcomes. Coordination, collaboration, and irnrnovation is critical on the state
level because most funding authority rests with the state and thus most control over progarns
and policies are on the state level. IfVermornt is to be true to its tradition of local control tlnen
they need to ﬁinction in a coordinated way to support the planrning and development undeway
in these local, community-based goups.

The focus of tire older Americans programs in the state of Vermont, is similar to that of
South Carolina—-independence. This theme resonates tlnrough a recernt passage of Public Act
160 "Shift the Balance Bill."32 This bill required reduced irnstitutional spending by eight to ten
percent over a four year period, and increased home-based care. They froze nursing home bed
construction and shifted the focus of ﬁnnds to the community-based health care alternatives.
This concerted effort to increase the quality of home and community based care also involved
the coordination of the service delivery system among state and local health and human service
agencies.

Not only does this state-local collaboration exist, but there is also an extensive cross-
agency relationship. All agency heads meet every Monday as an "Executive Policy
Connrnittee" to discuss cross-cutting issues and determine collaborative policy direction. In the
follow-up irnterview protocol the irnterviewees suggested that this top-level collaborative
agenda provided a uniﬁed purpose and direction for all of the state agencies and set the

expectation that cross-agency collaborative plarnning would take place at all levels.

 

3’2 The Redistribution of Long Term Care Expenditures: Shifting the Balance Act
(PA. 160) passed the Vermont legislature in 1996.

172
In Vermont the interviewees felt that no one leader or entity spurred collaboration and
innovative decision-making but that it simply was the culture of Vermont to work together
cooperatively. There is less mobility in Vermont than other states with many people stayirng in
their community for long time creating a close krnit family culture. The church is also a strong
factor. TownmeetingsarearegularpartoftheVermontcultureandusedasaforumto share
information and build consensus around issues and policies. Collaboration is an expected way

of doing business in Vermont.

Summary-What Makes A Maverick Innovator?

Lammers and Klingnan (1984) examined the variations in state based aging policies
over a twenty year period (1955-1975). They created an "index for innovation" that utilized a
variety of dependent variables, that were grouped into four categories: (1) the state's efforts at
income maintenance, (2) the state's social services programs, (3) the state's health and long-
term care delivery systems, and (4) the state's efforts at regulatory protection for the elderly.
Using regession arnalysis, connplemernted with a comparative case study involving eiglnt states,
they classiﬁed states irnto a four quadrarnt matrix: strong achieving states, underachieving states,
low achieving states, and maverick innovators based on this innovation index.

The state categorizations deﬁned in the Lammers and Klingnan study
(Lammers/Klingnan (1984) and Lammers (1989)) codiﬁed each of the ﬁfty states irnto one of
the following four categories:

Strong Achievement States: states possess dernogaplnic and socioeconomic

backgound to establish the aging issue as a problem, and have assertive political
traditions, and are developing irnrnovative responses to their aging populations;

173

Under-achieving States: states possess demogaphic and socioeconomic backgound

to establish the aging issue as a problem, and have assertive political traditions, but are

not developing innovative policies for their elderly;

Low Achieving States: states that traditionally make limited use of state goverrnnent-

ﬁnancial investment and policy development capacity, and they are not developing

irmovative programs/policies for their elderly; and

Maverick Innovators: states do not typically possess the policy capacity or the

political opernrness to develop innovative responses to their agng populations, and they

traditionally have a limited use of state goverrnnent, but they are developing irrnovative
policies for their aging population.

Using this ﬁamework for arnalysis, it would appear, based on the 1995 survey, that
California is an example of an "Underachieving State”, Indiarna of a "Maverick Innovator,"
South Carolina of a "Low Aclnievirng State," and Verrnornt of a "Strong Achieving State."
Althoughitisirnterestingthatthesestatesfallouttlnewaytlneydo,Iamuncertain,whein
looking at a "sriapsho " in time, if this classiﬁcation is relevant. Also, given the irnfonnation
gathered from the interviews, it is important to note the "timing” to any analysis. Based on
eﬂ‘orts from 1995-1997, it is clear that botln California and South Carolina would be considered
"Strong Achievers" if not "Maverick Irnrrovators."

Whatismoreirnportarntthanclassiﬁcationofthe statesisgaininginsiglnt intowhat
makes states become "Maverick Innovators” and how to encourage such development.
Clearly, in looking at the ﬁndings ﬁom the aggegate analysis and the comparative state
reviews, collaboration—within state government, as well as across the different levels of

goverrnnent—was a crucial element in explairning policy irnrrovation This ﬁnding augnernts the

study ofLammers in that he also found the pattern ofirmovation was not one ofexclusive state

17 4
involvernernt but rather a response that involved shared roles with other levels of government,
and sometimes with the private sector (Lammers, 1989). Collaboration and policy innovation
is distinctly linked.

In all four states, the irnterviewees suggested that government response to the shilling
demogaphics of the 215t century would need to be different from past efforts in meeting the
needs of aging citizens. As a society, we carnrnot continue to do "business as usual”-
goveimnent cannot aﬂ‘ord to be the answer to all problems—we don't have the money to do it.
Osborne argued this poirnt in his book, Modes of Demm.

"The ﬂindamental goal is no longer to create—or eliminate—government progarns; it is

to use govemmernt to change the nature of the marketplace. To boil it down to a

slogan, if the thesis was govemmernt as the solution and the antithesis was government

as the problem, the synthesis is govemmernt as partrner." (Osborne, 19882327).

The model for this "new govemmernt" or "post-bureaucratic society," is collaboration
and the creation of cooperative work enviromnents. The success of "collaboration" as
instigator of policy development and irmovation is apparernt irn all four states. Particularly when
we are addressing the issue of innovative decision-making processes collaboration appears to

be a key elernernt in turning states irnto "Maverick Irnrnovators. "

CHAPTER 6
CREATING A NEW STRUCTURE FOR INNOVATION:
THE IMPORTANCE OF COLLABORATION TO THE
INNOVATIVE DECISION-MAKING PROCESS
"Thus, analyzing innovation, and what factors facilitate or retard it, is
intrinsically valuable. At the state level the increasing competition among the
states lends added signiﬁcance to understanding innovation. "
Virginia Gray,
in New Perspectives on American Politics, 1994

Introduction

This dissertation tells a story about irmovative decision-making processes underway in
states as they prepare for the shilling demogaplnics of the 2lst century. It explores the
determinants of decision-nnking—demogaphic, socioeconomic, political and organizational-
and builds a theory of innovation which is cerntered on the process of decision-making and
long-range plarnning. Through this dissertation, important insights are gained regarding
governance structures and practices, especially the importance of collaboration and cooperative
work environments in stimulating innovation.

This dissertation assumes that states will continue to have a pronninerrt role in the

development of domestic social policy, and will continue to be the source of "vertical

17S

176
innovation33 in the future. The emphasis on "innovative process" in this dissertation challenges
the way in which many political scientists have gone about studying this issue in the past, and
moves the study of innovation closer to the agenda formation research. This dissertation
highlights the importance of policy entrepreneurs, policy networks and the development of
policy capacity within states as critical components to understanding state based innovation A
framework for studying decision-making processes in states is developed which can assist in

explaining state variation in irnrnovation.

Explaining State Variation in Innovation

As highlighted in chapter four, there are a variety of lessons learned regarding the role
of demographic, socioeconomic, political and organizational factors as determinants of
innovative processes in states. Based on the ﬁndings from the aggregate analysis, I conclude
that dernogaplnics was not a signiﬁcant factor in long-range plarnning for the aging of the baby
boomers. The currernt percentage of elderly in a state was not signiﬁcarnt in projecting the level

of aging policy irnnovation in that state. Those states that will undergo the largest growth in the

 

33 Gray (in Dodd and Jilsorn, 1994) suggested that irnrnovations will diﬂiise more rapidly
thanintlnepast, andtheirspreadwillbelesstiedtoregionalboundariesgiventlnecurrerntand
ever-growing technology available to state governrnernts, and the reliance on professional
networks and associations as information gathering mecharnisrns, making the study of diffusion
of inrnovation among states no longer relevarnt. These trends mean that state's exposure to new
ideas will become more similar and that the lag time between the ﬁrst and last adopters will
slnrink. She argued that the important focus ofﬁrture research is the process ofinnovation and
that ”vertical diﬂiision" of irnnovation was more likely the trend of the future. She supported
the suggestion by Richard Nathan that there would be less "horizontal diffusion of irnrrovation"
iii the future and more "vertical diﬂiision" with state irnnovations tricklirng up to the national
level.

177
percernt of elderly are doing the least long-range planning arnd innovative strategic policy
development. This suggests that these states will be least prepared for the aging of America.

Secondly, wealth, deternnined by per capita income, was a strong indicator of state
level innovative decision-making. This conclusion is not sruprising, given the long history of
"wealtln" as a determinant of policy development and innovation diﬂiision (Dye, 1966; Walker,
1969; Gray, 1973; Savage, 1978; Cannon and Baunn, 1981; Lammers and Klingnan 1984a;
Lammers, 1989; and Berry and Berry, 1990; 1992). However, "size" of a state, based on
population estimates, had no bearing on the inrnovative decision-making processes within a
state. Walker (1969) used "size" of a state as a proxy variable indicating large state
bureaucracies, and thus, argued that there were a signiﬁcant number of staﬂ‘ available to send
to policy network meetings and to attend association conferences. Therefore, it is possible that
" size" would be a relevarnt factor for diﬁ'usion, but not a signiﬁcant determinant of innovative
processes witlnin states.

Tlnirdly, on the political front, the ﬁnding regardirng the importance of urniﬁed executive
and legislative party control and innovative decision-making processes was anticipated. The
theory of "uniﬁed government" has been developed in the literature, and tlnere is some research
regarding the connection between party urniﬁcation and policy outcome. However, given the
ﬁndirng in this dissertation, it is irnportarnt that additional explorations of the potential role of
"urniﬁed party" as a determimnt of policy be continued.

The ﬁndirngs regarding political hberalism and innovative decision-making must be

regarded cautiously. There is a statistically signiﬁcarnt negative relationship between states

17 8
which consistently have Democratic executive leadership and innovative planrning and decision-
making processes. However, the two states which scored highest on the innovation index were
Vermont and Indiana and both currently have Democratic governors. It is arguable that this
variable is being unduly inﬂuenced by the southern states phenomenorn, and that ﬁrrther
research is necessary to determine the relationship between ideology and partisanship with state
policy inrnovation.

Finally, the ﬁndings regarding orgarnizational factors prove most interesting, and
possibly are most signiﬁcant, when considering the poterntial for adding to the political sciernce
discourse about policy development and inrnovation. Collaboration was consistently shown as
a signiﬁcarnt independernt variable in the OLS model in explaining and predictirng innovative
decision-making processes, as states prepare for the aging of America An open and
cooperative work enviromnent coupled with collaborative engagement between state agencies
results irn more irnnovative decision-making and long-range plarnning.

Equally as irnteresting, is the ﬁnding that bureaucratic structure inﬂuences innovative
decision making processes and long-range plarnning. The more orgarnizational status given to
the aging issue (cabinet level department status), the less likely to ﬁnd irmovative decision-
making processes underway. Although contrary to my initial hypothesis, this ﬁndirng that the
relationship between bureaucratic orgarnizational status and innovation is negative actually
complernernts the ﬁnding regarding collaboration. If aging issues are delegated to a single

department, then it appears that other agencies do not feel responsrble for or connected to

179
aging issues. Comprehensive state irrvolvemernt in aging issues is inhibited because only a
single agency has been assigned ownership of tire issue.

These ﬁndings from the aggegate analysis are helpful in telling a story about innovative
planning and aging policy development in states. In additiorn, the lessons from this quantitative
analysis are rounded out with the learrnings from the comparative case studies. The states of
California, Indiana, South Carolina and Vermont were outlier states in the aggegate analysis,
when considering the explanatory factor of collaboration. In looking at the ﬁndings from the
aggegate analysis, and complementing them with the comparative state reviews, collaboration-
-within state government, as well as across the different levels of government-was a crucial
element in explaining innovative decision-making processes underway in states.

The states involved in the comparative case study varied in size of state, regiorn,
demographics, wealth, and partisanship. The single most critical element in the four states
interviewed was the level of collaboration. Collaboration and policy innovation were distinctly
linked. It has been argued that collaboration enables better use of available resources and
improves the quality and rarnge of services (Melaville and Blarnk, 1992:12). In the era of "no
new taxes" and "anti-govemmernt public sentiment," collaboration has been cited as the wave of
the future of government by the interviewees, as well as in the literature.

In an era of tight budgets, leveraging a variety of resources and facilitating cooperation

are key ingediernts in successful goverrnnent irnitiatives. It is clear to many of the

current observers of the public sector that necessity brought about by budget trimrrning
is also givirng rise to a new spirit of collaboration. The notion of working together to
solve problems creatively eﬁciently and cost-effectively is a common theme among the

1995 Irnrnovations in American Govemrnernt Awards wirmers sponsored by the Ford

Foundation and administered by the Kennedy School of Governmernt at Harvard
Urniversity. (Jordan in Governing, 1995227).

180

In all four states, the interviewees suggested that “govemmernt responses to the shitting
demogaphics of the 21 st century would need to be diﬂ‘erent than past efforts at meeting the
needs of aging citizens.” They submitted that government cannot aﬂ‘ord to be the answer to all
problems. All of the interviewees stressed that "collaboration - doing business in new and
different ways with new and different partners - is the wave of the future. " However, they
stated that “collaboration is not typical of state governance practices, and the challernge before
state government is ﬁguring out how to change the culture of their orgarnization to meet the
demands of the 215t century.”

The Challenge to States: Change the Culture and Create Innovation

Through the years, the public sector tended to follow the prevailing paradign of
private management. In the 193 Os, Roosevelt's Comrrnittee recommended a structure patterned
largely aﬁer corporate America in the 19305. From the 19305 through the 19605, large, top-
down centralized bureaucracies were developed to take care of the public's business. These
hierarchical bureaucracies were patterned after the corporate structures in which tasks were
broken into simple parts, each the responsrbility of a different layer of employees, each deﬁned
by speciﬁc rules and regilations. With rigid preoccupation with standard operating
procedures, vertical chains of command and standardized services, these bureaucracies were
steady, but oﬂen slow and cumbersome.

Massive reorgarnization and restructuring has taken place witlnin private enterprise
throughout the 19805 and 19905. True to form, on all levels of government, there has been a

move to "refomn," "reengineer," "restructure," and "reinvernt" govemmernt. "Reorgarnization

181

sometimes appears to be a code word symbolizing a general ﬁustration with bureaucracy and
governmental intrusion" (March and Olson, 1983:290). In 1993, the National Performance
Review issued a report suggesting that in today's world of rapid change, lightening-quick
information technologies, tough global competition, and demanding customers, large, top-
down bureaucracies—public or private don't work (Gore, 1993). The currernt bureaucratic
structure of govemmernt has little reason to irnrrovate, or to simply improve the way it does
business. This report called for the developmernt of effective, entrepreneurial public
organizations.

Also in 1993, the National Connrnission on State and Local Public Service issued a
report entitled, Hard Truths/Toug; Choices, which stated, that "making democracy work is
what the state and local public service must be about" Winter, 1993 :vii). An obvious part of
addressing the problems that face society is examining the structure of government and
deternnirning how it can be better orgarnized to do its job more successfully. The National
Comnnission on State and Local Public Service suggested that there was a consensus among
botln citizerns and public ofﬁcials that state and local institutions of goverrnnent needed to
drastically improve their capacity and performance, if they were to meet the challenges of the
rapidly changing economic and social systems. The report proposed that these govemmernt
systems move away from the encrusted and outmoded systenrs of command and control that
often emphasized processes at the expense of mission and results. They argued tlnat executive

leadership was at the heart of change.

182

Government "reinvention" has come to mean many different things. It has become
synonymous with reorganizing, downsizing, rightsizing, and privatizing. According to
Reinvmg' Government, reinvention means "the fundamental transformation of public systems
and organizations to create dramatic increases in their effectiveness, eﬂiciency, adaptability and
capacity to innovate" (Osborne and Gaebler, 1992). Reinvention is about changing governance
structures—replacing bureaucratic systems with innovative, "self-renewing" systems. Osborne
and Plastrik propose ﬁve strategies to change the "government's DNA" focused on changing
purpose, incentives, accountability, power structure and the culture of public systems (Osborne
and Plastrik, 1997).

What does governance look like in this post-bureaucratic society? It has been called
"entrepreneurial govemmernt" focused on "quality," "learning," ”adapting," and "innovating"
(Osborne and Gaebler, 1993; Gore, 1993; and Osborne and Plastrik, 1997). Tom Peters is
very direct about nmnagement requirements in the private sector if companies wish to survive
in the information era. Peters is adamant about creating entrepreneurial environments which
are centered around teams of people working together collaboratively—"slnmk wo " (Peters
and Austin, 1985). Peters also argues that leadership is a critical component of creating
innovative processes in which people working within these contexts "own" the issue, the
problem, or the product.

Osborne and Plastrik suggest much of the same can be applied to the public sector.
They argue that bureaucratic systems were designed to be stable, but in the globally

competitive information age, these systems are doomed for failure (Osborne and Plastrick,

183

1997: 38). They, too, encourage a different governance model—focused on customers, based
on entrepreneurial leadership, employee empowerment and changing the "culture" of work by
creating collaborative work environments. Regardless of what you call it, or how it is being
done, governance structures appear to be ripe for change. Public systems seem to be slowly
shedding the binds of the industrial age and shitting to a new paradigm of governance to
ﬂourish in the information era.

Thus, the challenge before the states is how to work within their governmental
(bureaucratic) institutions to encourage, entice and elicit collaboration. The interviewees
stressed that “states must leverage the necessary process changes within their institutions while
not growing in scope or size, to create the policy innovation necessary to meet the challenges
presented by the shifting demographics of the 215t century.” Government exists to do things
that people want done. The determination of what the government shall do involves the
deﬁnition of the tasks which the bureaucracy shall perform (Hynernan, 1950). Collaboration
nriglrt be a model for "new goverrnnent" in a "post-bureaucratic society." .
Building a More Complete Theory of Innovation

As shown in the aggregate ﬁndings and the comparative state reviews, collaboration
plays a critical role in explaining the innovative decision-making processes in state government,
as they prepare for the shifting demographics of the 21st Century. The importance of
entrepreneurial leadership, capacity to plan, and the existence of policy networks or policy

communities is a recurring theme throughout the ﬁndings. In this globally competitive

184
information era, collaboration appears to be explicitly linked with the irmovative, adaptive
organization.

Leadership is a critical component of collaboratives. "Leadership is an agent of
change" (Rockman, 1994:144). In Indiana, in particular, a political entrepreneur—the
governor-played a critical role as change agent in innovative decision-making processes.
Fowler (1994) stressed the importance of political entrepreneurs in creating policy change. "It
is this dynamic quality that transforms political entrepreneurs in potential catalysts for change"
(Fowler, 1994: 298). Also validated in the comparative state reviews was the importance of
“planning netwo ” or “policy committees.” Both in Indiana and Vermont, the existence of a
“planning group” or “planning network” created the necessary collaborative environment to
stimulate irmovative decision-making processes.

Collaboration is a process tool which challenges the roles, responsibilities and the
standard operating procedures of many existing organizational structures. It levels playing
ﬁelds among workers and leaders, and gives "voice" to the customer or client base.
Collaboration enables innovation and adaptation to the changing enviromnent. Social capital
can be built through these new governance networks, and thus, the potential exists for re-
engaging the public with the administration of government, and in the development of social
policy.

It is arguable that collaboration is the most important organizational variable to
consider when attempting to explain and predict state level innovation. "Characteristics of the

way government decides what to do affect the characteristics of what it does" (Creighton

1 85

Campbell,1992 :28). Gaining insights into "collaboration" is especially necessary because it is
important that ﬁrture research on innovation build a variable of "collaboration" into this new
structure of innovation.
Conclusion

This dissertation pointed out the importance of the role of states in ﬁxture domestic
social policy development. It showed the potential implications of the changing demographics
of the 21 st century and illustrated the importance of this demographic metamorphosis on state
policy. This dissertation constructed a model for explaining and predicting state policy
innovation and built an innovation index based on political entrepreneurs, policy communities
or networks and general state policy capacity. The major ﬁnding in this research, found in both
the aggregate analysis and comparative case study, identiﬁed collaboration as a new dimension
of state level policy innovation.

Gray suggested that process studies and variance studies could learn ﬁom one another
(Gray, in Dodd and Jilson, 1994). This dissertation has done just that: taken ﬁ'om the
knowledge in the agenda formation and political entrepreneur literature and built it into the
examination of state variance in innovation in aging policy development. This single issue
studycaneasilybecriticizedasbeingonlyrelevantintheareaofaging. However, sinceitis
examining state governance structures and innovative decision-making processes, it is
reasonable to suggest that any issue in which states must develop policy to respond to any
pressing matter would beneﬁt from this dissertation's ﬁndings. This dissertation advances the

innovation theory-building process by linking a theory of agenda formation with innovation,

186

and identiﬁes a structure for innovation which includes a critical new dimension of

collaboration.

LIST OF REFERENCES

LIST OF REFERENCES

AARP Proﬁle. 1995. "A Proﬁle of Older Americans: 1995." Washington, D.C.:
Program Resources Department, American Association of Retired Persons.

Achen, Christopher H. 1982. Interpreting and Using Regression. London: Sage Press.

Allen, Jesse. 1993. "The Front Lines." In Women on the Front Lines Meeting the
Challenge of an Aging America, ed. Jessie Allen and Alan Pifer. Washington, D.C.: The
Urban Institute.

Atkins, G. Lawrence and et a1. 1994. Old Age in the 215t Century. Washington, D.C.:
The National Academy on Aging.

Axelrod, Robert. 1984. The Evolution of Cooperation. New York: Basic Books.

Baer, David and Lee Cohen. 1996. The State Economic, Demographic and Fiscal
Handbook 1996. Washington, D.C.: Public Policy Institute, American Association of
Retired Persons.

Baumgartner, Frank and Bryan D. Jones. 1993. Agendas and Instability in American
Politics. Chicago: University of Chicago Press.

Baybrooke, D. and GE. Lindblom. 1970. A Strategy of Decision: Policy Evaluation as
a Social Process. New York: Macmillan Publishing Co., Inc.

Berry, Frances Stokes. 1994. "Sizing Up State Policy Innovation Research." Policy
Studies Journal, Vol 22, No.3: 442-456.

Berry, Frances Stokes and William D. 1990. "State Lottery Adoptions as Policy
Innovations: An Event History Analysis." Public Administration Review 84 (2): 395-415.

. 1992. "Tax Innovation in the States:
Capitalizing on Political Opportunity." American Journal of Political Science 36: 715-
742.

 

187

188

Berry, William D. 1982. "An Alternative to the Capture Theory of Regulation: The Case
of State Public Utility Commissions." American Journal of Political Science 28: 524-558.

Beyle, Thad. 1989. "From Governor to Governor." In The State of the States, ed. C.
Van Horn. Washington DC. :Congressional Quarterly Inc.

. 1992. Governors and Hard Times. Washington, D.C.: Congressional
Quarterly Press.

Bingham, Richard D. 1977. "The Diffusion of Innovation among Local Governments."
Urban Aﬂairs Quarterly (13) (Dec): 223-232.

 

Binstock RH. 1972. "Interest Group Liberalism and the Politics of Aging." The
Gerontologist, 12: 265-280.

. 1991. "Aging, Politics, and Public Policy." In Growing Old in America
(4th ed). ed. B. Hess and E. Markson. New Jersey: Transaction Books.

 

Blau, Peter. 1963. The Dynamics of Bureaucracy. Chicago: The University of Chicago
Press.

Bonnett, Thomas and Richard Nathan. 1995. The Devolution Revolution. Washington,
D. C.: Council of Governors' Policy Advisors.

Bouvier, Leon and Carol J. De Vita. 1991. The Baby Boom-Entering Mid-life.
Washington, D. C.: Population Reference Bureau, Inc.

Brizius, Jack. 1994. Investment Budgeting. Washington D.C.: National Association for
Public Administrators.

Brudney, Jeffrey L. and F. Ted Herbert. 1987 . "State Agencies and Their Environments:
Examining the Inﬂuence of Important External Actors." Journal of Politics 49: 186-206.

Burns, Tom and GM. Stalker. 1961. The Management of Innovation. London: Tavistock
Publications Limited.

Campbell, Donald T. 1975. "Degrees of Freedom and the Case Study." Comparative
Political Studies, Vol. 8 #2:]78-193.

Campbell, John Creighton. 1992. How Policies Change: The Japanese Government and
The Aging Society. New Jersey: Princeton University Press.

189

Campbell, Paul R. 1994. "Population Projections of the United States, by Age, Sex, Race
and Hispanic Origin: 1993-2020." U. S. Bureau of the Census, Current Population
Reports, P25-IIII. Washington, D.C.: US. Government Printing Office.

Canon, Bradley C., and Lawrence Baum. 1981. "Patterns of Adoption of Tort Law
Innovations: an Application of Diffusion Theory to Judicial Doctrines." American
Political Science Review 75 (Dec): 975-987.

Chalmers, A.F . 1982. What is this thing called Science? Australia: Queensland Press.

Cherryholmes, Cleo. 1988. "Construct Validity and the Discourses of Research."
American Journal of Education, v96 n3: 421-57.

. 1990. "Reading Research. " Essay prepared for the Critical
Discourse and Educational Reform lecture series at the University of Delaware.

 

Chrystal, Stephen. 1982. America '5 Old Age Crisis. New York: Basic Books.

Chynoweth, J. et a1. 1992. Experiments in Systems Change. Washington D.C.: Council
of Governors' Policy Advisors.

Chynoweth, Judith. 1993. A Guide to Community-based Collaborative Strategic Planning.
Washington, D.C.: Council of Governors' Policy Advisors.

Chubb, John E. 1988. "Institutions, the Economy and the Dynamics of State Elections."
American Political Science Review 82: 133-154.

Clinton, William Jefferson. 1996. Remarks made at the National Governors Association
Meeting-January 31, 1995. In the Rockefeller Institute Bulletin: American Federalism
Today. New York: Rockefeller Institute of Government.

Cornman, John M. and Eric R. Kingson. 1995. "Trends, Issues, Perspectives and Values
for the Aging of the Baby Boom Cohorts. " Paper presented at the Fordham University
Conference: Many Faces of Aging-Challenges for the Future.

Cohen, E. S. 1990. "The White House Conference on Aging: an Anachronism. " Journal
on Aging and Social Policy, 2(3\4): 27-35.

Cohen, Michael, James Marsh and Johan P. Olson. 1972. "A Garbage Can Model of
Organizational Choice." Administrative Science Quarterly 17.

Cutler, Neal E. 1977. "Demographic, Social-Psychological, and Political Factors in the
Politics of Aging: A Foundation for Research in "Political Gerontology." American
Political Science Review 71: 1011—1025.

190

Cox, Carole E. 1996. Block Grants: An Overview of Where We 've Been and Where
We 're Going. Washington, D.C.: National Academy on Aging.

Dawson, John and Peter Stan. 1995. "Proﬁling Public Expenditures ". New York: Rand
Research Brief.

Day, Jennifer Cheeseman. 1993. "Population Projections of the United States, by Age,
Sex, Race and Hispanic Origin: 1993-2050." US. Bureau of the Census, Current
Population Reports, P25-1104. Washington, D.C.: US. Government Printing Office.

Derthick, M. 1987. "American Federalism Madison's 'Middle Ground' in the 1980's."
Public Administration Review 47(1): 66-74.

De Vita, Carol. 1989. America in the 21st Century: A Demographic Overview.
Washington, D.C.: Population Reference Bureau, Inc.

. 1996. The United States at Mid-Decade. Washington, D.C.: Population
Reference Bureau, Inc.

 

Dole, Robert. 1996. Response to Welfare and Medicaid ReformuRemarks at the National
Governors' Association meeting: February 6, 1996. In the Rockefeller Institute Bulletin:
American Federalism Today. New York:Rockefeller Institute of Government.

Downs, Anthony. 1967. Inside the Bureaucracy. Boston: Little Brown and Co.

Downs, George. 197 6. Bureaucracy, Innovation, and Public Policy. Massachusetts,
Lexington Books.

Downs, George and Lawrence Mohr. 1976. "Conceptual Issues in the Study of
Innovation." Administrative Science Quarterly, Vol 21: 700-714.

. 1979. "Toward a Theory of Innovation."
Administration and Society, Vol 10 No 4 (February): 379-408.

 

Dychtwald, Ken and Joe Flower. 1990. Age Wave: The Challenges and Opportunities
of an Aging America. New York: Bantam Books.

Dye, Thomas. 1966. Politics, Economics and the Public: Policy Outcomes in the
American States. Chicago: Rand McNally & Co.

. 1978. Understanding Public Policy. New Jersey: Prentice Hall Inc.

 

Dye, Thomas and Virginia Gray. 1980. The Determinants of Public Policy.
Massachusetts: Lexington Books.

191

Easterlin, Richard A. 1968. Population, Labor Force, and Long Swings in Economic
Growth. New York: National Bureau of Economic Research.

Easterlin, Richard, et a1. 1990. "Retirement Prospects of the Baby Boom Generation: A
Different Perspective." The Gerontologist, Vol 30, No. 6: 776-783.

Easton, David. 1965. A Framework for Political Analysis. New Jersey: Prentice Hall.

Eckstein, Harry. 1975. "Case Study and Theory in Political Science." in Strategies of
Inquiry, (ed.) Fred Greenstein and Nelson Polsby. Massachusetts: Addison-Wesley
Publishing Company. '

Elazar, Daniel. 1972. American Federalism: A View from the States. New York: Crowell
Publishing.

Elder, Charles D. and Roger W. Cobb. 1984. "Agenda Building and the Politics of
Aging." Policy Studies Journal, 13: 115-29.

Erikson, Robert S., John P. McIver and Gerald C. Wright, Jr. 1987. "State Political
Culture and Public Opinion." American Political Science Review 81: 797-813.

Eyestone, Robert. 1977. "Confusion, Diffusion and Innovation." American Political
Science Review 71: 441-447.

F ahey, Charles. 1996. "The Third Age. " Presentation at the June 1996 National Aging
Symposium: The Aging of America and State Policymaking: Creating a 2020 Vision.

Foster, Susan E. and Jack A. Brizius. 1993. "Caring Too Much? American Women and
the Nation's Caregiving Crisis." In Women on the Front Lines Meeting the Challenge of
an Aging America, ed. Jessie Allen and Alan Pifer. Washington, D.C.: The Urban
Institute.

Fowler, Linda L. 1994 "Political Entrepreneurs, Governing Processes, and Political
Change." In New Perspectives on American Politics, ed. Lawrence C. Dodd & Calvin
Jilson. Washington, D.C.: Congressional Quarterly Inc.

Frendreis, John. 1983. "Innovation: A Practice in Search of a Theory, or So What Does
All This Research Mean, Anyway?" Journal of Urban Aﬁ'airs 5 (2): 109-122.

Glick, Henry R. 1981. "Innovation in State Judicial Administration: Effects on Court
Management and Organization." American Politics Quarterly 9 (January): 49-69.

. 1992. The Right to Die -- Policy Innovation and Its Consequences.
New York: Columbia University Press.

 

192

Glick, Henry R. and Scott P. Hays. 1991. "Innovation and Reinvention in State
Policymaking: Theory and the Evolution of Living Will Laws." Journal of Politics, Vol.
53, No. 3.

Gore, Albert. 1993. Creating a Government That Works Better and Costs Less.
Washington, D.C.: The National Performance Review.

Gormley, William T. 1982. "Alternative Models of the Regulatory Process: Public Utility
Regulation in the States." Western Political Quarterly 25: 297-317.

Granato, Jim. 1991. "An Agenda for Econometric Model Building. " Political Analysis,
3: 123-154.

Gray, Virginia. 1973. "Innovation in the States: A Diffusion Study." American Political
Science Review 67 (December): 1174-1185.

. 1994. "Competition, Emulation, and Policy Innovation." In New
Perspectives on American Politics, ed. Lawrence C. Dodd & Calvin Jilson. Washington,
D.C.: Congressional Quarterly Inc.

 

Greenberg, George ct al. 1977. "Developing Public Policy Theory: Perspectives from
Empirical Research." American Political Science Review, V0171 (4) (December): 1532-
1543.

Gross, Richard. 1996. The States Forge Ahead Despite the Federal Impasse. Washington,
D.C.: Council of Governors' Policy Advisors.

Gujarati, Damodar. 1995. Basic Econometrics. New York: McGraw-Hill.
Hamilton, Lawrence. 1992. Regression with Graphics. California: Duxbury Press.

. 1994. Statistics with Stata. California: Duxbury Press.

 

Hamm, Keith E. 1980. "US. State Legislative Committee Decisions: Similar Results in
Different Settings." Legislative Studies Quarterly 5: 31-54.

Hanushek, Eric A. and John E. Jackson. 1977. Statistical Methods for Social Scientists.
San Diego: Academic Press, Inc.; Harcourt Brace Jovanovich, Publishers.

Harkness, Peter. 1996. "Civic Journalism in the 1990's. " Presentation at the 1996
Council of Governors' Policy Advisors 1996 Annual Meeting.

Harvey, Andrew. 1991. The Econometric Analysis of Time Series. Massachusetts:
Cambridge Press.

193

Hayes, Cheryl D. and Anna E. Daneggar. 1995. Rethinking Block Grants: Toward
Improved Intergovernmental Financing for Education and Other Children 's Services.
Washington, D.C.: The Finance Project.

Hayes, Michael. 1992. Incrementalism and Public Policy. New York: Longman.

Hays, Scott. 1996. "The States and Policy Innovation Research: Lessons from the Past
and Directions for the Future." Policy Studies Journal, Vol. 24, No. 2.:321-326.

Hechlo, Hugh. 1978. "Issue Networks and the Executive Establishment." In Public
Administration: Concepts and Cases, ed. J. Stillman. Boston: Houghton Mifﬂin.

Hobbs, Frank and Bonnie L. Damon. 1996. "65 + in the United States." US. Bureau of
the Census: Current Population Reports, Special Studies, P23-190. Washington, D.C.:
US. Government Printing Ofﬁce.

Hodgkinson, Harold. 1992. A Demographic Look at Tomorrow. Washington, D.C.:
Institute for Educational Leadership.

Hodgkinson, Harold and Janice Hamilton Outtz. 1992. The Nation and the States: A
Proﬁle and Data Book of America 's Diversity. Washington, D.C.: Institute for
Educational Leadership.

Hofferbcrt, Richard. 1971. "The Nationalization of State Politics." In State and Urban
Politics: Readings in Comparative Public Policy, ed. Hofferbcrt, R. and Ira Sharkansky.
Boston: Little, Brown and Company: 463-474.

Hyneman, Charles. 1950. Bureaucracy in a Democracy. New York: Harper and Brothers
Publishers.

Jacobson, Gary C. 1990. The Electoral Origins of Divided Government. San Francisco:
Westview Press.

Jewell, Malcolm. 1982. "The Neglected World of State Politics." Journal of Politics
44:638-657.

Jones, Landon Y. 1980. Great Expectations: America and the Baby Boom Generation.
New York: Coward, McCann & Geoghegan.

Jordan, Anne. 1995. "The Power of Collaboration." Governing, (November).
Washington, D.C.: The Congressional Quarterly, Inc.

Kemp, Kathleen A. 1978. "Nationalization of the American States: A Test of the Thesis."
American Political Quarterly: 6 (April): 237-47.

194
Kennedy, Peter. 1992. A Guide to Econometrics. Great Britain: MIT Press.

Key, V. O. Jr. 1949. Southern Politics in State and Nation. New York: Alfred A.
Knopf.

. 1964. Politics, Parties, and Pressure Groups. New York: Thomas Y.
Crowell and Co., ﬁfth edition.

 

King, Gary et a1. 1991. "Scientiﬁc Inference in Qualitative Research." Unpublished
paper.

Kingdon, John. 1984. Agendas, Alternatives and Public Policies. Boston: Little, Brown
and Company.

Kingson, Eric and Barbara A. Hirshorn and John Comman. 1986. Ties that Bind: The
Interdependence of Generations. Washington, D. C.: Seven Locks Press.

Kingson, Eric. 1992. The Diversity of the Baby Boom Generation: Implications for Their
Retirement Years. Washington, D. C.: American Association of Retired Persons.

Klingman, David and William W. Lammers. 1984. "The General Policy Liberalism
Factor in American State Politics." American Journal of Politics 28 (August): 598-609.

Lammers, William and David Klingman. 1984. State Policies and the Aging: Sources,
Trends and Options. Massachusetts: Lexington Books.

Lammers, William. 1989. "Prospects for Innovation in State Policies for the Elderly."
Journal of Aging and Social Policy, vol (3/4): 37-66.

Lammers, William and Liebig, PS 1990. "State Health Policies, Federalism, and the
Elderly." Publius: The Journal of Federalism: 20 (3): 131-148.

Lammers, William and Jon Pynoos. 1988. "State Roles in Aging Policy." Generations,
35-38.

Lavery, Jack. 1996. "The Economic Challenges and Opportunities of an Aging Society. "
Presentation at the National Symposium on the Aging of America and State
Policymaking: Creating a 2020 Vision.

Learner, Edward. 1985. "Sensitivity Analyses Would Help. " American Economic Review
Vol 75 (3):308-3l3.

Lewis-Beck, Michael. 197 7. "The Relative Importance of Socioeconomic and Political
Variables for Public Policy." American Political Science Review Vol 71 (2):559-566.

195

. 1980. Applied Regression. London: Sage University Press.

 

Light, Paul C. 1988. Baby Boomers. New York: W.W. Norton Company.

Lijphart, Arend. 1975. "The Comparable-Cases Strategy in Comparative Research."
Comparative Political Studies, Vol.8 No.2: 158-177.

Lindblom GE. 1959. "The Science of Muddling Through." Public Administration
Review: vol 19: 79-88.

. 1979. "Still Muddling, Not Yet Through." PublicAdministration Review:
vol 39 #6: 517-526.

 

. 1980. The Policy Making Process. New Jersey: Prentice Hall.

 

Lowery, David and Lee Sigelman. 1982. "Political Culture and State Public Policy."
Western Political Quarterly, 35: 376-384.

Lowi, Theodore J. 1964. "American Business, Public Policy, Case-Studies, and Political
Theory." World Politics, 677-715.

. 1969. The End of Liberalism. New York: W.W. Norton.

 

Lowry, Robert C., James E. Alt and Karen Ferree. 1996. "Fiscal Policy and Elections
in American States, " Paper prepared for the Midwest Political Science Association
Meeting, April 12-14, 1996: Chicago.

March, James and Johan Olson. 1983. "Organizing Political Life: What Administrative
Reorganization Tells Us about Government. " American Political Science Review 77: 281-
293.

. 1984. "The New Institutionalism: Organizational
Factors in Political Life." American Political Science Review 78:734-749.

 

March, J.G. and HA. Simon. 1959. Organizations. New York: John Wiley and Sons.

Melaville, Atelia and Martin J. Blank. 1992. What It Takes: Structuring Interagency
Partnerships To Connect Children and Families with Comprehensive Services.
Washington, D.C.: Education and Human Services Consortium.

Miller, Louisa and Jennifer Marks. 1982. "Population Estimates and Projections of the
United States, by Age, Sex, and Race: 1970-1981." Current Population Reports, P25-
917, US. Bureau of the Census. Washington, D.C.: US. Government Printing Ofﬁce.

196

Mintrom, Michael and Sandra Vergari. 1996. "Policy Networks and Innovation
Diffusion: The Case of State Education Reforms." Paper presented at the Annual Meeting
of the Midwest Political Science Association, Chicago, 1996.

Mohr, Lawrence. 1969. "Determinants of Innovation in Organizations," American
Political Science Review 63:1 1 1-126.

Moon, Marilyn. 1996. "The Patient's Advocate," The Washington Post. September 24,
1996, Washington DC.

Mooney, Christopher and Mei-Hsien Lee. 1995. "Legislating Morality in the American
States: The Case of Pre-Roe Abortion Regulation Reform." American Journal of
Political Science Vol 39 (3):599-627.

Mosher, F. 1982. Democracy and the Public Service. New York: Oxford University
Press.

Nathan Richard P. 1986. "The Administrative Presidency." In Bureaucratic Power in
National Policy Making. ed. Francis Rourke. Boston: Little, Brown.

. 1989. "The Role of the States in American Federalism." In The State
of the States. ed. C. Van Horn. Washington, DC: Congressional Quarterly, Inc.

 

. 1995. "Block Grants--Should We or Shouldn 't We? " Testimony for
the US. House Committee on Government Reform and Oversight, July 20, 1995.

 

1996. "The Devolution Revolution: An Overview." In the
Rockefeller Institute Bulletin: American Federalism Today. New York: Rockefeller
Institute.

 

National Governors' Association Resolutions. 1996. Presentation at the 1996 Winter
Meeting of the National Governors' Association. Washington, DC.

Nice, David C. 1994. Policy Innovation in State Government. Iowa: Iowa State
University Press.

Osborne, David. 1988. Laboratories of Democracy. Massachusetts: Harvard Business
School Press.

Osborne, David and Ted Gaebler. 1993. Reinventing Government. New York: Plume;
Penguin Group.

Osborne, David and Peter Plastrik. 1997. Banishing Bureaucracy: The Five Strategies for
Reinventing Government. Massachusetts: Addison-Wesley Publishing.

197

Patterson, Samuel. 1990. "State Legislators and the Legislatures." in Politics in the
American States, (ed.) Gray, Jacob and Albritton. Illinois: Scott, Foresman.

Peters, Tom and Nancy Austin. 1985. A Passion for Excellence. New York: Warner
Books.

Peters, Tom. 1988. Thriving on Chaos. New York: Alfred A. Knopf.

Peterson, Paul E. 1995. The Price of Federalism. Washington, D.C.: The Brookings
Institution.

Peterson, Paul E., Barry G. Rabe and Kenneth K. Wong. 1986. When Federalism
Works. Washington, D.C.: The Brookings Institution.

Peterson, Peter G. 1996. "Will America Grow Up Before it Grows Old?" New York:
The Atlantic Monthly.

Polsby, Nelson. 1984. Political Innovation in America. Connecticut: Yale University
Press.

Quinn, J.F . 1996. Entitlements and Federal Budget: Securing Our Future. Washington,
D.C.: National Academy of Aging.

Ramsey, J .B. 1969. "Test Speciﬁcation Errors in Classical Linear Least Squares
Regression Analysis." Journal of the Royal Statistical Society. Ser. B 31 (2) 350-71.

Ray, David. 1986. "Assessing the Performance of State Legislative Committees: A Case
Study and a Proposed Research Agenda." Western Political Quarterly 39: 126-137.

Rockman, Bert A. 1994. "New Institutionalism and the Old Institutions." In New
Perspectives on American Politics, (ed.) Lawrence C. Dodd and Calvin Jilson.
Washington, D.C.: The Congressional Quarterly.

Rogers, Everett M. 1962. Diﬂusion of Innovations. New York: The Free Press of
Glencoe.

Rogers, Everett M. and Floyd Shoemaker. 1971. Communication of Innovations - A
Cross-Cultural Approach. New York: Free Press.

Rose, Douglas. 1973. "National and Local Forces in State Politics: The Implications of
Multi-level Policy Analysis." American Political Science Review 67 (December): p 1162-
1173.

198

Rourke, F. 1984. Bureaucracy, Politics and Public Policy. New York: Harper Collins
Publishers.

Russakoff, Dale. 1996. "Students Learning the Cost of Overcrowded Classrooms," The
Washington Post. September 14, 1996, Washington DC.

Ryan, Bryce and Neal C. Gross. 1943. "The Diffusion of Hybrid Seed Corn in Two Iowa
Communities." Rural Sociology. 8: 15-24.

Salisbury, Dallas. 1996. "The Economic Challenges and Opportunities of an Aging
Society." Presentation at the National Symposium on the Aging of America and State
Policymaking: Creating a 2020 Vision.

Savage, Robert L. 197 8. "Policy Innovativeness as a Trait of American States." Journal
of Politics: 40 (February): 212-224.

Sharkansky, Ira. 1967. "Regional patterns in the expenditures of American States."
Western Political Science Quarterly 20(December): p 955-971.

Sharkansky, Ira and Richard Hofferbcrt. 1969. "Dimensions of State Politics, Economics
and Public Policy," American Political Science Review: 63 (Sept): p 867-79.

Sigelman, Lee and Nelson Dometrius. 1988. "Governors as Chief Administrators—The
Linkage Between Formal Powers and Informal Inﬂuence." American Politics Quarterly
16:157-170.

Songer, Donald R. et a1. 1986. "The Inﬂuence of Issues on Choice of Voting Cues
Utilized by State Legislators." Western Political Quarterly 39:118-125.

Squire, Peverill. 1988. "Career Opportunities and Membership Stability in Legislatures."
Legislative Studies Quarterly. 13:65-87.

Stanﬁeld, Rochelle L. 1995. "Holding the Bag? " National Journal, September 9,
1995:2206.

Steuerle, Eugene and Gordon Mermin. 1996. "Changes in the Composition of the Federal
Budget--the implications of the aging of America. " Presentation at the June 1996 National
Aging Symposium: The Aging of America and State Policymaking: Creating a 2020
Vision.

Stonecash, Jeff. 1980. "Sources of State Policy." In The Determinants of Public Policy,
(ed.) Thomas Dye and Virginia Gray. Massachusetts: D.C. Heath and Co.

199

Thompson, Victor. 1965. "Bureaucracy and Innovation," Administrative Science
Quarterly, vol 10:1-20.

Torres-Gil, Fernando. 1995. Presentation at the 1995 Annual Meeting of the National
Association of Area Ofﬁces on Aging, held in Los Angeles, California.

Torrey, Barbara Boyle, Kevin Kinsella and Cynthia M. Taeuber. 1987. "An Aging
World." International Population Reports Series P-95, No. 78, Washington D.C.:
Bureau of the Census, U. S. Department of Commerce.

Treas, Judith. 1995. Older Americans in the I990s and Beyond. Washington, D.C.:
Population Reference Bureau.

United States General Accounting Ofﬁce. 1995 . "Block Grants: Characteristics,
Experience and Lessons Learne . " GAO/HEHS-95-74.

Uslaner, Eric M. 197 8. "Comparative State Policy Formation, Interparty Competition,
and Malapportionment: A New Look at V.O. Key's Hypothesis." Journal of Politics: vol
40 p 409-432.

Von Mises, Ludwig. 1944. Bureaucracy. Connecticut: Yale University Press.

Walker, Jack L. 1969. "The Diffusion of Innovations Among the American States."
American Political Science Review: 63 (Sept):880-899.

Watkins, John. 1994. "Retirees as a New Growth Industry? Assessing the Demographic
and Social Impact." Review of Business, Vol 15, No. 3, Spring 199429-14.

Weber, Max. 1946. "Bureaucracy." In Perspectives on Public Bureaucracy. (ed.) Fred
Kramer. Cambridge, Massachusetts: Winthrop Publishers, Inc.

Weissert, Carol S. 1991. "Policy Entrepreneurs, Policy Opportunists and Legislative
Effectiveness." American Political Quarterly 19:262-273.

Welch, Susan and Kay Thompson. 1980. "Impact of Federal Incentives on State Policy
Innovation." American Journal of Political Science: vol 24 (Nov): p 715-729.

Winter, William F. 1993. Hard Truths/Tough Choices: An Agenda for State and Local
Reform. New York: The Nelson A. Rockefeller Institute of Government.

Wildavsky, A. 1975. Budgeting: a Comparative Theory of Budgetary Processes. Boston:
Little, Brown and Co.

. 1992. New Politics of the Budgetary Process. New York: Harper Collins

 

Publishers.

200

Wilson, James Q. 1966. "Innovation in Organization: Notes Toward a Theory." In
Approaches to Organizational Design, ed. James D. Thompson. Pennsylvania: University
of Pittsburgh Press.

. 1989. Bureaucracy: What Government Agencies Do and Why. New

 

York: Basic Books.

Wright, Gerald D., Robert S. Erikson and John P. McIver. 1985. "Measuring State
Partisanship and Ideology and Survey Data." Journal of Politics, Vol 47:469-489.

Yin, Robert. 1979. Changing Urban Bureaucracies. Massachusetts: Lexington Books.

. 1984. Case Study Research Design and Methods. California: Sage

 

Publications.

Zedlewski, Sheila R. and et. al. 1990. The Needs of the Elderly in the 215t Century.
Washington, D.C.: The Urban Institute.

APPENDIX A

APPENDIX A

LIST OF AGENCIES SURVEYED

State Unit on Aging

Department of Social Services

Department of Public Health

Department of Mental Health (including Disabilities)
Ofﬁce of Veterans Affairs

Insurance Commission

201

APPENDIX B

APPENDIX B

Aging of America and State Policymaking: Creating a 2020 Vision
Health & Human Support Survey

General Perceptions About Your State’s Agigg Policies

1. Aﬁer 2010, the number of Americans over age 65 will swell rapidly as the ﬁrst of the baby boomers
reach retirement age. Most demographers project that state and local governments will experience
substantial changes - in terms of services demanded and tax structure — as the baby boom cohort ages.
Throughout this decade and the next one, older Amerieans will form a signiﬁeantly increasing
percentage of our population. What is your assessment of the signiﬁeance of this demographic change
on your state? (circle one)

1. Minor 2. Moderate 3. Major

2. In the past ﬁve years, how well has your state government responded to the anticipated increase in the
65+ population?

1. Poorly 2. Adequately 3. Very Well

3. Please evaluate the current eapacity of the state to effectively meet the challenges and opportrmities of
these shifting aging demographics.

Ranking of eapacity
1. Minimal 2. Sufﬁcient 3. Superior

Enhancing and assuring the quality of life of older Amerieans including health eare and social
support services.

Assuring the economic independence of older Americans via social security supplements, ﬂexible
pension and retirement systems and supports.

Providing affordable housing options for older Americans.

Supporting transportation alternatives for older Americans allowing their affordable independent
living.

Adapting state and loeal tax and ﬁnance structures to adjust or accommodate for the changing
demographics.

202

203

4. We are interested in your assessment of how the growing aging population will affect your state in the
future. For the year 2010, what level of importance do you anticipate the following aging issues will
have to your state:

Rankings of importance in 2010
1. Minimal 2. Moderate 3. Signiﬁcant

Enhancing and assuring the quality of life of older Americans including health care and social
support services.

Assuring the economic independence of older Americans via social security supplements, ﬂexrble
pension and retirement systems and supports.

Providing affordable housing options for older Americans.

Supporting transportation alternatives for older Americans allowing their affordable independent
living.

Adapting state and local tax and ﬁnance structures to adjust or accommodate for the changing
demographics.

5. With the project changing demographics comes an implied change in the types of social support
services and options that may be needed How would you describe your state government’s efforts to
plan for these changes?

1. Poor 2. Fair 3. Good

6. Given the shifting tides of federalism, states might be ealled upon to take a lead role in designing,
developing and funding alternative options for social support services for older Americans. Has your
state developed any innovative responses to these possible new challenges?

Yes No

7. Ifyes, please brieﬂy identify them:

 

 

 

 

 

8. What is your current investment in human service programs for the older Amerieans? S

9. Have the aforementioned demographic changes inspired your state to complete a long-term cost
analysis for Medicare/Medicaid funding?

Yes No

10.

ll.

12.

13.

14.

15.

204
What is your state’s estimated future investment in human service programs for the older Amerieans
in:
2000: S
2005: S
2010: S
2015: S

2020: 3

Does your state have an overall strategy to address the rising costs of long-term health care for older
people?

Yes No
Ifyes, what is the primary focus of this strategy? (circle)

i limiting access to health care services (e.g., managed care or HMOS) for older Amerieans

ii. enhancing older Americans’ access to health promotion and prevention programs (e. g.,
through education and public awareness)

iii. reducing the public subsidy for health eare services for older Amerieans (e.g., to a percent of
cost or an actual limit)

iv. promoting long-term care private insurance options for older Amerieans

What are your current state funding sources for human services for the older Americans?

a)

 

13)

 

C)

 

What do you anticipate future state funding sources will be for human services for the older
Americans?

a)

 

b)

 

C)

 

Which legislative entity (e.g., committee) controls appropriations of funding for programs for the
older Amerieans?

 

16.

205

Which organizational entity (e.g., agency) manages social support programs for older people in your
state?

 

l7.

l8.

19.

20.

21.

22.

23.

24.

Have you surveyed your state’s nriddle-aged population in order to forecast what services and other
resources will be needed in the future?

Yes No

Are your services community-based and coordinated?

Yes No
How would you describe the level of collaboration between your agency and other state agencies and
departments on policies and programs for older Americans?

1. Poor 2. Good 3. Excellent
Within your state, how would you describe the level of collaboration among state agencies and
departments in developing a strategy for meeting the changing needs of the older Amerieans in the
next few deeades?

1. Poor 2. Good 3. Excellent
How would you describe the level of collaboration between you agency and loeal communities on
policies and programs for the current older Amerieans population?

1. Poor 2. Good 3. Excellent
How would you describe the level of collaboration with local communities in developing a strategy to
meet the future needs of the changing older Americans cohort.

1. Poor 2. Good 3. Excellent
How would you describe the level of collaboration between you state and the federal government on
policies and programs for the current older Amerieans population?

1. Poor 2. Good 3. Excellent

How would you describe the level of collaboration with the federal government in developing a
strategy to meet the future needs of the changing older Amerieans cohort?

1. Poor 2. Good 3. Excellent

206
25. On this scale (1=poor, 2=good, 3= excellent), how would you describe the ability of your human
service programs to meet the needs of:
older women?
older minorities?

economically disadvantaged older individuals?

26. Does your state subsidize the following programs for older citizens?

(i) In-home nursing assistance programs Yes No
(ii) Home-delivered meals programs Yes No
(iii) Homemaker services? Yes No
(iv) Visiting nurse services? Yes No
(v) Home health aides? Yes No
(vi) Adult day-care? Yes No

27. Does your state have a state-based supplemental Medicare program to assist older Americans to pay
for long-term care?

Yes No

28. If so, what is the source of funding for this program?

 

 

 

29. How would you describe the effectiveness of your state’s support and social services to older people
(e. g., homemaker/chore services, personal care, ﬁnancial services, out-of-home day care or respite
eare, protective services, casework, counseling, et.)?

1. Poor 2. Good 3. Excellent
30. How would you rate the “innovativeness” of your state’s human service policies for the current older
American population?

1. Not at all 2. Somewhat 3. Very

31.

32.

33.

34.

35.

36.

37.

38.

39.

207
How innovative is your state’s strategy for providing human services to the aging baby boomer
cohort? .
1. Not at all 2. Somewhat 3. Very

How effective has your state been in developing a continuum of community-based services for the
older Americans?

1. Not at all 2. Somewhat 3. Very
How would you describe your state’s ability to provide support to family members taking care of older
Americans’ relatives?

1. Not at all 2. Somewhat 3. Very
How would you describe your state’s planning for a coordinated support system to detect gaps in
services and develop new resources to meet the needs of a changing older American cohort?

1. Not at all 2. Somewhat 3. Very
Has your state developed policies to encourage the development of a private long-term care insurance
market?

Yes No

Does your state support programs speciﬁcally designed for victims of Alzheimer’s disease?

Yes No

Describe the quality and effectiveness of your state’s safety net for the older Amerieans population.

1. Poor 2. Good 3. Excellent

How would you describe your state’s ability to implement federal older Americans human service
programs?

1. Poor 2. Good 3. Excellent

How would you describe your agency’s working relationship with your state’s ofﬁce on aging?

1. Poor 2. Good 3. Excellent

208

40. How would you describe your state’s adherence to the provisions of the Older Amerieans Act?
1. Poor 2. Good 3. Excellent
41. How would you describe your state’s ability to assess the needs and determine the priorities of the
current older American population?
1. Poor 2. Good 3. Excellent
the future older American population?:
1. Poor 2. Good 3. Excellent
42. How would you describe the efforts of your state government to provide information, referrals, case
management, protective services, and programs related to elder abuse?

1. Poor 2. Good 3. Excellent

APPENDIX C

APPENDIX C

INTERVIEW PROTOCOL
FEBRUARY 1997
STATE:
DEPARTMENT:
INTERVIEWEE:

F allowing-up on a survey conducted by the Council of Governors' Policy Advisors in the
Summer of 1995, I have some specific questions regarding the long-term planning and
innovative strategies your state is developing as they prepare for the aging of the baby
boom population. This information will become a part of a publication which is planned
to be released by the Council. No direct response will be attributed to you, but only to
the state. These questions should only take ten minutes to respond to. You are free not to
answer any of the questions asked and you may discontinue the interview at any time.

1. Do you think that your state-either in your agency or within the Governor’s
ofﬁce-has adequate policy capacity currently to address the challenges presented by the
shifting demographic balance of the 215t century? If yes, please tell me about how your
state has developed this policy capacity?

2. Is there a strategic policy development process within your state which will assist
in preparing the state-agencies, administrators, and political leaders--for these shifting
demographics of the 21 st century?

209

 

210

3. Tell me why you feel that your state has an innovative strategic policy process in
place to address the challenges faced by your state because of this growing aging
population?

4. Does this policy process or planning group involve cross-agency collaboration in
which your state develops policy for the 21 st century and the aging of the baby boomers?
Please explain.

4a. If there is this collaborative effort underway, what do you attribute to its initial
start-up? For example, was there speciﬁc technical expertise, political leader or policy
entrepreneur which spearheaded this collaborative effort?

5. Did your state-agency, collaborative or planning group--receive any special
appropriation to ﬁrnd this collaborative eﬂ‘ort, either to initiate it or allow it to continue?

211

6. Has your state agency considered the impact that the aging of the baby boom
cohorts and their mass retirement will have on state services and resources? Please rate
your agency's level of involvement and consideration about the following issues in regards
to the well-being of the future senior population in your state, on a scale from 1 - 3.

1=not considered an topic that concerns this agency

2=considered, recognizing this topic as a viable initiative for the future

3=seriously considered and currently engaged in comprehensive, long-term
planning in this area

 

State-funded or Initiated Retirement
Savings Programs

 

Workforce Development Programs
and Policies to Reﬂect the Older
Worker

 

Economic and Community
Development Strategies for an Aging

Society

 

Housing Initiatives and Policy
Changes Reﬂecting the Changing
Demands

 

Transportation Policies

 

Education and Life Long Learning
Programs

 

Community Based Health Care Efforts

 

 

 

Tax Policy Changes

 

 

7. Please share with me any innovative program that you are aware of in your state
which we should highlight as a potential model for the country?

APPENDIX D

 

AITWHWDEXID

THE WHITE HOUSE

WA SHINOYON
January 5, 1994

The Honorable Evan Bayh
Governor of Indiana
206 State House

Indianapolis, Indiana 46204

Dear Cove 3

I congratulate you on the creative, Innovative and practical
approach of the Indiana Consolidated State Plan on Service to
Children and Families. This plan should enhance collaboration
along federal, state and local programs as well as between the
public and private sectors. Through the Indiana Policy Council
on Children and Paeiliee and the Step Ahead Councils, you have
created a aechanise which encourages cos-unity based planning

aanageeent working together to transfers the etate, federal and
local response to children and faailiee.

vice President Gore joins ee in the belief that the reinvented
relationship of all levels of govern-ant to the delivery of
services is essential to the process of cos-unity espowersent.
we are pleased that the develop-ant of cos-unity values and
goals is a priority under your plan, and we are particularly
enthusiastic about the faaily focused, coapreheneive and
preventive principles of service. We urge you to carefully
consider the ways in which public funding can be used to
leverage private funding. He also encourage you to establish
clear bench-arks of progress, evaluating and aeasuring success.

As you know, under the leadership of Carol Rasco of the
Doeestic Policy Council, several federal agencies and Ie-bere
of a sub-group of the Cos-unity Enterprise Board have been
available to your representative, Cheryl Sullivan, as the plan
was introduced. Along than were the National Econosic Council:
the Vice President's Office and the Depart-ants of Agriculture,
Education, health and husan Services, Housing and Urban
Developeent, Justice and Labor. They have reviewed the lane
and set several tises, and are hopeful that your initiat we
will provide thee with an oppor-tunity to learn sure about
successful service'integrations and about the barriers created

212

 

by categorical funding, eligibility requiresents and regulations.
In addition, they have also contacted their-tiqi°n.l 01116.3.
where appropriate and sent letters indicating that YO“! EOPZC'
sentative has set with us. This relationship will cantinue as
the Plan and processes develop and continue to OVOlVO. You will

have at each agency and office soseone available to you to answer '-

continuing questions that we will need to resolve.

As I have frequently said, 'governsents don't raise children.
fasilies do'. An esphasis on learning directlY tr“. fﬁlili.’
about their needs will lead to reforss that will enable fasilies
to becose stakeholders in their own future and that of their
children and cossunities. It is our hope that the reinvented.
service delivery to children and to fasilies will lead to
cosprehensive plans for econoslc and husan develop-ant, since

we believe that econosic self-sufficiency is essential to the
revitalization of cossunities.

.We hope that one seasure of success will be in preventing the
probiess which necessitated the need for these services.

we look forward to learning, through the Indiana Consolidated
State Plan, isportant lessons about effectiveness, econosy and
cooperation. The Cossunity Enterprise Board will provide an
effective forus in which to review your trials and'triusphs.
Best wishes in your initiative.

sincerely,

37%

213