.5 ..‘..._4.

I_
‘-.a
v...‘

-
—
-r‘): ‘b

. ‘b‘r‘4

_.

 

.—_._

  
  

’v t“

flyIIIr “4;;-
DI '1

    

I: ,. DD?“ . ,
DVD; D DD I} I ”DDDDDF D

I' .Dr‘ DI , ‘
[4| :DDDD‘D‘ I'LD'D 13D

   
   
   
   
 

    
 

 
  
   
  

          
    
  
  
    
  
     
  
  

  
 
   
  
 
   

. I
.1{. ". fign| 44D 5 .. I 9 in:
.DDDDID DUD? DDD'IDDD IDDDDQ DII'..-,IDD'~ '_-.D I. L ,. LD‘ D V D IDD .
{GUD-III’DDDEWMA'D "IIDD D‘I IDD‘D'I "- .1 'h DID.“ ‘DD'A' DDD‘DDD
. D. 3' . '.' I, .‘ D 'D.
DDVLDDDD’Dj’DW'J ,1. MD .‘4'44:If;“.I DD .414 II' .._ “SW" D DDDDDDDDDD D
.‘1 ' ID? ‘ I" "D l I‘ - DDDD- . I'D-:ID.1 DDDD
'.I_"ID DD}? ‘I '£.IDIDDDD I! D' ' 431‘, 1' '.-' DD DD.“ D“-
. 3d - .4 i .5" o,. , ,:4.:4“‘ 4;: .'_I:. ‘ 1i 3-

hDD'DD; 'I I EDDID£ uglfl‘ DD'D‘D ‘ . " DD4f ~ D D
DEIDID .

PDDED ‘ND DIIDIIIQIDI. p:
“N- D}; C'J D 'I ‘l I:

 
   
  
 
 
   

  

.."I‘3’Y".D I4:'-. {DIM
_DD'DI'.DD1DԤD DMSH; "-DDD,

    
 
 

        
    
   
   
  
     
  
     
   
    
 

  

 
      

‘ ‘DDD DDDDDDSIDD (D: DD .D-D‘DDID'DDID' D D D; .‘D I
I II».- EDDDD 4444».- 4» ‘ "’4“. 4‘»
., .s . - DD‘.D , .ID I II (DDD' I‘DIID DD D-DDSDIIDDD;
III» 4‘ 4 44.». I“ 4.424 4II=4I..4444"'..4.44
ZDD', JD. Dgzb DDD I'DI. I DDDD-éD' ' . . DD E_ D; I' v - D
‘7‘ ‘ {.DDED DIN-ID» D ... 3&4 ‘ 'I' l4 4IDII

D

 
   
    

I,’ WD II. {I .,
. g ,.hg 4 4'4“ 4 :1
“3194.1 DDDIDDD II-DID' ID". :. DDIDIIIu‘DD
' "it. 35114;}: ID! ‘ .DQD'. ..,...:I.
$414K“ D 4 31.31-94.33 4
.‘DDDpDDDD D1II‘EI“,DII'; DD‘D‘DDD'DD

          

 

 
  

D DIDID ‘.',

DOIDII‘:
DDDD'DDDDD
I"

 

 

. . nit-7t

'

. DI. I.» 3‘ '4: ‘4

I 12D:.| . _ I
.'I4' D‘IDDIDDDDDI. 4‘ I .jD4 '
u I'. {DIM ' '

    
   
 

    
 

     
 
    
  

   

    
   

 
    

.4 "3"
.‘DD D‘ I _D.
D.'DD7IDI1D. I
D,‘ D.".:DD D ,D
D I . I DI . .-D. .I II.
.I-'_.I‘; 'I II" 4,. - '«DD'D- |.-¥ ' .,
D. ID..D‘.DDD D _.",'; DDDDDIID ‘. . . . . . 3 1‘4 I 4‘:.I,D:4 DD‘DD'DDDI. IHDDDD-DDDDDDDI I151: ‘4: '
4.» II «1.4 U.fnu .w 44' ,II I“ 4‘ .. 4 , . 4 . b DDDDDD‘DDU w;;fia. I
I I ‘ ‘ ..‘ . I .’.-':I- .
lm‘ 'DII'.' _‘;D_‘DI.'.4 4} . '. ' . . ' DID DDDDDDDDD DDIII'D ' - D
4 I» 4. '4 . ‘. ‘ 4"..444I LI 44:1'D.
. D' 2".” DDD'DD'I“D II: :' . ‘D’DII 4‘4" I'. 51: ' '. .. I‘ ID ,4”! .fDD I'D»DDDDIDDIDDDDD‘IDDIDDWDDD
DDDDD D ‘DID‘DD_IDD.D,.DD D‘DIDDDDDHDDD ,'.,DD I I DIIDDID DDDDDDDD DDDDD DDDDc DDDDDDDiDDDD DD D
ID... .. . . ~» III-
I

 
 

.av-Z

I ,DDDD ;D‘I ""41ID. DID '4'DD'I. DDDDDDDDD‘DD . ’ DDDD “D' DDIDDD DfDDD' DD DDDDDDDDD‘ DD'D'1
DID 'I DDDDDDDDDD . D DDDDDDDDDDDID DDDD DDID DIDDDDID‘. D 'DID DIDD DDDD. DDDDDDDD D D. .DDDDDD DDDDD'I I If DDDD’. IDIIDI D DDDIDDD IDDDDD‘DD. DDDDDDDDDDDD DD DDDIDDDD‘DI DD'D I

. I .D DDDD'DII.IIDD.DD, DDI'IDMDDTD‘D I I D D DDEIIDZD’ . I.

DDD DDD4D' DDDD'DD D I DDD DDDDDDDDDD DDDDDDD DDDDDDDDDDDD DIDDDIDDD DD I D D D D DDDDDDDD DDDDD DDDDDD DDD
. ‘ I'D DIIDD D“IDD|., I' DDDDDID IDDI IDD DDDDI‘DDD D I. ,‘DDDDIIDID. DIDD

""" ‘D4 II. 'D‘“ W "'DDDD'IDI.'DII.1"'.ID‘.'IDI.
.D'IDDDD

“21‘..-

       
 

. .H
fl. ‘ -
l—th

,fl.‘ ID D4

Ia.» DID
. I.» .II . .. ....I»4..4» .. MIDI

THESIS

llll“Willi“!UlUllHllHHlHlHUJill“!!!lllllllll

301712 0076

This is to certify that the

dissertation entitled

The Incidence and Housing Market Effects of Michigan's

1994 School Finance Reforms
presented by

Jeffrey Paul Guilfoyle

has been accepted towards fulfillment
of the requirements for
Ph.D. Economics
degree in

 

 

,K/

Major professor

 

 

Date//%/ /;;%

MS U is an Affirmative Action/Equal Opportunity Institution 0.12771

 

 

 

 

LIBRARY

Mlchlgan State
University

 

 

 

PLACE IN RETURN BOX
to remove this checkout from your record.
To AVOID FINES return on or before date due.

 

I MTE DUE DATE DUE DATE DUE

Mttztofifl FEB 11an

a MEI. e4 2 9 '14

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1M chlMpfiS-p.“

 

THE INCIDENCE AND HOUSING MARKET EFFECTS OF MICHIGAN’S 1994
SCHOOL FINANCE REFORMS
By

Jeffrey Paul Guilfoyle

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Economics

1998

ABSTRACT

THE INCIDENCE AND HOUSING MARKET EFFECTS OF MICHIGAN’S 1994
SCHOOL FINANCE REFORMS

By

Jeffrey Paul Guilfoyle

Michigan’s 1994 school finance reforms dramatically changed the way public
schools in Michigan are financed. As part of these reforms, reliance on local property
taxes as a funding mechanism was greatly reduced. Lost revenues were replaced by
increasing a variety of state taxes, including a state property tax. The reforms also shified
control over school funding decisions from the local school districts to the State
government.

This dissertation examines two aspects of the recent reforms. The major concern
of this dissertation is to determine the effects of inter-community property tax and school
spending differentials on house prices. House sales occurring before and after the
reforms are examined to see how differences in school property tax rates and school
spending amounts between communities are reflected in the price of housing. The nature
of the reforms and the availability of good sales data allow this study to avoid many of
the difficulties encountered by other studies.

This dissertation also examines the overall tax incidence of Michigan’s recent
school finance reforms. The incidence of the reforms is measured for three types of
households: a senior citizen couple; a family of four; and a single resident. For each of

these groups the incidence is measured under several different housing and community

assumptions. The measured groups should represent the experiences of a large portion of

Michigan’s taxpayers.

To Wendy

iv

ACKNOWLEDGMENTS

Many people helped to make this dissertation possible and I would like to briefly
thank some of them. First, I would like to express my gratitude and sincere appreciation
to my advisor, Professor Ronald Fisher, for guidance, advice, counseling and
encouragement. I am greatly in his debt for all he has done for me during my time at
Michigan State. I would also like to than the other members of my committee, Professors
Charles Ballard and Leslie Papke, for their insight, support, and valuable input.
Additionally, I would like to thank Jeffrey Wooldridge for his fi‘equent advice and help in
solving, what were to me, perplexing econometric problems.

I have had the good fortune of making many friends at Michigan State, several of
whom also contributed to my academic success. I especially want to thank Kathleen
Beegle, Jen Brown, Dan Hansen, and Marianne Johnson.

I thank my parents and family for their continuing love and support. I also thank
Steve and Sherry Mileski for all of their help and support (and baby-sitting) during my
four years at Michigan State.

But most of all I wish to thank my wife Wendy and son Paul. My tenure in
graduate school was in many ways harder for them than it was for me. I want to thank
them both for their love and support, for helping me through the difficult times, and for

all they have done to help make my graduation possible.

TABLE OF CONTENTS

LIST OF TABLES ..................................................................................................... viii

LIST OF FIGURES ..................................................................................................... ix

INTRODUCTION ......................................................................................................... 1
CHAPTER 1

MICHIGAN ’8 CHANGES ............................................................................................ 4

I. Overview ....................................................................................................... 4

II. Michigan’s School Funding History ............................................................ 4

111. Tax Changes ................................................................................................ 7

IV. Changes in Revenue Allocation ................................................................ 11
CHAPTER 2

REVIEW OF PROPERTY TAX THEORY ................................................................ 15

I. Introduction ................................................................................................. 15

II. The Benefit View ....................................................................................... 16

III. The New and Classical Views .................................................................. 21

IV. Yinger’s Theory of Capitalization ............................................................ 25

V. Predictions for Michigan ............................................................................ 29
CHAPTER 3

EMPIRICAL CAPITALIZATION LITERATURE .................................................... 32

I. Introduction ................................................................................................. 32

II. Previous Empirical Work ........................................................................... 33

Aggregate Studies ................................................................................ 36

Micro Estimates ................................................................................... 39

Natural Experiments ............................................................................ 41

III. Opportunities in Michigan ........................................................................ 44
CHAPTER 4

CAPITALIZATION ESTIMATES ............................................................................. 45

I. Introduction ................................................................................................. 45

II. Michigan’s Reforms ................................................................................... 47

III. The Econometric Model ............................................................................ 52

IV. Mean Sales Price Estimates ...................................................................... 60

V. Dual Sales Estimates .................................................................................. 65

VI. Conclusion ................................................................................................ 69

vi

CHAPTER 5

THE INCIDENCE OF MICHIGAN ’3 SCHOOL REFORMS ................................... 84
I. Introduction ................................................................................................. 84
II. Incidence Definition ................................................................................... 86
111. Tax Incidence Estimates ........................................................................... 87
Income Assumptions ............................................................................ 87
Sales Tax Estimates ............................................................................. 89
Effects of Property Tax Changes on Renters ....................................... 92
Effects of Property Tax and School Spending Changes on
Existing Homeowners .......................................................................... 96
Effects of Property Tax and School Spending Changes on
New Homeowners .............................................................................. 100
Changes in the Income Tax ................................................................ 103
The Cigarette Tax .............................................................................. 104
IV. Total Tax Incidence ................................................................................ 105
V. Service Level Changes ............................................................................. 108
VI. Conclusion .............................................................................................. 111
APPENDIX ............................................................................................................... 134
REFERENCES .......................................................................................................... 153

vii

LIST OF TABLES

CHAPTER 1
Table 1.1 Revenue Replacement Alternatives ................................................ 14
CHAPTER 4
Table 4.1 Events Leading Up to Michigan’s 1994 School Reforms .............. 77
Table 4.2 Full Capitalization of $1 per Year Under Differing Assumptions..78
Table 4.3 Summary Statistics for Average Sales Price Sample ...................... 79
Table 4.4 Estimates Using Means Sales Price for Each Community ............. 80
Table 4.5 Summary Statistics Dual Sales Data ............................................... 81
Table 4.6 Dual Sales Estimates ....................................................................... 82
Table 4.7 Summary of Effects of Tax and Spending Changes ....................... 83
CHAPTER 5
Table 5.1 Estimates of Income for Various Household Types .................... 121
Table 5.2 Sales Tax Estimates for Various Demographic Groups .............. 122
Table 5.3 Calculations of Rent Savings from Michigan’s School Reforms 123
Table 5.4 Property Tax Credit Estimates -- Renters .................................... 124
Table 5.5 Estimated Property Tax Payments and Credits for
Homeowners existing and new) -- w/o Reforms ........................ 125
Table 5.6 Estimates of Reform Effects on Existing Homeowners .............. 126
Table 5.7 Estimate of Reform Effects on New Homeowners ...................... 127
Table 5.8 Income Tax Estimates for Various Demographic Groups ........... 128
Table 5.9 Percentage Smoker by Income for 1993 ...................................... 129
Table 5.10 Total Tax Incidence for Smokers ................................................. 130
Table 5.11 Total Tax Incidence for Existing Homeowners ........................... 132
Table 5.12 Total Tax Incidence for New Homeowners ................................. 133
APPENDIX
Table A.1 Oakland County Population and Housing Data ............................ 136
Table AZ List of Dual House Sales ............................................................... 139
Table A3 Oakland County School District Tax and Spending Data ............. 152

viii

LIST OF FIGURES

CHAPTER 4
Figure 4.1 Property Value Growth 1990 - 1996 .............................................. 73
Figure 4.2 Property Value Growth 1990 - 1996 .............................................. 74
Figure 4.3 School Spending Changes 1990 - 1995 .......................................... 75
Figure 4.4 Average Millage Rates 1990 - 1996 ............................................... 76
CHAPTER 5

Figure 5.1 Tax Changes as a Percent of Income
(Non-Smoker - Brandon Twp) ...................................................... 113

Figure 5.2 Tax Changes as a Percent of Income
(Smoker - Brandon Twp) .............................................................. 114

Figure 5.3 Tax Changes as a Percent of Income
(Non-Smoker - Bloomfield Twp) ................................................. 115

Figure 5.4 Tax Changes as a Percent of Income
(Smoker - Bloomfield Twp) ......................................................... 116

Figure 5.5 Tax Changes as a Percent of Income
(Non-Smoker - F emdale) ............................................................. 117

Figure 5.6 Tax Changes as a Percent of Income
(Smoker - F erndale) ..................................................................... 118

Figure 5.7 Tax Changes as a Percent of Income
(Non-Smoker - Pontiac) .............................................................. 119

Figure 5.8 Tax Changes as a Percent of Income
(Smoker - Pontiac) ....................................................................... 120

ix

INTRODUCTION

In 1993 and 1994, the State of Michigan dramatically overhauled the way it
finances public schools. The reforms changed several of Michigan’s taxes and the local
property tax rates of many school districts. The reforms also changed the way in which
revenues are allocated to Michigan’s public school districts. In this dissertation, I
examine two aspects of the reforms in detail. First, I measure the effects of the property
tax and school spending changes on house prices in Michigan. Second, I estimate the
overall incidence of Michigan’s reforms.

I measure the effects of the property tax and school spending changes on house
prices in Michigan, using data on house sales from Oakland County. I generate two
measurements of the capitalization effect, using two different samples. The first sample
consists of observations of the average sales price, tax rate, and level of school spending
for a number of communities in Michigan’s Oakland County. With this sample, I find
that a $1.00 property tax differential between communities results in a $9.93 difference in
the average sales price of houses. This finding is consistent with the results of two
similar studies (Gabriel 1981 and Rosen 1982) that examine property tax capitalization in
the context of California’s Proposition 13. Gabriel finds that a $1.00 property tax
differential leads to a $12.00 difference in the average sales price of houses, while Rosen

finds that a $1.00 property tax differential leads to a $7.30 price differential With this

 

2
sample, I do not find evidence that differences in per-pupil school spending across
communities are reflected in house prices.

The second sample consists of observations of over 700 houses that sold once
before and once afier Michigan’s reforms. I argue that this sample is superior to the
average-sales-price sample for a number of reasons. Using this sample, I find that a $1
property tax differential between communities leads to a $4.25 difference in house prices.
I also find that an increase in school spending by one community of $100 per pupil will
increase house prices in that community by 0.5 percent, all other things held constant.
The results of this sample are substantially different than the results found using the
average sales price of houses. I argue that the dual-sales sample is superior, and that the
results found using the average sales prices of houses may be biased.

I then use these findings, along with a number of other estimates, to estimate the
incidence of Michigan’s school finance reforms. I find that non-smokers who owned a
house at the time of the reforms generally received a large tax cut. Non-smoking
individuals who purchased their houses subsequent to the reforms generally saw their
overall tax burden increase as a result of the reforms. For non-smoking renters, the
results are mixed. If property tax savings received by landlords are reflected in lower
rents, renters generally came out ahead as a result of the reforms. If property tax savings
received by landlords are not reflected in lower rents, renters generally came out behind.
There is little evidence yet as to the effects of the reforms on rents in Michigan.

For Michigan residents who are smokers, the cost of Michigan’s reforms were
much higher. As part of the school finance reforms, the state excise tax on cigarettes was

raised from $0.25 to $0.75 per pack. This amounted to a relatively large additional tax

cost for smokers. For low-income households with heavy smokers, the increased
cigarette tax is a large burden.

This dissertation is composed of five chapters and an appendix. Chapter 1
contains a detailed description of Michigan’s recent reforms. Chapter 2 contains a review
of the theoretical property tax literature. Chapter 3 examines the empirical capitalization
literature. Chapter 4 contains estimates of the differential effects of Michigan’s property-
tax cuts and school spending changes on the prices of owner-occupied housing. Chapter
5 contains estimates of the overall incidence of Michigan’s tax and spending changes.
The appendix contains a subset of the data that were used to generate the estimates in this

dissertation.

Chapter 1

MICHIGAN ’8 CHANGES

I. Overview

This chapter contains a detailed description of Michigan’s school finance reforms.
The information contained in this chapter was gathered from four main sources: Courant
(1982), which describes Michigan’s property tax system prior to the reforms; Brazer,
Laren, and Sung (1982), which describes K-12 public school funding in Michigan prior to
the reforms; Addonizio, Kearney, and Prince (1995), which contains the history of
Michigan’s recent reforms and a description of the tax changes; and finally, Kearney
(1994), which describes school funding after the reforms and also details the new

property tax system.

[1. Michigan’s School Funding History

Prior to 1973, Michigan allocated aid to school districts in the form of a lump sum
(as long as a minimum property tax millage was levied). Because aid was in the form of
a lump-sum grant, it created only income effects and not price effects, with respect to
local revenue decisions.1 The gain to a school district from raising its property tax rate

was directly proportional to the property wealth of the community.

 

' See Fisher (1996) for a discussion of the advantages and disadvantages of different forms of grants.

4

5

In 1973, Michigan passed the Bursley Act and began using what is known as a
district power equalization (DPE) method for allocating aid to school districts. DPE
funding methods seek to equalize the amount of revenue that school districts can raise
with a given tax rate. The aid given to a school district is based on the property tax base
of the district and on the tax rate levied. There were several reasons for Michigan’s
switch to a DPE format. In 1970, a popular book, Private Wealth and Public Education,
advocating the DPE method was published by Coons, Clune, and Sugarrnan. In 1971, the
California Supreme Court ruled in Serrano v. Priest that California’s reliance on local
property taxes for school firnding violated the state and federal constitutions. This ruling
raised concerns as to whether Michigan’s school finance system was constitutional.

The Michigan state government had been trying to reform school spending prior
to the Bursley Act. A ballot initiative that would have limited the use of local property
taxes for school funding was rejected by the state’s voters in 1972. In December of 1972,
the Michigan Supreme Court declared Michigan’s school financing program to be
unconstitutional. Although this decision was later reversed, the reversal did not come
until afier the Bursley Act was passed.2

The passage of the Bursley Act did not result in equalized spending for
Michigan’s school districts. Feldstein (1975) showed that DPE aid allocation methods
generally do not result in equalized spending, because voters in different school districts
will generally choose different spending amounts for their district. In general, richer

school districts choose higher spending levels, even if a DPE system is fully

 

2 The source for the pre-Bursley Act history is Brazer, Laren, and Sung (1982).

6
implemented. By the time Michigan’s reforms were implemented, Michigan’s DPE
system was no longer working correctly; districts with high tax rates did not always have
high spending levels.3 Roughly one-third of Michigan’s school districts were “out of
formula” meaning they received no firnding aid from the state. Spending in the out-of-
forrnula districts was highly correlated with wealth. In addition, because out-of-formula
districts received no aid from the state, they had little incentive to support funding
increases for the in-formula districts.

By 1993—94, per-pupil spending in Michigan’s school districts was highly
correlated with the districts’ property tax wealth (see Courant, Gramlich, and Loch 1982).
Additionally, the variance in property tax rates for school operations was high, ranging in
1993-94 from 8 mills to 47 mills.4

Michiganders were clearly unhappy with their school finance system. The
system’s high reliance on the property tax and the inequitable school funding that resulted
prompted many efforts for reform. However, Michigan citizens did not approve of any of
the alternatives offered them, either. Between 1972 and 1993, Michigan voters rejected
all 12 of the reform initiatives presented to them.5

In 1990, John Engler, then a candidate for Governor of Michigan, promised
property tax relief. Three years into his term, despite three different ballot initiatives,

Michigan’s system remained intact, and Engler’s promise remained unfulfilled. In July

 

3 Source: Addonizio, Kearney, and Prince (1995).
‘ A mill is a tax rate of $1 per $1000 of taxable property. In Michigan property is assessed at 50 percent of
its cash value so that a 1 mill tax should produce $0.50 per $1000 of property.

5 Source: Addonizio, Kearney, and Prince (1995).

7
of 1993, State Senator Debbie Stabenow proposed eliminating the property tax as a
method of funding schools. The bill Stabenow introduced did not provide for the
replacement of lost revenues -- approximately $6.5 billion. The bill was quickly passed
and the governor and legislature began work on replacing the lost revenues. Governor
Engler saw this as an opportunity to revise the method by which revenues were allocated
to the local school districts. The ballot initiative that was eventually enacted did change
the method of allocating revenue. It also restored some of the property taxes that had

been eliminated, and changed several other Michigan taxes.6

111. Tax Changes

Prior to 1994, the property tax consisted of taxes levied by 264 cities,
approximately 270 villages, over 1200 townships, 83 counties, and about 600 school
districts. The property tax in Michigan was used to fund a variety of state and local
services. Prior to Michigan’s reforms, however, the majority of property tax revenues
went towards funding public schools. In 1980, 68.2 percent of all property taxes in
Michigan went to school districts.7

School property taxes fall into three categories: operation; building and site; and
debt retirement. The operation tax was used to fund the operation of schools. In 1991,
the operations millage ranged from 8.37 mills to 46.25 mills. The building and site

millage accounted for a far smaller percent of taxes, ranging from zero, for 474 of

 

6 Ibid.

7 Source: Courant (1982).

8
Michigan’s school districts, to a high of 7.27. The debt retirement millage ranged from
zero, for 96 of Michigan’s school districts, to a high of 9.0.8

Prior to 1994, taxing units at the county level could levy a fixed number of mills
(up to 18 in some cases) without obtaining voter approval. Usually school districts were
granted between 6 and 11 mills of this taxing authority. This amount was known as the
district’s “allocated” millage. School districts could raise additional funds by asking the
voters of their districts to approve additional “voted millages.”’

Michigan’s property tax is calculated using a property’s assessed value. Michigan
law requires property to be assessed at 50 percent of its cash value.10 Because certain
state policies are based on a county’s or district’s assessed property wealth, the state
monitors the accuracy of local assessment practices. Each city, township, or village
assesses the property in its jurisdiction. The county and state then each check to ensure
that the total assessed value of property in the community is equal to 50 percent of its
cash value. Courant (1982) argues that Michigan’s assessment quality is among the best

in the nation in terms of assessment accuracy.

 

8 Michigan State Board of Education Bulletin 1014.

9 Source: Kearney (1994).

'0 As part of the 1994 reforms, a distinction was made between assessed value and taxable value. Property
is still assessed at 50 percent of cash value. In 1994, taxable value and assessed value were the same for all
properties. Increases in the taxable value of property are capped at 5 percent a year or the rate of inflation,

whichever is less. The taxable value is set back to 50 percent of the cash value upon the sale of the

property.

9

In August 1993, Michigan passed Public Act 145. This act eliminated the use of
local property taxes for school operations. These property taxes had accounted for 66
percent of K—12 school revenues in the 1993-94 school year (approximately $6.5
billion)“ Replacing the lost revenues was made more difficult by two provisions of the
Michigan Constitution. First, the state was limited in the share of total state personal
income that can be collected as a tax. In 1993-94, Michigan was approximately $4.2
billion under the cap. Second, raising the state sales tax, the preferred alternative,
required voter approval.‘2

In December, 1993, the state proposed its new tax system. First, to come in under
the State’s constitutional tax limit, a portion of the local property tax was restored.
Second, the legislature produced a proposal that contained two funding alternatives. The
preferred alternative raised the needed revenues primarily through an increase in the
state’s sales tax. As noted above, raising the state’s sales tax required the approval of the
voters. Wary of the history of school finance ballot initiatives in Michigan, the
legislature also proposed a fall-back plan in case the sales tax increase was not approved.
This plan relied on raising the state income tax -- something that did not require voter
approval. Table 1.1 outlines the competing proposals.

In March of 1994, Michigan voters approved the ballot proposal by a large
margin. Several aspects of the ballot plan should be noted. First, in addition to raising

the state’s sales tax, the cigarette tax was also increased from 25 to 75 cents per pack, the

 

” Source: Kearney (1994).

’2 Source: Addonizio, Kearney, and Prince (1995).

10

Michigan lottery was slightly expanded, and the state’s income tax rate was reduced from
4.6 to 4.4 percent. In addition, the property tax credit available to renters was slightly
increased.'3

Although the use of the property tax in Michigan for school funding was greatly
reduced, it was not eliminated. Owner-occupied houses, known as “homesteads” in
Michigan, are taxed at a rate of 6 mills for school operations. Non-homestead property,
which is all taxable property other than owner-occupied housing, is taxed at a rate of 18
mills. Although taxed at a higher rate than homestead property, the property tax rate for
non-homestead property in most communities was reduced. In addition, school districts
remain responsible for their own debts and can levy millages to pay for them. School
districts that were spending more than $6500 per pupil in 1993-94 were required to levy
additional “hold harmless” millages, if they wished to maintain their high spending
levels. An assessment cap has been imposed that limits annual assessment increases on
property to the lesser of inflation or five percent. The cap is reset upon the sale of the
property. Although the property tax for school operations is relatively uniform now, prior
to the reforms it varied widely by district. Therefore, the size of the property tax cut

varied by district as well.‘4

 

‘3 lbid.

” Source: Kearney (1994).

11
1V. Changes in Revenue Allocation

In 1994, Michigan fundamentally changed its school funding methods. Funding
decisions were, for the most part, removed from the local districts. These decisions are
now made primarily at the state level. As part of these changes, Michigan replaced its old
DPE system with a foundation grant approach. Under a foundation grant approach, each
school district is guaranteed a minimum funding level.

The state of Michigan now guarantees that each district will receive a minimum
level of funds known as the district’s foundation grant allowance. Technically, the
foundation allowance for 1995 was $5,000. However, due to the large cost of moving all
districts up to the foundation grant at once, the state decided to implement the changes
gradually."

Districts that were spending below $4,200 per pupil in the 1993-94 school year,
were moved up to $4,200 in 1994-95, or by $250, whichever was greater. Districts
spending above $4,200 and below $6,500 were increased from the 1993-94 funding levels
according to a sliding scale. The formula for this scale is:

93-94 Revenue per pupil + ($250 - ($90*((93-94 rev per pupil - $4200)/$2300)))

For example, a district spending $5,000 per pupil in the 1993-94 school year

would see their per-pupil revenue increase by approximately $219, a 4.4 percent increase.

The effect of this formula was that districts closer to $4,200 received larger percentage

 

‘5 lbid.

12

increases than districts closer to $6,500. All districts with 1993-94 spending below
$6,500 were taxed at the same millage for school operations.l6

Districts spending above $6,500 in 1993-94 were allowed to maintain their high
spending levels, but were required to levy an additional “hold harmless” property tax.
These districts were allowed to increase their funding by $160 per pupil in the 1994—95
year.‘7

The basic foundation allowance is expected to increase annually. The changes in
funding for school districts will be based on changes in the basic allowance. The dollar
increase in the foundation allowance is calculated by taking the previous year’s
foundation allowance and multiplying it by a number known as the final index. The final
index is based on changes in the School Aid Fund and the changes in pupil head count.
The School Aid Fund is composed of a number of taxes earmarked for school spending.

Districts still spending below the foundation allowance after 1994-95 will be
moved up according to a sliding scale. This scale is used to ensure that districts below
the basic allowance are increased at a rate higher than the rate for districts already above
the foundation allowance. For districts already above the allowance, revenue is increased
by the same number of dollars that the basic foundation grant is increased. Thus, all of
these districts are increased by the same dollar amount. However, this dollar amount

represents a larger percentage change for lower-spending districts. Essentially, some of

the difference in dollars spent by districts is being held constant. However, as the

 

'6 Ibid.

'7 Ibid.

1 3

foundation allowance rises, this difference will represent a shrinking percentage of
spending.‘8

The State School Aid Act also provides for special and categorical grants. These
grants finance a number of programs, including special education, gifted and talented
programs, bi-lingual education, and at-risk education programs. In 1994, the State
specifically allocated $230 million in funds for districts with a high level of poverty.
These funds will help offset any differences in per-pupil costs faced by the districts. To
what extent these firnds match up with costs is a subject for further research. Finally, no
attempt is made by the state to adjust the foundation grant for cost differences due to
factors such as climate, wages, or other “non-student” factors.‘9

Michigan’s reforms centralize much of the decision making for local public
schools. Increased educational equity is gained at the expense of local choice. Courant,
Gramlich, and Loch (1995) refer to this potential problem in noting that many more of
Michigan’s citizens will be off their demand curve with respect to educational spending.
Many poor school districts will have access to far more school revenues than their
citizens would have chosen. Many rich districts will have access to less revenues than
their citizens would have chosen. This effect could lead to pressure to change the system

in the future.

 

'8 Ibid.

‘9 Ibid.

14

Table 1.1 Revenue Replacement Alternatives

 

 

Tax 1993-94 1994-95 Ballot Proposal 1994-95 Statutory
(Pre-Reforrn) (Approved April, 1994) Alternative
Local Property Tax All property: Homesteads: 0 All property:
34 mill Non-homesteads: 12 mills
average 18 mills
State Property Tax None All property: 6 mills Homesteads: 0
Non-homesteads:
12 mills
State Sales Tax 4% 6% 4%
State Income Tax 4.6% 4.4% 6%
Income Tax Personal $2,100 $2,100 $3,000
Exemption
State Real Estate None 2.0%a 1.0%
Transfer Tax
State Cigarette Tax 25 cents 75 cents 40 cents
(per pack)
Single Business Tax 2.35% 2.35% 2.75%
Interstate Telephone None 6% 4%
Tax
Keno Lottery None Plannedb Not Included

 

Note: The ballot proposal was the one eventually enacted.
a. The tax was subsequently lowered to 0.75%.

b. After enactment of the ballot proposal, Governor Engler indicated that he did not
intend to implement a Keno lottery but would instead seek Michigan inclusion in a multi-

state lottery.

Source: Addonizio, Kearney and Prince (1995).

Chapter 2

REVIEW OF PROPERTY TAX THEORY

I. Introduction

In this chapter, I review the theoretical economic literature of the property tax on
residential housing. The effects of a non-residential property tax are not examined.
Because empirically estimating the degree to which interjurisdictional tax differences are
capitalized into property values is the primary focus of my research, special emphasis is
given to the capitalization effects predicted by the various theories. If property tax
differentials are reflected in the prices of otherwise identical properties, the tax
differential is said to be capitalized. Full capitalization occurs when the prices of
otherwise identical properties differ by the full present value of the property tax
differential.

The degree to which property taxes are capitalized is an important element in
examining property tax policies for several reasons. First, the degree to which the tax is
capitalized has an effect on the incidence of the property tax. If a property tax change is
fully capitalized, the selling value of the asset is reduced by the present discounted value
of the tax. Under full capitalization, the owners of property at the time of a tax change
bear the full burden of the tax. They are unable to escape the tax by selling their
property, because the property’s value has been reduced by the full amount of the firture

tax stream. Second, capitalization is important for determining the efficiency of the

15

16
property tax. The degree to which the property tax is a distorting tax on capital is to some
extent determined by the degree of capitalization. The ability of local governments to
provide local public goods efficiently, and the amount of redistribution that occurs
through local public good provision, are also in part determined by the degree of property
tax capitalization.l Finally, different economic theories of property taxes have somewhat
different predictions concerning the extent to which property taxes are capitalized.
Examining the degree to which a property tax change is capitalized can help in the
evaluation of these theories.

This chapter is divided as follows. The next section discusses the “benefit” view
of the property tax. The third section discusses the “new” and “classic” views of the
property tax. The fourth section discusses the contributions to capitalization theory made
by John Yinger. The final section discusses the implications of these theories for

Michigan’s tax changes.

11. The Benefit View
Although a decentralized market system has many advantages, it generally will
not produce the optimal level of public goods. In his classic article, Samuelson (1954)

argues that “no decentralized pricing system can serve to determine optimally these levels

 

‘ Many models examining the efficiency of the property tax assume that the local public service level is
selected by voters who must use the property tax to finance expenditures. These models are no longer
valid for Michigan, at least with respect to school finance, due to Michigan’s centralization of local school
spending decisions. Information on the degree to which property taxes are capitalized is still useful,

however, in determining whether local decision making will lead to an efficient outcome.

17

of collective consumption.” (p. 388). Writing in response to Samuelson, Charles Tiebout
(1956) argues that while this may be true for public goods provided at the federal level, it
is not necessarily true for local public goods. Tiebout develops a list of assumptions
under which it was possible for local public goods to be provided at an efficient level.
Tiebout’s model consists of a large number of local governments that finance
expenditures via a head tax. Consumers reveal their preferences for local goods through
their choice of residence. Consumers preferring a high level of public services live in a
locality with high taxes and services. Consumers preferring a low level of services live in
a locality with low taxes and services.

Tiebout’s model contains a number of highly restrictive assumptions. One of
these assumptions is that local expenditures could be financed via a head tax. 2
Subsequent authors have examined the implications of replacing Tiebout’s head tax with
the more commonly observed property tax. This modeling change has led to the
“benefit” view of the property tax. In this section, I discuss the benefit View of the
property tax. This view argues that the property tax does not necessarily lead to
inefficiencies.

Hamilton (1976) argues for the benefit view of property taxation. He first notes

that, in a system of communities that are homogeneous with respect to house values, the

 

2 Other assumptions in Tiebout’s model include: consumers are fully mobile; consumers have perfect
knowledge concerning local revenues and expenditures; a large number of communities exists for
consumers to choose among; there are no community based employment restrictions; there are no
extemalities associated with local public service provision; and communities are sized so that they can

produce services at the minimum of their average cost curve.

18
property tax can be seen as a system of average cost pricing for public services. Hamilton
assumes in his analysis that all households in a community consume the same level of
public services and place the same value on these services. Homogeneous communities
can be achieved through a binding zoning requirement. That is, all households in a
community are required to consume a minimum amount of housing services, the tax on
which is exactly equal to the cost of providing public services to each household. No
household has any incentive to consume more than the minimum level of housing
services, because they would then be paying for more services than they receive (and
would be better off in another community). If the zoning requirement is in fact binding
and communities are homogeneous, public services are provided efficiently and the
property tax is non-distortionary (the zoning requirement keeps households from
adjusting their housing consumption in response to the tax).

The binding zoning requirement has been strongly criticized by Zodrow and
Mieszkowski (1983). They argue that these assumptions would turn any tax into a non-
distortionary tax. The choice of tax instrument at the local level would become irrelevant
with the appropriate zoning restrictions, as all taxes would become a non-distortionary
fee for public service.

Hamilton, however, extends his argument to communities that are not
homogeneous. He argues that capitalization effects can lead to average-cost pricing for
public services in non-homogeneous communities. In a neighborhood with inexpensive
and expensive houses and no capitalization, residents of the inexpensive houses enjoy
public services at a lower cost than their neighbors in the expensive housing. This makes

living in a relatively inexpensive house desirable. Therefore, people will compete to live

19
in fiscally advantaged housing, driving up the price of such housing. Thus, inexpensive
houses, although they have lower property taxes, sell for more due to the capitalization of
the fiscal benefits. Likewise, housing that is relatively luxurious compared to other
houses in the same community will sell at a discount.

If differences in fiscal surpluses are fully capitalized, then there is no advantage to
buying an inexpensive house in a rich neighborhood. The price of the inexpensive house
is increased by the present discounted value of the difference between its taxes and
service cost. This has potentially important implications. First, it means that property
taxes do not lead to horizontal inequity. Everyone gets exactly the services that they pay
for. Second, because people pay for exactly the services that they receive, they demand
an efficient level of public services.

For the property tax to be non-distortionary, however, it must also not distort
housing decisions. Hamilton (1983) argues that the property tax can be converted into an
efficient price for public services, if capitalization causes the following relationship to

hold.

1) v + r = C(H) + C(LPS)

The above equation states that efficiency requires that house value, V, plus taxes, T, must
equal the cost of providing housing, C(H), plus the cost of providing local public
services, C(LPS).

If the above relationship is to hold, then fiscally advantaged housing, which has a

higher value, must somehow cost more to produce. In Hamilton’s model (1976), fiscally

20
advantaged housing costs more to produce because its fiscal advantage is capitalized into
land values. Inexpensive housing is built on land zoned for that purpose. This land,
because it can be used for houses that will sell at a premium, sells for a higher price than
other land in the community. Hamilton implicitly assumes that property taxes are fully
capitalized into land values. He assumes that land is in fixed supply; therefore, the tax is
not distortionary.

This analysis would imply that land zoned for inexpensive housing in a
heterogeneous community would sell for more than land zoned for more luxurious
housing. Zoning ordinances are required to keep this land price differential in place.
Otherwise, the amount of land used for inexpensive housing will increase and the land
devoted to relatively expensive housing will decrease. If inexpensive housing is allowed
to expand until the land for inexpensive housing and the land for relatively more
expensive housing sells at the same price, households owning cheaper houses will pay
less for their public services than households owning relatively expensive houses.
Redistribution would occur through the provision of public services. Residents no longer
get only what they pay for, and the outcome can no longer be expected to be efficient.

Based on his analysis, Hamilton draws the following conclusions. First, with full
capitalization, there is no horizontal inequity. Consumers get the services that they pay
for. Second, efficient supplies of housing and public services exist when: land value is
the same in all homogeneous (with respect to house value) communities; in mixed-value
communities, land-value differentials exactly reflect the present value of fiscal surplus
differentials; the mean value of land (per acre) is the same in all communities, regardless

of their housing or service mix. Finally, the property tax generates an incentive for the

21
production of an inefficient amount of low—income housing. This incentive must be held
in check through zoning restrictions for efficiency to be maintained.
Hamilton cites some evidence that the above conditions may in fact hold. He
argues (1976) that empirical studies have shown that land zoned for low-income housing
does indeed sell at a premium. He also cites empirical studies (1983) that find a large

degree of property tax capitalization.

III. The New and Classical Views

An alternative to the benefit view of property taxes is the “new” view. The new
view of the property tax was developed by Procter Thomson (1965), Peter Mieszkowski
(1972), and Henry Aaron (1975). The discussion of the new view of the property tax
presented here is taken primarily from a review of the property tax literature by
Mieszkowski and Zodrow (1989) and a discussion of the new view in Zodrow and
Mieszkowski (1983).

The new view of the property tax generally examines the property tax in a
standard capital taxation framework. These models follow the Harberger (1962)
approach to examining the taxation of capital. Harberger models generally assume
perfect competition and that the overall capital stock in the nation is fixed. Additionally,
the capital stock is assumed to be perfectly mobile within the nation, so that the after-tax
return to capital in all sectors of the economy is the same.

Housing is viewed as the output of a production process that combines land and
capital in the production of housing services. These models examine the effects of a

property tax in one or more sectors. Although the results vary somewhat depending on

22
the model specifications, they often produce the result that the capital portion of a
uniform national property tax reduces the after-tax rate of return to all capital in the
nation by the amount of the property tax. The property tax impacts other sectors in the
economy because capital is perfectly mobile. When the property tax is imposed, it
reduces the return to capital in the housing sector. Capital flows from the housing sector,
increasing its pre-tax rate of return, into other sectors, reducing their rate of return. This
continues until all sectors have the same after-tax return to capital. Capital generally
bears the full brunt of taxation in these models, because it is assumed that the total capital
stock is supplied inelastically.

If the property tax were the only tax in the economy, it would have two effects.
First, the property tax would result in an inefficiently low allocation of capital to the
housing sector. Second, the overall rate of return to capital in the economy would be
reduced. This rate reduction would have an effect on the long-run capital formation in
the economy. This long-run effect is the subject of much debate in the literature and is
beyond the scope of this analysis.

Of course, there are other taxes on capital, most notably the corporate income tax.
Gravelle (1994) argues that housing, especially owner-occupied housing, is taxed at a
very low rate compared to other forms of capital. She cites studies that find that the
preferential treatment of housing accounts for about half of the distortions arising from
the misallocation of capital. Because of this, the property tax may actually help to correct
the misallocation of resources resulting from the corporate income tax.

The new view also has implications for the redistributional effects of the property

tax. The property tax finances local public expenditures by reducing the rate of return of

23
capital in all sectors of the economy. This means that, for a uniform national property
tax, capital owners bear the full burden of the tax on the capital portion of housing.
Because capital tends to be concentrated in the hands of relatively wealthy individuals,
the capital portion of the property tax is progressive. This result directly conflicts with
Hamilton’s findings that no redistribution occurs through the property tax.

The property tax, of course, is not uniform throughout the nation. Under the new
view, tax differentials between communities (that is, differences between local tax rates
and the national average tax rate) give rise to “excise” effects. A tax rate higher than the
average will be either shifted forward into higher housing prices, or shified backward to a
relatively immobile factor. Different models have different predictions as to the degree to
which this will occur. The predictions of these models generally depend on the elasticity
of substitution between land and capital in the production of housing, and assumptions
concerning the mobility of consumers and the degree to which land is in fixed supply.

For example, consider property that rents for $100 per year. If the discount rate is
10 percent, and we assume the asset is infinitely lived, the present discounted value of
this property, in the absence of taxation, is $1000. If we impose a property tax of $10 per
year and residents are completely immobile, we might see the tax fully forward shifted to
renters. This means that the cost of renting this property would rise to $110 and the value
of the property would remain unchanged. If renters are fully mobile, however, we might
see the tax fully backward shifted into the property’s price. Backward shifting means that
the property still rents for $100, but now the owner must pay $10 in tax. The net return is
$90 per year reducing the present discounted value of the property to $900. In the case of

full backward shifting of the tax, we say that the tax has been fully capitalized.

24

The predictions of the new view models are based on the assumptions made
concerning various parameter values. For example, Hobson (1986) finds that the degree
of property tax shifting depends in part on the relative sizes of the elasticity of
substitution between housing and other consumption goods, and elasticity of substitution
in housing production, as well as the mobility of the resident population.’

In their review of the property tax literature, Mieszkowski and Zodrow (1989)
argue that the new view is a more general case of the “classical” view of the property tax.
The classical view (Simon 1943, Netzer 1966) examines the property tax in a single
jurisdiction. This view argues that, since capital is perfectly mobile, it bears none of the
burden of the property tax. The capital portion of the property tax is assumed to be
entirely forward shifted in the form of higher housing prices, and the land portion is
assumed to be borne entirely by landowners because land is assumed to be inelastic in
supply. Under this View, the property tax is much more regressive than under the new
view, because a much higher proportion is borne by renters.

Mieszkowski and Zodrow argue that the classical view focuses exclusively on the
excise effects portion of the property tax. This arises because the classical view uses a
partial rather than a general equilibrium model. The full forward shitting found in the
classical view is just a special case of the possible results predicted by the new view.
Although the overall rate of return on capital appears unchanged in the classical view,
Mieszkowski and Zodrow argue that it is in fact reduced by an infinitesimal amount.

This reduction occurs because the higher property tax rate in the metropolitan region

 

3 The description of Hobson’s results are taken from Mieszkowski and Zodrow (1989).

25
reduces the national average by a small amount. This infinitesimal change multiplied by
the entire capital stock of the nation is in fact large relative to the revenue raised by the
tax.

Mieszkowski and Zodrow argue that, under the new view, there is a tendency
towards under-provision of local public services. This occurs because local jurisdictions
are reluctant to tax mobile capital. Additionally, local jurisdictions do not take into
account the possible extemalities associated with taxing mobile capital. Capital that flees
a jurisdiction to escape a local property tax benefits other jurisdictions. This reduces the
overall cost of taxing mobile capital in a way not taken into account by the taxing

jurisdiction. Again, they find that these results are sensitive to model specification.

IV. Yinger’s Theory of Capitalization

Yinger (1982) develops a model in which he derives a household’s bid for
housing. He argues that the amount a household is willing to pay for a unit of housing
services, in a particular jurisdiction, is based on the jurisdiction’s level of services and
taxes. Therefore, a household’s bid for housing services can be written as P = P(E,t),
where E is the level of local public services per household and t is the effective property

tax rate. The value of a house to a given household can be written as

2) V(E, t) = P(E,t)H/r

26

where H is the total amount of housing units consumed and r is the discount rate.4

Household utility is assumed to be a function of housing services, the level of
local public services, and a composite private good, Z. The household is assumed to
maximize this utility function subject to its budget constraint. In deriving the first-order
conditions, Yinger assumes that the household is choosing Z, H, E, and t. The amount a
household is willing to bid for housing, P(E,t), is derived by solving for P from the first-
order conditions of the utility maximization problem. Solving for P(E,t) requires solving
two first-order differential equations, and requires the assumption of a specific utility
function. Additionally, the solution to P(E,t) contains a constant of integration. I discuss
the method of solving for this constant shortly.

If a Cobb Douglas utility function of the form, U = c,ln(Z) + c21n(H) + c3ln(E), is
assumed, and if housing services are assumed to be a multiplicative function of housing

characteristics, XI to XM, the value of housing can be expressed in the following form:5

3) ln(V) = ln(v) + (c3/c2)ln(E) - ln(r + t) + 2a,ln(X,.)

Based on this derivation, Yinger draws a number of conclusions. First,

differences in service levels between jurisdictions will be inexactly capitalized. The

 

" Note that this formula implicitly assumes an infinite lifetime for housing. Given the long expected
lifetime of housing, Yinger argues that this should be a close approximation to the actual present
discounted value.

5 Yinger (1982) derives equation (3) in part by solving a differential equation. The v term represents the

constant of integration.

27
degree of capitalization will be based on taste parameters in the utility function. Second,
because there are no taste parameters on the tax term in the value equation, differences in
tax rates between communities will be exactly capitalized-— regardless of the tastes of
consumers.

Interestingly, Yinger also claims that his results imply that homogeneous
communities can be formed without zoning barriers. If PE, the derivative of the bid for
housing with respect to services, increases with income, Yinger argues that high-income
households will outbid low-income households for housing in high-service jurisdictions,
so that zoning barriers are not required to sustain the mean property tax base.

This conclusion, however, does not follow from Yinger’s model. If the amount of
housing required to enter a community were fixed, as in a zoning requirement, then it is
true that rich households would outbid poor households. Yinger’s model does not impose
this restriction, however. Households are free to choose the level of housing they
consume. Rich and poor households would probably select very different levels of
housing, given a tax and service package. The assertion that the rich will bid more for
one unit of housing is not relevant, since households will consume different amounts of
housing. Yinger’s model does not prevent the conclusion that poor families will try to
consume a small amount of housing in a high-service jurisdiction.

Yinger recognizes that if the price for housing is different in different
jurisdictions, suppliers will have an incentive to supply houses to jurisdictions with high
house values. But Yinger concludes:

“...the supply of land within a jurisdiction is fixed, so the
conversion of nonresidential land into residential land

cannot continue indefinitely. Once all profitable
conversion has occurred--that is, in long-run equilibrium--

28

local fiscal variables will be capitalized into house values.”
(1982,p.935)

Developers will not simply continue to add new jurisdictions, because there are other
factors affecting house values besides the service/tax package. Expanding further from
the city center will cause the price households are willing to pay for housing services to
fall, ceteris paribus. This is due to the increase in commuting costs for residents in more
distantly located suburbs. Profits for housing developers will fall as they continue to
develop further and further from the city center."

Yinger specifies an equilibrium condition for determining the furthest point of
community expansion. At the furthest point of expansion, the price of housing is exactly
equal to the opportunity cost of resources used in production. The price of housing in this
jurisdiction, along with the tax and service level, can be used to solve for the constant of
integration in Yinger’s housing bid model. Yinger (1985) argues that the property tax in
this base jurisdiction is distortionary with respect to housing consumption decisions.
Variations from this base tax rate, however, are not. These variations will be perfectly
capitalized into housing values. Therefore, relatively high property tax rates in one
jurisdiction will not repel capital, since the rate will be fully capitalized into immobile

factors.

 

° Yinger notes that some models of capitalization use flexible boundaries for communities, leading to
communities with the most favorable tax and service packages expanding into the territories of less
favorable communities. However, he argues that in practice community annexation is relatively rare and

community boundaries seldom change.

29

Yinger’s conclusions differ from those of the new view in that, in Yinger’s model,
it is the average tax rate in the metropolitan area that is distortionary. In the new view, it
is the average tax rate on capital in the nation that variations are measured against.
Mieszkowski and Zodrow (1983) have argued that Yinger’s model is in fact consistent
with the new view. Unlike the new view models, Yinger does not explicitly model
housing production. Agents are choosing a level of housing services that are presumably
made up of land and capital; but he does not explicitly model this. The baseline tax rate
distorts the housing decision; since the capital/land components are not explicitly
modeled, there is no prediction as to the effect on the overall return to capital. Yinger
does argue that deviations from the baseline tax rate (equivalent to the excise effects of
the new view) are completely capitalized. The new view writers are more agnostic about
the excise effects. In the new view, these effects generally depend on the model

parameters-especially assumptions concerning resident mobility.

V. Predictions for Michigan

The economic theory of property taxes does not have clear implications for the
effects of Michigan’s property tax cuts. Different theories of the property tax reach very
different conclusions. For example, the benefit view sees the property tax as a non-
distortionary fee for public services; the new view sees the property tax as a distortionary
tax on capital that is somewhat progressive; and the classical view sees the property tax
as being somewhat regressive.

The new view focuses on the average rate of property (or capital) taxation in the

nation. Courant (1982) has argued that, because Michigan only represents about four

30
percent of the US. economy, the general equilibrium effects can safely be ignored when
examining Michigan’s property tax. Therefore, the relevant effects of the Michigan
property tax cuts depend on the partial equilibrium or “excise” effects. Yinger argues that
only a “base” community’s tax rate will lead to a distortion, and all deviations from the
base in a metropolitan area will be capitalized. The new view argues that the national
average tax rate on capital is the comparison base, and is somewhat more agnostic in
predicting the effects of deviations from this average rate. The effect of Michigan’s
property tax out under the new View depends in part on the elasticities in the production
and consumption of housing.

The property tax theories do not have clear predictions for the effects of
Michigan’s school reforms. Residents who live in owner-occupied housing will clearly
receive a benefit in the form of lower annual taxes. What is less clear is the extent to
which the reduced taxes will be capitalized into the values of their houses. If the benefits
of reduced taxes are capitalized into house values, house owners will receive a capital
gain on their houses. Future buyers will receive a smaller benefit from the lower taxes,
because they will have to pay higher prices for their houses.

For renters, the incidence is harder to predict theoretically. With mobile capital,
renters were bearing the portion of the property tax that was shifted forward. A property
tax cut should relieve them of this burden. Land owners will benefit from a reduction in
the tax, if part of the tax had been shifted into the value of land. Renters may also benefit
from a reduction in the price of owner-occupied housing. Renters who were on the

margin between renting and buying a house, or who planned to buy a house in the future,

31
will benefit from any reduction in the annual cost of owning a home. The extent of these

benefits cannot be determined by theory and needs to be investigated empirically.

Chapter 3

EMPIRICAL CAPITALIZATION LITERATURE

I. Introduction

This chapter reviews the empirical literature on capitalization. The approaches
taken by other authors, their major findings, and some criticisms of the earlier studies are
all discussed in this chapter. This chapter draws heavily on Bloom, Ladd, and Yinger
(1983), and Yinger et a1. (1988). These works provide an excellent review of the large
number of empirical studies in this area. They also discuss some of the major obstacles
associated with estimating the degree of property tax capitalization.

This chapter is divided into three sections. The next section discusses the
methodology and results of some of the major capitalization studies. Rather than trying
to list all of the studies that have been done on capitalization, this section focuses on
listing some of the more important works in this area. The final section discusses

opportunities for estimating the degree of property tax capitalization in Michigan.

32

33

II. Previous Empirical Work'

This section discusses some of the findings of previous authors. The authors
discussed in this section are primarily interested in discovering the degree to which
property tax differentials are capitalized into the price of houses. As discussed in the
chapter on property tax theory, the property tax has two effects. First, the average level
of property taxes in a metropolitan area (or a nation, depending on the theory) is expected
to distort the housing decision. This average level of taxation may change the amount of
capital or land used in housing, and in addition, some portion of the tax may be
capitalized into the price of houses. The second effect results from differences in the
property tax rates on houses within a metropolitan area. These differences can either be
interjurisdictional, resulting from different tax rates levied in neighboring communities,
or intrajurisdictional, resulting from different tax rates levied on houses within a
community as a result of assessment practices.

As discussed in the previous chapter, the property tax theories are not in
agreement as to the degree to which property taxes are capitalized. The new view
approach generally models property taxation in a general equilibrirun framework similar

to the one used by Harberger (1962). The degree of capitalization predicted by the new

 

' Note that many of the problems with the studies noted here have also been identified and discussed by

Yinger et al. (1988).

34
view depends on assumptions concerning factor mobility, and the elasticities of
substitution between capital and land.2

An alternative approach is the one followed by Yinger (1982, 1985). Yinger uses
an urban model with a central business district. This model predicts that tax differentials
between communities will be fully capitalized. The model also predicts the full
capitalization of tax differences resulting from assessment errors.

These models generally assume complete information. That is, the models
assume that agents have complete knowledge of both the current tax rates and of what the
future path of tax rates will be. Of course in practice, complete information is unlikely.
In their study of intrajurisdictional capitalization in Massachusetts, Yinger et al. (1988)
find substantially less than full capitalization. They argue that uncertainty surrounding
the likely persistence of tax differentials in the communities they study reduces the
degree of capitalization.

The studies discussed in this chapter focus on estimating the capitalization of tax
differentials between communities, rather than estimating the effect of the average tax
rate on property values. There is generally little or no sample variation in the average tax
rate, making it difficult to measure the effect of the average tax rate on housing prices.3

Capitalization studies are generally interested in measuring the degree to which
property tax differences are capitalized. A property tax differential is said to be fully

capitalized, if the difference in price between two otherwise identical properties is equal

 

2 These theories are discussed in the previous chapter. For a more in depth discussion see Mieszkowski
and Zodrow (1989).

3 For an exception that does focus on the average tax rate see Wassmer (1993).

35
to the present discounted value (PDV) of the tax differential. The price differential
between two properties can usually be observed. Measurements of the PDV of this
differential, however, cannot. The PDV is calculated from the price differential using
two parameters, the discount rate and the time horizon. The discount rate and the time
horizon must be chosen by the researcher and the choices have varied greatly.4

Assumptions concerning the discount rate and time horizon can have a large effect
on the interpretation of the capitalization results. For example, King (1977) assumes a
discount rate of 5 percent and a time horizon of 40 years and states that 67 percent of
property tax differentials are capitalized. Yinger et al. (1988) recalculate the degree of
capitalization in King’s study using their preferred parameters of a 3 percent discount rate
and an infinite housing life, and find the degree of capitalization to be just 36 percent.
Therefore, any estimates of the degree of capitalization will be sensitive to the parameter
choices.

Yinger et al. (198 8) recalculate the capitalization findings for many of the studies
discussed below assuming a 3 percent real discount rate and an infinite housing life. This
convention is followed for the remainder of this chapter so that the results of the different
authors can be easily compared. The assumption of a 3 percent discount rate and infinite
housing life are rather conservative. The effect of varying these parameters on the

capitalization estimates is discussed in greater detail in the next chapter.

 

" An exception to this approach is the one taken by Do and Sirmans (1994). They begin by assuming full
capitalization and a 25 year time horizon and then estimate the discount rate to be 4 percent. Of course,

their estimate of the discount rate relies on the strong assumption of full capitalization.

36

The capitalization studies can be divided into three broad categories. First, there
are “aggregate studies” which use aggregated house price and tax figures such as the
median house price and tax rate for a community. This aggregate figure is often used in
cases where more detailed housing information is not available. Second, there are
“micro” studies that use individual houses as observations. These two types of studies
reflect a trade-off in the types of data that are available. The aggregate studies generally
contain a large number of communities, so that there is a large amount of sample
variation in the tax rate. However, these studies use an aggregated house value measure
that is of lower quality for estimating purposes than individual house observations.
Micro studies tend to have a higher quality dependent variable-«the actual sales price of
individual houses. These studies, however, tend to involve fewer communities so there is
less variation in the tax rate. Many micro studies look at only one community and focus
on intrajurisdictional capitalization.

Finally, there are studies that take advantage of large scale policy changes, which
often serve as “natural” capitalization experiments. These studies have been performed
using both aggregated data and individual house observations.

Aggregate Studies

Perhaps the seminal capitalization is study is Oates (1969). Oates uses 1960
census data to study 53 municipalities in northeastern New Jersey. Oates regresses the
median value of owner-occupied houses in each community on the effective property tax
rate, the annual expenditure per student in the public schools, and a number of additional
control variables including: the median number of rooms per house; a proxy for the age

of the housing stock; the distance of the community’s center from Manhattan; and the

37
number of poor families in the community. He finds that a higher tax rate depresses
house prices and that increased school spending increases the price of housing. Using a
discount rate of 5 percent and a 40 year time horizon, Oates finds that tax differentials are
fully capitalized. Under a 3 percent discount rate and an infinite house life, the
capitalization percentage is 61 percent.

Oates recognizes that the tax rate term in his regression may be endogenous.
Capitalization theory assumes that higher tax rates lead to lower housing prices.
However, a community with relatively low housing values needs a relatively high tax rate
to fund a given service level. Therefore, it is difficult to determine the direction of
causation between tax rates and housing values. This problem is fundamental to this
literature. For a study to credibly state that it has measured the effect of higher taxes on
property values, it must first convincingly show that it is measuring capitalization and not
the fact that property poor communities need high tax rates to fund services. The same
problem is present in reverse when measuring the effects of public expenditures.

To correct for the endogeneity of the tax and expenditure terms, Oates uses
two-stage least squares (ZSLS). To correctly perform ZSLS, Oates needs variables that
are correlated with the tax and expenditure terms, but do not in part determine house
prices. Some of the instruments Oates uses include: the median years of school
completed by adult males; the population density; the percentage change in population
between 1950 and 1960; the percentage of the population enrolled in K-12 schooling; and
the value of commercial property per resident. It is possible that the value of all of these
instruments has some impact on the price of housing in a neighborhood. Therefore, they

might not be suitable instruments. In fact, the coefficient on the tax term in Oates’s 2SLS

38
regression is the same as the coefficient on the tax term in the ordinary least squares
regression, suggesting that if there is an endogeneity problem, the ZSLS estimation does
not completely fix it.

Several subsequent authors attempt to improve upon Oates’s work. King (1977)
suggests that Oates mis-specifies the tax term in his regression. The dependent variable
in Oates’s study is the house price and the tax term used is the tax rate. King argues that ww-
if the house price is the dependent variable then the tax payment should be used as the
independent variable; because for a given tax rate increase, the dollar effect will be larger
on a higher priced house. Therefore, Oates’s specification understates the capitalization
for high value houses and overstates it for low value houses. King finds that correcting
this specification error reduces the capitalization found with Oates sample. With a 3
percent discount rate and infinite house life, Kings capitalization estimate is 36 percent,
an estimate 41 percent lower than Oates’s finding.

Rosen and Fullerton argue that per-student spending is not a good control for
service quality. They also use Oates’s sample to estimate capitalization after replacing
per-student spending with 4th grade student test scores. They find the capitalization rate
to be slightly lower than Oates (58 percent with a 3 percent discount rate and infinite
house life). However, their study contains the tax rate specification error identified by
King.

Several authors have attempted to use simultaneous equation systems when
estimating the degree of capitalization. Gronberg (1979) uses a six-equation model, with
separate equations included for the tax rate and public service expenditure. Using a

sample consisting of census data on 83 Chicago suburbs in 1970, Gronberg finds that

39
property tax differentials are not capitalized into house values. Dusansky, Ingber, and
Karatjas (1981) also use a simultaneous equation model. They examine 62 communities
in Long Island, NY using census data. This study also attempts to model and measure the
interaction between the rental price of apartments and housing prices. Their estimate of
the capitalization rate, assuming a 3 percent real discount rate and infinite house life, is
22 percent.

The studies using aggregated data all tend to have similar shortcomings. They
often use some form of the median or mean house price as the dependent variable. It is
not clear how well these average price figures predict the experience of individual houses.
These studies generally rely on census data, where the house price reflects the house
owner’s guess as to what the property is worth. It is possible that owner house price
predictions are not very accurate. These studies also run a large risk of omitted variable
bias. The median house price in a community is a function of many variables, yet these
studies often use only a handful of controls. Finally, these studies generally do not
convincingly handle the endogeneity problem. It is difficult to find variables that are
potentially correlated with the tax rate but not the price of housing. Therefore, with many
of these studies, the possibility that the estimates of tax capitalization are inconsistent
remains.

Micro Estimates
Krantz, Weaver, and Alter (1982) look at 243 single-family owner-occupied
homes which were sold in 6 Pennsylvania cities and their surrounding suburbs in 1979.

The data were gathered from Multiple Listing Service Records. They find the

40
capitalization rate to be 20 percent (using a 3 percent real discount rate and infinite
housing life).

One important aspect of this study is that, unlike the other studies discussed in
this chapter, Krantz, Weaver, and Alter do not attempt to correct for the simultaneity of
the tax variable. They argue that property taxes will adjust slowly to changes in a
community’s property values. Therefore, property taxes and house values are not
simultaneously determined and ordinary least squares can be used to estimate
capitalization. However, the authors provide no evidence to suggest that they are looking
at a long run equilibrium, where tax rates have fully adjusted to property values.
Therefore, the results of this study should be viewed with some skepticism.

Lea (1982) uses a simultaneous equation model to estimate capitalization. Lea
looks at 680 households from the Panel Study of Income Dynamics. Observations from
the 1968 survey are merged with local tax and expenditure data for cities and counties in
which the families are located. Lea finds the capitalization rate to be 26 percent.
Unfortunately, for survey confidentiality reasons, Lea was unable to determine the actual
municipality in which his households reside. Instead he only knows the county in which
the houses are located. Therefore, he uses average tax and expenditure data for the
county in which the household resides. Lea’s study may be more appropriate for
estimating the effects of the average tax rate in an area than for estimating the effects of
tax rate differentials.

Richardson and Thalheimer (1981) examine 861 house sales in Fayette County,
Kentucky in 1973 and 1974. These house sales occur in two bordering municipalities

that share a school district but have different property tax rates. Richardson and

41
Thalheimer argue that there are few differences in other municipal services between the
communities. They find the capitalization rate to be 15 percent (assuming a 3 percent
real discount rate and infinite house life).

The micro studies are free from some of the problems of the aggregate studies.
The observations are individual houses; therefore, the estimates are free from potential
aggregation problems. These studies also tend to have more control variables for both
housing and neighborhood characteristics. Therefore, omitted variable bias is less of a
concern. Some of these studies use actual sales prices rather than owner estimates of the
value, eliminating one more potential source of error. These studies, however, still must
contend with the endogeneity issue and face many of the same challenges here as the
aggregate studies.

Natural Experiments

As noted, the most difficult problem encountered by researchers studying
capitalization is the potential simultaneity between tax rates and house values. Several
authors have exploited particular policy changes to avoid this problem. State-wide
property tax reforms can be treated as exogenous to the local communities. Therefore, by
examining the change in property values that results from such a tax change, the degree
of capitalization can be measured without the usual endogeneity problems.

Rosen (1982) examines the effects of California’s Proposition 13 on house prices.
Proposition 13 was approved by California voters in 1978. The proposition stated that
the annual tax rate on real property could not exceed 1 percent of the cash value of the
property. The cash value of the property was defined as the county assessor’s evaluation

of the property’s value as stated on the 1975-76 tax bill. For property sold after this year,

42
the cash value was defined as the market price. Assessment increases were capped at 2
percent per year, with properties reassessed to market value upon sale. Rosen argues that
because of initial state bailouts supporting local public expenditures, Proposition 13 had
little effect on service levels in the short run.

Rosen regresses the change in a community’s average house sales price on the
change in a communities average tax payment and a number of other regressors. His
sample consists of 64 communities in the San Francisco Bay Area in the years 1978-79.
He finds the capitalization rate to be 22 percent (with a 3 percent real discount rate and
infinite housing life).

Gabriel (1981) examines the effect of Proposition 13 on the average sales price of
houses in the San Francisco Bay Area over the same time period as Rosen. He finds the
capitalization rate to be somewhat higher than Rosen (36 percent). Given that Gabriel
and Rosen examine the same geographic area over the same time period the large
difference in their results is troubling. The divergence of the estimates is most likely due
to a specification difference between the two studies. Gabriel suppresses the constant
term in his regression, while Rosen does not. The regressions in both studies have the
difference in average sales price of houses as the dependent variable, and the difference in
annual tax payments as an independent variable. Including a constant term in the
regression would capture any differences in the overall price level for housing in the

estimating region. Because he suppresses the constant term, Gabriel’s tax term is

43
probably capturing some of the effects of the change in the average tax rate.5 Therefore,
the two studies are not measuring the same thing.

A similar large scale tax change is exploited by Yinger et al. (1988). They
examine the effects of court-ordered property revaluation in Massachusetts in the early
19708. They examine individual home sales before and after revaluation. Their study
focuses on the degree of intrajurisdictional property tax capitalization. Their preferred
estimate is 21 percent in one community and 15.8 percent in another. They conclude that
the degree of intrajurisdictional capitalization is likely to vary by community. However,
their model predicts that intrajurisdictional capitalization should be complete. They
argue that their estimates fall sharply below this level because of uncertainty surrounding
the reforms. They note, however, that they do not have evidence to support this assertion.

Despite the large number of studies examining capitalization, there is no
consensus as to the degree to which tax differentials are capitalized. It is possible that the
degree of capitalization is not constant across geographic areas. Areas that are more fully
developed may have a higher rate of capitalization, because the supply of housing (or
land) is more inelastic. It is impossible to know if the differences in capitalization

findings are due to different capitalization rates or flaws in the empirical work.

 

’ It does not appear that Gabriel is deliberately trying to measure the effects of the average tax rate change.
In addition, the tax term in Gabriel’s regressions may also be capturing any secular increase in the demand

for housing that occurred over the sample period.

44

III. Opportunities in Michigan

The recent property tax reforms in Michigan present a unique opportunity to
estimate property tax capitalization. As discussed in the chapter on Michigan’s reforms,
Michigan reduced its reliance on the local property tax as a means of funding K-12 public
schools, replacing lost revenues with increases in other taxes. All Michigan school
districts saw substantial property tax cuts on owner-occupied housing. However, the size
of the reduction varied by district. The new property tax rate was set by the state
government. Therefore, the change in property taxes was exogenous, with respect to the
local communities.

Michigan also changed the amount of dollars available per pupil for each of its
school districts. Again, the spending amount was set by the state government. Therefore,
the school Spending change was also exogenous, with respect to the local communities.
The circumstances surrounding Michigan’s reforms may have prevented most taxpayers
from anticipating the change. Therefore, Michigan’s reforms present an excellent

opportunity to study capitalization.

Chapter 4

CAPITALIZATION ESTIMATES

I. Introduction

The effects of inter-community tax and spending differentials on house prices is
of some importance in the study of local public finance. The capitalization rate has
implications for the efficiency and redistributive aspects of using a local property tax to
fund local public services. Michigan’s recent school finance reforms created a natural
experiment in that property tax rates and service levels were both changed substantially in
a way that was exogenous to local communities. In this chapter, I use Michigan’s
experience to generate new estimates of the effects of interjurisdictional differences in
property taxes and spending levels on house prices.

Property taxes can affect house prices in two ways. First, the average level of
property taxation can depress house values in a metropolitan region.1 Second, deviations
from the average rate of taxation can be capitalized into house prices. If tax differentials
between communities are capitalized, a community that taxes property at a higher rate

will have lower property values, all others things constant. This type of capitalization is

 

‘ Yinger (1982) argues that it is the tax rate of a base-line community, rather than the average tax rate that
is important. Mieszkowski and Zodrow (1989) argue that it is the average level of property taxation in the

nation that is important. These distinctions are unimportant for the discussion that follows.

45

46
known as interjurisdictional capitalization. Full capitalization of a property tax
differential is defined as the case where the price of two otherwise identical properties
differs by the present discounted value of the tax stream differential between the
properties.

If a property tax change is fully capitalized, homeowners cannot move to escape
the tax. Likewise, if fiscal differentials are not fully capitalized, residents may switch
communities in an effort to benefit from a more desirable tax and service combination.
Similarly, if fiscal differentials are not fully capitalized, people may try to construct
relatively low cost housing in high service communities in order to benefit from the high
service level without paying for the full cost of the services. Incomplete capitalization of
fiscal differentials could be one factor contributing to suburban sprawl if richer residents
continually move to escape service “free riders.”

Michigan’s reforms lowered the average tax rate and changed each community’s
difference from the average. Therefore, these reforms should have had two effects. First,
the overall sales price of housing should have increased as a result of the reduction in the
average tax rate (assuming, of course, that the average property tax rate does affect house
values). Second, each community’s property values should have changed differently,
depending on the relative tax change and spending in the community. If tax differentials
are capitalized, communities with larger tax changes should have seen their property
values go up more than communities that received smaller tax changes. Likewise,
communities with larger service increases will see their property values go up more than

communities that received smaller service increases.

47

I estimate the effects of the policy change using two samples. One sample
consists of individual houses that sold both before and after the reforms. The second
sample consists of observations of the mean house sales price, tax rate, and amount of
school spending for a number of communities within a metropolitan area. With the dual-
sales sample, estimates of the effects of a $1.00 property tax differential on house prices
range from $4.25 to $5.20. I find that a $100 increase in per-pupil school spending raises
house prices by 0.4 to 0.5 percent on average. Using the mean-sales-price sample, I find
that a $1.00 differential in the property tax payment on the average house causes a $9.93
difference in the average house price. For the mean-sales-price sample, the estimate of
the effect of the change in school spending is not statistically significant. In addition, I
find that the average house sales price increased greatly after Michigan’s reforms. It is
possible that this increase was due to the reduction in the average tax rate, although
economic growth may also be factor. Much of the increase in the average sales price of
houses in rural areas appears to be the result of the construction of relatively expensive
houses in these areas.

The remainder of this chapter is divided into five sections: The next section
briefly discusses Michigan’s reforms; Section 111 contains the capitalization model used
to generate the estimates; Section IV contains the estimates generated using the mean
sales price data; Section V contains the estimates generated using the individual sales
data; and section VI contains concluding remarks.

II. Michigan’s Reforms
In 1994, Michigan profoundly changed the way it finances its public schools. The

property tax share of school operating revenues was reduced from 66 percent in 1993-94

48
to 32 percent in 1994-95. The property tax revenues were replaced primarily with an
increase in the state sales tax from 4 percent to 6 percent.2 Michigan also changed the
way it allocates revenues to school districts, switching from a district power-equalization
method to a foundation grant approach.’

Prior to 1994-95, the variation in school district property tax rates was large,
ranging in 1993-94 from 8 mills to 47 mills.4 There was also a wide variance across
districts in per-pupil spending, ranging from a high of $10,141 per-pupil to a low of
$3,173 per-pupil. The higher school taxes were not always in the districts with the higher
spending amounts. In fact, the average millage rate for the 10 lowest-spending districts
(per pupil) in Michigan in 1993 was 30.14 mills, while the average millage rate for the 10
highest-spending districts was only 26.08 mills.5

As part of the 1994 reforms, Michigan removed much of the authority local
districts previously had in making funding decisions. Funding decisions are now made
primarily at the state level. The role of the local property tax as a revenue source was
also greatly reduced, and lost revenues were replaced with an increase in the state sales
tax and a conversion of part of the local property tax into a state property tax. Districts

spending below $6500 per-pupil in 1993 had their owner-occupied housing property tax

 

2 See Kearney (1994) for a detailed description of the Michigan reforms.

3 For a discussion of these two funding approaches, see Reschovsky (1994).

‘ A mill represents a $1 .00 tax for every $1000 in taxable property value. Michigan property is assessed at
50 percent of its cash value so that a 1 mill tax on property with a cash value of $1000 would raise $0.50 in
revenue (if the property were correctly assessed at $500).

5 Source: Michigan State Board of Education Bulletin 1014.

49
rate for school operations reduced to 6 mills.6 Districts spending above $6500 were
allowed to maintain their higher spending levels if they levied an additional “hold
harmless” millage.

School districts spending below $4200 per-pupil in the 1993-94 school year were
moved up to $4200 in the 1994-95 school year. Districts spending above $4200 and
below $6500 were increased from their 1993-94 levels according to a sliding scale, with
higher-spending districts getting a smaller percentage increase. Districts spending above
$6500 per-pupil were allowed to increase their school spending by $160 per pupil in the
1994-95 school year. Future increases in funding depend on the grth of state tax
revenues, with previously low-spending districts expected to get a larger percentage
increase each year than the previously hi gh-spending districts.

Michigan’s reforms provide an excellent opportunity to assess the effects of
property taxes and school spending on home values. The tax and spending changes were
large and exogenous (at least initially) to the local communities. The changes were most
likely not anticipated by most homeowners. Governor Engler had campaigned in 1990
with a promise to reduce property taxes. However, Michigan voters had a long history of
rejecting school finance reform proposals, rejecting all 12 that were presented to them
between 1972 and 1994, including the first three presented during the Engler
administration. Table 4.1 contains a summary of the events leading up to Michigan’s
reforms.

k

6 Actual school millages in some of these districts will be higher because school districts also levy taxes to

make payments to retire debts. The school districts remained responsible for their own debts.

50

To measure the degree to which these tax and spending changes were capitalized,
I examine home sales in Oakland County, Michigan. Oakland County covers 910 square
miles immediately north of the city of Detroit. It has several characteristics that make it
an attractive location for estimation purposes. First, it has a large population, consisting
of just over one million people in 1990, divided among 410,000 households. Second, it is
an area of relatively high incomes. Oakland County advertises itself as the “third most
affluent county on the map.” This affluence translates into a high degree of home-
ownership, with 73 percent of households living in owner-occupied housing in 1990.
Finally, it consists of a large number of municipalities (61), none of which had more than
75,000 people in 1990.7

Figure 4.1 shows the growth in the real average sales price of houses in Oakland
County from 1990 through 1996.8 Oakland County averaged 20,000 house sales per year
during this period. Michigan’s reforms took place at the end of 1993 and the beginning
of 1994. Figure 4.1 shows substantial grth in sales prices after the reforms.

To get a more detailed picture of prices changes in Oakland County, I divide the

county into three regions: urban; semi-urban; and rural. I define urban areas as those

 

7 Data from 1990 census and Oakland County promotional literature. Note the number of school districts is
substantially less than the number of municipalities but the effective tax rate will depend on the assessment
ratio which varies by local municipality as well as on the statutory millage rate.

8 The houses used in this graph are houses that were zoned as either “residential improved" or “suburban
inrproved.” These two zoning categories are made up almost entirely of single family dwellings.
Commercial properties, condominiums, and lakefront property have different zoning classifications and are

not present in this sample.

51
communities with less than 20 percent of their land available for development. Semi-
urban areas are defined as those communities with between 20 and 50 percent of their
land available, and rural areas are those areas with more than 50 percent of their land
available for development.’ The urban and semi-urban areas account for approximately
40 percent of the sales each. The remaining 20 percent of the sales occurred in the rural
areas.

The growth trend for all three regions is depicted in Figure 4.2. All three regions
experienced little growth in prices before the reforms. After the reforms, all three regions
show a substantial increase in average sales price, with rural areas experiencing the
largest increase.

Of course, there is no way to be certain how much of this price growth occurred
because of the reforms and how much was due to other factors, such as economic growth.
The unemployment rate of the Detroit MSA, of which Oakland County is a part, fell
dramatically over this period. From a high of 10.2 percent in the first quarter of 1991 , the
unemployment rate fell steadily to a low of 3.8 percent in the fourth quarter of 1996.10

While the economic expansion began in 1992, property values did not start to rise
substantially in Oakland County until 1994. Although this rise in property values closely

matches the implementation of Michigan’s reforms, it is possible that the rise was a result

 

9 The percentage of land available was calculated using data from the Southeast Michigan Council of
Governments (SEMCOG) Community Profiles. Any land falling into the categories of agricultural, or
woodland/grassland/wetland was considered available.

1° Source: Michigan '3 Labor Market News various issues. (published by The Michigan Employment

Security Agency)

52
of the continued economic expansion in the Detroit area. Therefore, while the data are
suggestive they are inconclusive.

Figure 4.3 shows the changes in school spending between 1990 and 1995 for the
three regions. Figure 4.4 shows the changes in the average school millage rate. These
figures show that Oakland County residents saw an increase in school funding and a
dramatic reduction in property taxes due to the reforms. Each school district in Oakland
County experienced a different tax and spending change. I use the differential effect of
these changes on the prices of houses in these school districts to estimate the degree of

capitalization.

III. The Econometric Model

Capitalization of the property tax means that changes in the tax payment stream
over time affect house prices. In this chapter, I follow the capitalization model as defined
in Yinger (1982) and Yinger et al. (1988). These papers demonstrate that the
capitalization equation can be derived from an asset pricing model or from a utility
maximizing model. I follow the asset pricing approach here because the derivation is

simpler.

 

N R(a,E) ‘i T

(I) V : n=l (1 +1.)" n=l(1+i)n

Following the notation of Yinger et al. (1988), equation (1) states that the value of

a house, V, is equal the present value of the rental stream of service it generates, R(a,E),

53
where i is the real discount rate and N is the house life. T represents the annual tax
payment on the house. The annual rental price of the house, R, is a function of the
amenities of the house, a, and the level of government services in the community, E. The
house amenities are assumed to include both house and community features. House
features include characteristics such as the house size and square footage, and community
features include the house’s distance from parks and highways, and such quality-of-life
issues as pollution and crime. In generating my estimates, I assume that or remains

constant over the period of estimation.‘1

 

" To be more specific, the rental price of housing is a function of a, E, and the average tax rate, and
overall real housing price level in the community. Assuming an infinite house life, we can write equation
(1) as

(a) iV=R(a.E.P,r)-(T-I)

where P is the real housing price level and ris the baseline or average tax level, T is the tax payment on the
individual house and all other variables are as defined in the text. Equation (a) indicates that differences in
house prices across communities are in part due to how the communities tax rate differs from the average
tax rate. I can then rewrite equation (a) as

(b) iV=(R(a'. E, P, r)- z)-T

and then

(c) iV=R’(a, E, P, z) -T

where R has been redefined to include the additional baseline tax term. P and rare omitted from equation
(1) in the text because they are constant across all housing observations in my samples. Therefore, they are

not individually estimated; but, are instead subsumed into the year dummy variables in the regressions.

54
Equation (1) assumes that home values are reduced by the full present value of the
stream of future tax payments. Equation (2) shows the formulation if tax capitalization is

less than full.

 

(2) V: N R(a,E)_‘ZV: flT

"=1 (1 + l')" n=l(1+i)n

A ,6 of 0.5, for example, indicates that 50 percent of the tax differential is
capitalized. F ifty-percent capitalization implies that a tax increase would reduce house
values by 50 percent of the present value of the tax increase. Estimates of ,6 are going to
be highly dependent on assumptions concerning the relevant house life and discount rate.
To make my results comparable with Yinger et al., I follow their assumptions of an
infinite home life and a 3 percent real discount rate. Under these assumptions, full
capitalization implies that a permanent $1.00 change in annual property taxes in a district,
holding taxes and spending in all other districts constant, would change house prices by
$33.33.

Alternatively, Oates (1969) assumes a house life of 40 years and a 5 percent
discount rate. A $1 .00 tax change would be said to be fully capitalized under these
assumptions if it changed the house price by $17.16. Therefore, the assumptions of a 3
percent discount rate and infinite house life represent will produce a more conservative
estimate of the degree of capitalization.

Assumptions concerning the real discount rate and house life have little effect on

the regression estimates other than scaling them. Of course, the dollar change in property

55
values that actually occurred does not depend on assumptions regarding the discount rate
or housing life. What these assumptions do affect is the interpretation of the results. The
dollar amount that represents full (100 percent) capitalization decreases with the assumed
discount rate and increases with the assumed house life. A summary of the dollar totals
representing full capitalization of a $1 tax change under varying discount rate and house
life assumptions is presented in Table 4.2. To avoid confusion, I report the dollar
changes resulting from the tax changes, as well as the capitalization rate.

The effective tax rate is defined as the annual tax payment divided by the sales
price of the house. The annual tax payment is the assessed price of the house multiplied
by the statutory millage rate. Therefore, the effective tax rate also can be stated as the
assessment ratio multiplied by the statutory rate, where the assessment ratio is the
assessed value of the house divided by the cash value of the house.

The effective tax rate is an endogenous variable since it is calculated using the
house’s price. The endogeneity problem is best illustrated by an example. Suppose that a
house sells for a higher price than it normally would due to a random shock. This higher
price will also be associated with a lower effective tax rate, since the house price is in the
denominator of the effective tax rate. Therefore, the random element to house prices will
make the effective tax rate seem low for houses that sell at a premium and high for
houses that sell at a discount, resulting in an upwards capitalization bias.

To avoid this problem, I use the average assessment ratio for each community,
rather than the actual ratio for each house when calculating the effective tax rate. I define
the average assessment ratio as the average assessed value for the community divided by

the average sales price for the community. By using the average ratio, I am ignoring the

56
possible capitalization of within-community assessment errors, which is known as
intrajurisdictional capitalization. Instead, by using the average ratio, I can look at the
effects of tax and spending differences across communities, which is known as
interjurisdictional capitalization.

I can, therefore, rewrite the tax term in equation (2) as

(3) T=tV

where t is the average effective tax rate.12 Based on equation (3) and under the

assumptions of an infinite house life, I can rewrite the capitalization equation as

 

'2 Another issue with the tax rate is the deductibility of property taxes from the federal income tax. The
following argument is based on Yinger et al. 1988. The basic capitalization equation I use is:

_£_fltV

(a) V . .
l 1

 

Assume that the marginal tax rate faced by the homeowner is 3. After the property tax deduction, the net
tax payment is (1 - s)tV. Yinger et al. argue that it is also important to consider the effects of taxation on
the discount rate. The opportunity cost of housing is the return given up by the investor when he or she

invests in a house. The return that matters is the return after taxes. Let i' be the gross of tax return. Then

the net return is (1-s) i’. I can rewrite equation (a) as

_ R _fl(1—s)tV
—(1-s)i' (1—s)i’

 

(b)

Simplifying equation (b) produces

V: R
(1-S)[i'+flt]

 

(C)

57

V_R(ar,E)

(4) " (Hat)

Taking the natural logs of both sides of equation (4), produces

(5) an=lnR(a,E)-ln(i+,6’t)

For estimation purposes, I rewrite this model as

(6) ln V, = a, + floschool, -— ln(i + ,6, t, ) + 2 D, a“, + a,

The annual rental rate of a house, R in equation (5), is assumed to be based on time-
invariant house and neighborhood characteristics, a,, and school spending, which does
vary over the sample period. In all specifications, the a, are assumed to be arbitrarily
correlated with the other right-hand-side variables. The effect of school spending on

home values is measured using per-pupil expenditures. Note that I am assuming an

 

Taking logs we get
u) mV=mR—ma—n—mw+fln

The tax term is time constant and will be differenced out of the estimates.

58
exponential function for the effects of school spending on house prices. '3 Per-pupil
school expenditures serve as a proxy for the real variable of interest, the amount and
quality of schooling services available in the community.” In using this model, I have
implicitly assumed that the effects of school spending are constant across communities.
Therefore, the coefficient, flo, should be interpreted as the average effect of school
spending on house prices.

The D, represent a set of year and quarter dummy variables. Year dummies are
included to capture the overall price level in Oakland County and quarter dummies are
included to capture any seasonal price differences. The year dummy variables will
capture any overall rise in price level for Oakland County homes that resulted from the
policy change, as well as any change in the price level that resulted from other county-
wide economic factors, such as real construction cost increases or increased demand due
to economic growth. Therefore, ,4, and ,6, measure the effects of tax and spending
differences across communities. The coefficients should be interpreted as a measure of
the effect of changing the tax or spending amount in one community, holding the tax and

spending levels of all other communities constant.

 

'3 The model was also estimated using the log of per pupil expenditures with similar results. I judged the
semi-elasticity model to provide the better fit. The actual school spending variable used is the revenue per-
student available to the district, from local, state, and federal sources. This figure was taken from The
Michigan Department of Education Bulletin 1014. Figures for the 1996-97 school year are not yet
available, so I have assumed that real spending remained constant between 1995-96 and 1996-97.

” Note that other public services can be safely omitted as long as they either remained constant over the

relevant period or had changes that were not correlated with the other variables of interest.

59
The real discount rate, 1, is assumed to be 3 percent. I have converted all dollar
values into constant 1990 dollars using the national Consumer Price Index (CPI-U).
Equation (6) is non-linear, so it cannot be estimated using ordinary least-squares
(OLS). Assuming initial parameter values of zero for all parameters, the linear

approximation to equation (6) can be written as
_ 161 In
(7) 1n V, _ a, + ,6, ln school, — ——,—— + 2 D, a, + a,
I

If the ori were known, equation (7) could be estimated using OLS. The ori are not known,
however, so they must be removed prior to estimation. I estimate equation (7) after

taking first-differences, equation (8), and by fixed-effects, equation (9).15

A .
(8) A In Vi, =floA In School” — ’6—‘-t'—’- +2 AD, 5,, + A8,,
1

161(tir —Z)

l

(9) 1n V, — 1117,. = ,6’0(ln School“ — In School, ) — +

Z<Dk —D_k)§k +51: ‘5—1

¥

'5 Because there are only two time periods in the dual sales sample first differencing is equivalent to fixed

Effects.

60

IV. Mean Sales Price Estimates

In this section, I present estimates of the degree of interjurisdictional
capitalization generated using the mean house sales price sample. I examine the effects
of Michigan’s reforms on the mean house sales price for each of Oakland County’s
municipalities. The 49 cities, townships, and villages that averaged more than 15 sales
per quarter between 1990 and 1996 are included in this sample. The sample consists of
quarterly data from 1990 through 1996. Observations from the last quarter of 1993 and
the first three quarters of 1994 have been dropped. This period represents a time when
there was a great deal of uncertainty surrounding Michigan’s reforms. Including these
quarters in the regressions reduces the capitalization estimates and the overall fit of the
model.

This sample is attractive for several reasons. First, the number of sales used to
generate the average figures is large. Oakland Corinty averaged approximately 20,000
home sales per year between 1990 and 1996. Second, the sample covers a relatively long
time period with observations several periods before and after the policy reforms. This
sample uses all home sales in Oakland County, rather than a subset of the houses that
sold.” The two studies that use Proposition 13 data (Gabriel 1981 and Rosen 1982) also

use the average community sales price as the dependent variable. Therefore, the

k

‘6 To be specific the sample includes all properties that were zoned as “residential improved” or “suburban

inrproved.” Houses with special zoning classifications such as lakefront homes were not included.

61
estimates generated with this sample can be compared easily to the results of these
studies.

Using aggregated data has some severe drawbacks, however. The capitalization
model I use for estimation is defined at the house level. The aggregate version of this
model should have the average of the log of the individual home prices as the dependent
variable. I use the log of the average house price as the dependent variable instead. The
size and direction of the effect of this difference is unknown.

Michigan’s policy change affected taxes and spending by school district. In
Michigan, school districts are a separate level of government from municipalities.
Oakland County’s school districts do not line up exactly with the borders of the
municipalities. The property tax assessed on a house in Michigan is the combination of
all state, county, municipality, and school district taxes. I assume that municipalities that
lie mostly within one school district were entirely within the school district. For
municipalities more or less split evenly by two school districts, I use the average values
of the school tax and spending figures.

There is also no way to be certain that the policy change did not affect the
composition of the homes that sold. An increase in the mean sales price of a community
may correspond to a situation in which more expensive houses are being sold rather than
an increase in house price. The policy changes may also have affected the mean price by
influencing the type and location of new construction. This is a problem for all studies
that rely on aggregate data.

Table 4.3 presents summary statistics for the mean-sales-price sample. The

regression results for the mean-sales-price sample are presented in Table 4.4. The

62
estimates are generated using the linear approximation described in equation (7). The
OLS estimate of equation (7) is presented in column I of Table 4.4. Recall that the ori are
unobserved and so are not estimated. The effective tax rate has been divided by the
assumed discount rate of 3 percent prior to estimation. Therefore, the coefficient on the
tax term is interpreted as the capitalization rate under a 3 percent discount rate and
infinite house life. The OLS estimate for the capitalization rate is 140 percent. The
school spending coefficient represents the percentage change in housing prices resulting
from a $100 per pupil increase in school spending. The OLS estimate predicts that such
an increase would raise house prices by 0.9 percent.

Recall that the ori may reflect time constant characteristics of a community that are
correlated with both the tax rate and price of housing in a community. If so, the ori must
be removed prior to estimation or the regression results will be inconsistent. I estimate
equation (7) using both first differences and fixed effects.17 The results of both of these
estimation techniques are reported in Table 4.4. The coefficients on the tax and spending
terms decrease substantially suggesting that the OLS estimates are indeed inconsistent. I
test for serial correlation using the residuals of the differenced equation. The coefficient
on the lagged residuals is -0.52 suggesting there is no serial correlation in the
undifferenced model. Therefore, the fixed effects estimator is the more efficient. The

reported standard errors are robust to first-order serial correlation and heteroskedasticity.

 

'7 The fixed effects estimates are generated by time-demeaning the observations for each municipality prior

to estimation.

63
I use robust errors to account for any heteroskedasticity that might result from the use of
grouped data.

The coefficient on the normalized tax rate in the fixed-effects estimate is -0.2978,
indicating a 29.78 percent capitalization rate. With the maintained assumptions of a 3
percent discount rate and infinite house life, full capitalization occurs if a $1.00 tax
differential results in a $33.33 price differential. The fixed-effects estimates indicate that
a $1.00 tax differential between two communities would result in a $9.93 property value
differential (29.78 percent of $33.33), on otherwise identical properties. Rosen (1982)
finds that a $1.00 tax differential leads to a $7.30 property value differential, while
Gabriel (1981) finds a value of $12.00. Therefore, this estimate falls between the two.

Both the fixed-effects and first-differences estimates of the effects of school
spending are not statistically different from zero. There are several factors that may be
contributing to this. First, recall that school spending is a proxy for the real variable of
interest, which is the quality and amount of schooling services provided. Second, there
may be a great deal of uncertainty in the communities themselves as to the effects of the
additional school spending. As discussed in Wassmer and Fisher (1996), there are a
number of ways Michigan’s schools can spend their new revenues and not all of them
(new school buses, for example) will improve student learning and performance. There is
also a great deal of uncertainty surrounding the future funding levels of Michigan’s
school districts (again, see Wassmer and Fisher (1996)). As the impact of the spending
changes on school quality becomes clearer, the effect on property values is likely to

become more pronounced.

64

To test the robustness of these estimates, I also estimate the effects of the policy
change using a “random growth model” (Heckman and Hotz 1989). Under this model,
each community is allowed to grow at a separate rate. This grth rate is assumed to be
potentially correlated with the other right-hand-side variables. For example, districts that
received the largest tax cuts could have been growing at a slower rate than the districts
that received smaller tax cuts. The random growth model is presented in equation (10),
where p represents the time period and 7, represents the grth coefficients for the

individual townships.
t.
(10) anuzai+7ip +flolnSchooli, —£‘i—i+szdk+eh

I am still including the year dummies in equation (8), so that the x’s represent deviations
of the individual communities from the overall price growth in Oakland County.
Equation (8) is first differenced to eliminate the time-constant unobservable component.
The )gare then removed from the differenced equation using standard fixed-effects
techniques. The results of this estimation are presented in column IV of Table 4.4. As
can be seen by comparing columns 111 and IV of Table 4.4, the random grth model
produces results almost identical to the results found by first-differencing the first model.
Therefore, with the mean-sales-price sample, deviations from the overall price trend in

Oakland County seem to be unimportant for estimation purposes.

65

V. Dual Sales Estimates

There are three drawbacks to attempting to estimate capitalization with the mean-
sales-price sample. First, the capitalization model does not aggregate well. The
dependent variable should be the average of the log of the sales prices, but I am forced
use the log of the average sales price. Second, the school district boundaries do not line
up exactly with the municipal boundaries, so using municipalities represents an
approximation. Third, there is no way to be certain that the policy change did not affect
the composition of the houses that were sold. It is possible that the tax change affected
both the houses that sold and the location and type of new construction.

In this section, I present estimates that correct these three problems. I exanrine a
set of individual houses sold once in 1992 and then again in 1996. These sales straddle
Michigan’s policy reforms. By looking at individual houses, I avoid the aggregation
problem that is present in the mean-sales-price estimates. Many factors that affect house
prices, such as the distance of the house from parks or the urban center, can be assumed
to be constant over such a short time period. I attempt to eliminate the small number of
houses that underwent major additions or renovations between the sales from the sample,

by using building permit data obtained from the local municipal governments.18 Houses

 

‘8 Building permit data is mostly kept by the individual cities, townships, and villages and the availability
and quality of the data varied by municipality. Therefore, it is possible that some houses that underwent
major renovations remained in the sample. As long as these houses are few in number or the renovations

are not correlated with the tax change they should not bias the estimates.

66

that were not owner-occupied or that were not the homeowner’s primary residence are
also removed fi'om the sample.19 In total 74 houses (9 percent of the total matched sales)
are removed prior to estimation. The problem of school district boundaries not being
coterrninous with municipal boundaries is also eliminated, since each house’s school
district is known.

A potential drawback of the dual-sales data is that these homes represent houses
that sold twice in four years. I identify approximately 700 houses that sold both in 1992
and then again in 1996. Only the first three quarters of these two years were compared,
and the 700 houses represent approximately 5 percent of the total number of houses that
sell in any three quarter period.20 Houses that sell twice in such a short period may be
systematically different in some way from the housing stock at large. For example, it is
possible that the consumers of such houses anticipate selling them soon after purchase
and thus have a shorter time horizon than the average home purchaser. The purchasers of
such homes might have different preferences for services than the population at large.
For example, they might be more likely to have small children and prefer a larger amount
of school spending. Therefore, it is possible that the effects of the policy changes on

these houses are different from the effects on the overall population of houses.

 

'9 Homeowners primary residences in Michigan are known as “homesteads” for taxation purposes and after
the reforms they were taxed at a lower rate than rental properties and secondary residences. Each house
used in the dual sales estimates is a homestead.

2" Only the first three quarters of the years were compared due to the expense of the data collection.

67

Table 4.5 contains summary statistics for the dual sales sample. The growth
variable is the percentage change in the average house sales price for a house’s
community between 1990 and 1992. It is included in the regressions after differencing to
help reduce the noise in the estimates. The improvement dummy variable is equal to one
for those houses that had improvements done to them, where the improvements were not
substantial enough to warrant removing the house from the sample. A variable
representing the percentage change in the number of households in a house’s community
between 1990 and 1996 is included as a measure of housing construction between 1990
and 1996.

Recall that the capitalization equation (7) contained an unobserved vector of
components, or,. The dual-sales sample consists of two time periods and the unobserved
component is removed by first-differencing. To account for the fact that each community
may have its own price grth trend, 1 include the percentage change in the average sales
price of houses in the regression. This variable represents a somewhat crude measure of
price growth for existing houses, since its value may be influenced by new construction.
The results of this regression are reported in Table 4.6 column I. The coefficient on the
normalized tax rate represents the percentage of the tax differential that is capitalized, in
this case 6 percent. Note that coefficient is not statistically different from zero.

Column 11 contains the regression results with the inclusion of the percentage
change in households in a house’s community between 1990 and 1996. The inclusion of
this variable makes the equation more structural. Column 11 is essentially the inverse
demand curve for housing, with the percentage change in households representing the

change in quantity. The inclusion of the quantity variable improves the regression results

68
substantially. The estimated capitalization rate is now 11 percent, implying that a $1.00
tax differential leads to a $3.73 price differential between houses. This estimate is
statistically significant at a 5 percent level (p=0.029). The estimate of the effect of school
spending implies that a $100 per-pupil increase in spending increases house prices by
0.56 percent. This estimate is significant at a 1 percent level (p=0.001).

Including a quantity variable in the regression, however, raises the possibility of
an endogeneity problem, since the price of housing may be in part determining the
quantity. Testing for endogeneity and correcting the problem if it exists requires a supply
shifter. I use the percentage of land available in a community for development as the
supply shifter. A Hausman test for endogeneity indicates that the quantity variable is in
fact endogenous Qr=0.028).

Column III contains the two-stage least squares (2SLS) estimates using the
percentage of land available as an instrument. These estimates indicate a capitalization
rate of 12.7 percent, implying a $1 .00 tax differential results in a $4.23 price differential.
A $100 increase in per-pupil spending, holding spending in all other districts constant, is
estimated to raise house prices by 0.51 percent. Both variable are significant at a 1
percent level (p=0.014 and 0.003 respectively).

The percentage change in households between 1990 and 1992 was largely
determined by the availability of land. The simple correlation between the change in
households and percentage of land available is 0.8. Urban communities, those with
between 0 and 20 percent of their land available, saw a 1.3 percent increase in households

between 1990 and 1996. Semi-urban communities, those with between 20 and 50 percent

69

of their land available, saw an average increase of 10.8 percent, and rural communities,
those with more than 50 percent of their land available, saw an increase of 22.8 percent.

Figure 4.2 shows the increase in the average house sales price for each of these
regions. Rural areas have the largest increase in average price. Column IV of Table 4.6
shows the dual-sales estimates where the price growth for each of the three regions is
estimated holding all other factors fixed. The growth rate is estimated by including a
dummy variable indicating whether the house’s community is semi-rural or rural. With
this specification, the constant term represents the grth rate of urban areas and the
dummy variables capture differences from this grth rate. These coefficients reveal that
for existing houses, urban areas actually experienced the largest price grth on average,
almost three times the price grth of rural areas. This suggests that the price growth in

Figure 4.2 is biased by the effects of relatively expensive new construction.

VI. Conclusion

In this chapter, I have used home sales data from Oakland County, Michigan, to
examine the effects of Michigan’s school funding reforms on house values. I estimate the
capitalization of interjurisdictional differences in property tax and school spending using
two samples. A summary of the capitalization estimates is presented in Table 4.7. With
the first sample, I examine the effects of the policy changes on the mean sales price of
houses in the different communities. Using this sample, I find that a $1.00 tax
differential would cause a $9.93 property value differential. This estimate falls between
the estimates of Rosen (1982) and Gabriel (1981) who find values of $7.30 and $12.00

respectively. These studies use mean-sales-price data to study interjurisdictional

CE

51;

ha
651
sm

pm;

199:
This
inert

Can

7O
capitalization, in the context of California’s Proposition 13. My estimate of the effect of
school spending differences on house values was not significant for this sample.

Using a second sample, consisting of houses that sold twice, straddling the policy
changes, I find a much lower rate of property—tax capitalization. The high estimate of the
effects of a $1.00 tax change is $5.20. I also find the effects of school spending to be
significantly different from zero, with estimates showing that a $100 increase in per-pupil
spending would raise house values 0.4 to 0.6 percent.

The large difference in the estimates provided by the two samples is troubling.
One possibility is that the new construction that occurred in Oakland County, which I
have argued had a large effect on the average sales price, is biasing the tax capitalization
estimates for the aggregate data upwards. A second possibility is that tax capitalization is
smaller for the dual-sales data than for the housing stock at large. It is possible that the
purchasers of such houses have shorter time horizons than the purchasers of the housing
stock at large.

An additional question of interest concerns the relative strengths of the price
decrease caused by additional taxation and the price increase caused by increased school
spending. For this purpose, consider the following, admittedly crude, experiment.
Suppose that a hypothetical district in Oakland County wished to raise school spending
by $100 per-pupil in the 1990-91 school year. Assume that this district had the county’s
1990-91 median state equalized value (SEV) of property wealth per-pupil of $1 15,897.
This district would have needed to increase tax rates by 0.863 mills in order to finance the
increase in spending. In 1990, the median price of an owner-occupied house in Oakland

County was $95,400. If the median-priced house was assessed at 50 percent of its cash

71

value, as the law in Michigan requires, the tax on it would be $41.17 a year. At a
capitalization rate of 15.6 percent, the high dual sales result, this tax would reduce the
median house’s price by $214.08. At a capitalization rate of 29.8 percent, this tax would
reduce the median house’s price by $410.47. Assuming that a $100 increase in per-pupil
spending increases house prices by 0.4 percent, the increase in school spending would
raise the median house price by $381.60. Although this calculation contains some rough
approximations, it does suggest that the effects of higher property taxes on house values
may be offset to a large degree by the effect of increased school expenditure on property
values.21

In the case of Michigan’s reforms, windfall gains were experienced by existing
home owners. It appears that homeowners experienced an increase in the value of their
houses as a result of the reduction in the overall property tax rate, although this increase
might be in part due to Michigan’s recent economic expansion. Rural areas experienced
the largest increase in average sales price, although this seems to be due largely to the
value of new construction. For existing housing, urban areas experienced the largest

price increase.

 

2' The median house value for Oakland County is taken from the 1990 census. The SEV figure used in the
calculation is the median of the SEV values for each of Oakland County’s school districts for the 1990-91
school year. This figure was taken from Bulletin 1014 published by the Michigan State Board of
Education. The majority of Oakland County’s school districts were “out of formula” in 1990-91, meaning
that state aid did not lower the price of an additional dollar of schooling for those districts. I assume that

this hypothetical district is also out of formula.

72

Interjurisdictional differences in the tax cuts were capitalized as well. Because the
tax differentials were not fully capitalized, residents have an incentive to move to the
communities that received the largest tax cuts. Incomplete capitalization implies that
future buyers can receive some of the benefits of the tax cuts.

Residents of previously low-spending school districts also received a benefit from
the increase in school spending. The higher spending levels increased house values and
will probably result in more schooling services being provided to residents of these
districts. The extent to which the increased spending increases schooling services is a
topic that needs to be investigated further. Residents of the school districts that spent the
most prior to the reforms did not fare as well. These residents are likely to see decreases
in their real school funding levels as time passes (see Wassmer and Fisher 1996). This
will likely decrease school service levels, as well as property values, in those
communities. The effects of the school spending changes on property values are likely to
become more pronounced, as the uncertainty regarding future funding levels and the
effects of these funding levels becomes resolved. Residents of these high-spending
communities also received smaller tax cuts, since they were required to levy an additional
millage to maintain their spending levels.

Finally, it should be noted that this analysis ignores the effects of the increase in
other taxes in Michigan, especially the state’s sales tax, that occurred to as part of
Michigan’s reforms. Also, further research is needed to investigate the effects of the

property tax cuts on renters, a group that was not included in this study.

Real Percentage Increase Over 1990 Prices

15

-10

73

All of Oakland County

 

 

 

1 1 I I l

 

 

I T

1990 1991 1992 1993 1994 1995 1996

Figure 4.1 - Property Value Growth 1990 - 1996

Real Percent Increase Over 1990 Prices

74

+ Pct. Growth Urban +Pct. Growth Semi—Urban

+ Pct. Growth Rural

 

-5 -_

 

1 L 1 L I

 

 

-10
1990

I I I T r

1991 1992 1993 1994 1995

Figure 4.2 - Property Value Growth 1990 - 1996

1996

Real Percentage Increase Over 1990 School Spending

75

+ Urban School + Semi-Urban School

+ Rural School

 

15

10*

-54

l I l I

 

 

 

-10

1990

I I I T

1991 1992 1993 1994

Figure 4.3 - School Spending Changes 1990 - 1995

1995

Unweighted Average Millage

+ Urban Millage

+ Rural Millage

76

+ Semi-Urban Millage

 

45

 

 

 

1

l

 

I

 

l 1

 

 

I

1991

1992

I

1993

 

1994 1995 1996

Figure 4.4 - Average Millage Rates 1990 - 1996

77

Table 4.1 Events Leading up to Michigan’s 1994 School Reforms

 

Date Event

 

 

1972 to 1989 Nine ballot proposals to either reduce property taxes and/or reform
school funding presented to Michigan voters. All are defeated.

1990 Gubernatorial candidate John Engler promises to cut property taxes if
elected.

November 1992 Two ballot initiatives are presented to voters

0 Governor Engler’s “Proposal C” that would have reduced property
taxes and capped assessments

o The legislature’s “Proposal A” that would have limited assessment
growth.

Both initiatives are defeated

June 1993 Governor Engler and the legislature agree to a new ballot proposal
that links school finance reform and property tax cuts. “Proposal A”
on the 1993 ballot is defeated.

July 1993 State Senator Debbie Stabenow introduces legislation to entirely
eliminate the use of the property tax for school firnding.

August 1993 Governor Engler signs Public Act 145. $6.5 billion in property taxes
are eliminated.

December 1993 The governor and legislature present their plans for replacing the lost
revenue, including a limited role for the property tax, and
restructuring school finance.

March 1994 Voters approve “Proposal A” on the 1994 ballot.

 

 

Source: Addonizio, Kearney, and Prince 1995.

 

78

Table 4.2 Full Capitalization of $1 per Year Under Differing Assumptions

 

Real Discount Rate
House Life m 3% 51;; 1%
10 Years $9.47 $8.53 $7.21 $7.02
40 Years $32.84 $23.11 $17.16 $13.33
100 Years $63.03 $31.60 $19.85 $14.27
Infinite $100.00 $33.33 $20.00 $14.29

 

This table shows the dollar amount representing the full capitalization of a $1 .00 tax differential between
communities under varying discount rate and house life assumptions.

79

Table 4.3 Summary Statistics for Average Sales Price Sample

 

Mean Std. Dev. M_in m
Avg. House Price $94,765.00 $38,837.00 $14,983.00 $251,947.00
Log(Avg. House Price) 11.37 0.44 9.61 12.44
Per-pupil School Spending $5,789.00 $1,393.00 $3,897.00 $10,489.00
Effective School Tax Rate 1.36% 0.62% 0.28% 2.39%

 

Note: All dollar totals are reported in constant 1990 dollars. Sample observations are from
the 49 Oakland County municipalities that averaged over 15 sales a quarter between 1990
and 1996. Observations from the last quarter of 1993 and the first 3 quarters of 1994 have
been dropped. This time period reflects a period of uncertainty over Michigan’s reforms.
Total observations used in estimation: 1176.

Table 4.4 Estimates Usifl Mean Sales Price for Each Community

 

 

Fixed First Random
OLS bft’ects Differences Growth
Normalized Tax —l.4348 -0.2978 -0.2513 02451
Rate’ (0.4153) (0.0820) (0.1238) (0.1259)
School Spending 0.0094 0.0000 0.0028 0.0030
Per Pupil (0.0028) (0.0016) (0.0034) (0.0035)
(100’s of dollars)
Y9] -0.0284 -0.0368
(0.0142) (0.0131)
Y92 0.0216 -0.0069 0.0620 0.0620
(0.0156) (0.0155) (0.0497) (0.0497)
Y93 0.0211 0.0145 0.1395 0.1393
(0.0173) (0.0151) (0.0907) (0.0906)
Y94 -0.5573 -0.0549 0.0845 0.0866
(0.1819) (0.0363) (0.1109) (0.1108)
Y95 —0.4714 0.0149 0.1719 0.1738
(0.1790) (0.0358) (0.1464) (0.1461)
Y96 -0.3825 0.0948 0.3014 0.3032
(0.1745) (0.0320) (0.1877) (0.1873)
Constant 11.5771 -0.0110
(0.3037) (0.0087)
Obs. 1176 1176 1127 1127

 

Note: The dependent variable is the log of the mean sales price for homes, for each
city/town/village in Oakland County that averaged more than 15 sales per quarter
between the first quarter of 1990 and the fourth quarter of 1996. Robust standard
errors are used to allow for heteroskedasticity and first-order serial correlation.
Dummy variables indicating the quarter of the year are included in the regression
but are not reported. 1"The effective tax rate is divided by the discount rate (3
percent) prior to estimation.

81

Table 4.5 Summary Statistics Dual Sales Data

 

Mean ' Std. Dev. M! M

Avg. Price 1992 $103,296.20 $60,580.63 $13,901.76 $389,249.30
Log(Avg. Price 92) 11.39 0.55 9.54 12.87
Per Pupil Spending 1992 $6,101.12 $1,523.36 $3,897.13 $9,909.18
Effective Tax Rate 1992 1.76% 0.28% 0.99% 2.69%
Avg Price 1996 $119,242.30 $63,455.85 $15,754.56 $447,761.20
Log(Avg. Price 96) 11.56 0.50 9.66 13.01
Per Pupil Spending 1996 $6,682.06 $1,296.92 $4,741.91 $9,659.28
Effective Tax Rate 1996 0.63% 0.25% 0.29% 1.33%
% Chng Hshlds 90-96 9.22% 9.10% 0.3% 35.3%
Price Growth 90-92 -1.68% 6.34% -18.00% 20.00%

 

Note: All dollar totals are reported in constant 1990 dollars. Total number of
observations used in estimation: 709.

O

Table 4.6 Dual Sales Estimates

82

 

 

I u m m
(OLS) (OLS) (ZSLS) (OLS)
Normalized Tax -0.0604 -0.1120 -0. 1274 -O.156l
Rate’ (0.0504) (0.0513) (0.0519) (0.0522)
Spending Per 0.0073 0.0056 0.0051 0.0040
Pupil (100’s of 3) (0.0017) (0.0017) (0.0017) (0.0017)
Community Price 0.3326 0.3642 0.3736 0.3285
Growth 90 - 92 (0.0902) (0.0895) (0.0897) (0.0913)
Improvements 0.0499 0.0475 0.0467 0.0418
(0.0183) (0.0181) (0.0181) (0.0179)
Pct. Chng. -0.2635 -0.3420
Households 90-96 (0.0632) (0.0727)
Semi-Urban -0.0620
(0.0132)
Rural -0.0774
(0.0160)
Constant 0.1010 0.1154 0.1196 0.1229
(0.0202) (0.0202) (0.0203) (0.0203)
Obs. 709 709 7 09 709
R2 0.1154 0.137 0.147 0.1544

 

Note: Standard errors are in parentheses. The dependent variable is the log of the
house’s sales price. Dummies indicating the quarter of the houses sale in each
year are included in the regression but are not reported above. The number of
observations represents the number after differencing. ’The tax term in the OLS
estimates is divided by the discount rate (3 percent) prior to estimation. The
percentage of land available for development in the community is the instrument

for the ZSLS estimate.

83

Table 4.7 Summary of Effects of Tax and Spending Changes

 

 

 

 

Dual Sales Aggregate Data
Fixed mt Random

1 H m m Effect Difference Growth
Effect ofa $1 $2.00 $3.73* $4.25" $5.20“ $9.93" $8837“ $8.17
Tax
Differential
Tax 6.0% 11.2% 12.7% 15.6% 29.8% 25.1% 24.55%
Capitalization
rate'r
% Effect of 0.7%" 0.6%“ 0.5%" 0.4%* 0.0% 0.3% 0.3%
additional
$ 100 per

Jrupil

 

Notes: * Indicates the estimate was significantly different from zero at the 5% level, ** at a 1% level.
*Assumes 3 percent discount rate and infinite house life.

Chapter 5

THE INCIDENCE OF MICHIGAN ’8 SCHOOL REFORMS

I. Introduction

In 1994, Michigan reformed its method of funding public schools. Michigan
greatly reduced its reliance on local property taxes and lowered the state’s income tax by
a small percentage. The lost revenues were replaced by increasing the state’s sales tax
from 4 to 6 percent, by increasing the cigarette tax from $025 to $0.75 per pack, and by
small increases in various other taxes. Michigan also changed the way revenues are
allocated to its public school districts. Although the reforms were, for the most part,
revenue neutral, they did not affect all state residents equally. The tax changes impacted
state residents by changing both the tax payments that citizens had to make and by
changing the prices of certain goods, most notably housing. This chapter presents an
estimate of how these reforms impacted Michigan’s taxpayers.‘

Michigan’s reforms did not affect all communities equally. Rural school districts,
which tended to have relatively low per-pupil spending amounts prior to the reforms, in
general experienced large property tax cuts and large spending increases. Urban areas,
which tended to spend more per pupil than the rural areas, also experienced large property

tax cuts. Real per-pupil spending in urban districts generally increased, but the amount of

 

‘ Michigan’s school finance reforms are described in Addonzio, Kearney and Prince (1995).

84

85
the increase varied. Urban areas that had relatively high spending prior to the reforms
saw much smaller increases than those urban areas that had relatively low spending.
Semi-rural or suburban areas were often the highest spending school districts prior to the
reforms. These districts received smaller tax cuts in percentage terms, as they were
required to levy “hold harmless” millages to maintain their higher spending amounts.
Many of these previously high spending districts experienced real spending decreases
after the reforms.

To measure the incidence of Michigan’s reforms, the tax and spending changes
experienced by four communities in Oakland County are examined. Brandon Township
is used as an example of a rural district, the City of Pontiac as an example of an urban
district with relatively high pre-reform school spending, the City of Femdale as an
example of an urban district with relatively low pre-reform spending, and Bloomfield
Township as an example of a suburban district.2

Within each of these communities, the effects of the tax and spending changes on
three hypothetical household types are considered: a senior citizen couple; a family with
two adults and two children, and a single person living alone. For each of these groups,
the effects of the changes under three housing assumptions are examined: that the

household rents; that the household owns a house and purchased it prior to the school

 

2 Michigan’s reforms affected school districts rather than municipalities. However, in most cases the
school district and community boundaries are similar. The school districts assumed for the communities
outlined are as follows: for Brandon Township the Brandon School District; for Bloomfield Township the
Bloomfield Hills School District; for Femdale the Femdale School District; and for Pontiac the Pontiac

School District.

86
reforms, and that the household owns a house, but purchased it subsequent to the school
reforms. These different groups should represent the experiences of a large portion of

Michigan’s residents.

11. Incidence Definition

In measuring the incidence of Michigan’s tax changes, it is important to
distinguish between the statutory incidence and the economic incidence. The statutory
incidence refers to who is required by law to pay the tax. In the case of Michigan’s sales
tax, the statutory incidence is on retailers. They are the ones who must write the check to
the state government. The economic incidence refers to who actually pays for a tax. In
the case of Michigan’s sales tax, retailers may pass some or all of the tax onto consumers
by charging higher prices. If consumers face higher prices as a result of a sales tax
increase, they are in essence paying part of the tax, even if the law states that the tax is on
retailers.

To measure the incidence of Michigan’s reforms, the tax payments made by
individuals in 1996 are compared with the tax payments they would have made, had the
reforms not been implemented. The economic incidence of the tax is used as opposed to
the statutory incidence. That is, when calculating tax payments, an attempt is made to
account for any shifting of the taxes that may have occurred.

There are some important aspects of the tax changes that are not examined. First,
the change in the excess burdens of the various taxes is not considered. The excess
burden is the cost of taxes to consumers over and above the tax payments. Taxes change

the relative prices of goods causing consumers to substitute away from taxed goods. This

87
substitution effect is what causes the excess burden of the taxes. Changes in taxes may
lead to changes in employment, capital allocation, and wages. These “general
equilibrium” effects are not considered. Finally, the tax incidence measurements
represent a single point in time. No attempt is made to measure the lifetime incidence of

Michigan’s tax changes.

111. Tax Incidence Estimates

Estimating the tax incidence of Michigan’s reforms requires numerous
assumptions and calculations, as well as data on income, tax rates, expenditure patterns,
and housing costs. The tax and spending incidence is calculated for four communities:
Brandon Township, Bloomfield Township, the City of Femdale, and the City of Pontiac.
For each of these communities, three household types are considered: a senior citizen
couple; a family with two adults and two children; and a single person living alone. For
each of these groups the effects of the changes are examined under three housing
assumptions: that the household rents; that the household owns a house and purchased it
prior to the school reforms; and that the household owns a house but purchased it after the
school reforms were implemented.

To calculate the tax incidence, an estimate of the income and housing costs for
each of these groups is needed. Based on these data, the effects of the policy changes on
each of these groups are calculated.

Income Assumptions
To estimate the incomes of the various demographic groups, data from the 1990

US. Census of Population and Housing are used. For each of the three communities, six

88
measures of income are needed: the income of the three household types assuming they
lived in owner-occupied housing, and the income of the three household types assuming
they were renters.

The following procedure is used to estimate the income of these groups. First, for
each community the median incomes for households headed by a senior, family
households, and non-family households are used as a base. I assume these figures
represent the median 1990 incomes for the three household types: a senior citizen couple;
a family with two adults and two children, and a single person living alone, respectively.
From the census data, the median income of renters, homeowners, and the overall median
income for each community is determined. I then assume that the ratio of income for
renters to the overall median is the same for all three household types and that the ratio of
income for homeowners to the overall median is the same for all three household types.
The 1990 incomes for all demographic groups are then calculated based on these
assumptions.

The income calculations are best illustrated with an example. In 1990, the median
income for family households in Brandon Township was $47,434. The median income
for households in owner-occupied housing in Brandon Township was $46,871, and the
overall median income in Brandon Township was $45,139. Therefore, the ratio of house-
owner income to the overall median income in Brandon Township was 1.038 to 1. To
calculate the income for families living in owner-occupied housing in Brandon Township
in 1990, multiply $47,434, the overall median income for families living in Brandon

Township, by 1.038 to get an income estimate of $49,236.

89

After calculating the 1990 incomes, the next step is to estimate the income of
these groups for 1996. The Current Population Survey (CPS) is used to calculate the real
percentage growth in mean income for each income quintile between 1990 and 1996.
The CPS shows that the income growth experienced by each quintile of the overall
income distribution was not the same over this time period. Higher income individuals
experienced much higher income grth between 1990 and 1996 than did lower income
individuals. Between 1990 and 1996, the lowest earning quintile in the US. experienced
a 3.5 percent reduction in real income, while the highest quintile experienced a 10.8
percent increase in average'real earnings.

These data are not directly applicable to Michigan, however. The real median
income in the US. declined by 1.2 percent between 1990 and 1996. The real median
income in Michigan increased by 9.14 percent over the same period, and per capita
income increased by 11.5 percent. To adjust the US. income quintiles to match the
Michigan experience, I add 10 percentage points to each of the percentage changes for the
US. income quintiles. This percentage is used to calculate the 1996 income for each
group in 1990 dollars. The figures are converted to 1996 dollars using the CPI-U. The
income estimates for each group are presented in Table 5.1.

Sales Tax Estimates

As part of its school finance reforms, Michigan increased its sales tax from 4 to 6
percent. Michigan’s sales tax, known in Michigan as the “Sales and Use Tax,” is not
comprehensive in coverage. Michigan exempts most services, grocery foods,
prescription drugs, and many other items from its sales tax. The Michigan Office of

Revenue and Tax Analysis estimates that sales tax exemptions reduced state revenues by

90

over $2.5 billion dollars in 1993. Total sales tax collections in 1993 amounted to just
over $3.5 billion. The Michigan sales tax is assessed on retailers; it is safe to assume,
however, that some or all of the tax is passed on to consumers. 3

The sales tax can be divided into two components: the part assessed on goods
purchased by Michigan consumers, and the part assessed on goods purchased by
producers and non-residents. Generally, tax incidence studies assume that the part of the
sales tax assessed on goods purchased by consumers is borne fully by consumers (see
Pechman 1985 or Blume 1982 for an example using the Michigan Sales and Use Tax). A
recent empirical study by Poterba (1996) supports the validity of this assumption.’

The effect of the sales tax assessed on goods purchased by producers is less clear.
Using the Consumer Expenditure Survey, Ring Jr. (1989) estimates the amount that
Michigan residents spent on taxable goods in 1979.5 He estimates that total consumer
spending on taxable goods by Michigan residents accounted for approximately 59 percent
of Michigan’s sales tax revenues. He speculates that the remainder of sales tax revenues
were generated through purchases by producers and non-residents.

The incidence of the remaining 41 percent of taxed expenditures is much more
difficult to measure. Presumably, much of the sales tax on goods purchased by non-
residents is exported. The sales tax paid by producers may be partly exported to non-

resident shareholders. Some fraction, however, may also be reflected in lower wages for

 

3 Source: Michigan ’5 Sales and Use Tax, Office of Revenue and Tax Analysis, Michigan Department of
Treasury 1994.
‘ Poterba (1996) finds that retail prices rise by approximately the amount of a sales tax increase.

5 Ring Jr. (1989) performs this analysis for all 50 states.

91
Michigan’s workers, a lower return on capital invested in Michigan, or in higher prices
for some goods.

To estimate the incidence of the sales tax change, I assume that my demographic
groups bear the full burden of the tax assessed on the goods that they purchase, and none
of the burden of the tax assessed on producers. Assuming that the demographic groups
bear none of the burden of the producer portion of the sales tax is, admittedly, somewhat
unrealistic. There is, however, insufficient evidence to support any other assumption.
However, it should be noted, that this assumption may cause the estimated tax share of
some groups to be underestimated.

The share of income that different income classes in Michigan spent on taxable
goods is calculated in Michigan ’5 Sales and Use Taxes: 1994. These calculations are
derived from the Bureau of Labor Statistic’s Consumer Expenditure Survey. The income
figures are adjusted for inflation and then used to predict how much the various
demographic groups used in this chapter spent on taxable items in 1996. I assume that
consumers do not substitute away from taxed goods as a result of the sales tax increase.

The sales tax estimates are presented in Table 5.2.

92
Effects of Property Ex Changes on Renters

Michigan’s recent school finance reforms affected renters in several ways. In
Michigan, renters are allowed to apply for the state’s property tax credit, which is a
refundable credit on the state’s income tax. Residents are allowed to receive the credit if
their property tax payments exceed a Certain percentage of their income. As part of the
school financing reforms, the fraction of rent that is considered to go towards property
taxes for the purpose of calculating the credit was increased from 17 to 20 percent.

The reforms also affected school property tax rates. Prior to the reforms no
distinction was made between the various classes of property. As part of the 1994
reforms, a distinction was made between owner-occupied residences, known as
“homesteads,” and all other forms of taxable property, which became known as “non-
homestead” property. The property tax rate was reduced on both types of property. The
reduction, however, was generally larger for homestead property than for non-homestead
property.

The reduction in the property tax rate on non-homestead property reduced the tax
payments that landlords needed to make. Some of this tax decrease may have been
shifted onto renters in the form of lower rents. If renters did experience a decrease in rent
as a result of the reforms, the decrease would also affect their eligibility for Michigan’s
property tax credit. For renters, the size of the tax credit they might be eligible for is
based in part on their annual rent. Finally, the change in service levels may also be

reflected in rental prices.

93

There is no consensus in the economic literature as to the extent to which
landlords are able to shift property taxes onto renters in the form of higher rent.6 If
renters are fully mobile, landlords cannot pass tax increases onto renters, because the
renters would simply move to a community with lower rents.7 Carroll and Yinger (1994)
have argued that property taxes could lead to higher rents if they were used to finance
service increases. Residents may be willing to pay more in rent for better services.
Carroll and Yinger estimate that landlords are able to shift $0.11 of a one dollar property
tax increase onto tenants. However, these results are not directly applicable to the
Michigan example, because in Michigan the property tax changes were not directly tied
to service changes.

Because the literature does not offer a clear conclusion as to the effects of
property tax changes on rents, and because I do not have data describing what actually
happened to rents in Michigan, I estimate the effect of Michigan’s property tax cuts under
a variety of assumptions: that the property tax changes were firlly passed onto renters;

that none of the property tax changes were passed onto renters; that half of the property

 

6 See Carroll and Yinger (1994) for a review of the literature and a discussion of the arguments
surrounding tax shifting in the case of rental housing.

3 If capital is full mobile, the capital portion of Michigan’s property tax will reduce the return to capital in
all sectors. A tax on Michigan property is negligible when compared to the overall capital stock in the US.
(see Courant 1982). Therefore, the capital portion would not be shifted onto renters; it would instead be
borne by all capital owners. The land portion of the property tax could either be shifted into rents or borne

by landlords.

 

 

 

94

tax changes were passed onto renters; and following Carroll and Yinger (1994), that
$0.11 per $1.00 dollar in property tax changes was passed onto renters.

Carroll and Yinger’s results suggest that very little of the property tax decrease
would have been shifted onto renters. Estimates of the change in rent prices for the
Detroit-Ann Arbor MSA, taken frOm the CPI Detailed Report, show that real-rent prices
declined slightly between 1992 and 1996, decreasing approximately 3.1 percent. This is a
slightly larger decrease than was seen by other midwestem cities. Rents decreased by
approximately 2 percent in real terms in the Cincinnati-Hamilton area, increased by
approximately 1/2 percent in the Cleveland-Akron-Lorain area, and increased by
approximately 1.5 percent in the Chicago—Gary-Lake Co. area, over the same period.8
These findings suggest that rents may have decreased as a result of Proposal A; but if
they did, the decrease was fairly small.9

To calculate the rent for each of the demographic groups, the percentage of
income that went towards rent for the household’s income type and location is used. For
example, the estimate of the household income for a renter family in Brandon Township,
in 1990, is $25,614 (see Table 5.1 for income estimates). An average of 22.4 percent of
income went towards rent, for renter households with an income of $25,614, in Brandon

Township, in 1990. Therefore, the estimate for the annual rent for this household is

 

3 The real percentage change in rent prices is calculated by comparing the change in rent and the change in
the overall price level for the specified geographic area.

9 It should be noted that even if tax savings are fully shifted they would reduce rents by only a small
percentage. For instance, for Brandon Township, the assumption of full shifting leads to an estimate of a 4

percent reduction in annual rents (the largest effect of the three municipalities).

95
$5,738 and the estimate of monthly rent is $478. The rent prices are adjusted to 1996
dollars using the CPI- U.

Michigan allows renters to qualify for its property tax credit. The property tax
credit is a refundable credit on the state’s income tax that residents are eligible for, if their
property tax payments exceed a certain percentage of their income. For the purposes of
the credit, renters were allowed to assume that 17 percent of their rent payments went to
cover the property tax prior to the reforms, and 20 percent subsequent to the reforms. I
assume that 20 percent of the rent price went to pay the landlord’s property tax on the
apartment prior to the reforms.10 I assume that the property tax payment that landlords
paid for apartments decreased by the same percentage as the percentage decrease in the
non-homestead total property tax rate. Based on the estimated dollar total of the property
tax reduction, the new rent and the new property tax credit that renters are eligible for is
calculated under the different shifting assumptions outlined above.

I also assume that the increase in services did not result in higher rents. The
empirical literature has little to say about how a change in per-pupil school spending
would affect apartment rents. For the case of apartments marketed towards seniors and
single people, it is probable that increases in school spending have little or no effect on
rents. For families, services may have an effect, but is not clear how large the effect

would be. Therefore, I assume that the service changes were not shifted into apartment

 

‘0 Rubinfeld and Vishny (1982) speculate that the 17 percent figure used to calculate the percent of rent
that went towards property taxes was taken from Netzer (1966). Netzer cites a survey of over 29,000
apartments. This survey estimates that in the northeastern U.S., real estate taxes represent approximately

20 — 25 percent of rent. In the Midwest, taxes generally represented between 15 and 20 percent of rent.

96

rents. The effects of the property tax changes on rents and on income tax credits are
presented in Tables 5.3 and 5.4.
Effects of Propem Tax and School Spending Changes on Existing Homeowners

Michigan’s reforms substantially benefited individuals who owned a house at the
time of the reforms. The reduction in local property tax rates reduced the annual tax
payments that homeowners needed to make, generating a large tax savings. The
reduction in property tax rates reduced the annual cost of owning a home, making owning
a house or buying a bigger house more attractive. This caused an increase in the demand
for owner-occupied housing, resulting in capital gains for existing homeowners. The
changes in school spending levels also affected housing prices. Houses in school districts
that received spending increases, increased in price relative to houses in commrmities
that did not receive spending increases.

To calculate the pre-reform house price for each demographic group, I assume
that the ratio of the income of the demographic group to the median income of
homeowners was the same as the ratio of the house price of the demographic group to the
median house price.ll For example, the median house price, in 1990, for Brandon
Township was $94,700. The median income of households residing in owner-occupied
housing in Brandon Township was $46,871. The estimate of the household income of
families residing in owner-occupied housing in Brandon Township is $49,236.

Therefore, the ratio of family income to the median income for house owners is 1.05046

 

” This assumption implies that the income elasticity of housing is 1.0. Harmon (1988) finds the long-run

income elasticity of housing to be 1.0 using data from the Panel Study of Income Dynamics.

97
to 1. To calculate the house price for families in Brandon, the median house price is
multiplied by this ratio to get $99,478. The CPI-U is used to adjust this figure to 1996
prices.

To calculate the change in property tax payments, I use the effective school tax
rate in 1992 as the pre-reform rate and the effective tax rate in 1996 as the post reform
rate.12 I use the same house price for both cases. Although Michigan’s reforms increased
the price of many houses, the reforms also included an assessment cap which limited the
rate of assessment growth to 5 percent or the rate of inflation, whichever was less.
Because of this, house assessments could not be increased to account for any real capital
gains which resulted from Michigan’s reforms.

The tax payment calculations are complicated by two factors. First, for taxpayers
who claim itemized deductions on their federal income taxes, property tax payments are
deductible. Second, Michigan’s income tax allows taxpayers to receive a property tax
credit, if a filer’s property tax payments exceed a certain percentage of their income.
Both of these factors must be considered when calculating the effects of a property tax
change.

To calculate the effects of the federal income tax deduction, I assume that all of
my household types that owned houses claimed itemized deductions on their income tax.

I assume that the total income tax deductions for the household were equal to the average

 

'2 The effective tax rate is the annual tax payment divided by the price of the house. This differs from the
statutory tax rate if house assessments are not perfect. I use the average effective tax rate for the

community. See Guilfoyle ( 1997) for an explanation of how these tax rates are calculated.

98
deductions for their income bracket.13 Using information on the household’s income,
deductions, and exemptions, the marginal federal income tax rate the household faced in
1996 is calculated. The total property tax payment is multiplied by the marginal federal
income tax rate to estimate the property tax savings resulting from the itemized
deduction. The amount of prOperty tax credits on the Michigan income tax, is calculated
using the Homestead Property Tax Claim Form for the 1996 Michigan income tax.

The appreciation in house prices that resulted from Michigan’s reforms, is
calculated using Guilfoyle (1997). This study uses a sample of 700 individual houses in
Oakland County that sold once before and once after Michigan’s reforms. The study
estimates the effects of the tax and spending differentials between communities on house
prices. This study can be used to predict the total real price change for houses between
1992 and 1996.” Estimates of the effects of Michigan’s school financing reforms are

presented in Tables 5 .5 and 5.6.

 

'3 Source: Statistical Abstract ofthe United States 1995 -- Table No. 535 “Individual Income Tax Returns--
Itemized Deductions and Statutory Adjustments, by Size of Adjusted Gross Income: 1992.”

‘3 Guilfoyle (1997) focuses on estimating the effects of tax and service differentials between communities.
The estimates do not distinguish between changes in house prices due to Michigan’s economic expansion
and changes due to the reduction of the average tax rate. For the purposes of predicting the effects of
Michigan’s reforms, I assume that house prices in Michigan would have grown at the same rate as house
prices in the Midwest had the reforms not been implemented. Once this assumption is made, Guilfoyle
(1997) can be used to estimate the effects of Michigan’s reforms on house prices. The median sales price
of one family houses, in the midwest, increased by 9.23 percent (Source: Statistical Abstract of the US. --
1997 Table No. 1191. “Existing One-Family Houses Sold and Price, by Region: 1970 to 1996”) over the

period examined by Guilfoyle (1997).

99

The estimates in Table 5.6 show that existing homeowners experienced large
capital gains as a result of Michigan’s reforms. Houses in Bloomfield Township saw a
real gain in price of approximately 3.5 percent, houses in Brandon Township and the City
of Pontiac saw real gains of approximately 14 percent, and houses in the City of Femdale
saw gains of approximately 17 percent." These capital gains estimates are not estimates
of the total change in the price of houses in these communities; rather, they are estimates
of the capital gains due to Michigan’s reforms. House prices in Michigan also changed
due to construction, economic growth, and other demand factors for the community.
When all factors are included, prices in Brandon Township are predicted to have
increased by approximately 15.5 percent in real terms between 1992 and 1996; houses in
the City of Pontiac saw an increase of approximately 20 percent; houses in Bloomfield
Township saw an increase of approximately 6.57 percent, and houses in Femdale saw an
increase of approximately 29 percent.16

Michigan’s reforms resulted in a large one-time capital gain for homeowners.
The tax incidence estimates presented in this chapter compare actual tax costs in 1996
with what they would have been had the reforms not been implemented. Including the
entire capital gain as a tax benefit for existing homeowners in 1996 would overstate the

annual incidence of the reforms, while excluding the capital gains item the incidence

 

3’ The exact predicted capital gain rate used in later calculations is 13.72 percent for Brandon Township,
3.46 percent for Bloomfield Township, 13.51 percent for the City of Pontiac, and 17.10 percent for the
City of Femdale. These calculations include the effect of school spending changes on house prices.

'6 The total percentage change in price was also calculated using the results from Guilfoyle (1997).

100
estimates would understate the benefit to existing homeowners. Therefore, an estimate of
the annual value of the one-time capital gain is needed.

I assume that the capital gain is traded in for a 30 year annuity with an annual
interest rate of 7.5 percent per year. The annuity is assumed to have 12 payments per
year. The annual benefit of the capital gain is calculated as the sum of the 12 annual
annuity payments. This annuity is essentially a reverse mortgage, where the homeowner
receives payments instead of making them.

The annuity term and interest rate are consistent with the mortgage assumptions
used to calculate the tax incidence for households purchasing their houses subsequent to
the reforms. These households must pay a higher price for their homes as a result of the
reforms. The higher house price results in a higher annual mortgage payment for these
households. By assuming existing homeowners trade their capital gains for an annuity,
the annual cost to new homeowners due to housing appreciation is comparable to the
annual benefit received by existing homeowners.l7
Effects of Propem Tax and School Spending Changes on New Homeowners

The effects of Michigan’s reforms on homeowners who purchased their homes
after the reforms were implemented were somewhat different than the effects on
homeowners who owned their home prior to the reforms. Michigan’s reforms led to a

real increase in the price of housing, with an increase in price of 3.5 percent in

 

'3 The cost and benefit are not exactly the same because new homeowners can deduct mortgage interest
from their federal income taxes, making their cost somewhat lower. Note, that I am also assuming that
homeowners do not pay income taxes on their annuity. Most homeowners will probably leave the capital

gain as equity in their house. Therefore, they will not have to pay federal income taxes as a result of it.

1 01
Bloomfield Township, an increase in price approximately 14 percent in Brandon
Township and the City of Pontiac and an increase in price of approximately 17 percent in
the City of Femdale. Because of these price increases, individuals who purchased their
homes subsequent to the reforms had to pay higher prices for them, and as a result will
see smaller savings from the reforms.

New homeowners pay property taxes on the full value of their house, because
houses are reassessed upon sale. This differs from the tax treatment of homeowners who
owned their house at the time of the reforms. For these homeowners, assessment
increases were restricted to the rate of inflation, so that increases in home prices due to
the reforms could not be taxed. As part of Michigan’s reforms, a real estate transfer tax
was created. House sales are taxed at 0.75 percent of the value of the house being
transferred.

In calculating the incidence of Michigan’s reforms, the expenses and savings
individuals experienced in 1996 are compared with what they would have experienced,
had the reforms not been passed. The experience of new homeowners in the absence of
the reforms would be the same as the experience of existing homeowners in the absence
of reforms. Therefore, the data in Table 5.5 describe the estimates for new homeowners
in the absence of the reforms, as well as the experience of existing homeowners.

The annual property tax payments are still lower for new homeowners after the
reforms; but, the reduction is smaller than the reduction for existing homeowners. The
reason is that for new homeowners, the entire house price is subject to taxation. To
calculate the new house price, the estimates of the capital gains from Table 5.6 are added

to the base house price. The annual tax payment is the effective tax rate multiplied by the

102
new house price. The federal property tax deduction and the Michigan property tax credit
are also calculated.

The higher housing prices also result in higher annual mortgage payments for new
homeowners. The additional annual mortgage payment is calculated assuming a thirty
year mortgage at a mortgage interest rate of 7.5 percent.18 I assume that sellers bear the
full burden of the real estate transfer tax. The real estate transfer tax is relatively small
(0.75 percent of the sales price) and assumptions concerning who bears the burden of this
tax do not have a large effect on the estimates.

Mortgage interest is deductible from the federal income tax. Since the portion of
a mortgage payment that goes towards interest decreases annually, the federal income tax
deduction will decrease as well. The mortgage deduction is calculated assuming that the
loan repayment is in the first year. Estimates of the effects of Michigan’s reforms on new

homeowners are presented in Table 5.7.

 

'8 The average interest rate for fixed rate mortgages in the US, in 1993, was 7.3 percent; in 1994 it was 7.9
percent. (Source: Statistical Abstract of the US. -- 1995. Table No. 805 “Characteristics of Conventional

First Mortgage Loans for Purchase of Single-Family Homes: 1980 to 1994.”)

103
Changes in the Income Tax

As part of Michigan’s school reforms, the state’s income tax was reduced from
4.6 percent to 4.4 percent. Calculating Michigan’s income tax is straightforward. In
1996, income tax filers were allowed to deduct an exemption of $2,500 per household
member. Family households are assumed to have four members, senior households two,
and households with a single occupant one. After deducting exemptions, the household’s
income is multiplied by the tax rate to calculate the income tax paid.

Calculating the taxable income is somewhat more complicated for seniors.
Seniors are allowed an extra exemption of $900 per household member. Additionally,
Social Security income is not taxed under Michigan’s income tax. Finally, in 1996
seniors were allowed to deduct pension income of up to $31,920 for a single filer or
$63,840 for a married filer, from their income when computing the tax.13

Seniors are assumed to earn all of their income from Social Security and pension
earnings. I assume that seniors received the average payment for a married couple
collecting Social Security. All remaining income is assumed to be from pension
earnings.

Household types described as homeowners are assumed to claim itemized
deductions on their federal income tax. For these individuals, state income tax payments
qualify as a deduction. The federal deduction is calculated based on the household’s

estimated federal marginal tax rate. Individuals described as renters are assumed to claim

 

'9 The amount of pension income that seniors were allowed to deduct was increased dramatically in 1994.
In addition, beginning in 1994 seniors were allowed to deduct interest, dividend, and capital gains income

(subject to a cap). These changes, however, were not part of Michigan’s school reforms.

104
the standard deduction. Thus, they were unable to claim state income taxes as a
deduction in their federal income tax calculations. The income tax calculations are
presented in Table 5.8.
The Cigarette Tax

As part of Michigan’s school reforms, the Michigan excise tax on cigarettes was
raised from 25 to 75 cents per pack. The average retail price of a pack of cigarettes
jumped by slightly more than 50 cents after the imposition of the tax (and has remained at
this higher level) suggesting that the cigarette tax increase was essentially passed on to
consumers in the form of higher prices.20

The effect of the increase in Michigan’s cigarette tax on households depends
primarily on the number of smokers in the household. In 1992-93, approximately 25
percent of the adult population in Michigan were smokers. The rate of smoking among
Michigan residents declines with income (see Table 5.9). However, a substantial fraction
of individuals were smokers in each income category.

The increase in the cigarette tax that resulted from Michigan’s reforms was
substantial. The 50 cent increase was equal to approximately 30 percent of the 1993
average retail price of a pack of cigarettes. Michigan smokers could reduce their tax
liability in several ways. First, they could decrease the amount of cigarettes they smoke.

Due to the addictive nature of smoking, however, the price elasticity of cigarettes is

 

2° Source: Michigan '3 Cigarette and Tobacco Taxes 1996, Office of Revenue and Tax Analysis,

Department of Treasury.

105

relatively low at 0.5.21 Michigan residents could also try to evade the tax by consuming
smuggled cigarettes. The cigarette excise tax rates are substantially lower in Indiana and
Ohio. Michigan’s Department of Treasury estimates that approximately 4 percent of
cigarette tax revenues are lost due to smuggling.22

For the purpose of the incidence calculations, a fairly simplistic smoking
assumption is used. Smokers are assumed to smoke one pack of cigarettes per day, both
before and after the imposition of the tax. The annual cost of the increase in the cigarette
tax is $182.50 for a one pack per day smoker. When reporting the total tax incidence for
different household types, the incidence is reported assuming the household has no
smokers, and the incidence is reported assuming the household had one smoker who
smokes one pack per day. The tax effects for households with additional smokers, or
households with smokers who smoke more or less than a pack a day can be readily

calculated by multiplying $0.5 0 by the number of cigarette packs consumed.

IV. Total Tax Incidence
The total tax effects for each group are presented in Tables 5.10 - 5.12. The total
tax effect as a percentage of income is presented for all communities and demographic

groups in Figures 5.1 - 5.8. The largest beneficiaries of the tax changes were homeowners

 

2' Source: Sung, Hu, and Keeler (1994). The authors find that the short-run price elasticity of - 0.40 and a
long-run elasticity of -0.48.
22 Source: Michigan 's Cigarette and Tobacco Taxes 1996, Office of Revenue and Tax Analysis, Michigan

Department of Treasury.

106
who owned their homes at the time of the reforms. These homeowners benefited from a
large reduction in their annual property taxes.

The lower property taxes had two effects for existing homeowners. First, the
property tax cuts reduced the annual property tax payments existing homeowners needed
to make. Second, a fraction of the prOperty tax cuts were capitalized into house values.
This resulted in a large one-time capital gain for many homeowners. To calculate the
annual incidence of Michigan’s reforms, this one-time capital gain has been converted
into an annual annuity.

The size of the property tax cut varied by school district. Residents of previously
high spending districts often saw smaller property tax cuts, since their districts were often
required to levy hold-harmless millages. Among existing homeowners, all groups
benefited from Michigan’s reforms. Non-smoking households in Brandon Township, the
City of Femdale, and the City of Pontiac saw annual payment savings of 2 percent of
personal income or more. Existing homeowners in Bloomfield Township also benefited,
although the annual gains were somewhat smaller.

The large gains of existing homeowners came at the expense of new home
purchasers. Because some of the property tax cuts are capitalized into house prices,
existing homeowners are able to take some of the future benefit of the lower property
taxes with them when they sell their house. The capital gains of existing homeowners
represent higher housing prices faced by new homeowners. For most households, the
annual cost of owning a house was still smaller after the reforms. However, the reduction
in the annual cost of owning a house was substantially smaller than the reduction seen by

existing homeowners. As a result, the reduction in the annual cost of owning a house was

107
not sufficient to offset the increased annual cost of the sales tax. All new homeowner
household types paid more in taxes after the reforms were implemented.” The increase
in annual payments for non-smoking new homeowners ranged between 0.30 and 0.60
percent of the household’s annual income. For smoking households, the tax
consequences were even worse, with some new homeowners seeing an increase in their
annual payments greater than 1 percent of their income.

For renters, the picture is less clear. Renters tended to have lower incomes than
homeowners, and lower income individuals spent a larger fraction of their incomes on
taxable goods. Therefore, the increase in the sales tax hurt renters more in percentage
terms. The effects of the property tax cut were not clear. There is little evidence on the
effects of property taxes on rents. If the property tax cuts were fully shifted onto renters,
non-smoking renter households come out ahead from a tax standpoint in Brandon
Township and the Cities of Femdale and Pontiac. Bloomfield Township saw very small
decreases in the property taxes levied on apartments, so renters there would still come out
behind.

If the property tax cuts were not shifted into rents, then virtually all renter
households pay more after the reforms. Further, because of their relatively low income,
renters who smoke were especially hard hit by the reforms. Renter households with one

smoker saw an annual payment increase of over 1 percent of income in some cases.

 

33 Note that the term “tax” is used somewhat loosely here. The higher mortgage payments faced by new

homeowners is included in this tax calculation.

108

V. Service Level Changes

Michigan’s school reforms changed both how revenue for school districts was
raised and how it was allocated to school districts. In this section, I briefly discuss the
incidence of the school spending changes.

Prior to Michigan’s reforms, revenues were allocated using a district power
equalization (DPE) approach. DPE systems attempt to equalize the effective property tax
base school districts have access to, by allocating aid to poorer districts. Districts are
then allowed to choose their own level of funding.“ The combination of district choice
and a DPE system that did not fully correct for wealth differences caused a wide variance
in the pre-reform spending levels in Michigan school districts, with wealthier districts
generally spending more per pupil.

As part of the 1994 reforms, Michigan switched to a foundation grant approach to
allocating revenue. Under Michigan’s foundation grant system, all schools spending
below the basic grant amount of $5000 per pupil, in 1993-94, were moved up to that

1.25 The base grant is increased yearly based on grth in the “school aid fund,” the

leve
group of taxes allocated to school spending. Districts spending above the foundation
grant were allowed to maintain their old spending amounts. Very high spending districts,

those spending over $6500 per pupil in 1993-94, were required to levy additional “hold-

harmless” millages to maintain their spending levels. Local choice was removed from the

 

2‘ See Reschovsky (1994) for a discussion of DPE and other school funding mechanisms.
2’ To ease the cost of switching to the new system, low spending districts were not immediately raised to

the new higher spending amount. See Kearney (1994) for an in-depth explanation of the new system.

109
school spending process and school district spending amounts are now set by the state
government.

As a result of these changes, rural districts, which tended to have per pupil
spending amounts well below the foundation grant, often saw very large increases in their
revenue available per pupil. ,The spending changes seen by urban districts had more
variation. Low spending urban districts saw increases in funding; but, some higher
spending urban districts saw little change in their funding levels. Many wealthy suburban
school districts, which generally spent the most per pupil prior to the reforms, saw their
real spending levels decrease.

There are obviously many school districts whose experiences do not neatly fit into
one of the categories above. Despite this, I believe that this categorization is a useful way
of examining the effects of Michigan’s reforms. The districts chosen for this analysis
were picked because they are representative of the types of districts described above. The
Brandon School District, a rural district, saw a real spending per pupil increase by 18
percent, between 1992 and 1996. Pontiac Schools, an urban district, saw a real increase
of 10 percent per pupil, Femdale Schools, a low spending urban district prior to the
reforms saw real spending increase by approximately 30 percent, and the Bloomfield
Hills School District (the district for much of Bloomfield Twp, a suburban location) saw
real spending fall by 2.5 percent per pupil.

Of course, to some extent, increased educational spending benefits everybody. To
the extent that increased spending increases the output of public schools, that is increases
the education of children, society is benefited with a more educated electorate, more

skilled work force, reduced crime, and so forth. The most direct benefit, however, clearly

110
goes to households with school-aged children. Brandon schools were able to spend close
to $900 more per pupil; Pontiac schools had approximately $650 more per pupil available
to them; and F emdale Schools had almost $1600 more per pupil. The Bloomfield Hills
school district, on the other hand, saw a reduction of approximately $300 per pupil in
resources. Michigan’s new foundation grant system does not take into consideration cost
differences between communities.26 Therefore, the dollar increases have even more
impact in rural districts, where costs are presumably lower.

There are two important caveats that need to be considered when examining the
changes in school spending in Michigan. First, higher spending does not necessarily lead
to better schools, much depends on how the money is spent.27 Second, the residents of a
community may not value the services fully. In previously low spending districts,
spending was raised to a level significantly higher than the one chosen by the electorate.
It is possible that these districts had lower spending because residents of the districts did
not place a high value on educational spending. What the money is spent on may also
effect the value residents place on it. For example, school districts may choose to spend
their extra resources on trying to retain students who are at risk of becoming a drop-out.
Parents with children who are not a high drop-out risk may place little value on such
efforts.

To summarize, families in Brandon Township saw approximately $875 more per

year, per pupil, spent on educating their children. Families in Pontiac saw approximately

 

3" The system does allocate more revenue to districts considered to have a large number of “at-risk”
students. Again see Kearney (1994) for a description of special funds appropriations.

23 See Wassmer and Fisher (1996) for a discussion of this.

111
$650 more per pupil and F emdale saw roughly $1600 more per pupil. The dollar cost of
these additional services is far in excess of the additional tax cost any of these families
saw. Some families in these municipalities, most notably existing homeowners, saw
large overall tax cuts.

Families in Bloomfield Township saw a reduction in their school spending of
roughly $300 per pupil. Many residents of Bloomfield Township, therefore, saw an
increase in their taxes and a cut in their service levels. It should be noted, however, that
Bloomfield Township residents are, on average, far wealthier than the residents of the
other communities. Even with the spending cuts, school spending in Bloomfield

Township remains at one of the highest levels in the State of Michigan.

VI. Conclusion

Michigan’s recent school reforms resulted in changes in a number of the state’s
taxes, including the sales, property, and income tax. School spending was also
significantly changed in a number of districts. This chapter presents an estimate of the
degree to which these changes effected various Michigan household types. The effects of
the changes on seniors, families, and single individuals are examined. For each of these
groups, the incidence of the changes on renters, homeowners at the time of the reforms,
and homeowners who purchased their homes alter the reforms were implemented is
estimated. Finally, each of these groups is examined in two urban communities, a rural
community, and a semi-urban community. These community types broadly represent the

experiences of many Michigan households.

112

In general, in urban and rural communities, existing homeowners received large
overall tax cuts. New homeowners often saw small increases in their taxes over what
they would have been had the reforms not been implemented. The tax effect renters saw
is largely dependent on the degree to which property taxes are shifted. These
communities generally saw increases in the money per pupil available to their schools.

Semi-rural or suburban locations, often the highest spending prior to the reforms,
saw slight tax decreases for existing homeowners. All other household types generally
saw moderate tax increases. School spending in the communities that were spending the
highest amount prior to the reforms, generally decreased in real terms.

It should be noted that although the analysis presented in this chapter applies to a
large fraction of Michigan households, several important topics were not investigated.
The effects of the tax reforms on landlords depends in part on the degree to which rents
changed. The effects of the portion of the sales tax increase that effected producers was
not measured in this chapter. Finally, the effect of Michigan’s reforms on wages, capital
allocation, the return to capital, and overall economic growth were not considered nor

was the change in the overall excess burden resulting from Michigan’s tax system.

4.00%

3.00%

2.00%

1.00%

0.00%

-1.00%

-2.00%

-3.00%

-4.00%

113

 

 

IExisting HomeOwners Senior
lExisting Homeowners Single
EEINew Homeowners Family

ERenters (Full Shifting) Senior
IRenters (Full Shifting) Single
H Renters (No Shifting) Family

1:1 Existing Homeowners Family
INew Homeowners Senior
ElNew Homeowners Single
"Renters (Full Shifting) Family
IRenters (No Shifting) Senior
flRenters (No Shifting) Single

 

 

Figure 5.1 - Tax Change as a Pct. of Income (Non-Smoker - Brandon Twp).

 

4.00%

3.00%

2.00%

l .00%

0.00%

-1.00%

-2.00%

-3.00%

-4.00%

114

 

 

IExisting Homeowners Senior
IExisting Homeowners Single
EEINew Homeowners Family

ERenters (Full Shifting) Senior
IRenters (Full Shifting) Single

IRenters (No Shifting) Family

ElExisting Homeowners Family
INew Homeowners Senior
EINew Homeowners Single
[IRenters (Full Shifting) Family
IRenters (No Shifting) Senior

IRenters (No Shifting) Single

 

 

Figure 5.2 - Tax Change as a Pet. of Income (Smoker - Brandon Twp).

 

115

 

 

 

 

I Existing Homeowners Senior 13 Existing Homeowners Family
I Existing Homeowners Single INew Homeowners Senior
EEINew Homeowners Family ENew Homeowners Single
ERenters (Full Shifting) Senior [IRenters (Full Shifting) Family
IRenters (Full Shifting) Single IRenters (No Shifting) Senior
HRenters (No Shifting) Family IRenters (No Shifting) Single

4.00%

3.00%

2.00%

1.00%

0.00%

-1.00%

-2.00%

-3.00%

-4.00%

 

Figure 5.3 - Tax Change as a Pet. of Income (N on-Smoker - Bloomfield Twp).

4.00%

3 .00%

2.00%

l .00%

0.00%

-l .00%

-2.00%

-3 .00%

4.00%

116

 

 

I Existing Homeowners Senior
I Existing Homeowners Single
EEINew Homeowners Family

ERenters (Full Shifting) Senior
IRenters (Full Shifting) Single
IRenters (No Shifting) Family

13 Existing Homeowners Family
INew Homeowners Senior
ElNew Homeowners Single
[IRenters (Full Shifting) Family
IRenters (No Shifting) Senior
IRenters (No Shifting) Single

 

 

Figure 5.4 - Tax Change as a Pet. of Income (Smoker - Bloomfield TWp).

 

4.00%

3.00%

2.00%

1.00%

0.00%

~1.00%

-2.00%

-3.00%

-4.00%

117

 

 

IExisting HOmeowners Senior
IExisting Homeowners Single
EEINew Homeowners Family

ERenters (Full Shifting) Senior
IRenters (Full Shifting) Single
IRenters (No Shifting) Family

1:] Existing Homeowners Family
INew Homeowners Senior
ElNew Homeowners Single
[IRenters (Full Shifting) Family
IRenters (N o Shifting) Senior
IRenters (No Shifting) Single

 

 

Figure 5.5 - Tax Change as a Pet. of Income (Non-Smoker - Femdale)

 

4.00%

3 .00%

2.00%

1.00%

0.00%

-l.00%

-2.00%

-3.00%

-4.00%

118

 

 

IExisting Homeowners Senior
IExisting Homeowneis Single
EilNew Homeowners Family

ERenters (Full Shifting) Senior
IRenters (Full Shifting) Single
IRenters (No Shifting) Family

1] Existing Homeowners Farrrily
INew Homeowners Senior
EINew Homeowners Single
[IRenters (Full Shifting) Family
IRenters (No Shifting) Senior
IRenters (No Shifting) Single

 

 

Figure 5.6 — Tax Change as a Pet. of Income (Smoker - Femdale).

 

4.00%

3.00%

2.00%

1 .00%

0.00%

-1.00%

-2.00%

-3.00%

-4.00%

119

 

 

I Existing Homeowners Senior
IExisting Homeowners Single
EEINew Homeowners Family

ERenters (Full Shifting) Senior
IRenters (Full Shifting) Single
IRenters (No Shifting) Family

U Existing Homeowners Family
INew Homeowners Senior
EINew Homeowners Single
[IRenters (Full Shifting) Family
IRenters (No Shifting) Senior
IRenters (N o Shifting) Single

 

 

Figure 5.7 - Tax Change as a Pet. of Income (Non-Smoker - Pontiac).

 

4.00%

3.00%

2.00%

1.00%

0.00%

-l.00%

-2.00%

-3.00%

-4.00%

120

 

 

I Existing Homeowners Senior
IExisting Homeowners Single
EINew Homeowners F arrrily

BRenters (Full Shifting) Senior
IRenters (Full Shifting) Single
IRenters (No Shifting) Family

Cl Existing Homeowners Family
INew Homeowners Senior
BNew Homeowners Single
IRenters (Full Shifting) Family
IRenters (No Shifting) Senior
IRenters (No Shifting) Single

 

 

Figure 5.8 - Tax Change as a Pct. of Income (Smoker - Pontiac).

 

121

.3259 soon 26: macaw... oEom ”082
3.53% coca—anon E250 use £550 093 622%

 

 

 

 

 

 

N25; 094$ $3 _ 25.3 ON: Swim “swam
$0.on N_N.:N s: N toe; oNed NNNNNN NEE”.
:N.N ; ENQN; $2 _ N Na 8.; tho; Seam
52.5%
SNNNN NNNdNN S: N 33w; 2: SEN; swam
25.on ENS SN. _ _ v OSNNN o2: NNNNNN been.
w ENNN N028 some N NRNNN 23 Ed; 5:3
. . 3:302:02
385m usage 0128.. co 3.6
NNNNNN NNNdNN $2 N 93.; £3 83% 235
N _.NNN NNNNNN s? N NSENN £3 NNNNNN NEE
cNodNN $.sz S? N NNNéN £2 3.on seam
hump—um
NNoNNN ONNNNN s? N ENNNN 0N _ ._ 25.va swam
N s. 5 Mafia sow. : v SNNNN oN _ ._ NNNNNN been
25.on ENNN s? N NNNNNN nNZ NEON» 55m
5:302:03
385m swipe emcee; .8 EC-
2:. _ NN $.25 32 N sad; 81o 0:.on swam
NNN.oNN 03.0% $3 _ a. N _ NS.» 8% 5.8» read
oNNNNN 035% SN. _ x a. ow _ . :a 9:3 NVNSN 5.5m
. hut—MM
oNvNNN o _ 33.. 0%.. i v N2.on N2: SENN 29.5
. NONNEN 95.2; SEN N NNNNSN 33 RN 3% SEE
NNNNN ; 3 _ .2 ; SEN N NcNNON N2: NNNNNN Baum
uogcgo:
385m 2:: EEEEE 9; BEESE
3N; No3. ; S: N N 3% $3 $33 swam
ENS 4N _ .NNN is N a. _ EN 82 «Nata r55
:2 5 N23. 5 s? N EN; 82 NSNNN seem
how—5M
o _ oNNN NSSNN s? N 338 N3. 83% “ism
NNOSN ESNNN .\..N._ _ v 3.23 NS; 5.3 NEE”.
NNNNNN NNNNNN some N 23% NNS N 5.va 5.50m
LUESOOEOI
Eoozum c0985. 85% 531%.
5380
$2.8 08: 93:8 8o: 2:85 2:55 2:85 25 55:23 0am
9:85 cam: 0E8:— com: Rom emu oEeoE ooo. Baa—tam decoy—16:30 088:— 5602

 

89:. 22.8.5: 32...; 3.— 053..— ..e «8.533 —.m 033—.

122

.8959 :03 022 8.5m: 980m 6qu
2552330 9553‘ can Max—3.1 xah can 353.: .8 350 “oasom

 

 

 

 

 

 

N_ S oNNN vNNN 3.3 .x.N.NN N25; 29%
£5 NNNN NNNN NoNdN <53 N858 been.
$3,. 2% ONNN 23.3 $9? _ .NN ; 55m
hour—um
a; NNNN NNNN $2» 91.3 83% 235
VNNN N8. ; meow _ :55 ..\&.NN Nada Nzefl
.NNN 33 New» M33; .x. _ .NN NSNNN 5:3
50350050:
30325045203 oflEom .6 36
oo ; .NNN owNN N _ now ON; .oN NNNeNN 29.5
SNN NNNN N3 23 ; ..\.._.NN ENNN N=E£
mm a 22 NNNN NSNN ego? 08.on seam
8.5.x
SS 8% NNNN NNN.N_N .x; .NN NNN..NNN 23%
oNNN N 5. ; ONCN Need; o\szN N _ N. _ NN N=E£
NNNN NNNN waw NNoN ; ON; .NN 8°.on seem
3:360:52
Eoocum BEES“: oavEum ho EC
2.; NNNN NNNN 3mg e52 NNN._NN 23m
oNNN 9i NSN N23; cN.N.NN NNN,oNN Nzefl
NcNN axe. ; NNNN NN _ .N ; aNNNN ONNNNN 55m
hug—BM
SNN N8. ; NNON NoN.N_N ONONNN 03.5 swam
NNNN €th 35. a No _ . 3N c\,.N.NN No}: _ N NEE
35 onNN NNN. 3 83$ °\,.N.NN NNN.NN a Baum
hugogoz
385m 2:: 29:58.8 9; Euceogm
oN a 83 oNNN NEW 02.3 $3; “ism
ENN NoNN NNNN co 1N; e13. SNNNN beam
_N ; NNVN .ch oNNNN ego? _NN.£N 5:3
SEE:
NNNN NNNN 0va 5.3 e: .NN 23$ 29%
83 Nos. ; NNNN No? ; ..\°N.NN N853 res".
NNNN _NNN v _ NN N32 5 ON; .oN NVNNNN Seam
505500501
Eoofim 53:9 93% 5655
28 3N 30 9% New: 225; C.o 2;: 22:5 8 93:8 25:
5 039.8:— xmh. mug—am 5C. mug—mm “Guam mus—OD “cam 9:005 .x. ©00_ 0:500:— $39:sz

 

«N520 gingham—:09 Sata> .5.— 33525 nah. 8.5 «d 033—.

123

macaw—330 PEAS/w 98 £550 ooam ”830m

 

 

 

 

 

_NN OON ON NO; xNNNN- NSN ONON NONON 2N5
_NN EN ON NNNN .NNNNN- NNNN SEN ONO,NN NEE;
ONN NO_ N ON :NN o\cNNNN. NNNN _NO._N NNN.NN Baum
NNN OEN ON NNNN o\,.NN.NN. ONO_N SEN
ONN NNON ON NNNN oNONNNN. VNNN SEN
NNN NEN ON OONN .NNNNN- NNNN NOEN : N.NN seem
. aflooaom DEF—QEBNUEQK mo END
NN ON ON NON o\oONO. OOO._N NNNgN NON.NN ”mam
ON NNN ON ONN NON? OOEN ONN._N ON_.ON 365
ON ONN ON NNN “NON? NNO._N OOONN 81$ 88%
NNN NEN ON NONN $3.8- NOON NON,_N
NNN OEN ON NONN ONSNN- NOON NNN._N
ONN 5N ON NONN XSNN- OONN NNN._N
NOEEN NOEEN NOEEN NOEEN $2882.82 2 85 25 2 gem 23V 2 95 23
388m 2 ram oNoN =an 5d. Begum and E O85 xah .9235 33 3:52
NcocNESNN< Naota> 8ch own—EU gum Begum Mm .ONm 1352‘ “6:95 Box 8am
Nw5>mm End 3:52

 

2583— .85m 9535:: 5?... NuEZNm 23% NO 33.2530 Wm ~35.

2

1

 

 

60258 :03 32 £383: 088 ”202
Ocean—3:5 553x can mchoU cog bosom
ONN NON ONN NON ONNN ONNN _ONN NNNN NNON OOON ONON _OON NOON N _ NN 2N5
EN 2 5 EN; NON :ON NONN NSN _NNN NOEN NON._N :EN NNN._N ONON OON.; OOENN
ON 5 ON _ N _O _ N N_ _N OONN OONN NONN ONON OOO. _N ONO. O N _OO. _ N OOO._N NONN 23 Beam
NN_N ON 5 ON _ N O. _N NNNN O_ NN ONNN OOON NNO._N ONOON NOO._N NOO._N OONN NNEN 2N5
. OO_N OON :_N NON OO_N OON _:N NON NOZN ONN._N :EN ONN._N ON OON.; NOEON
O_~N OO _ N OOON OO_N NOON NOON _OON N_ON OON. _ N OOZN NON. _ N NOZN NONN NOSN 5.5m
EN ON _ N NEN ON .N NNON NNON ONON ONON _ON. _ N NON. _ N OON,_N NON,_N SNN ONN._N O_N:_N
ON ON ON ON ON ON ON ON NON. _N OON.; OON. _ N NN_ . _ N ON ONN._N NOSE
ON ON ON ON ON ON ON ON NOO. _ N OOO, O N OOO. _ N ONO. _ N ON 34.; Baum
gagfiaaagg
OEN NO _ N OO _ N OON NNON OOON OOON OOON OON. _ N OON. N N OON. .N ONN._N O _ NN ON_._N 2N5
25 83 SN NNN 2: 8; :5 ONN OON._N OON.; OON._N N_N,_N ON 215 NOEON
ON _ N ON; ON .N 25 OOON OOON NNON OOON NNSN SEN NNSN ON_._N OOON ONO._N SEN
gag
NOEEN NEOEN NOEEN NEOEN NEOEN NEOEN NOEEN NOEEN NOEEN NOEEN NEOEN NOEEN ONEONOm O3: 35.8% 23
.523. _ _ :O: OOON ==O 285 : Na: OOON =OO EEO _ _ N3: 93 :5 NBC 5: N385. :85
A533. 30% 3:83— NNomv O8 5;. moi
£920 x3. 56¢on E 0308:. N620 xfl. €39; No oEESNm 8895 szU 5.. xfl. “8:95 .No No.g_.mm

 

:38: .. 8.55.3.— ..85 a: 3.2.2.. N as;

125

.3958 :83 26: 33:5: 080m ”802
Ocean—830 N.No£=< 95 £550 oaa 605cm

 

 

 

 

 

 

omm cow 30m $2 aamavmm mmwNWNm flmfim
Sm 5; 03:5 $2 @363 moaJWN 328m
85 mm; Com . O\Om_ w3<.mm ommdmm Baum
Econom ulmcaonc outcom .No 36
ON NN_N ONNN GO: NOO.NNN NNONNN ONNOON
om ow _ m wmm. 5 $9 qumm $3.me 388m
ow 2 5 worm OON— Noadmm acqmmm Baum
ON _ N ONNN NNN,_N 0O: OONNNN SO.O_ ON 2N5
om vwm. N m m3 4% £1 m 833; owmdmmm $83
cm 33% 03.3 O\O_ m wmmdflm oatonmm 880m
Eoozom NEE 22.22025 93. 22.3805
OOON OOON NON: GO: ONO.ONN OOONON OONaN
3% Nova 39$ O\Om_ mmodcm vind— 5 338m
m _ vm 8% Own. _ a $2 www.mmm v3.8a 880m
ANOEoOom 33v Amcbofim 015 $5.83” 35 2nd Ed. 0885 ANOEQOM 036
5.80 :28:an NSF Begum EEwNO—Z BE 3503
xfl. moi x5. Begum @038:wa
OONEOE

 

3:83— 2? I 05: “End—533 N53335:: 3.. 3:55 an: 3:259:— nah 5.53.5 @8253“.— m.m «Bah

 

.3988 :03 26: £3.58 080m 682
€33 038230 98 3032330 OLSEN/x “Ooh—om

 

126

 

 

OOON NOON NOOONN ON ON 2 .N NNN.NON 2N5
NONJN NNNN NNO.ON ON ONN OOON NOO._NN beam
OOON OOON O_ N.NN ON ONN NO; ONN.ONN seem
Econom omcconc canon mo 35
NNSN OONN NONOON ON ONN OONN NNNNNN 2N5
NOONN OOON OONON ON NON NNON ONO.ONN NOSE
. NOON OOON ONO.NN ON ONN NO_N OOO.ONN OONOON
NOON NNNN NOO.NN ON ON _ N ONO._N SOON NN 2N5
OON.; OONN ONO.ONN ON OONN NOO.ON ONN.OONN NOSE
OOO._N ONNN NNO.ON ON _ONN NNN.ON OOO.OO~N seem

  
 

 

 

285m, £5 23885
NOO._N ONON ONN.ON ON ONN OONN OO_.NON 2N5
ONO.ON OON, _ N NON.O _ N ON NON ONON OON.ON ON :85
OOZN OOON NOONN ON NON ONNN OOO.OON seem
ONOSON 58 338 NO 56 ONNEONBN O3 Naneum 3O OBOONON E ONOENON OE
Evian 02.3 3295 N620 £55350 x3. 30:5 85 025$
335$ 3352‘ .ONm 331355 xOH dam x3. Begum
ONN OONEONE

 

Nhoaiouficmfincaxm .8 325% 8.3.3.: no 835:3— e.m «Bah.

 

.3958 :89 3a: 838:: 25% “80 Z
Aug: 030.230 98 macaw—:28 NLo£=< ”330m
48% 2:85 35va 05 ENE $225 owmwtofi mo 8.326% 2: N8 $5.3th Noam Evian uwmwtoE ~959ch 3:328?

 

127

 

 

OO_N ONON ON ON NOON OONOON 2N5
OOON OONN ON ONN NNON NON.ONN NEE
OO O N ONNN ON OON NON N NNOOON seem
NNN NOON ON NNN _NON ONO.OON . . Dana
OO_N OOON ON OON OONN NON.OON Omefl
NON OOON ON ONN NO_N O_O.ONN Baum
O _ N - .OOON ON NSN OOO._N OOO.N_ ON 2E5
NOON ONON ON ONN NON.ON OOO.NONN NEE
NOON OONN ON ONNN OOO.ON NNN,ONON BEN

  
 

 

 

£85m NEE 2».chon

ON - OOON ON SN OOON OOONON swam
N_N OOSN ON OO_N :ON OOOONON been.
OK; - meow om v3 mama Nwmdow Baum
NwEZOm .Ncugmm @583— 3; ONE—0.3m Ed ANENOMDM \BV ONE—chum EL

.58an owmwtoE N680 £8.50:on x5. Begum outn— 8:03

3:53.. F5223 xmk doi xmh Begum

32 5322

 

30:38:82 .5 Z :c 38:”.— Eheom .3 835sz hm 033—.

 

128

 

 

go 6958? .3555 32 can 9.58 x3 .2 eg— .m:m=oU 002 ”350m 63:20.. :03 05; 85mm oEom ”202

.98 9:85 .803.“ on. :0 05852? 93 2:85 888 Mo 22838 >5 .8 .0: 8a mesuc 5:. 0535 58222.

 

 

 

 

 

 

m 8 88 88 898 8 8 8 98.8 _ 838 29.8
88 $8 88 N858 8 8 8 9938 e N858 9:88
8 8 8 8 N88 98.8 23.8 95.8 N :38 .528
whoa—5M
88 28 88 98.88 8 8 8 98.8 _ 838 28:8
88 88.8 .88 $9.98 8 8 8 9938 v 9558 9:88
8 8 8 8 828 98. 8 o 8.38 95.8 N 8 8.88 5.58
whoa—BO
3868 088.: glaze; 8 >5
«8 88 v8. 8 88.88 8 8 8 98.8 _ 83.8 28:8
88 8 _. 8 8 _. 8 v2.88 8 8 8 993 8 v 8 E8 9:58
8 8 8 8 898 98.8 23.8 98.8 N 80.88 .258
$585M
88 8N. 8 N8. 8 838 8 8 8 98.8 _ 88.88 29.8
:8 98. 8 8o. 8 N 8. :8 8 8 8 9558 v E. _ 8 9:58
8 8 8 8 8988 98. 8 23.8 98.8 N N858 5.58
30:30
38:8 28.8% 28:5 8 3.6
08 $8 88 838 8 8 8 98.8 _ 858 28:8
98 8:8 8N8 8de8 8 8 8 9858 v 8888 9:58
8 8 8 8 08.9.8 98. 8 o _ 3. _ 8 898 N 88.88 .258
EBSM
88 88.8 88.8 @888 8 8 8 98.8 _ 8,..N8 28:8
3 8 98.8 98.8 88.8 8 8 8 8 9938 v 88.88 been
M8 98.. 8 NS. 8 828 88.88 98. 8 N. :38 95.8 N 8N8 8 SE8
88:30
28:8 2...: 2858.9 93. 28588
88 o _ 8 N88 2.3 8 8 8 8 98.8 _ $8.88 28:8
:8 89 8 N8. 8 85.8 8 8 8 993 8 v 9.88 9:58
8 8 8 8 N8. 8 98. 8 23.8 98.8 N 838 5E8
8220M
98 SN. 8 08. 8 238 8 8 8 98.8 _ 238 288
88 898 N28 8988 8 8 8 95.08 v 8988 9:58
8 8 8 8 ONE 8 98. 8 23.8 898 N 88.88 .288
8:38 99. v3 99. o: 2:85 68.98 v2:): 5:988 2:85 28> 28958 68:
5; .98.... .xfl. 0388;. .5558 .820 can we bizoow coca—53m Mo 53:52 0:58:
_m:cc< 5.5.31 Hmom 5.53% 0.5 camp—on— —a._oom 8mm vggtmm

 

REED Sanfluchn 2.8.!» .3.— 83533— nah. 2:3..— afi 025—.

129

Table 5.9 Percentage Smoker by Income for 1993

 

Income lass Percent Smoker
Less than $10,000 36.3
$10,000 - $19,999 26.7
$20,000 - $34,999 27.3
$35,000 - $50,000 23.8
Greater than $50,000 17.3

 

Source: Michigan Office of Revenue and Tax Analysis

I!
1‘

Aowmm :82 no 325.80 oBa-C

 

130

 

88 S8 28- S8- 8 :8 28:8
88 88 _8 - NN8 - 8 88 9:88
88 _8- 8 88- 8 9.8 .28
onccom mo NzU
88 _8- 38- 88- 8 88 28:8
88 88 98 - _ _ 8 - 8 $8 9:58
58 N8 - 8 88 - 8 88 SE8
2.658 8.0 95
9.8 28 88 - . 88 - 8 88 28:8
:8 88 S8 - 8 8 88 9:58
$8 88 8 8 8 88 SE8
g. .S. 28588
88 _ 8 88 - SN 8 - 8 88 28:8
88 88 N8 - t 8 - 8 $8 3E3“.
8 8 88 - 8 8 8 - 8 88 SE8
gh- coccflm
8:828 Boa
88 88- 28- 88- 88. :8 28:8
88 :8 - _8 - 98 - 88 - 88 9:58
8- 88. 8 E8- 38- 88 SE8
umccom ho NED
8N 8 - 88 - 38 - 2 8 - 88- 88 28:8
88 8.8 - o8 - 88 - 88- 88 2.5.8
:8 - 88 - 8 N08 - 9.8- 88 SE8
SE.SE.-I8...S do
8 8 8 - 98 - 8 8 - 88 - 88 28:8
88 3 8 S8 - 8 08 - 88 9:68
88 _ 88 8 8 N8 - 88 SE8
an; 28.808
88 88 - 98 - o8 - 9.8 - 88 288
S8 8 8 - :8 - N8 - No8 - 38 9:58
N8 - 88 - 8 8 8 - .88 - 88 SE8
gh- cog—8m
«55.8 :5
.Caonmv €8.95 .85 omens-u new—5:0 “.620 End :0 uwcmsu
Zebu x“... “ooh-E 5:. 9: 9:8:— xah moi 09:95 5:. xmh 83m
.58 .98 S 888 -S .85 88 80 Sam .S .85

 

33:0“ .3.— ouco—zuc— =.—. :23. .26 033—.

131

Saw a x03 _ 88—988 on? 5x088 28 was 238.8: 38:85

 

 

88 8- 28- 88- _8- :8 28.8
88 S8 .8- 88- _8- 88 3E5
9.8 98 - 8 88 - .8 - 88 SE8
. OMS—hon ho sz
88 v8- 38- 88- 88- 88 288
88 .8 88 - 88 - o8 - 88 3E5
8 8 88 - 8 28 - 88 - 8 8 SE8
25:8 8 86
88 _8 88- :8- 8- 88 28:8
88 88 88 - 8 8 - 88 3E3”:
988 88 8 8 8 - 88 SE8
nah-8.8.8. '88
28 88 - 88 - 88 - 88 - 88 28:8
88 88 :8 - 2 8 - N8 - $8 3E8
88 88 - 8 88 - 88 - 88 SE8
gh- covcem
825.8 8:
z 8 88 - 8 8 - 88 - 88 - :8 288
88 88- .8 - o. 8 - :8 - 88 3E5.
888 88 - 8 88 - 88 - 88 SE8
umzcom ‘6 NWT-v
88 88- 38- . 88- 88- 88 28:8
28 o8 - 8.8 - 88 - 8 8 - 38 8:68
88 88- 8 88- 88- 88 SE8
28.5“: 88. C...
8 8 8 - o8 - 88 - 88 - 88 28:8
88 88 88 - 8 88 - 88 5E5.
888 88 8 8 S8 - 88 SE8
m3 28:85
«8 28- 88- 88- 88- 88 28:8
88 88 - :8 - 88 - 88 - 38 8E8
38 88- 8 8.8- 88- 88 SE8
gh- SEEN—m
«5.2.8 3
839:8 $8328 .55 quE-u owEE-u ”Eu-C Eva co owcmzu
Set-m xx... goo-cm 5:. xfl- 9:50:— :8... moi owcnzu xfl- 5d. 82am
3.: .92- ? SSE 8o SSE 8: -S .85 -S SSE

 

92.523“! 28:01 .3.— 3522: nah. .33. e-m 03.:-

132

.fivUGn—OH GOOD 2’2 when—HHS: 080m 5qu

5% Ba x08 _ mow—08m of» 20:83: 5 3x25 28 88:35

.8850 53 9305 Eva: vows—2:

8m £55 xab 5885 533:2 05 5 x8 2:02: 7225.“ 05 80¢ 53 atomoa 05 .«0 328.6% 05 9 26 mowEEo 53 0885.

 

 

 

 

 

 

ma - 83 - Rm - 83 - a; 23m
2% - 8a.; - o; - Sq; - $3 335
m _ a - mo: - om as: - 5% seem
Economlomccomv ow. Eel; we Elna
oi - 23 - Sm - 3.; - 8% $5
23% - an; - Em - 83m - 3mm 55$
cam - mg - 2 £3 - SS Baum
8S - o5 - ma- 2% - 3% 23¢
23m- med. 3;- cad- 83 .265
$3 - 8o: - ma - $9; - 8% - 8:3
285m 2:: Eoceose 9; Eaceosm
25.; - SN: - a3 - wax; - «Rm 035m
wig - 5mg - 8m - 29$ - 8mm been
35 - 8% - cm 5.; - RS
"GB—9:3 Coy—08:55 Ema—£0 x3. owcnnu ego
69cm xah 69cm USP 2:85 xah doi x5. 83m
.95 :2; Mo scum go scum .8 ~85

 

9.2.3859: wnzmmxm 3.— 3.8an— xah .38. :.m «Bah

133

 

5% 8a xoaq fl moxofim on? 8:95 28 mm: 228:2.— mufismmf

.flouto x8 .90ch Eva: coca—oi 8m EEC «8H

€305 cmwfiuzz 2t 8 53 0885 :fiovfl 2: 80¢ 53 .mtomoa 05 .3 amazonvov 05 9 26 mowcnco x8 088:?
dorm—=23 £5 E vows—2: 8a $35 $590: no 3582 PameEZ Boa 3:38 :23 mowfifio
333m («0 88:0 2a 802 docflooaaa wagon 8 26 2am $085880: Bu: :8ng owawtoE :28:ch 05 39:05.

 

 

 

 

 

 

2% v3 5 - 8; - a; 23m
8% Em Em - 8S - 3mm :65
RS :5 em 3; - 5% seem
Afloor—Um UNCCOAC Data—Om .«0 >30
83 a; Sm - 3m - SS 23m
3% we; 5 - cm; - 33 SEE
Na $5 cm 5 - NEW seem
Eoofim 285$ 286$ mo 36
v3 :3 a; - N; 3% ~35
:3 mg a; - SS - mmwm 3E5
$3 8% m; - 3S - 8; - 8:3
385m 2:: @380qu San Eoceosm
8% mg a3 - 5 3% gnaw
:i fig 8a - m; - 8mm been
23 a; om 3 SS seam
mfiuonwv fioonméoé Newcasu “8% Emcwno owcmso
tutu x3. “outm— xah 0685 fig .35 5d. 83m
33 BE .6 scam ‘8 scum Mo 35

 

Eocicoficfl 3o Z .8.“ 8522: 55. 33,—. «fim ~33.

APPENDIX

APPENDIX

This appendix contains a subset of the data that were used inthis dissertation.
Table A.1 contains summary information on the various cities, townships, and villages of
Oakland County. The data contained in this table are mostly from the US. census. It is
included to give the reader a brief overview of the various communities in Oakland
County.

Table A2 contains a listing of the 709 houses that were used to perform the dual
sales capitalization estimates. The dual sales were generated by manually matching lists
of house sales in Oakland County. The unmatched house sales lists were produced by the
Oakland County Equalization Office. Each house is identified by its sidwel], a unique
identifier assigned to the house by municipal authorities. Also included in this list are the
city, town, or village, in which the house is located, the school district the house is
currently zoned for, the 1992 and 1996 sales price of the house, and the 1992 and 1996
state equalized value of the house.

The final table of this appendix, A.3, contains summary information on Oakland
County’s school districts. For each the school district in Oakland County, the school
millage rate for the fiscal years 1991-92 through 1996-97 are included. Also included are
the per-pupil revenues for each school district for fiscal years 1991-92 through 1995-96.
Per pupil revenue data for 1996-97 are not yet available. Data on the Northville, Romeo,

and Warren school districts are also included, although these districts lie primarily within

134

135
other counties. They are included because at least one of the dual sales houses was zoned

for each of these districts.

 

 

136

 

 

-1.

.3m 33 wmm .38 938» M5 .38 80.5% .2092 33 mm
.\.m.o $2 23 .32 2.32. 32 .32 833 Beam away on
six :2 25.2 .33 £33 .33 .38 83°; 852.83 H
.\....o 82 2% .33 $33 :3 .32 83:” 2.83 gases: 2
.32 $2 $3 .35 «23% mt; $03 8me 3.5 H
.3: m5 :3: .38 5.3 :3 .3§ 83% 2.33: m
.3m Nae God... .32 23% 2.3. .3: 83$ flames: mm
.3wm 22 85. .3; £35 33 .3Q O33; 82220 0
.3m :2 £3 .38 253 23 .3: 835 .35 saga .5
:3 3.2 38.2 sis 33$ 82: .30 8wa 288$ E
.39 E: 33R :98 £33 Ea $0.2 83?; £5 aewéafi 8
.\.~._ $2 22: .38 9.3% Max... .32 82c; aoaqafim o...
.32 $2 5.2 .38 $23 2; .38 832% 8.8.88 m
.\.Z . $2 5.2 .33. ~32” Sow .3m 83% 8.3.6 2
.3m :2 83 .38 23% mi. 5.2 832; 8.32.6 oh
.\.v.£ £2 32: .3a 83% a3 .33 83% 8.5.5 a
.3. 82 £3. .33 soda #3 .32 83%» £5 23885 2
.3. $2 max? .38 33% $3: .32 cowzmm 2&885 0
.3m $2 832 .3E $33 $2 .32 83m; 52355 we
.3m S2 83 x; .38; as. $3... ooofia énfiafiwam E.
.32 $2 23: .33 $33 83. .32 81mm; .Embasm E
33 $2 832 .33 83$ 3% $3.. 833 serum 3
.3: :2 08.: .33 3.3.3 $3 9.»? 838 £5 535 8
.3: NE 9:... .38 £33 :5 .3S 83c; 8.63 <

8.8% 5:93.85 38: 8%: €2.2— 3 38: 2.333. 893 25.2 SE... B
mug—6:023: emu—O: dam Ema—3900 0:59:— 33: Eman— uumhh 095: 556,—.
”QED o\e hao> =38: Lon—3c wean =38: ¢m¢u Mam-5: «cachom flan—.02

 

3.5 M435: can noun—nae.— .GESU vaniao _.< 93:.

137

 

 

o\om._ 33 v3.— o\oo.cw 3v.3.w mmw o\ow.o o2 .vom 815 52am vw

 

.3: 8: 58 .38 83:.» 83.8 .38 88.88 2.888 3
.3m 8: W38 .88 $38 8.8 .3: 2.38 2.888 8
.32 8: 8:3 .38 88.88 22% .38 8.88 83 58m 8
.83 R: 83.8 .88 8.88 818 .3: 83:. .30 :83: N.
.3: :2 88... .38 83.8% 85 .88 838 83: a
.38 ME: . 88.8 .3: 83.8 5.8 .38 88.3.; 2:: .2863: 8
.3. N8: 83 .38 88.8». 83 .38 838 .2.£8m 8
8...: 82 81: .38 8848 M38 .38 838 2.88 8
.30 8: EN .38 83.8 83 .\._ .c 83.8 .82 838: 8
.38 .9: v8.8 .38 88.2.8 83 . .38 83.8 28.5 m
8...: 82 N8; .38 838» ME. .32. 8388 2:825 on
.38... NE 88.: .88 8888 v8.5 .38 888 8.5 0
.3m 8: 83 .\._ .8 .88: m 98 .38 8188 83 2220 8
.38 22 RS .38 . :38 23 .38 8385 8.33 2
.3o 8: 8.5... .38 8on $8.: .3. 838 .8; 8.0 mm
.88... 82 81mm .38 :38 5.2 .88 88.2; :62 8
.3. NS: .88 .38 838 :8; 8% 83m; .5382 8
.38 8: v8.8 .38 838 83 .38 80.8; 2.82 4
.\._.~ 8: 81mm .38 $28 83: .3: 8.88 93.5 8.8.2 3.
.3: 8: 83. .38 838 E} .3: 85: a 83 v.
.32 8: S... .38 83.8 N: .38 838 2.83 a...
.\...._ 8: 83. .38 388 86.. .8...“ 8.88 885 9:53 8
.3_ N8: .88 .38 838 83 .38 888 8:0 83 .5
8.88 882.25 88: 88: 35.8. 3 88: 28:95. 88: 25.2 52F 5
«20:02.0: Own—O: .QOA— flea-590 0:50:— 33: 231— eomhnm 095: 556,—.
wag—U e\e La0> :3—52 .3530 «0h :38: can “Emmacm ”Euchom Gum—no:

 

San aim—5! can gun—.55— 3..on unaEaO _.< 03am.

138

 

 

 

 

[Fl-In?
aco§o>oo no =Q§OU GawEom—Z «maofizom tam mama—DU .m.D ooo— Hoousom

oi .m $2 23 £85 33% 25 «3.: 82% 3.5 85:03 am
5.8 :2 83 £33“ 35$ mt? 9&3 8§$ 59:3 3
sow: 22 R92 §.% 33% SN.” :52 83% 33 233 >
£82 £2 23% $03 $33 wRdN £13 Sci; 20:885 303 x
33 82 898 <2 i . 810$ «Venom $2: 8%; 28533 3
$3 $2 Ed $98 93.2% $3 $92 83% 3.3 333 a
0%? $2 $3 $0.3 33$ 83 .x: .2 83% Eexo .0 as; 2
$0.8 $2 82 £30 23$ 83 33m 83% 28:2 mo ”mg; 3:
$3 $2 a? 0%.: 33% $3 82% 833 33: mo ow2=> E
$0.9 NE vwwfi $35 Emma 818 can: 8qu EH mm

3.23. 5:92.25 8%: 823 $5.5 as 83: «32.56 83: 2:2 5:; E
mEOSGE—O: 02:»: A33 813090 0890—: mam—.3 nun—a..— uumhh 025: :30?

“5.0 .x. 23> 5:52 3:30 «on 5:52 23— wEmsol How—om 5:32

 

3.5 ”£25: can coca—Egon 3589 252.5 ~.< «Bah

Table A.2 List of Dual House Sales

139

 

 

Town

Sldwell ID Town School Dlstr. 92 Price 92 SEV 96 Prlce 96 SEV

05-22-251-024 A Addison Oxford $95,000 $51,000 $134,900 $58,580
05-35-200-002 A Addison Romeo $1 10,000 $59,700 $165,000 $61,350
05-36-300-006 A Addison Romeo $80,000 $25,300. $80,000 $47,720
14-05-326-01 l 2 Auburn Hills Pontiac $65,000 $19,800 $78,000 $28,380
14-1 1-452-047 2 Auburn Hills Pontiac $54,000 $24,100 $85,000 $30,990
14-14-205~003 2 Auburn Hills Pontiac $69,800 $34,300 $88,000 $37,030
14-14-253-023 2 Auburn Hills Pontiac $69,900 $22,000 $98,000 $28,280
14-35-305-002 2 Auburn Hills Avondale $81,000 $37,950 $125,000 $46,100
14-35-326-009 2 Auburn Hills Avondale $87,000 $37,950 $1 10,250 $45,070
14-35-326-037 2 Auburn Hills Avondale $77,000 $31,350 $113,000 $39,710
14-35-376-007 2 Auburn Hills Avondale $82,900 $34,700 $125,500 $41,130
14-35-453-036 2 Auburn Hills Avondale $90,000 $38,000 $125,000 $48,640
14-36-427-015 2 Auburn Hills Avondale $55,226 $23,100 $68,000 $32,300
25~07-356-008 4 Berkley Berkley $33,000 $23,600 $64,500 $31,330
25-07—405-01 5 4 Berkley Berkley $89,000 $30,500 $103,000 $38,180
25—07-410-026 4 Berkley Berkley $85,000 $49,300 $137,000 $56,700
25—07-432—028 4 Berkley Berkley $131,000 $53,800 $179,900 $73,870
25—07-476-01 7 4 Berkley Berkley $1 12,000 $50,300 $159,000 $63,730
25-16-353-027 4 Berkley Royal Oak $82,500 $27,800 $1 14,900 $33,230
25-17-106-019 4 Berkley Berkley $71,500 $23,900 $1 10,750 $30,450
25-17-258-025 4 Berkley Berkley $87,875 $34,800 $121,000 $44,400
25-17-302-006 4 Berkley Berkley $44,000 $19,500 $58,200 $22,000
25-17-303-008 4 Berkley Berkley $89,000 $33 .300 $140,000 $44,670
25-17-354-017 4 Berkley Berkley $58,400 $29,100 $87,000 $38,940
25-17-403-016 4 Berkley Berkley $79,000 $27,100 $105,100 $32,990
25-17-408-075 4 Berkley Berkley $95,000 $25,300 $1 19,000 $34,670
25-17-432-026 4 Berkley Berkley $96,500 $33,900 $121,000 $43,940
25-17-452-032 4 Berkley Berkley $78,000 $39,800 $131,000 $53,900
25-17-456-002 4 Berkley Berkley $73,000 $31 .000 $95,000 $43,630
25-18-101-027 4 Berkley Berkley $62,000 $26,400 $80,000 $34,350
25-18-103-020 4 Berkley Berkley $68,400 $26,900 $106,000 $36,060
25-18-106-002 4 Berkley Berkley $64,000 $28,000 $100,000 $36,690
25-18-152-005 4 Berkley Berkley $69,500 $28,100 $90,000 $33,900
25-18-176-007 4 Berkley Berkley $72,500 $30,700 $85,000 $38,530
25-18-202-013 4 Berkley Berkley $51,000 $22,100 $84,000 $27,630
25-18-252-023 4 Berkley Berkley $75,000 $33,400 $104,900 $45,960
25-18-278-002 4 Berkley Berkley $73,900 $28,000 $1 19,900 $38,870
25-18-303-012 4 Berkley Berkley $67,500 $30,900 $93,000 $38,170
25-18-307-032 4 Berkley Berkley $81,000 $31,200 $1 19,000 $39,500
25-18-478-001 4 Berkley Berkley $81,000 $41,100 $1 15,000 $44,150
24-01-283-007 TH Berverly Hills Birmingham $1 17,500 $43,600 $126,500 $54,890
24-01-151-002 TH Beverly Hills Birmingham $124,000 $65,400 $155,900 $69,320
24-01-228-015 TH Beverly Hills Birmingham $91,600 $40,800 $124,900 $44,530
24-01-228-024 TH Beverly Hills Birmingham $123,000 $60,700 $164,000 $67,200
24-01-254-026 TH Beverly Hills Birmingham $128,000 $65,400 $161,000 $78,730
24-01-282-004 TH Beverly Hills Birmingham $140,000 $67,600 $212,000 $84,830
24-01-433-001 TH Beverly Hills Birmingham $177,000 $76,500 $190,000 $83,980
24-01-456~002 TH Beverly Hills Birmingham $125,900 $66,800 $150,000 $59,400
24-02-151-010 TH Beverly Hills Birmingham $257,500 $101,100 $315,000 $123,100
24-02-378-021 TH Beverly Hills Birmingham $165,000 $70,200 $217,500 $86,990
24-09-203-005 TH Beverly Hills Birmingham $194,500 $84,600 $217,500 $102,550
19-25-151-035 8 Birmingham Birmingham $165,000 $80,500 $180,765 $84,870
19-25-428-008 8 Birmingham Birmingham $270,000 $170,100 $300,000 $207,010
19-25-452-008 8 Birmingham Birmingham $190,500 $85,300 $233,000 $87,410
19-25-476-004 8 Birmingham Birmingham $169,000 $88,200 $249,000 $99,220
19-26-328-017 8 Birmingham Birmingham $250,000 $134,700 $305,700 $135,380

 

Table A.2 List of Dual House Sales

140

 

 

Town

Sidwell ID Town School Distr. 92 Price 92 SEV 96 Price 96 SEV
19-35-277-036 8 Birmingham Birmingham $175,000 $89,300 $235,000 $1 13,100
19-35-327-035 8 Birmingham Birmingham $179,000 $82,700 $230,000 $70,630
19-36-229-016 8 Birmingham Birmingham $151,000 $60,200 $150,000 $69,620
19-36-256-009 8 Birmingham Birmingham $153,000 $74,300 $182,100 $63,780
19-36-278-001 8 Birmingham Birmingham $153,000 $79,700 $257,000 $75,580
19-36-326-029 8 Birmingham Birmingham $135,000 $60,600 $164,000 $74,430
19-36-329-016 8 Birmingham Birmingham $175,000 $80,200 $212,500 $86,410
19-36-330-012 8 Birmingham Birmingham $150,000 $67,400 $180,000 $80,060
19-36-401-043 8 Birmingham Birmingham $148,500 $64,300 $223,500 $71,740
19-36-402-022 8 Birmingham Birmingham $173,500 $77,800 $207,000 $71,580
19-36-429-039 8 Birmingham Birmingham $83,000 $42,400 $101,500 $45,190
19-36-430-045 8 Birmingham Birmingham $95,000 $48,400 $157,250 $60,170
19-36-451-023 8 Birmingham Birmingham $1 19,000 $50,000 $132,575 $63,370
19-36-482-030 8 Birmingham Birmingham $60,000 $33,300 $91,000 $39,280
20-30-355-004 8 Birmingham Birmingham $267,000 $1 13,700 $350,000 $1 14,960
20-30-426-01 1 8 Birmingham Birmingham $129,000 $57,800 $170,000 $69,310
20-30-427-027 8 Birmingham Birmingham $126,000 $61,200 $178,000 $74,440
20-31-101-027 8 Birmingham Birmingham $140,000 $61,300 $280,000 $76,660
20-31-177—030 8 Birmingham Birmingham $85,000 $31,700 $108,000 $43,680
20-31-179-032 8 Birmingham Birmingham $102,500 $40,000 $106,500 $53,820
20-31-351-006 8 Birmingham Birmingham $93,000 $32,700 $117,000 $41,930
20-31-403-012 8 Birmingham Birmingham $105,000 $49,300 $133,000 $60,000
20-31-453-039 8 Birmingham Birmingham $1 12,000 $43,600 $132,500 $60,510
19—01-102-024 C Bloomfield Avondale $235,000 $20,900 $320,000 $123,250
19-03-327-005 C Bloomfield Bloom. Hills $73,500 $32,900 $122,200 $38,240
19-13-177-016 C Bloomfield Bloom. Hills $226,000 $100,600 $295,000 $121,530
19-17-351-013 C Bloomfield Bloom. Hills $158,000 $80,500 $236,000 $92,740
19-19-252-009 C Bloomfield Birmingham $226,000 $104,600 $250,000 $1 13,370
19-19-427-033 C Bloomfield Bloom. Hills $383,000 $164,900 $540,000 $192,390
19-26-353-008 C Bloomfield Bloom. Hills $220,000 $106,700 $375,000 $131,540
19-27-429-009 C Bloomfield Bloom. Hills $312,000 $148,100 $425,000 $188,040
19-29-101-005 C Bloomfield Bloom. Hills $160,000 $76,100 $165,000 $86,640
19-29-227-003 C Bloomfield Bloom. Hills $182,000 $77,000 $183,000 $94,120
19-30-126-023 C Bloomfield Birmingham $190,000 $84,500 $265,000 $94,410
19-34-103-026 C Bloomfield Birmingham $195,000 $81,800 $241,000 $92,480
19-34-402-009 C Bloomfield Birmingham $200,000 $74,700 $257,500 $93,810
19-01-104-003 C Bloomfield Twp Avondale $246,625 $74,400 $360,000 $142,030
19-18-428-016 C Bloomfield Twp Bloom. Hills $255,000 $99,200 $267,000 $1 12,790
19-19-352-001 C Bloomfield Twp Birmingham $1 15,000 $31,000 $131,500 $86,160
19-29-227-026 C Bloomfield Twp Bloom. Hills $244,000 $91,700 $250,000 $112,640
19-31-477-019 C Bloomfield Twp Bloom. Hills $395,000 $153,200 $460,000 $169,550
19-32-203-015 C Bloomfield Twp Bloom. Hills $180,000 $76,100 $210,000 $85,570
03-08-476-015 D Brandon Twp Brandon $96,500 $41,400 $125,000 $52,000
03-14-151-004 D Brandon Twp Brandon $154,000 $58,900 $225,000 $78,400
03-19-278-003 D Brandon Twp Brandon $59,000 $23,500 $76,500 $29,000
03-29-200-034 D Brandon Twp Brandon $150,000 $67,200 $187,000 $83,600
03-29-403-001 D Brandon Twp Brandon $58,000 $20,200 $83,500 $28,800
03-35-103-001 D Brandon Twp Brandon $94,500 $47,700 $90,000 $51,200
20-33-128-015 16 Clawson Clawson $84,000 $37,900 $107,900 $46,860
20-33-130-026 16 Clawson Clawson $85,000 $39,800 $1 15,000 $50,540
20-33-201-015 16 Clawson Clawson $95,000 $37,200 $126,000 $41,960
20-33-276-002 16 Clawson Clawson $87,000 $38,100 $105,000 $74,380
20-33-453-037 16 Clawson Clawson $68,750 $37,200 $90,000 $42,970
20-34-356-01 3 16 Clawson Clawson $62,000 $20,000 $70,000 $21,990
25-03-151-045 16 Clawson Clawson $80,000 $32,400 $1 10,000 $37,470
25-04-131—018 16 Clawson Clawson $79,000 $29,500 $1 19,000 $38,160

 

‘|__; —h—- m

Table A.2 List of Dual House Sales

141

 

 

Town

Sidwell ID Town School Distr. 92 Price 92 SEV 96 Price 96 SEV

25-04-152-014 16 C lawson C lawson $65,000 $23,600 $81,000 $27,420
25-04-228-026 l6 Clawson Clawson $79,000 $37,100 $100,000 $46,100
17-10-227-035 E Commerce Walled Lake $140,000 $10,500 $175,500 $60,300
17-11-151-007 E Commerce Walled Lake $82,900 $35,100 $95,000 $39,200
17-14-400-041 E Commerce Walled Lake $105,000 $41,200 $137,500 $47,300
17-16-401-029 E Commerce Walled Lake $128,250 $57,200 $162,000 $72,100
17-23-326-003 E Commerce Walled Lake $68,500 $29,500 $102,900 $40,000
17-23-476-021 E Commerce Walled Lake $125,000 $66,800 $165,000 $82,300
17-24-102-015 E Commerce Walled Lake $123,000 $49,200 $145,000 $60,500
17-01-205-002 E Commerce Twp Walled Lake $45,000 $14,900 $85,000 $27,900
17-06-200-032 E Commerce Twp Huron Valley $129,000 $62,000 $160,000 $74,900
17-10-326-003 E Commerce Twp Walled Lake $118,000 $46,300 $137,000 $65,000
17-12-151-030 E Commerce Twp Walled Lake $135,000 $49,900 $160,000 $67,500
17-12-177-009 E Commerce Twp Walled Lake $87,000 $37,000 $1 10,000 $50,600
17-16-252-020 E Commerce Twp Walled Lake $112,000 $47,000 $133,000 $58,100
17-24-102-006 E Commerce Twp Walled Lake $125,500 $48,800 $156,500 $62,600
17-26-277-016 E Commerce Twp Walled Lake $89,900 $40,700 $131,000 $48,800
17-10-255—005 E Commerce Twp. Walled Lake $88,500 $25,000 $100,000 $34,000
17-16-127-023 E Commerce Twp. Walled Lake $254,913 $68,700 $385,000 $142,600
17-21-277-053 E Commerce Twp. Walled Lake $115,000 $52,000 $139,900 $55,300
17-25-101-028 E Commerce Twp. Walled Lake $58,000 $22,700 $75,000 $28,900
23-26-301-046 20 Fannington Fannington $89,500 $38,100 $1 17,000 $47,500
23-26-304-004 20 Farmington Fannington $74,500 $35,900 $105,200 $42,940
23-26-352-014 20 Fannington F armington $81,000 $34,600 $98,000 $37,030
23-26-353-020 20 Fannington Farmington $73,000 $33,400 $1 12,000 $40,150
23-27-106-025 20 Fannington Fannington $99,900 $44,100 $140,000 $48,060
23-27-328-01 1 20 Farmington Fannington $120,000 $58,300 $148,000 $66,440
23-27-351-009 20 Fannington Fannington $92,000 $47,800 $102,000 $59,950
23-28-205-016 20 F armington Farmington $149,000 $68,900 $180,000 $77,610
23-28-226-007 20 Farmington Fannington $130,000 $57,500 $161,000 $70,860
23-28-228-007 20 Farmington Fannington $91,500 $33,300 $1 15,500 $39,410
23-28-254-003 20 F armington Fannington $169,900 $72,900 $189,900 $78,450
23-28-428-037 20 Fannington Fannington $145,000 $65,000 $165,000 $72,980
23-34-352-017 20 Fannington Fannington $1 17,000 $51 .400 $142,000 $63,680
23-02-176-038 22 Fannington Hills Fannington $91,500 $38,100 $130,000 $45,510
23-03-202-022 22 Farmington Hills Farmington $176,100 $76,700 $184,500 $83,490
23-03-303-005 22 Farmington Hills Farmington $167,500 $80,100 $200,000 $84,880
23-03-402-032 22 Fannington Hills Fannington $143,500 $59,100 $190,550 $70,610
23—03-403-033 22 Fannington Hills Fannington $133,000 $62,000 $180,000 $73,780
23-04-128-004 22 Fannington Hills Fannington $198,000 $87,300 $222,000 $102,860
23-04-226-031 22 Farmington Hills Fannington $176,500 $82,300 $215,000 $88,940
23-04-406-007 22 Fannington Hills Fannington $133,000 $56,100 $169,000 $66,650
23-04-453-011 22 Fannington Hills Farmington $134,000 $56,200 $153,000 $65,260
23-06-431-012 22 Fannington Hills Walled Lake $203,000 $89,800 $231,000 $108,260
23-06-451-018 22 Fannington Hills Walled Lake $201,000 $90,400 $231,000 $11 1,160
23-06-451-026 22 Fannington Hills Walled Lake $208,500 $96,200 $238,500 $108,240
23-07-155-047 22 Fannington Hills Fannington $259,900 $120,700 $286,000 $137,270
23-07-276-007 22 Fannington Hills Fannington $315,900 $33,900 $465,000 $184,360
23-07-277-003 22 Fannington Hills Fannington $349,200 $27,500 $472,000 $190,360
23-07-351-004 22 Fannington Hills Fannington $354,900 $146,400 $385,000 $164,240
23-08-403-029 22 Fannington Hills Fannington $170,000 $75,000 $243,000 $89,160
23-08-430-017 22 Fannington Hills Fannington $143,000 $65,200 $185,000 $77,010
23-09-152-003 22 Fannington Hills Fannington $196,500 $87,800 $204,000 $102,460
23-09-227-009 22 Farmington Hills Farmington $197,500 $98,100 $261,000 $109,490
23-09-305-027 22 Fannington Hills Farmington $162,500 $77,000 $189,000 $86,410
23-09-351-021 22 Fannington Hills Fannington $150,000 $69,300 $176,000 $79,300

 

 

Table A.2 List of Dual House Sales

142

 

 

Town

Sidwell ID Town School Distr. 92 Price 92 SEV 96 Price 96 SEV

23-09-429-010 22 Fannington Hills Farmington $146,000 $60,500 $173,000 $71,520
23-10-226-021 22 Farmington Hills Fannington $138,000 $59,500 $165,000 $67,890
23—10-301-016 22 Fannington Hills Fannington $121,000 $55,900 $165,500 $68,300
23-10-329-002 22 Fannington Hills Fannington $131,000 $62,900 $172,000 $74,830
23-12-126-032 22 Fannington Hills Fannington $179,500 $81,600 $205,000 $94,710
23-13-179-002 22 Fannington Hills Fannington $87,000 $35,600 $104,500 $40,480
23-13-206-008 22 Fannington Hills Fannington $144,000 $71,800 $184,500 $85,080
23-14-378-012 22 Farmington Hills Fannington $114,000 $50,500 $135,000 $56,380
23-16-402-013 22 Fannington Hills Farmington $284,000 $85,000 $331,650 $97,120
23-23-378-013 22 Fannington Hills Fannington $94,500 $43,900 $123,000 $48,380
23-23-402-026 22 Fannington Hills Fannington $156,500 $70,800 $180,000 $88,780
23-23-428-011 22 Fannington Hills Farmington $78,000 $26,900 $80,000 $30,170
23-23-477-004 22 Fannington Hills Fannington $83,000 $30,900 $95,000 $31,910
23-24-227-006 22 Fannington Hills Fannington $149,000 $68,600 $166,000 $77,730
23-25-177-013 22 Farmington Hills Fannington $152,000 $76,000 $187,500 $95,000
23-26-252-003 22 Fannington Hills Fannington $107,500 $43,100 $139,900 $50,350
23-26-329-052 22 Fannington Hills Fannington $85,500 $32,800 $109,000 $37,680
23-26-427-009 22 Farmington Hills Farmington $98,000 $39,700 $134,900 $47,520
23-26-454-020 22 Fannington Hills Farmington $81,500 $37,800 $1 13,000 $45,630
23-33-279-050 22 Fannington Hills Fannington $85,000 $33,900 $107,500 $38,450
23-33-430-012 22 Fannington Hills Fannington $75,000 $25,500 $98,000 $27,680
23-35-230-002 22 Fannington Hills Fannington $60,000 $18,400 $79,900 $27,890
23-35-230-023 22 Fannington Hills Fannington $68,000 $24,500 $60,000 $35,850
23-36-156-012 22 Farmington Hills Farmington $70,000 $29,400 $102,500 $34,940
23-36-204-013 22 Fannington Hills Farmington $88,000 $37,700 $122,000 $43,370
25-26-152-028 24 Femdale Hazel Park $39,900 $17,300 $67,000 $20,280
25-26-152-030 24 Femdale Hazel Park $38,900 $17,200 $65,000 $19,380
25-26-351-018 24 Femdale Hazel Park $41,000 $17,800 $56,000 $19,660
25-26-352-024 24 Femdale Hazel Park $29,000 $15,300 $47,000 $16,840
25-27-202-044 24 Femdale Femdale $40,000 $13,000 $52,500 $18,130
25-27-283-016 24 Femdale Femdale $20,500 $14,400 $44,500 $17,620
25-27-328-050 24 Femdale Femdale $60,900 $19,700 $80,900 $24,830
25-27-329-019 24 Femdale Femdale $31,250 $19,600 $85,500 $25,240
25-27-404-025 24 Femdale Femdale $43 .900 $18,600 $54,900 $20,830
25-27-454-004 24 Femdale Femdale $38,300 $17,600 $92,000 $20,530
25-28-452-014 24 Femdale Femdale $62,500 $25,500 $75,000 $34,450
25-33-127-041 24 Femdale Femdale $55,900 $20,800 $102,900 $28,990
25-33-128-026 24 Femdale Femdale $51,000 $18,800 $75,000 $23,750
25-33-202-027 24 Femdale Femdale $46,500 $20,500 $88,000 $32,660
25-33-277-001 24 F emdaie Femdale $46,500 $18,900 $82,500 $22,700
25-34-108-046 24 Femdale Femdale $38,900 $17,400 $3 8,900 $23,980
25-34-132-023 24 Femdale Femdale $47,000 $20,300 $99,999 $30,130
25-34-328-008 24 Femdale Femdale $33,000 $19,500 $72,000 $27,460
25-34-353-043 24 Femdale Femdale $41,500 $20,200 $65,000 $26,930
25-35-306-063 24 Femdale Femdale $35 .000 $18,100 $35,000 $23,000
25-35-451-018 24 Femdale Hazel Park $25,800 $12,800 $36,900 $14,620
02-10-200-017 G Groveland Brandon $180,000 $82,700 $262,000 $93 .920
02-24-226-004 G Groveland Brandon $147,000 $76,900 $164,000 $77,310
02-24-201-006 G Groveland Twp Brandon $83,000 $31,600 $129,900 $41,180
25-25-129-003 28 Hazel Park Hazel Park $52,000 $23,500 $61,700 $29,650
25-25-303-016 28 Hazel Park Hazel Park $38,000 $13,400 $49,900 $16,960
25-25-330-022 28 Hazel Park Hazel Park $46,900 $18,800 $58,000 $25,310
25-25-379-010 28 Hazel Park Hazel Park $41,500 $13,800 $58,000 $19,080
25-26-204-024 28 Hazel Park Hazel Park $58,000 $19,600 $77,250 $25,290
25-26-285-018 28 Hazel Park Hazel Park $35,900 $14,000 $55,300 $16,290
25-35-283-003 28 Hazel Park Hazel Park $38,800 $13,800 $55,000 $15,790

 

 

Table A.2 List of Dual House Sales

143

 

 

Town

Sidwell iD Town School Distr. 92 Price 92 SEV 96 Price 96 SEV
25-35-427-036 28 Hazel Park Hazel Park $25,000 $13,100 $40,000 $15,790
25-35-427-038 28 Hazel Park Hazel Park $15,000 $11,000 $29,000 $15,650
25-35-428-042 28 Hazel Park Hazel Park $45,000 $19,000 $62,000 $23,630
25-35-430-005 28 Hazel Park Hazel Park $19,500 $1 1,900 $33,001 $15,860
25-35-476-038 28 Hazel Park Hazel Park $26,000 $10,100 $26,000 $12,770
25-36-132-035 28 Hazel Park Hazel Park $38,500 $14,200 $48,000 $17,680
25-36-254-014 28 Hazel Park Hazel Park $20,000 $13,400 $38,500 $14,840 .
25-36-329-018 28 Hazel Park Hazel Park $31,900 $13,600 $39,000 $16,470
25-36-335-006 28 Hazel Park Hazel Park $39,500 $15,300 $51,900 $20,380
25-36-452-025 28 Hazel Park Hazel Park $26,000 $12,600 $32,200 $14,330
25-36-456-027 28 Hazel Park Hazel Park $37,500 $16,600 $42,000 $20,100
11-02-251-013 H Highland Huron Valley $144,000 $13,500 $190,000 $82,030
1 1-12-132-039 H Highland Huron Valley $73,900 $26,900 $102,000 $36,520
11-11-304-017 H Highland Twp Huron Valley $123,000 $48,400 $141,800 $61,770
11-11-379-003 H Highland Twp Huron Valley $98,000 $43,600 $129,900 $58,620
11-12-302-045 H Highland Twp Huron Valley $71,000 $29,400 $90,000 $32,420
1 1-09-428-011 H Highland Twp. Huron Valley $53,000 $22,000 $75,000 $25,080
11-12-201-007 H Highland Twp. Huron Valley $62,000 $26,700 $89,900 $32,440
11-12-476-020 H Highland Twp. Huron Valley $79,000 $38,700 $93,425 $39,090
01-27-477-015 1 Holly Twp Holly $77,000 $44,000 $110,000 $45,600
01-32-276-010 1 Holly Twp Holly $87,500 $47,500 $89,500 $56,100
01-33-276-01 1 1H Holly Village Holly $51,900 $16,200 $51,900 $28,200
25-20-227-016 32 Huntington Woods Berkley $115,000 $42,400 $155,000 $55,120
25-20-229-012 32 Huntington Woods Berkley $82,000 $43,200 $172,500 $61,140
25-20-303-002 32 Huntington Woods Berkley $96,000 $47,300 $152,900 $57,880
25—21-106—019 32 Huntington Woods Berkley $172,000 $78,100 $215,000 $81,250
08-14-476-019 J Independence Clarkston $1 15,000 $45,900 $160,000 $72,900
08-21-178-006 J independence Clarkston $121,800 $57,200 $162,500 $60,500
08-22-351-037 J independence C larkston $122,000 $47,400 $145,000 $50,300
08-23-101-001 J independence Clarkston $1 10,000 $44,300 $123,000 $51,000
08-28-154-007 J independence Clarkston $92,900 $42,100 $1 18,500 $55,500
08-32-403-003 J independence Waterford $120,500 $53,900 $140,000 $57,000
08-34-402-021 J independence C larkston $81,500 $38,700 $1 18,500 $43,100
08-01-354-012 J independence Twp Clarkston $33,000 $21,300 $41,904 $23,000
08-12-328-043 J independence Twp C larkston $77,900 $26,300 $109,900 $37,300
08-17-230-008 J independence Twp Clarkston $315,000 $127,100 $330,000 $148,300
08-18-178-003 J independence Twp Clarkston $128,900 $73,100 $168,000 $66,700
08-22-351-039 J independence Twp Clarkston $1 12,000 $47,400 $133,000 $55,200
08-26-301-012 J independence Twp Clarkston $75,000 $27,100 $102,900 $35,500
08-26-353-006 J independence Twp Clarkston $72,900 $24,500 $92,500 $32,500
08-28-153-012 J independence Twp Clarkston $1 12,000 $37,700 $1 19,700 $60,600
08-31-201-006 J independence Twp Clarkston $59,700 $30,200 $90,000 $32,800
08-34-252-003 J independence Twp C larkston $1 15,500 $52,500 $140,900 $57,500
08-34-327-01 1 J independence Twp Clarkston $75,900 $31,300 $99,000 $35,900
08-34-329-004 J independence Twp Clarkston $60,000 $28,200 $75,000 $33,300
08-34-403-033 J independence Twp Clarkston $59,000 $28,400 $93,000 $31,300
09-02-457-008 OL Lake Orion Lake Orion $33,500 $12,500 $38,500 $14,200
24-13-106-007 40 Lathrup Village Southfield $1 12,000 $44,600 $160,000 $54,050
24-14-276-012 40 Lathrup Village Southfield $112,500 $44,800 $139,000 $56,400
24-23-280-042 40 Lathrup Village Southfield $94,500 $43,500 $137,000 $55,460
25-14-255-009 4O Lathrup Village Southfield $135,900 $63,300 $189,890 $83,510
21—03-276—009 K Lyon South Lyon $136,000 $66,500 $185,000 $83,830
21-05-300-063 K Lyon Twp South Lyon $1 11,000 $49,100 $154,000 $59,240
25-12-453-013 44 Madison Heights Lamphere $60,000 $33,900 $90,000 $38,300
25-13-304-024 44 Madison Heights Madison $79,500 $33,900 $101,000 $43,320
25-13-354-024 44 Madison Heights Madison $48,000 $18,100 $59,700 $22,290

I -""'O- ’

 

 

l.

 

Table A.2 List of Dual House Sales

144

 

 

Town

Sidwell lD Town School Distr. 92 Price 92 SEV 96 Price 96 SEV

25-13-359-004 44 Madison Heights Madison $31,000 $18,600 $56,000 $23,890
25-14-230-034 44 Madison Heights Lamphere $78,500 $36,700 $95,500 $42,020
25-14-385-017 44 Madison Heights Royal Oak $36,000 $17,300 $61,000 $19,750
25-23-205-026 44 Madison Heights Madison $55,000 $19,500 $73,000 $26,350
25-23-429-009 44 Madison Heights Madison $45,000 $21,900 $63,000 $25,130
25-24-304-011 44 Madison Heights Madison $37,500 $25,100 $48,000 $18,650
25-11-232-028 44 Madison Hgts Lamphere $43,000 $18,600 $43,000 $24,440
25-12-177-014 44 Madison Hgts Lamphere $80,000 $40,300 $123,500 $48,620
25-12-426-008 44 Madison Hgts Lamphere $60,000 $24,700 $88,000 $30,880
25-12-430-030 44 Madison Hgts Lamphere $83,000 $35,800 $104,500 $39,540
25-13-177-015 44 Madison Hgts Lamphere $75,000 $31,700 $98,000 $39,540
25-13-254-006 44 Madison Hgts Lamphere $84,500 $32,700 $105,000 $36,960
25-13-279-027 44 Madison Hgts Madison $84,000 $35,200 $96,000 $39,420
25-13-303-035 44 Madison Hgts Madison $74,000 $36,300 $96,500 $38,850
25-13-451-029 44 Madison Hgts Madison $27,000 $14,900 $55,000 $20,800
25-23-253-005 44 Madison Hgts Madison $37,000 $17,000 $61,000 $20,040
25-23-427-029 44 Madison Hgts Madison $59,000 $23,000 $67,000 $29,790
25-24-128-031 44 Madison Hgts Madison $57,900 $21,900 $74,900 $26,640
25-24-132-014 44 Madison Hgts Madison $61,500 $19,600 $79,300 $25,340
25-24-456-018 44 Madison Hgts Madison $51,500 $19,600 $52,000 $23,410
16-01-100-039 L Milford Twp Huron Valley $142,000 $88,100 $193,010 $109,060
16-02-376-020 LM Milford Village Huron Valley $72,000 $33,100 $106,900 $40,500
16-02-377-039 LM Milford Village Huron Valley $76,301 $31,200 $95,000 $43,230
16-10-429-003 LM Milford Village Huron Valley $78,000 $27,400 $95,000 $34,870
16-10-476-019 LM Milford Village Huron Valley $73,450 $29,900 $124,000 $35,120
16-11-178—011 LM Milford Village Huron Valley $131,000 $38,600 $175,000 $50,430
16-14-201-026 LM Milford Village Huron Valley $75,100 $36,700 $99,000 $43,170
22-33-402-016 48 Northville Northville $320,000 $130,300 $330,000 $154,550
22-21-427-034 50 Novi Novi $194,000 $72,000 $260,000 $87,250
22-21-451-038 50 Novi Novi $147,000 $60,800 $173,000 $74,450
22-22-203-023 50 Novi Novi $62,700 $41,800 $126,000 $52,150
22-23-454-008 50 Novi Novi $133,750 $56,900 $154,900 $66,350
22-24-3 76-005 50 Novi Novi $1 10,000 $50,600 $145,000 $57,350
22-25-105-009 50 Novi Novi $108,500 $42,200 $127,000 $52,600
22-25-203-002 50 Novi Novi $129,000 $61,250 $158,900 $71,400
22-26-226-002 50 Novi Novi $1 13 .500 $46,200 $127,900 $52,800
22-27-201-012 50 Novi Novi $193,500 $88,050 $254,000 $97,150
22-27-303 -001 50 Novi Northville $195,500 $94,400 $227,000 $1 12,350
22-34-154-022 50 Novi Northville $158,000 $63,600 $164,000 $75,350
22-34-176-001 50 Novi Northville $145,000 $61,700 $180,000 $70,000
22-34-176-023 50 Novi Northville $185,000 $82,100 $225,000 $95,900
22-36-127-01 1 50 Novi Novi $144,000 $62 .000 $177,900 $72,750
25-19-178-026 52 Oak Park Berkley $94,000 $37,600 $140,000 $50,500
25-19-205-001 52 Oak Park Berkley $86,500 $36,300 $98,600 $43,500
25-19-327-011 52 Oak Park Berkley $82,000 $39,300 $131,500 $51,000
25-19-455-010 52 Oak Park Berkley $76,500 $32,400 $89,900 $50,200
25-28-151-019 52 Oak Park Femdale $67,000 $24,000 $93,000 $32,400
25-29-326-003 52 Oak Park Oak Park $33,500 $20,200 $40,000 $22,700
25-29-406-001 52 Oak Park Femdale $35,000 $16,800 $68,500 $22,300
25-29-430-028 52 Oak Park Femdale $48,000 $20,000 $67,000 $25,700
25-29-453-007 52 Oak Park Femdale $46,000 $15,900 $71,000 $20,500
25-30-208-015 52 Oak Park Oak Park $37,900 $13,200 $56,099 $17,600
25-30-332-037 52 Oak Park Oak Park $62,500 $29,300 $62,500 $36,800
25-30-335-011 52 Oak Park Oak Park $35,000 $21,000 $79,900 $28,500
25-31-201-015 52 Oak Park Oak Park $57,900 $18,300 $74,900 $27,300
25-31-203-014 52 Oak Park Oak Park $45,500 $20,500 $79,000 $28,900

 

Table A.2 List of Dual House Sales

145

 

 

Town

Sidwell ID Town School Distr. 92 Price 92 SEV 96 Price 96 SEV

25-31-276-019 52 Oak Park Oak Park $40,000 $22,000 $85,000 $31,100
25-31-476-014 52 Oak Park Oak Park $52,900 $18,100 $66,000 $26,500
25-32-203-032 52 Oak Park Oak Park $30,500 $12,100 $43,000 $14,500
10-02-158-014 N Oakland Twp Romeo $45,000 $20,700 $81,200 $23,580
10-02-159-020 N Oakland Twp Romeo $116,000 $47,100 $135,000 $56,870
10-24-101-006 N Oakland Twp Rochester $1 19,000 $50,100 $158,500 $55,130
10-27-177-007 N Oakland Twp Rochester $247,888 $82,300 $324,900 $130,620
10-34-128-01 i N Oakland Twp Rochester $375,000 $153,600 $429,000 $176,030
10-34-228-014 N Oakland Twp Rochester $96,000 $48,700 $146,000 $52,890
09-01 -477-047 0 Orion Lake Orion $65,000 $29,400 $92,000 $34,100
09-10-429-032 0 Orion Lake Orion $60,000 $24,300 $118,500 $30,600
09-10-429-039 0 Orion Lake Orion $63,000 $23,300 $107,900 $25,800
09-11-316-018 0 Orion Lake Orion $80,500 $32,600 $166,000 $43,900
09-21-3583-01 1 0 Orion Lake Orion $143,000 $65,000 $174,000 $71,800
09-30-377-007 0 Orion Lake Orion $104,000 $42,400 $128,900 $51,900
09-06-201-005 0 Orion Twp Lake Orion $105,000 $52,900 $151,000 $55,000
09-16-276—006 0 Orion Twp Lake Orion $84,950 $11,100 $107,500 $44,800
09-21-352-006 0 Orion Twp Lake Orion $130,000 $56,400 $145,000 $60,600
09-21-358-026 0 Orion Twp Lake Orion $160,000 $67,000 $193,500 $79,300
09-26-403-030 0 Orion Twp Lake Orion $160,000 $96,000 $175,000 $97,800
09-26-429-013 0 Orion Twp Lake Orion $145,900 $64,900 $164,000 $66,800
09-28-376-001 0 Orion Twp Lake Orion $68,500 $23,400 $92,900 $26,400
09-29-255-022 0 Orion Twp Lake Orion $96,900 $41,000 $123,000 $44,800
04-28-204-016 P Oxford Oxford $70,000 $44,000 $129,000 $50,530
04-05-276-034 P Oxford Twp Oxford $130,000 $64,300 $149,900 $74,010
04-28-355-034 P Oxford Twp Oxford $61,900 $35,300 $89,900 $42,630
04-30-301-002 P Oxford Twp Oxford $111,000 $42,900 $129,000 $53,400
04-22-378-006 PO Oxford Village Oxford $68,000 $36,100 $100,000 $40,980
04-26-227-015 PO Oxford Village Oxford $140,000 $76,600 $167,500 $76,000
25-28-251-004 60 Pleasant Ridge Femdale $90,000 $41,400 $191,500 $53,370
14-07-453-004 64 Pontiac Pontiac $69,800 $28,100 $94,500 $36,020
14-16-403-012 64 Pontiac Pontiac $36,000 $19,500 $63,000 $23,940
14-17-129-003 64 Pontiac Pontiac $48,900 $18,200 $55,650 $21,110
14-17-351-010 64 Pontiac Pontiac $32,000 $15,900 $42,500 $20,440
14-17-408-006 64 Pontiac Pontiac $16,150 $1 1,000 $19,000 $12,790
14-19-208-009 64 Pontiac Pontiac $30,800 $15,100 $30,800 $17,970
14-19-429-007 64 Pontiac Pontiac $28,500 $7,100 $35,000 $8,420
14-20-331-009 64 Pontiac Pontiac $20,000 $14,200 $32,500 $14,750
14-21-207-009 64 Pontiac Pontiac $35,000 $10,300 $48,000 $13,250
14-21-251-014 64 Pontiac Pontiac $24,000 $1 1,600 $40,000 $14,690
14-30-328-024 64 Pontiac Pontiac $65,000 $22,300 $85,000 $28,310
14-31-208-006 64 Pontiac Pontiac $70,000 $26,000 $84,500 $33 .080
14-31-227-005 64 Pontiac Pontiac $31,500 $21,800 $52,000 $29,850
15-1 1 -1 60-004 68 Rochester Rochester $109,000 $48,100 $134,000 $65,150
1 5-1 1-377-007 68 Rochester Rochester $75,000 $32,700 $86,500 $33,740
15-15-128-010 68 Rochester Rochester $91,000 $30,100 $1 1 1,500 $39,480
15-15-129-014 68 Rochester Rochester $85,500 $41,600 $123,000 $52,400
15-15-253-007 68 Rochester Rochester $103,500 $29,200 $130,000 $39,120
15-15-276-004 68 Rochester Rochester $56,000 $23,900 $78,000 $29,720
15-02-301-003 70 Rochester Hills Rochester $288,000 $127,200 $332,000 $152,020
15-03-127-003 70 Rochester Hills Rochester $147,500 $58,400 $168,500 $68,430
15-04-329-014 70 Rochester Hills Rochester $146,000 $67,200 $184,900 $79,560
15-05-126-016 70 Rochester Hills Rochester $205,000 $104,600 $265,000 $114,800
15-05-202-016 70 Rochester Hills Rochester $209,000 $93,400 $272,000 $1 14,060
15—05-202-01 7 70 Rochester Hills Rochester $225,900 $101,700 $269,888 $115,310
15-05-203-038 70 Rochester Hills Rochester $230,650 $20,600 $278,000 $120,450

 

Table A.2 List of Dual House Sales

146

 

 

Town

Sidwell lD Town School Distr. 92 Price 92 SEV 96 Price 96 SEV

15-05-204-023 70 Rochester Hills Rochester $236,010 $20,600 $270,000 $124,320
15-05-232-008 70 Rochester Hills Rochester $184,450 $83,800 $223,500 $91,890
15-05-482-007 70 Rochester Hills Rochester $190,000 $94,900 $245,000 $109,990
15-06-151-044 70 Rochester Hills Rochester $420,000 $31,000 $450,500 $224,840
15-06-151-046 70 Rochester Hills Rochester $350,000 $72,700 $429,000 $184,080
15-06-152-002 70 Rochester Hills Rochester $370,000 $11,900 $402,500 $187,660
15-06-152-013 70 Rochester Hills Rochester $335,000 $69,700 ' $405,000 $178,090
15-06-152-019 70 Rochester Hills Rochester $350,000 $33,500 $388,000 $167,790
15-06-179-011 70 Rochester Hills Rochester $304,000 $150,400 $337,500 $139,100
15-06-207-002 70 Rochester Hills Rochester $193,500 $96,600 $249,000 $105,900
15-06-254-001 70 Rochester Hills Rochester $216,000 $99,200 $254,000 $106,470
15-06-254-015 70 Rochester Hills Rochester $195,062 $95,900 $263,500 $103,940
15-06-301-003 70 Rochester Hills Rochester $244,000 $98,800 $265,000 $111,260
15-06-354-027 70 Rochester Hills Rochester $364,000 $89,800 $399,999 $175,690
15-07-376—01 8 70 Rochester Hills Rochester $209,000 $87,900 $235,000 $101,160
15-07-377-036 70 Rochester Hills Rochester $210,000 $83,000 $232,900 $97,640
15-08-329-012 70 Rochester Hills Rochester $168,000 $77,500 $205,000 $90,000
15-11-102-001 70 Rochester Hills Rochester $88,000 $22,800 $122,000 $38,220
15-14-326—012 70 Rochester Hills Rochester $152,900 $74,700 $185,888 $81,900
15-14-327-004 70 Rochester Hills Rochester $149,000 $61,300 $172,500 $66,980
15-14-352-001 70 Rochester Hills Rochester $111,500 $49,600 $147,000 $56,780
15-15-353—038 70 Rochester Hills Rochester $210,750 $98,500 $232,900 $101,480
15-16-303-037 70 Rochester Hills Rochester $112,000 $55,400 $136,000 $59,480
15-16-327-007 70 Rochester Hills Rochester $153,000 $73,900 $209,900 $82,830
15-17-128-021 70 Rochester Hills Rochester $125,000 $57,600 $174,600 $64,160
15-17-151-002 70 Rochester Hills Rochester $170,000 $84,700 $189,000 $94,830
15-17-452-004 70 Rochester Hills Rochester $137,900 $56,600 $153,900 $66,170
15-19-401-007 70 Rochester Hills Rochester $285,000 $136,000 $340,000 $145,190
15-22-329-009 70 Rochester Hills Rochester $115,000 $50,100 $140,000 $57,490
15-22-402-001 70 Rochester Hills Rochester $117,000 $52,100 $140,500 $57.1 10
15-22-427-009 70 Rochester Hills Rochester $113,500 $52,700 $139,000 $55,960
15-23-252-012 70 Rochester Hills Rochester $147,500 $19,500 $197,000 $86,000
15-23-254-003 70 Rochester Hills Rochester $179,500 $89,400 $233,000 $103,010
15-23-304-007 70 Rochester Hills Rochester $159,583 $12,800 $208,000 $89,950
15-25-252-002 70 Rochester Hills Rochester $170,000 $25,500 $235,000 $87,550
15-26-276-007 70 Rochester Hills Rochester $127,000 $53,600 $150,900 $62,300
15-28-402-095 70 Rochester Hills Avondale $88,000 $38,500 $115,000 $43,570
15-35-252-021 70 Rochester Hills Rochester $180,000 $72,100 $195,000 $79,280
15-35-352-060 70 Rochester Hills Avondale $112,000 $51,200 $138,000 $68,220
15-35-377-047 70 Rochester Hills Avondale $82,000 $34,400 $75,000 $43,480
15-35-476-007 70 Rochester Hills Rochester $72,900 $33,900 $106,500 $42,120
15-36-256-022 70 Rochester Hills Rochester $89,900 $7,300 $116,000 $50,170
06-16-427-003 R Rose Holly $96,000 $50,000 $128,000 $51,900
25-03-253-022 72 Royal Oak Royal Oak $78,500 $25,100 $125,000 $37,320
25-03-326-020 72 Royal Oak Royal Oak $77,900 $28,800 $114,000 $39,860
25-03-352-017 72 Royal Oak Royal Oak $81,000 $37,300 $1 15,900 $45,880
25-03-405-034 72 Royal Oak Royal Oak $50,000 $25,500 $93,000 $31,190
25-03-477-042 72 Royal Oak Royal Oak $82,000 $36,000 $1 19,500 $43,600
25-04-402-022 72 Royal Oak Royal Oak $82,000 $38,600 $1 27,000 $45,190
25-04-402-023 72 Royal Oak Royal Oak $93,000 $37,200 $127,000 $48,880
25-04-404-006 72 Royal Oak Royal Oak $96,000 $37,200 $146,000 $46,980
25-05-453-012 72 Royal Oak Royal Oak $60,000 $27,500 $86,300 $34,240
25-05-454-016 72 Royal Oak Royal Oak $63,500 $24,900 $79,700 $29,950
25-06-102-005 72 Royal Oak Royal Oak $86,500 $40,900 $122,000 $53,890
25-06-204-003 72 Royal Oak Royal Oak $92,000 $43,500 $144,500 $56,810
25-06-206-008 72 Royal Oak Royal Oak $79,000 $32,500 $127,750 $41 .240

 

Table A.2 List of Dual House Sales

147

 

 

Town

Sidwell iD Town School Distr. 92 Price 92 SEV 96 Price 96 SEV

25-06-230-022 72 Royal Oak Royal Oak $85,000 $35,400 $107,000 $45,590
25-06-254-015 72 Royal Oak Royal Oak $106,700 $45,900 $145,500 $58,080
25-06-327-024 72 Royal Oak Royal Oak $1 12,700 $61,300 $159,900 $65,100
25-06-431-031 72 Royal Oak Royal Oak $76,900 $32,500 $108,000 $43,050
25-06-433-013 72 Royal Oak Royal Oak $78,000 $28,500 $1 12,000 $39,190
25-07-104-009 72 Royal Oak Royal Oak $74,000 $31,700 $95,000 $32,940
25-08-181-021 72 Royal Oak Royal Oak $106,000 $31,700 $125,000 $38,010
25-08-401-01 1 72 Royal Oak Royal Oak $139,000 $51,000 $192,000 $71,080
25-08-427-004 72 Royal Oak Royal Oak $105,000 $47,800 $158,000 $62,560
25-08-433-013 72 Royal Oak Royal Oak $101,000 $50,900 $144,000 $61,790
25-08-484-023 72 Royal Oak Royal Oak $81,000 $39,000 $126,500 $49,720
25-09-352-003 72 Royal Oak Royal Oak $129,900 $39,700 $180,000 $57,400
25-09-408-015 72 Royal Oak Royal Oak $85,900 $38,700 $132,000 $48,830
25-09-478-009 72 Royal Oak Royal Oak $105,000 $31,600 $145,000 $45,540
25-10-103-010 72 Royal Oak Royal Oak $84,000 $36,200 $129,000 $42,140
25-10-105-002 72 Royal Oak Royal Oak $66,000 $29,500 $100,000 $38,970
25-10-129-030 72 Royal Oak Royal Oak $50,000 $20,400 $1 15,000 $33,340
25-10-129-081 72 Royal Oak Royal Oak $76,900 $30,400 $124,900 $42,580
25-10-208-019 72 Royal Oak Royal Oak $89,900 $40,800 $126,500 $44,710
25-10-477-037 72 Royal Oak Royal Oak $68,000 $26,100 $1 10,000 $35,240
25-14-377-013 72 Royal Oak Royal Oak $73,000 $31,800 $105,000 $38,850
25-15-103-022 72 Royal Oak Royal Oak $67,000 $28,100 $90,000 $36,900
25-15-103-041 72 Royal Oak Royal Oak $77,500 $27,300 $1 19,900 $34,020
25-15-126-039 72 Royal Oak Royal Oak $71,500 $24,800 $92 .000 $30,600
25-15-205-021 72 Royal Oak Royal Oak $76,000 $32,900 $1 13,000 $47,300
25-15-209-007 72 Royal Oak Royal Oak $81,000 $36,900 $122,450 $44,410
25-15-209-022 72 Royal Oak Royal Oak $76,000 $32,700 $1 19,900 $40,260
25-15-303-007 72 Royal Oak Royal Oak $45,000 $28,400 $109,900 $35,410
25-15-327-003 72 Royal Oak Royal Oak $70,000 $26,100 $1 10,500 $32,560
25-15-376-014 72 Royal Oak Royal Oak $36,000 $27,300 $95,500 $35,460
25-15-376-017 72 Royal Oak Royal Oak $66,500 $21,800 $99,000 $36,790
25-15-453-010 72 Royal Oak Royal Oak $83,000 $26,900 $1 1 1,000 $37,130
25-15-476-007 72 Royal Oak Royal Oak $86,000 $27,900 $11 1,900 $39,610
25-15-476-021 72 Royal Oak Royal Oak $90,500 $33,900 $123,000 $46,920
25-16-252-015 72 Royal Oak Royal Oak $105,900 $39,000 $138,000 $50,320
25-16-379-030 72 Royal Oak Royal Oak $72,000 $32,200 $133,320 $47,960
25-21-253-01 2 72 Royal Oak Royal Oak $69,000 $27,500 $1 19,500 $45,840
25-22-182-017 72 Royal Oak Royal Oak $79,200 $31,400 $108,000 $46,230
25-22-227-002 72 Royal Oak Royal Oak $62,100 $26,000 $138,000 $32,290
25-22-458-018 72 Royal Oak Royal Oak $56,000 $22,500 $69,000 $32,950
25-23-159-035 72 Royal Oak Royal Oak $65,500 $32,000 $83,500 $39,380
25-23-177-002 72 Royal Oak Royal Oak $74,000 $26,900 $103,500 $40,740
25-23-304-026 72 Royal Oak Royal Oak $77,500 $28,100 $105,000 $37,250
25-23-309-021 72 Royal Oak Royal Oak $72,500 $27,300 $105,000 $34,770
21-19-451—010 80 South Lyon South Lyon $73,000 $33,500 $99,900 $39,840
21-20-452-001 80 South Lyon South Lyon $113,000 $51,600 $138,000 $63,440
24-10-404-004 76 Southfield Southfield $240,000 $1 13.450 $270,000 $135,570
24-1 1-204-003 76 Southfield Birmingham $1 18,900 $44,750 $140,000 $53,560
24-11-303-014 76 Southfield Birmingham $1 1 1,100 $41,700 $131,500 $49,460
24-1 1-304—01 9 76 Southfield Southfield $86,500 $38,750 $1 14,000 $46,820
24-1 1-330-005 76 Southfield Southfield $67,000 $33,900 $99,000 $39,650
24-1 1-332-022 76 Southfield Southfield $79,000 $37,050 $123,700 $49,760
24-11-333-017 76 Southfield Southfieid $84,000 $33,100 $107,000 $47,210
24-1 1-352-040 76 Southfield Southfieid $1 18,000 $47,200 $148,000 $59,920
24-1 1-477-021 76 Southfield Southfield $55,000 $19,850 $87,500 $24,090
24-1 1-479-003 76 Southfield Southfieid $77,000 $33,700 $109,000 $44,630

 

Table A.2 List of Dual House Sales

148

 

 

Town

Sidwell iD Town School Distr. 92 Price 92 SEV 96 Price 96 SEV

24-12-233-004 76 Southfield Birmingham $79,000 $37,050 $1 1 1,000 $50,000
24-12-431-006 76 Southfield South field $52,500 $20,950 $72,000 $25,840
24-12-478-013 76 Southfield Southfield $76,000 $32,550 $96,500 $39,870
24-13-130-044 76 Southfield Southfieid $63,500 $25,500 $75,000 $31,660
24-13-177-012 76 Southfield Southfield $68,000 $27,500 $82,500 $32,290
24-13-228-017 76 Southfield Southfieid $59,000 $26,300 $69,000 $33,130
24-13-231-004 76 Southfield Southfieid $53,000 $20,400 $67,800 $25,920
24-13-377-008 76 South field Southfield $48,000 $27,000 $74,000 $31,200
24-15-178-016 76 Southfield Southfield $132,000 $51,050 $125,000 $63,780
24-15-227-031 76 Southfield Southfield $97,000 $42,450 $132,800 $51,610
24-15-351-030 76 Southfield Southfield $55,000 $21,900 $72,500 $28,940
24-15-376-015 76 Southfield Southfield $105,000 $42,350 $135,000 $54,870
24-19-351-018 76 Southfield Southfield $43,500 $22,450 $76,900 $31,880
24-21-327-017 76 Southfield Southfield $105,000 $52,750 $182,000 $66,950
24-25-255-022 76 Southfield Southfield $85,000 $38,000 $1 13,000 $51,170
24-25-278-008 76 South field Southfieid $70,000 $34,600 $124,000 $44,780
24-25-302-028 76 Southfield Southfield $73 .000 $29,650 $122,500 $39,070
24-26-276-018 76 Southfield Southfield $78,000 $41,500 $78,000 $51,120
24-28-152-001 76 Southfield Southfield $1 1 3,000 $42,500 $164,500 $57,740
24-28-427-015 76 Southfield Southfieid $1 10,000 $44,500 $145,000 $56,430
24-29-102-01 l 76 Southfield Southfield $87,000 $46,850 $1 14,000 $54,330
24-29-352-018 76 Southfield Southfield $86,000 $45,300 $122,000 $53,920
24-31-307-023 76 Southfield Southfield $81,000 $45,000 $1 15,000 $51,650
24-32-202-017 76 Southfield Southfield $40,000 $19,100 $69,700 $21,640
24-34-151-026 76 Southfield Southfield $60,000 $28,300 $79,000 $34,710
24-34-152-007 76 Southfield Southfield $84,000 $44,100 $1 12,000 $52,560
24-34-153-003 76 Southfield Southfield $74,000 $24,200 $1 10,500 $31,870
24-34-454-021 76 Southfield Southfield $40,000 $16,600 $57,900 $19,310
24-35-106-003 76 Southfield Southfield $133,000 $59,250 $1 15,900 $41,100
24-35-129-015 76 Southfield Southfield $1 15,000 $50,900 $160,000 $63,230
24-35-153-060 76 Southfield Southfield $53,000 $28,350 $85,000 $30,640
24-36-326-015 76 Southfield Southfield $88,000 $28,900 $1 12,500 $40,130
07-08-476-002 U Springfield Holly $79,900 $35,300 $1 15,250 $48,600
07-23-101-005 U Springfield C larkston $1 19,900 $46,200 $140,000 $57,400
07-25-326-048 U Springfield Clarkston $1 19,000 $32,700 $170,500 $66,300
07-26-301—004 U Springfield Clarkston $79,000 $36,300 $1 15,000 $42,100
20-02-301-036 88 Troy Troy $171,750 $19,600 $204,000 $80,910
20-03-177-017 88 Troy Avondale $1 14,500 $48,000 $127,750 $57,060
20-05-252-01 1 88 Troy Avondale $179,000 $86,600 $245,500 $101,530
20-05-453-005 88 Troy Troy $167,000 $84,500 $224,900 $93,450
20-06-351-007 88 Troy Bloom. Hills $157,000 $58,700 $185,000 $74,640
20—06-401 -019 88 Troy Troy $210,000 $129,400 $315,000 $128,830
20-08-128-007 88 Troy Troy $122,000 $55,400 $144,000 $62,640
20-08-153-008 88 Troy Troy $156,000 $73,800 $199,800 $89,700
20-1 1 -104-020 88 Troy Troy $200,000 $19,000 $265,000 $122,140
20-11-178-012 88 Troy Troy $132,000 $59,300 $165,000 $68,960
20-12-180-003 88 Troy Troy $155,000 $67,300 $185,000 $76,630
20-12-376-017 88 Troy Troy $180,000 $78,700 $2 30,000 $98,740
20-12-429-009 88 Troy Troy $130,500 $59,800 $165,000 $71,090
20-13-204-002 88 Troy Troy $103,000 $47,700 $134,000 $55,820
20-13-207-022 88 Troy Troy $103,000 $48,800 $130,000 $56,940
20-13-227—012 88 Troy Troy $1 12,000 $51,900 $145,000 $57,590
20-14-203-010 88 Troy Troy $135,000 $57,500 $1 35 .000 $66,310
20-17-104-002 88 Troy Troy $276,000 $44,700 $382,000 $154,280
20-17-431-008 88 Troy Troy $232,000 $100,300 $247,500 $1 15,630
20-17-476-046 88 Troy Troy $261,500 $5,225 $325,500 $138,600

 

 

Table A.2 List of Dual House Sales

149

 

 

Town

Sidwell iD Town School Distr. 92 Price 92 SEV 96 Price 96 SEV

20-18-152-005 88 Troy Troy $131,600 $68,100 $165,000 $73,500
20-19-152—009 88 Troy Birmingham $127,500 $62,300 $169,900 $75,120
20-20-302-002 88 Troy Troy $303,333 $20,000 $392,500 $162,360
20-20-327-0 1 4 88 Troy Troy $308,184 $20,000 $400,750 $159,850
20-21-102-001 88 Troy Troy $58,600 $34,300 $66,400 $42,590
20-21-231-009 88 Troy Troy $141,000 $19,000 $165,000 $71,090
20-23-153-017 88 Troy Troy $126,500 $59,500 $165,200 $66,880
20-23-203-031 88 Troy Troy $132,000 $62,600 $165,500 $75,820
20-23-251-015 88 Troy Troy $128,000 $56,000 $171,000 $63,830
20-23-276-004 88 Troy Troy $130,000 $58,600 $169,800 $68,660
20-23-406-015 88 Troy Troy $132,000 $61,600 $158,900 $68,410
20-23-427-004 88 Troy Troy $130,000 $62,200 $185,000 $71,620
20-24-102-004 88 Troy Troy $2 20,000 $19,000 $262,000 $1 13,500
20-24-427-006 88 Troy Troy $1 18,000 $50,500 $174,500 $55,910
20-24-479-012 88 Troy Troy $103,000 $45,500 $136,500 $54,480
20-25-178-001 88 Troy Warren $85,000 $42,600 $1 15,000 $47,230
20-25-2 54-004 88 Troy Warren $1 12,000 $51,300 $134,000 $58,770
20-25-256-003 88 Troy Warren $96,500 $45,600 $1 19,000 $52,760
20-25-280-022 88 Troy Warren $100,000 $79,000 $125,000 $59,090
20-25-306-018 88 Troy Warren $123,500 $13,000 $162,997 $63,460
20-25-306-035 88 Troy Warren $123,180 $13,000 $165,000 $65,370
20-27-403-023 88 Troy Troy $101,000 $45,300 $134,250 $53,650
20-30-153-002 88 Troy Birmingham $1 18,000 $51,100 $142,000 $59,780
20-35-301-003 88 Troy Royal Oak $96,000 $39,600 $1 19,500 $48,510
01-33-430-016 1H Village of Holly Holly $48,500 $20,500 $64,900 $24,000
17-26-178-006 92 Walled Lake Walled Lake $82,000 $38,500 $1 18,500 $43,920
17-26-306-022 92 Walled Lake Walled Lake $68,000 $24,800 $94,000 $32,870
17-35-154-019 92 Walled Lake Walled Lake $79,000 $38,900 $116,000 $45,540
17-35-180-024 92 Walled Lake Walled Lake $137,000 $55,700 $165,000 $68,950
13-01-151-008 W Waterford Waterford $1 12,290 $1 .000 $144,000 $66,320
13-01-251-019 W Waterford Waterford $139,000 $1 1,100 $168,900 $85,380
13-01-252—015 W Waterford Waterford $1 12,900 $1 1,100 $149,000 $66,120
13-01-304-001 W Waterford Waterford $106,000 $45,800 $122,000 $59,780
13-01-326-001 W Waterford Waterford $105,000 $54,300 $126,000 $63,000
13-01-328-005 W Waterford Waterford $1 10,000 $52,300 $129,000 $60,270
13-03-102-027 W Waterford Waterford $107,000 $41,900 $126,000 $44,020
13-03-427-036 W Waterford Waterford $77,000 $42,000 $94,000 $48,340
13-04-154-009 W Waterford Waterford $95,000 $34,500 $122,000 $42,760
13-04-428-012 W Waterford Waterford $68,000 $29,100 $102,900 $37,490
13-04-477-026 W Waterford Waterford $39,500 $16,100 $43,000 $16,980
13-05-130-01 8 W Waterford Waterford $58,900 $24,200 $73,500 $29,090
13-05-203-039 W Waterford Waterford $96,000 $48,300 $125,000 $55,690
13-06-176-007 W Waterford Clarkston $81,900 $45,000 $1 14,3 50 $59,510
13-07-478-027 W Waterford Waterford $58,000 $31,800 $71,500 $39,200
13-08-428-018 W Waterford Waterford $95,500 $49,100 $139,500 $56,610
13-08-453-006 W Waterford Waterford $94,000 $23,000 $155,000 $32,730
13-08-480-006 W Waterford Waterford $94,000 $48,300 $128,600 $57,170
13-09-104-018 W Waterford Waterford $74,900 $36,300 $1 17,000 $49,660
13-09-1 81-010 W Waterford Waterford $66,200 $23,900 $92,000 $29,990
13-10-252-01 1 W Waterford Waterford $65,500 $23,100 $81,000 $26,150
13-10-277—005 W Waterford Waterford $67,000 $27,600 $93,000 $36,420
13-1 1-127-017 W Waterford Waterford $79,900 $28,700 $92,900 $32,920
13-14-128-034 W Waterford Waterford $85,000 $34,600 $105,000 $43,440
13-15-351-001 W Waterford Waterford $95,200 $42,500 $127,000 $47,670
13-15-402-001 W Waterford Waterford $74,900 $34,600 $98,500 $47,800
13-19-378-01 5 W Waterford Waterford $143 .000 $9,400 $163,760 $75,860

 

 

 

150
Table 141 List of Dual House Sales
Town

Sidwell lD Town School Distr. 92 Price 92 SEV 96 Price 96 SEV
13-20-127-023 W Waterford Waterford $87,500 $29,400 $108,000 $42,660
13-20-251-026 W Waterford Waterford $78,000 $14,300 $106,000 $47.3 80
13-21-103-005 W Waterford Waterford $94,000 $36,500 $1 15,000 $49,490
13-21-280-018 W Waterford Waterford $67,000 $27,700 $83,800 $35,380
13-21-452-002 W Waterford Waterford $83,900 $30,100 $1 12,900 $40,540
13-21-477-045 W Waterford Waterford $62,900 $23,600 $92,000 $33,970
13-22-477-002 W Waterford Waterford $49,000 $24,700 $69,000 $27,770
13-22-480-012 W Waterford Waterford $139,000 $49,100 $172,000 $69,260
13-23-178-043 W Waterford Waterford $89,000 $3 9,900 $131,500 $50,430
13-23-378-006 W Waterford Waterford $74,750 $33 .400 $105,000 $40,450
13-25-151-016 W Waterford Waterford $88,000 $37,500 $100,000 $42,700
13-25-403-003 W Waterford Waterford $60,500 $26,000 $79,000 $34,360
13-25-406-024 W Waterford Waterford $26,000 $18,700 $71,400 $26,360
13-26-108-009 W Waterford Waterford $79,900 $3,200 $105,000 $39,550
13-26-227-008 W Waterford Waterford $81,000 $31 .700 $122,000 $39,390
13-27-201-026 W Waterford Waterford $106,000 $43,200 $136,500 $57,840
13-28-104-01 8 W Waterford Waterford $64,000 $21,600 $89,000 $30,040
13-28-107-01 2 W Waterford Waterford $77,900 $35,000 $106,000 $40,660
13-28-128-01 l W Waterford Waterford $57,000 $20,000 $79,500 $27,900
13-28-152-018 W Waterford Waterford $65,500 $24,200 $87,500 $32,800
13-30-476-009 W Waterford Waterford $62,000 $33,600 $96,000 $40,160
13-32-200-046 W Waterford Waterford $1 10,000 $74,100 $154,000 $71,530
13-32-477-001 W Waterford Waterford $190,650 $12,500 $249,900 $1 15.1 10
13-33-178-023 W Waterford Waterford $103,900 $40,700 $125,900 $50,990
1 3-34-157-003 W Waterford Waterford $70,900 $30,900 $97,000 $41 .460
13-34-328-009 W Waterford Waterford $93,500 $45,000 $1 19,900 $63,230
13-34-331-037 W Waterford Waterford $135,000 $56,100 $140,000 $61,060
13-34-427-029 W Waterford Waterford $71,000 $32,200 $90,000 $40,270
13-35-129-01 1 W Waterford Waterford $88,000 $32,800 $102,500 $38,890
13-35-159-018 W Waterford Waterford $65,000 $17,700 $83,000 $24,130
13-35-252-003 W Waterford Waterford $92,000 $51,700 $125,000 $74,900
13-35-451-01 1 W Waterford Waterford $80,000 $29,600 $86,000 $43,890
18-01-402-009 X W. Bloom. Pontiac $94,000 $38,850 $92,000 $47,160
18-04-101-014 X W. Bloom. Waterford $205,000 $95,200 $215,000 $99,000
18-04-251-013 X W. Bloom. W. Bloom. $155,000 $69,275 $200,995 $80,820
18-04-252-007 X W. Bloom. W. Bloom. $165,000 $77,150 $193,000 $86,320
18-04-253-005 X W. Bloom. W. Bloom. $166,500 $75,750 $215,000 $89,320
18-05-101-038 X W. Bloom. Waterford $100,000 $36,025 $130,500 $50,540
18-05-202-010 X W. Bloom. Waterford $70,000 $26,375 $83 .000 $37,180
18-05-202-017 X W. Bloom. Waterford $64,000 $33,750 $103,000 $59,730
18-05-252-016 X W. Bloom. Waterford $77,000 $29,845 $93,711 $36,590
18-05-476-01 1 X W. Bloom. W. Bloom. $260,000 $127,450 $294,000 $135,720
18-05-482-006 X W. Bloom. W. Bloom. $310,143 $32,100 $313,500 $147,520
18-06-151-009 X W. Bloom. Walled Lake $75,000 $30,025 $112,500 $39,710
18-07-326-010 X W. Bloom. Walled Lake $202,000 $82,725 $228,000 $111,450
18—12-428-002 X W. Bloom. Bloom. Hills $160,000 $87,175 $245,000 $91,640
18-14-402-002 X W. Bloom. Bloom. Hills $203,000 $91,325 $234,900 $103,450
18-17-129-030 X W. Bloom. W. Bloom. $79,900 $36,650 $100,938 $42,150
18-17-131-001 X W. Bloom. W. Bloom. $128,500 $56,025 $146,500 $63,670
18-17-154-034 X W. Bloom. W. Bloom. $102,000 $38,000 $149,900 $52,890
18-17-302-032 X W. Bloom. Walled Lake $84,000 $37,250 $108,000 $47,070
18-17-351-030 X W. Bloom. Walled Lake $60,000 $28,275 $85,000 $29,320
18-18-226-031 X W. Bloom. Walled Lake $310,000 $150,000 $310,000 $157,750
18-18-228-002 X W. Bloom. Walled Lake $230,000 $100,725 $250,000 $127,710
18-18-306-116 X W. Bloom. Walled Lake $107,500 $54,575 $145,000 $63,030
18-18-404-028 X W. Bloom. Walled Lake $88,000 $33,625 $1 15,000 $41,660

._—..——-—.-'=“‘

 

 

Table A.2 List of Dual House Sales

151

 

 

Town

Sidwell iD Town School Distr. 92 Price 92 SEV 96 Price 96 SEV

18-18-454-029 X W. Bloom. Walled Lake $35,000 $24,000 $85,000 $30,180
18-18-484-001 X W. Bloom. Walled Lake $55,900 $23,625 $72,000 $27,410
18-20-327-005 X W. Bloom. Walled Lake $305,000 $66,925 $365,000 $176,400
18-26-152-001 X W. Bloom. W. Bloom. $145,000 $67,025 $181,950 $83,570
18-26-454-021 X W. Bloom. W. Bloom. $189,000 $100,925 $257,000 $129,860
18-27-254-010 X W. Bloom. W. Bloom. $183,000 $82,325 $227,500 $94,860
18-28-253-20 X W. Bloom. W. Bloom. $151,250 $69,925 $182,000 $79,860
18-28-431-009 X W. Bloom. W. Bloom. $165,000 $71,475 $205,000 $79,270
18-29-136-125 X W. Bloom. Walled Lake $155,000 $33,925 $210,000 $84,630
18-36-202-007 X W. Bloom. Birmingham $191,000 $84,950 $220,000 $92,070
18-36-228-016 X W. Bloom. Birmingham $102,000 $45,400 $139,000 $54,910
12-06-377-057 Y White Lake Holly $58,000 $24,400 $79,900 $28,970
12-08-100-040 Y White Lake Holly $1 10,500 $45,500 $154,900 $56,730
12-18-206-002 Y White Lake Huron Valley $99,000 $40,900 $127,000 $48,810
12-22-401-030 Y White Lake Huron Valley $136,750 $60,000 $185,000 $70,910
12-23-178-018 Y White Lake Huron Valley $100,000 $41,800 $1 18,000 $48,560
12-33-202-01 1 Y White Lake Huron Valley $142,900 $34,200 $175,000 $68,690
12-34-353-014 Y White Lake Huron Valley $85,000 $6,900 $100,000 $46,660
12-35-230-24 Y White Lake Walled Lake $81,000 $38,600 $109,500 $47,460
12-35-231-044 Y White Lake Walled Lake $65,000 $21,500 $81,400 $29,120
12-26-226-018 Y White Lake Twp Walled Lake $157,000 $12,500 $253,000 $82,140
12-31-426-021 Y White Lake Twp Huron Valley $119,000 $37,700 $146,900 $59,590
12-35-479-013 Y White Lake Twp Walled Lake $74,500 $35,000 $87,000 $34,110
17-28-106-099 96 Wixom Walled Lake $138,000 $66,600 $167,510 $73,790
17-28-377-010 96 Wixom Walled Lake $122,000 $53,800 $144,000 $62,820
17-29-378-018 96 Wixom Walled Lake $96,000 $43,200 $133,000 $53,180
17-29-381-014 96 Wixom Walled Lake $1 16,900 $48,800 $154,900 $59,660
17-30-152-001 96 Wixom Walled Lake $169,667 $14,200 $205,000 $90,650
17-30-177-020 96 Wixom Walled Lake $158,545 $33,800 $196,000 $87,610
17-30-251-033 96 Wixom Walled Lake $1 17,000 $49,500 $138,500 $62,180
17-31-326-016 96 Wixom Walled Lake $137,500 $51,000 $162,900 $70,000
17-32-451-003 96 Wixom Walled Lake $88,500 $37,300 $1 16,800 $41,630
17-32-454-013 96 Wixom Walled Lake $127,850 $1 1,000 $166,000 $71,430
17-21-484-025 EW Wolverine Lake Walled Lake $80,000 $31,800 $105,000 $41,400
17-27-180-004 EW Wolverine Lake Walled Lake $71,500 $28,600 $94,900 $36,300
17-27-278-013 EW Wolverine Lake Walled Lake $98,500 $37,300 $105,000 $37,500
17-27-279-014 EW Wolverine Lake Walled Lake $104,000 $44,400 $138,000 $56,300

 

Source: Author's Compiliation
Note: The Sidwell represents the unique identifier for the property for legal and taxation purposes.

 

152

 

.5550 0:2an 95 0:038 82.030 32;

.3256 5:on 08822 03 flow—6.5 .8000 55>? 0.3 8:50 2:. 6.5.0.5 30:00 .0550 a??? a “0 8.5%. .808 305552 2F ”8oz

 

 

 

 

 

050.0% 000.0% 005.5% 00 _ .5% 000.5% 5.0— 000— 00.0 _ 00.00 _ 5.00 05.00 .553
0 _ 0.0% 02.0% 055.0% 5 _ 0.0% 000.0% _ _.0 00.0 00.0 00.00 00. _0 Q. _0 00501
000.5% 000.0% 000.0% 055.0% 000.0% 00.0 000— 000— 05.00 0 _ .00 0 ~ .00 0:33:02
000.5% 050.5% 00 15% 000.5% 000.5% 000— 00.0~ 000— _0.00 00.00 00.00 200585 803
000.0% 20.0% 000.0% 000.0% 000.0% 00.0 00.0 00.0 50.00 00.00 _0.00 030.533
_05.5% _00.5% 000.5% 000.0% 005.0% 00.0 _ 000— —0.0_ 00.00 00.00 05.00 03 00:03
500.0% 500.5% 000.5% 00 _ .5% 500.0% _ _ .0 _ 00.0 _ 000— 00 .00 00.00 00.00 00.5.
55.0% 000.0% 000.0% 000.0% 00.0% 00.00 00.00 00.00 05.00 00.00 00.00 22.2500
500.0% 000.0% 050.0% 000.0% 000.0% 000— 00.: 00.: 00.00 00.00 00.0 :93 5:00
00_ .0% 000.0% 00 _ .0% 50 _ .5% 005.0% 00.0 _ 000— 5.0— 50. _ 0 00.00 00.00 020 .93."—
000.0% 000.0% 53.0% 055.0% 500.0% 000— 00.0_ 000— 00.00 00.00 05.00 56058.0
500.0% 000.0% 050.0% 000.0% 050.0% 00.0 00.0 00.0 00.00 00.00 05.00 80:00
02.0% 000.0% 000.0% 00.0% 0_ _.0% 00.0— 00.0_ 000— 50.00 50.00 _0.00 0.8.05
000.5% 00.5% 53.0% 000.0% 005.0% 00.3 00.0. 000— 00.00 00.00 00.00 0.30 030
00 0.5% 000.5% 000.0% 0_ 0.0% 000.0% 0 _ .5 _ 00.: 00.0. 00.00 00.00 00.00 .502
00 _ .0% 000.0% 000.0% 000.0% 000.0% 00.0_ 0 _ . Z 0 _._ _ 00.00 00. .0 00. _0 ”.8622
000.0% 00.5% 0 _ 0.0% 30.0% 80.0% 00. _0 00. _0 00.0_ 00.00 00.00 00.00 050053
000.0% 00 0.0% 03.0% 000.0% 000.0% 000— 00.0 _ 00._ _ 00. 00 00.00 00.00 car—O 8.3
505.0% 000.0% 00.0% 00.0% 000.0% 00.: 00.0 00.0 00.00 00.00 00.00 02:5 :05:
055.0% 000 .0% 000.0% 000.0% 000.0% 00.0 _ 05.0_ 00.5 00.00 50.50 50.00 0:0:
_05.0% 20.0% 000.0% 000.0% 02.0% 00.0 00.0 00.0 00.00 00.00 00.00 #80 383
000.0% 000.0% 000.0% 050.0% 505.0% 00. 0 0 00.0 _ 00.5 00.00 00.00 00.00 0.00500
000.0% 000.0% 000.0% 000.0% 000.5% 00.00 00.00 00.00 _0.00 _0.00 00.00 28058.80
00.0% 000.0% 0 _ 0.0% 000.0% 000.0% 00.0 00.0 _ 0.0 _5. _0 00.00 50.00 gasps—U
050.0% _00.0% 000.0% 000.0% 000.0% 00. _ _ 00. _ _ 05.5 00 .00 00 .00 00.00 028120
00 _ .5% 000.0% 00.0% 505.0% 0 _ 0.0% 50.0 _5.0 00.5 0 _ .00 0 0.00 00.00 0:335:20
000.0% 000.0% 00.0% 000.0% 000.0% 00.0_ 00.0. 00.0_ 00.00 50.00 50.00 cow—"mum
000.. _% 0.0.05 000.— _% 000.2% 005.2% 00.5_ 00.5. _0.0_ 00.00 50.00 00.00 2:: 2008005
000.0 _ % 000.0 _ % 0 05.0% 000.0% 000.0% 55.0 _ 50.00 00.00 00.50 00.00 00.00 62055.00—
000.0% 000.0% _ 00 .0 % 000.0% 00 5.0% 00.0 _ 00.0 _ 00 .0 00.00 00. 00 00.00 002500
000.5% 000.0% 000.0% 000.0% 0 _ 0.0% 00.0 _ 00.0_ 050— 00.50 0 0.00 0 0.00 o_00=o><
00.00 00:00 00-00 00-00 0040 50.00 00.00 0000 00-00 00-00 00-; 0:32 «3.5m:— .0230
2.093% =9:— 30 8.5— own—=2 .00—.00

 

as: ӣ28m .23 a... 3:35 _85m 0550 e525 3. as;

 

REFERENCES

 

REFERENCES

Aaron, Henry J. 1975. Who Pays the Property Tax? Washington DC: Brookings
Institution.

Addonizio, Michael F., C. Philip Kearney, and Henry J. Prince. 1995. “Michigan's High
Wire Act.” Journal of Educational Finance. 20(Winter): 235-269.

Bloom, Howard 8., Helen F. Ladd, and John Yinger. 1983. “Are Property Taxes
Capitalized Into House Values?” in Local Provision of Public Services: The
T iebout Model After Twenty Five Years, ed. George R. Zodrow, 145-63. New
York, NY: Academic Press, Inc.

Blume, Lawrence E. 1982. “The Sales and Use Taxes.” in Harvey E. Brazer and
Deborah S. Laren, Eds., Michigan 's Fiscal and Economic Structure. Ann Arbor:
The University of Michigan Press.

Brazer, Harvey E., Deborah S. Laren, and Frank Yu-Hsieh Sung. 1982. “Elementary and
Secondary School Funding.” In Michigan '3 Fiscal and Economic Structure, ed. H.
Brazer and D. Laren, 411-446. Ann Arbor: The University of Michigan Press.

Carroll, Robert J. and John Yinger. 1994. “Is the Property Tax a Benefit Tax?: The Case
of Rental Housing.” National Tax Journal. 47(2): 295-316.

Coons, John E., William H. Clune III, and Stephen Sugarrnan. 1970. Private Wealth and
Public Education. Cambridge, MA: Harvard University Press.

Courant, Paul N. 1982. “The Property Tax.” In Michigan '3 Fiscal and Economic
Structure, ed. H. Brazer and D. Laren, 470-529. Ann Arbor: The University of

Michigan Press.

Courant, Paul N., Edward M. Gramlich, and Susanna Loeb. 1995. “Michigan's Recent
School Finance Reforms: A Preliminary Report.” American Economic Review.
85(2): 372-77.

Do, Quang A. and CF. Sirmans. 1994. “Residential Property Tax Capitalization:
Discount Rate Evidence From California.” National Tax Journal. 47(2): 341-48.

Dusansky, Richard, Melvin Ingber, and Nicholas Karatjas. 1981. “The Impact of
Property Taxation on Housing Values and Rents.” Journal of Urban Economics.
10: 240-55.

153

 

154

F eldstein, Martin S. 1975. “Wealth Neutrality and Local Choice in Public Education.”
American Economic Review. 65(1): 75-89.

Fisher, Ronald C. 1996. State and Local Public Finance. Chicago: Richard D. Irwin,
Inc.

Gabriel, S. A. 1981. “Interjurisdictional Capitalization Effects of Proposition 13 in the
San Francisco Bay Area.” National Tax Association Proceedings. 263-71.

Gravelle, Jane G. 1994. The Economic Effects of Taxing Capital Income. Cambridge,
Massachusetts: MIT Press.

Gronberg, Timothy J. 1979. “The Interaction of Markets in Housing and Local Public
Goods: A Simultaneous Equations Approach.” Southern Economic Journal. 46:
445-59.

Guilfoyle, Jeffrey P. 1997. “The Effect of Property Taxes and School Spending on x.
House Prices: Evidence from Michigan’s Proposal A.” Unpublished Working
Paper.

Hamilton, Bruce W. 1976. “Capitalization of Intrajurisdictional Differences in Local
Tax Prices.” American Economic Review. 66(5): 743-53.

. 1983. “Is the Property Tax a Benefit Tax?” in Local Provision of Public
Services: The T iebout Model After Twenty Five Years, ed. George R. Zodrow, 85-
107. New York, NY: Academic Press, Inc.

 

Harberger, Arnold C. 1962. “The Incidence of the Corporate Income Tax.” Journal of
Political Economy. 70(3): 215-40.

Harmon, Oskar R. 1988. “The Income Elasticity of Demand for Single-Family Owner-
Occupied Housing: An Empirical Reconciliation.” Journal of Urban Economics.
24(2): 173-85.

Heckman, James J. and V. Joseph Hotz. 1989. “Choosing Among Alternative
Nonexperimental Methods for Estimating the Impact of Social Programs: The
Case of Manpower Training.” Journal of the American Statistical Society.
84(408): 862-75.

Hobson, Paul. 1986. “Land Rents, Optimal Taxation and Local Fiscal Independence in
an Economy with Local Public Goods.” Journal of Public Economics. 15(1): 59-
85.

Kearney, Philip C. 1994. A Primer on Michigan School Finance 3rd Edition 1994. Ann
Arbor: Educational Studies Program, The University of Michigan.

 

155

King, A. Thomas. 1977 “Estimating Property Tax Capitalization: A Critical Comment.”

Journal of Political Economy. 85(2): 425-31.

Krantz, Diane P., Robert D. Weaver, and Theodore R. Alter. 1982. “Residential
Property Tax Capitalization: Consistent Estimates Using Micro-Level Data.”
Land Economics. 58(4): 488-96.

Lea, Michael J. 1982. “Local Tax and Expenditure Capitalization: Integrating Evidence
From the Market and Political Processes.” Public Finance Quarterly. 10(1): 95-
117.

Michigan State Board of Education. 1991. Bulletin 1014: Michigan K—12 School
Districts Ranked by Selected Financial Dat «1991-92. Lansing: The State
Board.

Mieszkowski, Peter. 1972. “The Property Tax: An Excise Tax or a Profits Tax?”
Journal of Public Economics. 1(1): 73-96.

Mieszkowski, Peter and George R. Zodrow. 1989. “Taxation and the Tiebout Model:
The Differential Effects of Head Taxes, Taxes on Land Rents, and Property
Taxes.” Journal of Economic Literature. 27: 1098-1 146.

Netzer, Dick. 1966. Economics of the Property Tax. Washington DC: Brookings
Institution.

Oates, Wallace E. 1969. “The Effects of Property Taxes and Local Public Spending on
Property Values: An Empirical Study of Tax Capitalization and the Tiebout
Hypothesis.” Journal of Political Economy. 77(6): 957-71.

Office of Revenue and Tax Analysis, Michigan Department of Treasury. 1994.
Michigan ’s Sales and Use Taxes 1994. Lansing, MI.

—. 1996. Michigan 's Cigarette and Tobacco Taxes 1996. Lansing, MI.

Pechman, Joeseph A. 1985. Who Paid the Taxes, 1966—85? Washington, DC: The
Brookings Institution.

Poterba, James M. 1996. “Retail Price Reactions to Changes in State and Local Sales
Taxes.” National Tax Journal. 49(2): 165-76.

Reschovsky, Andrew. 1994. “Fiscal Equalization and School Finance.” National Tax
Journal. 47(1): 185-97.

Richardson, DH. and R. Thalheimer. 1981. “Measuring the Extent of Property Tax
Capitalization for Single Family Residences.” Southern Economic Journal.
482674-89.

 

 

156

Ring, Raymond J. 1989. ”The Proportion of Consumers’ and Producers’ Goods in the
General Sales Tax.” National Tax Journal. 42(2): 167-79

Rosen, Harvey S. and David J. Fullerton. 1977. “A Note. on Local Tax Rates, Public
Benefit Levels, and Property Values.” Journal of Political Economy. 85(2): 433-
40.

Rosen, Kenneth. 1982. “The Impact of Proposition 13 on House Prices in Northern
California: A Test of the Interjurisdictional Capitalization Hypothesis.” Journal
of Political Economy. 90(1): 191-200.

Rubinfeld, Daniel L. and Robert W. Vishny. 1982. “Property Tax Reduction in
Michigan.” in Harvey E. Brazer and Deborah S. Laren, Eds, Michigan ’s Fiscal
and Economic Structure. Ann Arbor: The University of Michigan Press.

Samuelson, Paul A. 1954. “The Pure Theory of Public Expenditures.” Review of
Economics and Statistics. 36(4): 387-89.

Simon, Herbert A. 1943. “The Incidence of a Tax on Urban Real Property.” Quarterly
Journal of Economics. 59(3): 398-420.

Sung, Hai-Yen, Teh-Wei Hu, and Theodore E. Keeler. 1994. “Cigarette Taxation and
Demand: An Empirical Model.” Contemporary Economic Policy. 12(3): 91-100.

Thomson, Procter. 1965. “The Property Tax and the Rate of Interest.” in The American
Property Tax. ed. George S. Benson et al. 111-98. Claremont, CA: Lincoln
School of Public Finance.

Tiebout, Charles M. 1956. “A Pure Theory of Local Expenditures.” Journal of Political
Economy. 64(5): 416-24.

US. Bureau of the Census. Statistical Abstract of the United States 1995 and 1997.
Washington, DC.

Wassmer, Robert W. 1993. “Property Taxation, Property Base, and Property Value: An
Empirical Test of the ‘New View’.” National Tax Journal. 46(2): 135-59.

Wassmer, Robert W. and Ronald C. Fisher. 1996. “An Evaluation of the Recent Move
to Centralize the Finance of Public Schools in Michigan.” Public Budgeting and
Finance. 16(3): 90-112.

Yinger, John. 1982. “Capitalization and the Theory of Local Public Finance.” Journal
of Political Economy. 90(5): 917-43.

 

 

157

 

. 1985. “Inefficiency and the Median Voter: Property Taxes, Capitalization,
Heterogeneity, and the Theory of Second Best.” in Perspectives on Local Public
Finance and Public Policy, Vol. 2. ed. J .M. Quigley. Greenwich, Connecticut:
J AI Press.

Yinger, John., Howard S. Bloom, Axel B—ursch-Supan, and Helen F. Ladd. 1988.
Property Taxes and House Values: The Theory and Estimation of
Intrajurisdictional Property Tax Capitalization. San Diego, CA: Academic Press,
Inc.

Zodrow, George R. and Peter Mieszkowski. 1983. “The Incidence of the Property Tax.”
in Local Provision of Public Services: The T iebout Model After Twenty Five
Years, ed. George R. Zodrow, 109-29. New York, NY: Academic Press, Inc.

 

 

"‘lllllllllllllllEs