   

        
     
    

WI

1

W

W

WHIIHHHHIHIIH

1
~

4: IIIIM

$1 ocﬂ

 

 

 

 

LIBRARY
Michigan State
L nversity

 

 

This is to certify that the
thesis entitled

MEASURING THE INDIRECT FISCAL IMPLICATIONS OF
LAND USE PLANNING AND NATURAL AMENITIES ON
PROPERTY TAX REVENUES: A METHODOLOGICAL
FRAMEWORK

presented by
MICHAEL C. FORSYTH

has been accepted towards fulfillment
of the requirements for the

MS. degree in Urban and Regional Planning_

m ///’;J

Major Professor’ 3 Signature

/iL/oz,/M

Date

 

 

MSU is an Affirmative Action/Equal Opportunity Employer

_ .. -A—.--I----I-n-o--n-o-

 

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 Ki/Proj/Acc&Prele|RClDaIeDueindd

MEASURING THE INDIRECT FISCAL IMPLICATIONS OF LAND USE
PLANNING AND NATURAL AMENITIES ON PROPERTY TAX REVENUES:
A METHODOLOGICAL FRAMEWORK

By

Michael C. Forsyth

A THESIS
Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of
MASTERS OF SCIENCE
Urban and Regional Planning

2008

ABSTRACT
MEASURING THE INDIRECT FISCAL IMPLICATIONS OF LAND USE
PLANNING AND NATURAL AMENITIES ON PROPERTY TAX REVENUES:
A METHODOLOGICAL FRAMEWORK
By
Michael C. F orsyth

The limitations of ﬁscal evaluation methods are addressed by using a modiﬁed
hedonic pricing technique to document and quantify the indirect contributions of
surrounding land use patterns and natural amenities to property tax revenues. The
hedonic pricing function uses multiple regression analysis to decompose the assessed
“true cash value” of single and multi-family residential, commercial and industrial
building and land values based on their respective structural and location-based
characteristics. The method is tested in seven case study communities in Oakland
County, Michigan.

Structural and location-based characteristics incorporated within hedonic models
produce statistically signiﬁcant results, explaining between 35.8% to 68.8% of the
variance in assessed land values and 80.6% to 91.3% of the variance in assessed building
values. Methodologicaloutcomes demonstrate that planning policies and regulations that
deﬁne the spatial composition of land use development patterns indirectly add and
subtract from the property tax base of case study communities. Beneﬁts, challenges and
limitations associated with the application of hedonic pricing techniques to ﬁscal analysis

are presented and implications to planning practice are discussed.

 

ACKNOWLEDGEMENTS

I want to express my sincere appreciation and gratitude for the time, expertise and,
patience put forth by the members of my thesis committee. I would also like to thank
everyone at the Department of Urban and Regional Planning and the Land Policy
Institute for providing me with an invaluable learning experience during a most
rewarding past three years at Michigan State University.

In particular, I’d like to offer special thanks to Dr. Zenia Kotval for sharing her
expertise in academia and planning practice, instilling a sense of pragmatism in my
thinking, and always looking out for my well-being; to Mark Wyckoﬁ‘ whose work ethic,
passion, and seemingly endless knowledge of planning is truly an inspiration to which I
aspire; and to Dr. Patricia Machemer for her support, good humor, and patience in
dealing with me as I disappeared to the other side of the country. I would also like to
thank Nigel Griswold and Yohannes Haliu for their invaluable econometric expertise and
support, without which this thesis would not have been possible; GIS guru and good
friend, Travis Perkins, for his assistance in spatial modeling; Chris Moshier, Charlotte
Burkhardt, Rob Scripture and Dave Heiber with Oakland County for their expertise and
outstanding assistance with data collection; and Dr. Soji Adelaja for teaching me the
power of personal vision, leadership and econometrics. Finally, I want to offer my great
thanks to all of my friends and family who put up with me while I worked on this thesis.
Very special thanks to my mother and father for their caring love, support and inspiration
to make a difference in the world. And last but not least, to my dearest Rachel Moote

who stood by me from start to ﬁnish... I will always love you more than my thesis.

iii

TABLE OF CONTENTS

LIST OF TABLES .................................................................................... vi
LIST OF FIGURES .................................................................... f .............. ix
CHAPTER 1: INTRODUCTION

1.1 Introduction and Problem Statement ...................................................... l

1.2 Research Questions and Approach ........................................................ 3

1.3 Research Objectives .......................................................................... 4
1.4 Research Hypotheses ........................................................................ 5
1.5 Thesis Overview ............................................................................. 6

CHAPTER 2: PROPERTY TAX ASSESSMENT IN MICHIGAN: IMPLICATIONS
FOR LAND USE PLANNING AND THE BOTTOM LINE

2.1 Property Tax Assessment Theory and Practice in Michigan ........................... 8
2.2 Introduction to Property Tax Analysis in Case Study

Commrmities .................................................................................. 8
2.3 Measuring Assessed True Cash Value ................................................... 17
2.4 Analysis of the True Cash Value of Building Improvement Values in

Case Study Communities .................................................................. 21
2.5 Land Value in Case Study Communities ................................................ 25
2.6 Composition of the Property Tax Base in Case Study Communities ............... 28
2.7 Fiscal Implications of Michigan’s Property Tax Equation ........................... 30
2.8 Informing Hedonic Pricing Analysis of Indirect Fiscal Impacts through

Applications of Michigan General Property Tax Act ................................. 31

CHAPTER 3: MEASURING THE INDIRECT FISCAL IMPACTS OF THE BUILT
AND NATURAL ENVIRONMENT

3.1 Measuring Indirect Fiscal Impacts through Hedonic Pricing ........................ 36
3.2 Empirical Estimates and Regression Results ........................................... 47
3.2.1 Basic Property Attributes ......................................................... 49

3.2.1.1 Construction Class ..................................................... 51

3.2.1.2 Age ....................................................................... 51

3.2.1.3 Square Footage .......................................................... 52

3.2.1.4 Lot Size .................................................................. 52

3.2.1.5 Number of Structures .................................................. 53

3.2.2 Residential Building Attributes ................................................. 53

3.2.2.1 Residential Exteriors ................................................... 54

3.2.2.2 Residential Building Style ............................................ 55

3.2.3 Socio-Economic Characteristics ................................................ 55

3.2.4 Adjacent and Surrounding Land Use Patterns ................................. 57

3.2.5 Adjacent and Surrounding Natural Amenities ................................. 63

3.2.6 Adjacent and Surrounding Capital Facilities .................................. 66

3.2.7 Transportation Infrastructure and Access ...................................... 69

iv

3.3 Quantiﬁcation of Implicit Property Tax Revenues Attributed to
Surrounding Land Use Patterns and Natural Amenities .............................. 74

CHAPTER 4. SUMMARY OF FINDINGS, LIMITATIONS, SUGGESTIONS FOR
FUTURE RESEARCH AND APPLICATIONS TO THE PLANNING PROFESSION

4.1 Summary of Findings ...................................................................... 82
4.2 Limitations and Suggestions for Future Research ..................................... 86
4.2.1 Functional Form of the Hedonic Model ........................................ 87
4.2.2 Design of Independent Variables ................................................ 88
4.2.3 Temporal Dimensions ............................................................ 90
4.3 Applications to Planning Policy and Practice ........................................... 9O
APPENDICES ....................................................................................... 93
REFERENCES .................................................................................... 105

Table 1
Table 2
Table 3

Table 4

Table 5

Table 6

Table 7

Table 8

Table 9

' Table 10

Table 1 1

Table 12
Table 13

Table 14

Table 15

Table 16

Table 17

LIST OF TABLES

Population Trends in Case Study Communities .............................. l 1
City of Pontiac Land Use and Property Value Patterns, 2007 ............... 12
City of Royal Oak Land Use and Property Value Patterns, 2007 .......... 13
Charter Township of Commerce Land Use and Property Value

Patterns, 2007 ...................................................................... 14
Village of Wolverine Lake Land Use and Property Value Patterns,

2007 ................................................................................. 14
Milford Charter Township Land Use and Property Value Patterns,

2007 ................................................................................. 15
Village of Milford Land Use and Property Value Patterns, 2007 .......... 16
Charter Township of Oakland Land Use and Property Value Patterns,
2007 .................................................................................. 17
Basic Determinants of City Assessed Residential Building Values,

2007 .................................................................................. 22
Basic Determinants of Township and Village Assessed Residential
Building Values, 2007 ............................................................ 24
Basic Determinants of Commercial, Industrial and Multi-Family

Building Values in Case Study Communities, 2007 ........................... 25
Commercial and Industrial Depreciation Rates ............................... 26
Vacant City Land Values Based on Highest and Best Use, 2007 .......... 27
Basic Components of Property Tax Revenue in

Case Study Cities, 2007 ........................................................... 28
Basic Components of Property Tax Revenue in

Case Study Townships and Villages ............................................. 29
Description of Property Variables Used in Regression Analysis ........... 39
Social Characteristics of Case Study Communities, 2000 ................... 40

vi

Table 18

Table 19

Table 20

Table 21
Table 22
Table 23
Table 24

Table 25

Table 26

Table 27

Table 28

Table 29
Table 30
Table 31

Table 32

Table 33

Table 34

Table 35

Housing and Income Characteristics in Case Study Communities,

2000 ................................................................................. 41
Description of Socio-Economic Variables Used in

Regression Analysis ............................................................... 42
Description of Adjacent Land Use Variables Used in

Regression Analysis ............................................................... 43
Description of Proximity Variables Used in Regression Analysis. . .. . .....45
Building Value Model Summaries .............................................. 48
Land Value Model Summaries .................................................. 48

Linear Regression Coefﬁcient Estimates for Basic Property Attributes...50

Linear Regression Coefﬁcient Estimates for Residential Building
Charactenstrcs54

Linear Regression Coefficient Estimates for Socioeconomic Attributes..56

Impacts of Land Use Patterns on City Single Family Residential

Property Values58
Impacts of Land Use Patterns on Township and Village Single Family

Residential Property Values ...................................................... 59
Impacts of Land Use Patterns on Multi-Family Property Values ........... 6O
Impacts of Land Use Patterns on Commercial Property Values. . . . . . . . .....61
Impacts of Land Use Patterns on Industrial Property Values ............... 62

Impacts of Natural Amenities on Single Family Residential
Property Values in Cities ......................................................... 64

Impacts of Natural Amenities on Single Family Residential
Property Values in Townships and Villages ................................... 65

Impacts of Natural Amenities on Non-Residential and Multi-Family
Property Values .................................................................... 66

Impacts of Capital Facilities on Single Family Residential
Property Values .................................................................... 68

vii

 

Table 36

Table 37 '

Table 38

Table 39

Table 40

Table 41

Table 42

Table 43

Table 44

Table 45

Table 46

Table 47

Impacts of Capital Facilities on Non-Residential and Multi-Family
Property Values .................................................................... 69

Impacts of Transportation Infrastructure on City Residential
Property Values .................................................................... 70

Impacts of Transportation Infrastructure on Residential
Property Values in Townships and Villages ................................... 71

Impacts of Transportation Infrastructure on Multi-Family
Property Values ..................................................................... 72

Impacts of Transportation Infrastructure on Commercial
Property Values ..................................................................... 73

Impacts of Transportation Infrastructure on Industrial
Property Values .................................................................... 74

City Total Implicit Single Family Property Tax Revenues
Attributed to Surrounding Land Use Patterns ................................. 75

Township and Village Total Implicit Single Family Property Tax
Revenues Attributed to Surrounding Land Use Patterns ..................... 76

Township and Village Total Implicit Single Family Property Tax
Revenues Attributed to Surrounding Water-based Natural Amenities ...... 77

Township and Village Total Implicit Single Family Property Tax
Revenues Attributed to Surrounding Natural Areas ........................... 7 8

City Total Implicit Single Family Property Tax Revenues
Attributed to Surrounding Natural Amenities ................................. 78

Total Implicit Non-Residential and Multi-Family Property Tax
Revenues Attributed to Surrounding Land Use Patterns ..................... 79

viii

Figure 1

Figure 2

Figure 3

Figure 4

LIST OF FIGURES
Selected Case Study Communities ............................................. 10
Methodological Framework ...................................................... 32

Summary of Indirect Fiscal Impacts Attributable to Surrounding
Land Use Patterns in Case Study Communities .............................. 83

Summary of Indirect Fiscal Impacts Attributable to Surrounding
Natural Amenities in Case Study Communities .............................. 85

ix

CHAPTER 1
INTRODUCTION

1.1 Introduction and Problem Statement

The ﬁscal implications of land use planning and regulation are not well
understood by planners (Huddleston, 2005). The limitations to ﬁscally-informed planning
practice are based in the constraints of “value” as deﬁned by a monetized economy and
the shortcomings of traditional ﬁscal evaluation techniques that do not fully account for
the value of public, comprehensive land use planning in a community. Traditional
methods such as ﬁscal impact analysis and other closely related quantitative techniques,
measure the direct, current, and public costs and revenues associated with development to
the local jurisdictions where growth is taking place (Burchell, 1978). Current ﬁscal
analysis techniques are not designed to account for the value interactions between land
use patterns (Kotval and Mullin, 2006), the spatial dimensions of growth (Dekel, 1995)
and the cumulative outcomes of local government planning and investment (Heikkila and
Davis, 1997).

Under current valuation ﬁameworks, the economic and ﬁscal value of planning
merely coincides with the direct market value of development, and does not account for
the indirect “value added” to the development by the advantages and disadvantages of
location created by land use planning. The location-based beneﬁts of place that inﬂuence
property values represent the true purpose and value of planning to a community and its
residents. Without a method to quantify the “value-added” of planning to property values
in a community, planners will never be able to truly understand, communicate, or

measure the

 

ﬁscal outcomes of planning policies and regulations. This thesis attempts to address this
problem by introducing a new methodological framework for ﬁscal analysis.

As the foundation of this thesis, the limitations of conventional ﬁscal techniques
are explored through the revenue side of the ﬁscal equation. The “value of planning” is
quantiﬁed by measuring the indirect contributions of surrounding land use patterns and
natural amenities to private property values and ultimately the property tax base of local
communities. The property tax accounts for the largest source of local government
revenue in the United States (Michigan State Assessors Board, p. 1-1) and is assessed
based on an estimate of the true market value of real property. Therefore, the property tax
represents a vital and direct connection between private, economic value and public,
ﬁscal value.

The property tax captures the economic and ﬁscal value of planning in two
distinctive ways. First, zoning and building codes directly inﬂuence the value of real
property and property tax revenues by deﬁning the functions and development
characteristics of land and land improvements. The price paid for real property and the
property taxes captured under land use regulations represents the extent to which planners
have evaluated the property tax revenue outcomes of planning.

The value of real property and property tax revenues are also indirectly inﬂuenced
by the location-based advantages and disadvantages created by the spatial distribution of
land uses and zoning regulations. The price of a building or land represents not only a
bundle of physical characteristics but also a set of location speciﬁc attributes (Chesire
and Shepherd, 1995). In a monetized economy, the beneﬁts of place cannot be explicitly

bought, sold, or owned by a private entity but rather, are deﬁned by the willingness to

pay for market goods (i.e. a building or land parcel) that maintain a set of speciﬁc
location attributes (Field, 2001 & Taylor, 2003).
1.2 Research Questions and Approach

This thesis asks at a basic level, “what is the indirect value of planning and how
can it be measured?” For the purpose of this thesis, the value of planning is deﬁned by
the spatial distribution of land use patterns and natural amenities that indirectly inﬂuence
property values and property tax revenues.

To measure the indirect value of planning, a modiﬁed “hedonic pricing” method
is tested. “The hedonic method for non-market valuation is an econometric technique that
uses market transactions to ascertain the value of characteristics associated with
variations in products. By observing the price differentials between two product varieties,
the monetary trade-offs individuals are willing to make with respect to variations are
inferred indirectly” (Taylor, 2003, p. 332). In the spirit of improving the practice of ﬁscal
analysis, assessed true cash value rather than market transactions are used to ascertain the
value contributions of surrounding land use patterns and natural amenities to property
values and property tax revenues in seven case study communities in Oakland County,
Michigan. To improve the science of ﬁscal analysis through hedonic pricing, three
methodological research questions are explored. (1) Is the hedonic pricing method an
effective tool to document and measure the indirect impact of location-based attributes on
property values and property tax revenues? (2) What adjustments must be made to the
hedonic pricing model to effectively measure the ﬁscal impacts of land use patterns and

natural amenities? (3) What spatial variables represent the planned environment, how

should they be incorporated within the hedonic-ﬁscal analysis, and what are the sources
of data?

To ultimately improve planners’ understanding of the ﬁscal implications of land
use patterns and natural amenities, four research questions are used to interpret the
outcomes of methodological techniques: (1) Which surrounding land use and natural
amenity patterns signiﬁcantly impact property tax revenues? (2) How do signiﬁcant value
relationships differ between different land uses? (3) What are the ﬁscal implications
associated with signiﬁcant value relationships? (4) And ﬁnally, what are the implications
to planning policy and practice?

1.3 Research Objectives

The overall purpose of this thesis is to document and quantify the indirect ﬁscal
implications of existing land use and natural amenity patterns to a community’s property
tax base in order to understand the value-added of comprehensive land use planning to a
community and its residents. The ﬁscal implications of planning policies and regulations
are evaluated through ﬁndings that reveal how the proximity to speciﬁc land uses affects
the value of other land uses and how certain mixes of land uses add or detract from
property tax revenues. Three primary objectives of this research aim to help improve the
planner’s ability to document and measure the indirect value of planning.

The ﬁrst objective of this thesis is to document the ﬁscal implications of property
tax theory, law and assessment practice to a community’s property tax base and inform
the hedonic modeling of property tax revenues. To achieve this objective, Michigan’s
“property tax equation” is applied to deconstruct the property tax base of seven case

study communities in Oakland County using 2007 assessment data provided by the

Oakland County Equalization Department. Analyses of property taxes are informed by
legislative guidelines mandated by the Michigan General Property Tax Act and Michigan
Tax Assessors Training Manual.

The second objective of this thesis aims to improve ﬁscal analysis techniques
through hedonic pricing. A hedonic pricing technique applies multiple regression analysis
to measure relationships between property values and the surrounding planned
environment for single and multi-family residential, commercial and industrial land uses.
Regression coefﬁcients reveal statistically signiﬁcant causal relationships between
property values and spatial variables and assign an implicit value to variables. Signiﬁcant
coefﬁcients are then transformed using Michigan’s property tax equation to measure the
indirect ﬁscal impacts.

It should be noted that while a plethora of hedonic studies and applications exist,
this study may be one of the ﬁrst applications of the hedonic pricing technique to
measure implicit property tax revenues attributed to the planned environment. The goal
here is not to master the hedonic pricing technique as it applies to ﬁscal analysis, but
rather to document methodological outcomes, challenges, limitations and lessons learned.

Finally, this thesis aims to improve the planners understanding of property tax
revenues and the fiscal dimensions of policies that impact the property tax base
indirectly. Findings will help planners better understand, communicate and plan for ﬁscal
sustainability and will help improve coordination and mutual understanding between
planning and property assessment professions.

1.4 Research Hypothesis

A testable hypothesis exists with each independent variable incorporated within
the hedonic pricing model. For any given independent variable, the null hypothesis is that
no signiﬁcant relationship exists between the assessed true cash value of real property
and the independent variable analyzed, all else equal.

From a higher elevation, surrounding land use patterns are hypothesized to cause
both increases and decreases in the assessed value of real property. Causal relationships
between property values and surrounding land use patterns are predicted to be unique to
the land use being analyzed. Distance to the land use pattern is hypothesized to inﬂuence
variations in value relationships between property values and spatial variables. Natural
amenities, particularly lakes, are believed to signiﬁcantly inﬂuence surrounding property
values, especially residential values in rural communities. Due to the nature of property
appraisal, the physical characteristics of building improvement are predicted to outweigh
the value of location-based attributes. The location-based advantages and disadvantages
that inﬂuence property values may be minimal on individual properties, but the
cumulative impacts of these value relationships will likely be associated with signiﬁcant
ﬁscal impacts on a community-wide basis.

1.5 Thesis Overview

Chapter Two begins with a thorough analysis of Michigan property tax theory and
assessment practice based on the Michigan General Property Tax Act and the Michigan
Tax Assessors Manual. Case study communities are introduced and analyzed. Direct
ﬁscal impacts of land use patterns are measured and the key differences between case
study communities are analyzed based on urban, suburban and rural characteristics.

Implications of property tax assessment are related to land use planning patterns.

Analysis of tax assessment sets the stage modeling relationships between assessed values
and the planned environment in Chapter Three.

Chapter Three introduces the theoretical concepts of hedonic pricing and deﬁnes
the methodology used in this thesis. A six step process for estimating the hedonic pricing
function through multiple regression analysis is presented which includes (1) deﬁnition
of values estimated, (2) selection of dependent variables, (3) selection of independent
variables, (4) the sample geographic and real estate market frame (5) functional form of
the hedonic model and (6) issues of spatial dependence and correlation. Each step in the
estimation of the hedonic price ﬁrnction details the issues, challenges and beneﬁts of
using assessor’s data within the ﬁamework of Michigan’s property tax equation. Results
of the hedonic pricing model are then presented. Interpretation begins by documenting
statistically signiﬁcant impacts on private real property values. Signiﬁcant relationships
between real property values and surrounding land use and natural amenity patterns are
then translated to measure the ﬁscal impact of these relationships.

The thesis concludes in Chapter Four with a summary of ﬁndings. Limitations of
the hedonic pricing technique and suggestions for future research and improvements are
emphasized in the conclusion. Finally, opportunities and challenges of applying these

techniques to the planning profession are discussed.

CHAPTER 2
PROPERTY TAX ASSESSMENT IN MICHIGAN:
IMPLICATIONS FOR LAND USE PLANNING AND THE BOTTOM LINE
2.1 Property Tax Assessment Theory and Practice in Michigan

In Michigan, the General Property Tax Act, PA 206 of 1893 (M.C.L Sec. 211.1-
211.157) establishes the legal foundation for taxation where, “all property, real and
personal, within the jurisdiction of this state, not expressly exempted, shall be subject to
taxation” (Sec. 211.1). Michigan uses an ad valorem tax system where taxes are based on
the market value of property, commonly referred to “true cash value.” Section 27 (1) of
PA 206 deﬁnes “true cash value” as the usual selling price of a property at the time of
assessment, being the price that could be obtained for the property at private sale.

Sec. 27(1) highlights the vital role that land use patterns and natural amenities
play in the determination of true cash value, establishing that “the assessor shall consider
the advantages and disadvantages of location; quality of soil; zoning; existing use;
present economic income of structures. . among other factors.

The planner inﬂuences the determination of assessed true cash value in two
important ways. Land use planning and regulation inﬂuences the type, characteristics and
location of development, which directly impacts assessed real property values and in
large part determines the makeup of the property tax base from a land use perspective.
The spatial arrangement of land uses and public infrastructure creates advantages and
disadvantages associated with location, therefore indirectly impacting assessed real

property values and property tax revenues.

2.2 Introduction to Property Tax Analysis in Case Study Communities

In order to document direct revenue impacts of existing land use patterns and the
ﬁscal implications of Michigan’s General Property Tax Act, the property tax bases of
seven communities in Oakland County, Michigan are analyzed The case study approach
uses 2007 raw assessment data provided by the Oakland County Equalization
Department, 2007 spatial data provided by the Oakland County Department of
Community and Economic Development, and the principles of assessment practice
mandated by the Michigan General Property Tax Act and the Michigan Tax Assessors
Training Manual in order to research relationships between property tax assessment and
land use development patterns.

Case study communities are introduced in “Figure 1: Selected Case Study
Communities,” which provides a map of case study communities and is supplemented by
a population analysis in “Table 1: Population Trends in Case Study Communities.” Land
use patterns and direct sources of property value and property tax revenue are

documented in each case study community.

 

Figure 1: Map of Case Study Communities

 

 

l
I

Village of
Milford

   

   
   
  

 

Village of
Wolverine La e

IMF—t = Royal Oak
I
. .. ‘ I

1 I City of
I

 

Legend

Case Study Communities

I i Oakland County Jurisdictions N!
0 5 IO 20 Miles

I I I I l I I I l

 

 

 

Oakland County is located in southeast Michigan and is the second most
populated county in Michigan. Oakland County is composed by a diverse range of urban,
suburban and rural communities. Typical of many metropolitan areas across America,
suburban and exurban communities in Oakland Cormty have experienced an increase in
population and development investment, while central cities have experienced decline
and disinvestment. To best analyze the indirect ﬁscal implications of planning in different
types of communities, case study jurisdictions were chosen based on three primary
criteria: (I) to represent the urban transect, (2) capture grth patterns in the county as a
whole, and (3) the availability of assessors data. The cities of Pontiac and Royal Oak,
villages of Wolverine Lake and Milford and the charter townships of Commerce, Milford
and Oakland were chosen for ﬁscal analysis based on these criteria.

“Table 1: Population Trends in Case Study Communities” demonstrates that the
central cities of Pontiac and Royal Oak have lost between seven and eight percent
(decline of 4829 to 5348 persons respectively) of their total population from 1990 to
2000, while the suburban and exurban communities of Commerce, Milford and Oakland
Charter Township have experienced population gains ranging between 26 and 50 percent.
The Village of Wolverine Lake located adjacent to Commerce Township has experienced
minimal population loss while the Village of Milford small population gains.

Table 1. Population Trends in Case Study Communities

 

 

 

ngulation Population Change
Community 1990 2000 Total Change Percent Chang_e_
City of Pontiac 71,166 66,337 -4,829 -7%
City of Royal Oak 65,410 60,062 -5,348 -8%
Charter Township of Commerce 26,955 34,764 7,809 29%
Village of Wolverine Lake 4,727 4,415 —3 12 -7%
Charter Township of Milford 12,121 15,271 3,150 26%
Village of Milford 5,511 6,272 761 14%

Charter Township of Oakland 8,227 13,071 4,844 59%

ll

The City of Pontiac, centrally located in Oakland County, maintains a diverse tax
base associated with a mix of residential and non-residential land uses shown in “Table 2.
City of Pontiac Land Use and Property Value Patterns.” Density plays an important role
in determining direct property tax revenues in a community. Dense commercial and
residential development is a source of signiﬁcant property value and tax revenue within
Pontiac. For a central city, uncharacteristically large amounts of vacant land exist in
Pontiac. In the majority of these cases, vacant lands are residential lots with abandoned or

razed structures, representing a signiﬁcant loss of potential taxable value.

Table 2. City of Pontiac Land Use and Property Value Patterns
Estimated True Cash Property Tax % Total

 

 

 

Land Use Acres Value Revenue Revenue
Commercial 776 $626,696,729 $10,136,787 20.2%
Industrial 1,627 $580,618,599 $10,545,030 21.0%
Mobile Home Park 68 $9,989,966 $182,864 0.4%
Multi-Family 507 $199,709,] 18 $3,122,264 6.2%
Public/Institutional“ 1,342 $41,683,830 $597,261 1.2%
Railroad 61 $521,257 $13,408 0.0%
Recreation/Conservation 704 $440,050 $6,178 0.0%
Single Family, 1 to 2.5 acres 136 $16,852,028 $192,799 0.4%
Single Family; 14,000 to 43,559 sf 447 $122,855,095 $1,595,231 3.2%
Single Family, 2.5 to 4.9 acres 63 $4,045,514 $40,218 0.1%
Single Family, 5 to 9.9 acres 12 $606,479 $4,130 0.0%
Single Family, 8,000 to 13,999 s.f. 732 $378,303,248 $5,157,346 10.3%
Single Family, Less than 8,000 s.f. 1,767 $1,210,847,853 $16,312,785 32.5%
Single Family, More than one unit per acre 125 $20,934,325 $151,317 0.3%
Transportation, Utility, Communication 361 $14,245,262 $260,697 0.5%
Vacant 1,860 $133,416,576 $1,903,255 3.8%
Water" 10 $2,000 $53 0.0%
Undeﬁned 337 $21,807 $179 0.0%
TOTAL 10,935 $3,361,789,736 $50,221,800 100.0%

 

"‘ Public/Institutional contains both taxable and tax exempt properties. Most taxable public/institutional properties
are hospitals

*" In a few cases, water land use desginations represent minor inconsistencies in labeling between Equalization and
Comnmnity Flaming Departments

The City of Royal Oak is the oldest of the case study communities. Like Pontiac,

Royal Oak demonstrates a diverse tax base representative of a central city shown in
“Table 3. City of Royal Oak Land Use and Property Value Patterns.” Royal Oak relies

more heavily on commercial land uses than industrial land use as a source of property tax

12

revenue when compared to the City of Pontiac. High density residential uses on lots less

than 14,000 square feet account for over 65% of City’s tax base.

Table 3. City of Royal Oak Land Use and Property Value Patterns

 

 

 

 

Estimated True Cash Property Tax % Total
Land Use Acres Value Revenue Revenue
Commercial 377 $681,132,313 $12,846,577 15.8%
Industrial 165 $136,334,075 $2,707,263 3.3%
Multi-Family 382 $313,104,944 $6,555,839 8.1%
Public/Institutional 554 $15,882,146 $353,886 0.4%
Railroad 1 $0 $0 0.0%
Recreation/Conservation 583 $6,955,173 $127,088 0.2%
Single Family, 1 to 2.5 acres 12 $2,879,253 $31,864 0.0%
Single Family, 14,000 to 43,559 s.£ 408 $283,720,369 $3,906,187 4.8%
Single Family, 8,000 to 13,999 s.f. 912 $882,953,349 $12,657,625 15.6%
Single Family; Less than 8,000 s.f. 2,204 $2,858,545,942 $41,577,753 51.1%
Single Family, More than one unit per acre 1 $808,490 $15,793 0.0%
Transportation, Utility, Communication 55 $12,916,575 $243,706 0.3%
Vacant 43 $17,727.784 $310,851 0.4%
TOTAL 5,698 $5,212,960,413 $81,334,432 100.0%

 

Commerce Township, 3 growing suburban community rich in natural amenities,
demonstrates a diverse range of land uses. “Table 4. Charter Township of Commerce
Land Use and Property Value Patterns,” demonstrate that high density residential land
uses account for over 60% of the township’s tax base while commercial and industrial
land uses account for a combined 20% of the community’s tax base. Commerce
Township’s total property tax base is over $2 million greater than the city of Pontiac and

collects the highest total property tax of all case study townships.

l3

Table. 4 Charter Township of Commerce Land Use and Property Value Patterns

 

 

 

Estimated True Cash Property Tax % Total
Land Use Acres Value Revenue Revenue
Agricultural 209 $4,939,139 $31,684 0.1%
Commercial 495 $386,959,092 $6,667,902 12.7%
Extractive 303 $5,284,314 $97,588 0.2%
Industrial 563 $265,848,691 $4,788,735 9.1%
Mobile Home Park 209 $25,275,747 $542,138 1.0%
Multi-Family 177 $19,226,350 $394,256 0.8%
Public/Institutional 554 $9,751,557 $180,344 0.3%
Recreation/Conservation 4,881 $79,544,484 $614,407 1 .2%
Single Family, 1 to 2.5 acres 982 $212,487,610 $2,575,319 4.9%
Single Family, 10 acres or greater 331 $13,156,811 $129,224 0.2%
Single Family, 14,000 to 43,559 s.f. 2,538 $1,629,236,145 $19,805,773 37.8%
Single Family, 2.5 to 4.9 acres 400 $36,131,729 $422,133 0.8%
Single Family, 5 to 9.9 acres 322 $20,815,832 $212,748 0.4%
Single Family, 8,000 to 13,999 s.f. 1,128 $1,023,507,435 $12,162,154 23.2%
Single Family, Less than 8,000 s.f. 121 $155,654,543 $1,620,374 3.1%
Single Family, More than one unit per parcel 222 $4,210,280 $55,654 0.1 %
Transportation, Utility, Communication 152 $4,977,816 $73,256 0.1%
Vacant 2,213 $197,073,090 $1,970,915 3.8%
Water 70 $5,000 $4 0.0%
Undeﬁned 18 $0 $0 0.0%
TOTAL 15,888 $4,094,085.665 $52,344,607 100%

 

The Village of Wolverine Lake is a small community located adjacent to

Commerce Township. “Table 5. Village of Wolverine Lake Land Use and Property Value

Patterns” demonstrate that over 90% of the community’s property tax base consists of

residential uses.

Table 5. Village of Wolverine Lake Land Use and Property Value Patterns

 

 

 

 

Estimated True Cash Property Tax % Total
Land Use Acres Value Revenue Revenue
Commercial 21 $8,800,729 $194,613 3.2%
Multi-Family 40 $7,138,015 $154,393 2.6%
Public/Institutional 1 1 $0 $0 0.0%
Recreation/Conservation 75 $567,980 $6,149 0.1%
Single Family, 1 to 2.5 acres 11 $3,353,645 $43,927 0.7%
Single Family; 14,000 to 43,559 s.f. 220 $122,391,615 $1,774.781 29.4%
Single Family, 2.5 to 4.9 acres 3 $1,023,275 $9,869 0.2%
Single Family, 8,000 to 13,999 s.f. 257 $192,925,238 $2,862,887 47.4%
Single Family; Less than 8,000 s.f. 47 $60,257,906 $873,277 14.4%
Vacant 137 $9,836.61 1 $123,970 2.1%
Undeﬁned 0 $0 $0 0.0%
TOTAL 822 $406,295,014 $6,043,867 100%

 

14

Milford Township is a rural but suburbanizing community located on the western
edge of Oakland County. Development in the township consists primarily of medium
and low density residential land uses shown in “Table 6. Milford Charter Township Land
Use and Property Value Patterns.” Single family land uses on lots 1 to 5 acres represent
55 percent of the property tax base. Recreational and conservation land use and well as

industrial uses account for a signiﬁcant portion of land use.

Table 6. Milford Charter Township Land Use and Property Value Patterns

 

 

 

 

Estimated True Cash Property Tax % Total
Land Use Acres Value Revenue Revenue
Agricultural 189 $4,015,462 $12,731 0.1%
Cormnereial 80 $18,362,887 $304,590 1.5%
Extractive 436 $5,055,324 $42,524 0.2%
Industrial 1,922 $197,036,148 $3,951,331 18.9%
Mobile Home Park 154 $15,132,067 $309,681 1.5%
Multi-Family 6 $507,147 $10,949 0.1 %
Public/Institutional 89 $3,855,441 $80,707 0.4%
Recreation/Conservation 5,729 $22,404,960 $84,650 0.4%
Single Family, 1 to 2.5 acres 2,363 $462,258,536 $6,676,398 31.9%
Single Family, 10 acres or greater 1,008 $29,421,787 $333,683 1.6%
Single Family, 14.000 to 43,559 s.f. 223 $94,956,008 $1,375,983 6.6%
Single Family, 2.5 to 4.9 acres 3,099 $344,963,753 $4,875,107 23.3%
Single Family, 5 to 9.9 acres 1,636 $88,304,276 $1,170,379 5.6%
Single Family, 8,000 to 13,999 s.f. 34 $27,501,182 $393,043 1.9%
Single Family, Less than 8,000 s.£ 9 $8,534,547 $109,338 0.5%
Transportation, Utility, Communication 192 $3,888,133 $54,489 0.3%
Vacant 2,877 $97,208,947 $1,129,663 5.4%
Undeﬁned 1 $60,375 $0 0.0%
TOTAL 20,049 $1,423,466,980 $20,915,248 100%

 

The Village of Milford is a mature community located in the heart of Milford
Charter Township. “Table 7. Village of Milford Land Use and Property Value Patterns”
demonstrates that Milford Village is the urban hub of Milford. Commercial land uses
account for nearly a quarter of Milford Village’s tax base compared to only 1.5 percent of
Milford Township’s tax base. The majority of residential land uses are on lots ranging

from 8,000 square feet to one acre in size revealing a much denser residential

15

development pattern than Milford Township. Overall, residential uses account for 60

percent of the Village’s total tax base.

Table 7. Village of Milford Land Use and Property Value Patterns, 2007

 

 

 

Estimated True Cash Property Tax % Total
Land Use Acres Value Revenue Revenue
Commercial 91 $113,981,026 $2,038,583 22.7%
Industrial 23 $5,505,062 $114,361 1.3%
Multi-Family 142 $34,422,720 $419,729 4.7%
Public/Institutional 121 $1,005,200 $24,626 0.3%
Recreation/Conservation 1 73 $72,960 $545 0.0%
Single Family; More than one unit per 0 $123,792 $2,235 0.0%
parcel
Single Family; Less than 8,000 s.f. 43 $47,498,046 $760,539 8.5%
Single Family; 8,000 to 13,999 s.f. 246 $210,047,380 $3,314,595 36.9%
Single Family; 14,000 to 43,559 s.f. 210 $111,664,932 $1,768,763 19.7%
Single Family; 1 to 2.5 acres 70 $16,890,362 $255,880 2.8%
Single Family; 2.5 to 4.9 acres 27 $2,554,979 $35,917 0.4%
Single Family; 5 to 9.9 acres 25 $901,447 $15,973 0.2%
Single Family; 10 acres or greater 22 $654,513 $11,812 0.1%
Transportation, Utility, Communication 2 $208,892 $2,859 0.0%
Vacant 166 $16,820,836 $228,117 2.5%
Water 2 $0 $0 0.0%
TOTAL 1,365 $562,352,147 $8,994,534 100.0%

 

Oakland Charter Township is an afﬂuent and growing rural community located on

the eastern border of Oakland County. Nearly half of the tax base and over $1.2 billion of

estimated property value exists in residential properties of one quarter to one acre lots.

The development market in Oakland Township is driven by a demand for high quality,

low density residential development.

16

Table 8. Charter Township of Oakland Land Use and Property Value Patterns, 2007

 

 

 

Estimated True Cash Property Tax % Total
Land Use Acres Value Revenue Revenue
Agricultural 257 $9,793 ,420 $23,818 0. 1%
Commercial 79 $34,972,013 $709,567 2. 1%
Industrial 4 $2,719,694 $61,507 0.2%
Mobile Home Park 87 $8,228,109 $185,980 0.5%
Multi-Family 88 $149,670 $2,038 0. 0%
Public/Institutional 329 $2,678,757 $31,595 0. 1%
Recreation/Conservation 5494 $77,700,543 $793,950 2.3%
Single Family, More than one unit per
parcel 75 $955,039 $13,003 0.0%
Single Family; Less than 8,000 5.1: 33 $47,51 1,296 $630,927 1.8%
Single Family, 8,000 to 13,999 s.f. 83 $108,426,743 $1,459,837 4.2%
Single Family; 14,000 to 43,559 s.f. 1665 $1,261,753,077 $16,735,049 48.7%
Single Family, 1 to 2.5 acres 1621 $418,351,320 $5,305,177 15.4%
Single Family; 2.5 to 4.9 acres 1908 $268,934,522 $3,291,668 9.6%
Single Family, 5 to 9.9 acres 1753 $127,335,950 $1,491,606 4.3%
Single Family; 10 acres or greater 3244 $94,327,957 $1,043,030 3.0%
Transportation, Utility, Communication 57 $977,212 $19,186 0.1%
Vacant 5245 $202,153,350 $2,591,626 7.5%
Undeﬁned 1 $0 $0 0.0%
TOTAL 22,022 $2,666,968,672 $34,389,566 100%

 

2.2 Measuring Assessed True Cash Value

Assessors employ a speciﬁc set of economic principles and legally established

methods to determine TCV for any given property. There are three common methods

used to determine the value of property in the appraisal profession: the cost approach,

sale comparison approach and the income approach (Eckert et al., 1990). In Michigan,

assessors primarily use cost and sales comparison approaches to appraise real property.

The cost approach to assessment serves as the starting point for determining TCV. The

cost approach, sometimes called the “summation approach,” consists of adding together

land value and improvement value. The International Association of Assessing Officers

(1990) state the formula for the cost approach as:

V=LV+TV

l7

Where V is the market value, LV is the land value and IV is the depreciated
improvement value. Improvement value is deﬁned as the cost new (CN) of improvements
less depreciation (D), resulting in a ﬁnal equation of.

V = LV + (CN — D)

The cost approach to value is based on the economic principle of substitution
(Eckert et al., 1990). The principle of substitution assumes that an informed buyer will
pay no more for a property than it would cost to create an exact replica of the property
with the same utility, amenities and function (Michigan State Assessors Board [MSAB],
2002, p. 9-2). Here, the principle of substitution draws an important distinction between
replacement cost and reproduction cost, which is fundamental to assessing practice in
Michigan.

The State Tax Commission recommends that reproduction cost be used as the
starting point for assessing the “cost new” of building improvements within the
framework of the cost approach (MSAB, 2002, 9-6). The Michigan State Assessors
Board (2002) deﬁnes reproduction cost as, “the cost of construction at current prices of
an exact duplicate or replica using the same materials, construction standards, design,
layout, and quality of workmanship, embodying all the deﬁciencies, superadequacies, and
obsolescence of the subject building” (9-6). This differs ﬁ'om replacement cost which is,
“the cost of construction at current prices of a building having utility equivalent to the
building being appraised but built with modern materials and according to current
standards, design, and layout” (9-6).

Michigan assessing departments utilize standard “cost schedules” and

“depreciation rates” as the basis for estimating initial improvement value (MSAB, 2002,

18

p.9-3). Assessors apply standard cost schedules adopted by the Michigan State Tax
Commission to estimate the reproduction cost (or cost new) of building improvements.
Cost schedules specify cost “multipliers” for speciﬁc building type, characteristics and
class which are applied to comparable properties being assessed. Adjustments are then
made by the assessor to account for various aspects of design and workmanship in
construction.

Standard depreciation rates are applied to the reproduction cost of the
improvement to account for the loss of value and utility attributed to the age of the
structure. The result is an improvement value that represents cost new less depreciation.
Standard depreciation rates are adopted by the Michigan State Tax Commission for non-
residential and residential structures. Commercial and industrial structures depreciate at a
rate of 1.5 percent to 4 percent each year depending on the quality of construction
(MSAB, 2002, p. 9-13). All residential structures depreciate at a constant rate of one
percent each year all else constant. (MSAB, 2002, p. 9-56) Along with depreciation rates,
the State Tax Commission also deﬁnes an “effective life” for improvements. The
eﬁ‘ective life of a building improvement is reached when it no longer physically
depreciates as long it is used. The typical effective life of a residential building is 55
years. The effective life of commercial and industrial buildings can range from 60 to 25
years depending on the quality of materials used in construction (MSAB, 2002, pp.9-15,
9-56). It should be noted that in addition to physical depreciation from natural aging,
functional and economic obsolesce can accelerate depreciation while additions made to

the property in some cases can counteract the assessment of depreciation.

19

The cost approach and the application of standard cost schedules and depreciation
rates have two primary beneﬁts to the practice of assessment: eﬁiciency and equity. The
Michigan State Assessors Board (2002) states, “the cost approach is well adapted to a
mass appraisal environment, in which many properties must be appraised together, and is
currently the most widely utilized of the three approaches to value”(p. 4-15). In mass
appraisal, where efﬁciency and accuracy are the primary objectives, cost schedules and
depreciation rates have been integrated within computerized BS&A assessing software,
allowing for the automated application of cost multipliers and depreciation rates based on
the building attributes and class. The cost new, less depreciation values computed by
assessing software are commonly referred to as the “manual value” in the property
appraisal profession

To adequately assess building improvements based on their “true cash value,”
Michigan assessors adjust the cost new less depreciation to account for local real estate
market conditions using an Economic Condition Factor (ECF). The ECF for a given
property is based on recent sales of comparable properties. The Michigan State Assessors
Board (2002) speciﬁes the equation for determining an ECF as:

ECF = (SV/(CN-D))

The ECF for a recently sold property is the ratio of improvement sale value (SV)
to improvement manual value (CN-D). Economic Condition Factors are calculated for all
arm’s length transactions in a county and are then applied to comparable properties to
adjust the manual value to a more accurate assessed market value. Calculations of ECFs
require numerous recent sales that occur during the same time period to ensure accuracy

(MSAB, 2002, p.9—27). Assessors break down their community into various

20

“neighborhoods” of comparable properties based on similar economic and physical
characteristics, which helps guide application of ECFs to appropriate comparables.

The Michigan State Assessors Board states (2002) that, “care must sometimes be
taken to limit the size of the ECF area so as not to encompass neighborhoods which do
not belong together. This could lead to an ECF which properly values the ECF area as a
whole but incorrectly values the neighborhoods which make up the whole” (9-28). The
Michigan Assessor’s Manual does not establish a standardized practice for determining
ECF neighborhood boundaries. Often times ECFs can be spread across an entire county,
in order to ﬁnd comparables for unique properties.

Based on the application of the sales comparison and cost approach to property
assessment practice in Michigan, the equation for calculating the true cash value of a
property can be alternatively deﬁned as:

TCV = LV + ((CN-D) * ECF)
2.4 Analysis of the True Cash Value of Building Improvements in Case Study
Communities

Property appraisal methods used to estimate the assessed true cash value of
building improvements maintains important implications for a community’s bottom line
that are driven by the quality and age of current building stock as well as local real estate
market conditions. Tables 9 through 11 examine the determinants of assessed true cash
improvement value (referred to as building value in tables) in case study communities.
For each land use classiﬁcation in case study communities, the average estimated true
cash value, manual value, age and economic condition are presented for each

construction class. Construction class is based on a scale of A through D with A

21

representing the highest construction quality and D representing the lowest quality.
Tables highlight key similarities and differences between the quality and age of building
stock as well as the real estate market conditions that exist in cities, villages and
townships.

Table 9. Basic Determinants of City Assessed Residential Building Values, 2007

 

 

 

Community Residential Residential Average Average Aveage Average
Class Parcels Estimated TCV Manual Value Age ECF

Pontiac -
A 2 $523,892 $516,642 48 1.01
B 48 $266,262 $206,506 43 1 .23
BC 300 $188,002 $191,889 20 0.98
C 5148 $108,679 $97,293 31 1.13
CD 1 1642 $66,572 $55,823 39 1.19
D 476 $39,941 $34,707 45 1.15

Royal Oak
A 1 $968,982 $328,468 57 2.95
B 32 $492,306 $212,169 42 2.39
BC 83 $455,160 $212,335 26 2.22
C 19812 $148,124 $71,149 46 2.13
CD 815 $107,567 $51,734 49 2.17
D 50 $78,895 $31,740 57 2.47

 

“Table 9. Basic Determinants of City Assessed Residential Building Values,”
reveal that major differences exist in current residential real estate markets in the City
Royal Oak and Pontiac, highlighted by Economic Condition Factors (ECF). While older
residential buildings in Royal Oak have experienced a slightly higher degree of
depreciation than those in Pontiac (shown by average manual values), single family
residences have sold at a much average higher price. Average residential ECFs in Royal
Oak are all well over 200 percent, demonstrating that residential selling prices are two
times greater than the cost new, less depreciation. The majority of residential buildings in
Pontiac are built in the CD construction classes compared to class C in Royal Oak,
therefore demonstrating that the City of Pontiac has a lower overall quality of building

stock than Royal Oak. While Pontiac maintains ECFs of over 100 percent in most cases,

22

real estate market conditions add only 13 to 19 percent to residential manual values
compared to 113 to 150 percent for the majority of homes in Royal Oak.

A breakdown of average manual values demonstrates that relatively similar
residential reproduction costs exist in both cities, when holding age constant.
Construction class appropriately demonstrates that higher quality construction results in
buildings that are more costly to reproduce.

“Table 10. Basic Determinants of Township and Village Assessed Residential
Building Values” demonstrate noticeable difference between the age of city, village and
township residential properties, especially in high quality construction classes.
Residential development built within the last ten years represents a signiﬁcant source of
value in case study townships and villages.

The average manual value in Commerce, Oakland and Milford Townships are
extremely high within A through BC construction classes, representing some of the most
valuable residential property in Oakland County. In the case of some of the higher priced
construction classes, residential properties have an average ECF of less than 100%,
demonstrating that homes are selling for less than their depreciated reproduction cost. In
cases where manual values are low to start with, which is the case in many older central
cities, ECFs less than 100 percent would cripple the property tax base, but in afﬂuent

communities it has a minimal effect.

23

Table 10. Basic Determinants of Township and Village Assessed Residential Building Values,

 

 

 

 

 

 

2007
Community ﬁsidential Residential Average Average Aveage Average
Class Parcels Estimated TCV Manual Value Age ECF
Commerce Township
A 29 $1,024,369 $825,772 10 1.25
B 2098 $324,487 $325,400 10 0.99
BC 2330 $222,288 $216,060 14 1.03
C 5666 $122,780 $111,912 26 1.09
CD 545 $51,181 $44,945 37 1.14
D 7 $30,087 $25,077 45 1.20
Wolverine Lake Village
A 5 $719,250 $621,273 14 1.16
B 53 $359,410 $295,036 18 1.22
BC 135 $233,248 $186,052 24 1.25
C 1271 $105,452 $97,668 28 1.07
CD 149 $54,805 $49,631 33 1.12
D 0 $0 $0 0 0
Milford Township
A 63 $893,526 $876,245 8 1.01
B 1208 $340,571 $346,309 12 0.97
BC 552 $206,128 $203,632 20 0.99
C 958 $137,925 $129,457 25 1.05
CD 105 $78,833 $61,668 37 1.27
D 60 $52,759 $34,394 43 1.54
Milford Village
A 1 $912,190 $760,158 3 1.20
B 218 $276,506 $299,853 13 0.93
BC 220 $190,852 $200,279 15 0.99
C 1113 $117,616 $92,564 32 1.30
CD 217 $82,177 $59,862 37 1.39
D 23 $56,388 $35,315 44 1.60
Oakland Township
A 557 $852,308 $888,513 10 0.94
B 2333 $354,043 $401,252 , 11 0.88
BC 726 $205,478 $21 1,874 21 0.97
C 1398 $143,047 $145,078 26 0.98
CD 110 $65,353 $66,416 40 0.98
D 27 $52,251 $40,914 43 1.24

 

Michigan’s method of estimating the true cash value of building improvements,

which emphasizes the role of age and quality, have less pronounced impacts on non-

residential uses when compared to residential uses. “Table 11. Basic Determinants of

Assessed Commercial, Industrial and Multi-Family Building Values” demonstrate the

24

ﬁscal implications of building class, age, and economic conditions for multi-family,

commercial and industrial uses in the case study communities.

Table 11. Basic Determinants of Assessed Commercial, Industrial and Multi-Family Building
Values in Case Study Communities, 2007

 

 

 

 

 

ﬁling Number of Average Average Aveage Average
Class Parcels Estimated TCV Manual Value Age ECF
Commercial
A 45 $2,445,340 $2,211,322 14 1.18
B 9 $2,656,288 $2,282,402 48 1.16
C 1649 $593,849 $523,061 48 1.16
D 320 $443,387 $395,783 54 1.13
User Deﬁned 35 $363,554 $332,343 47 1.12
Industrial
A 6 $992,394 $1,126,235 13 0.94
B 1 $526,562 $626,860 6 0.84
C 494 $1,686,034 $1,803,117 33 0.98
D 20 $505,916 $537,821 27 1.01
User Deﬁned 78 $833,744 $861,585 20 1.04
Multi-Family
A 2 $3,522,120 $3,518,964 5 1.16
B 25 $2,305,475 $2,389,490 10 0.98
BC 19 $195,815 $152,079 5 1.32
C 23 1 $496,373 $429,246 29 1.16
CD 1 $129,501 $98,480 47 1.31
D 445 $959,601 $842,946 62 1.17
User Deﬁned 36 $1,353,555 $1,145,222 41 1.18

 

Economic conditions seem to be supporting a relatively healthy market for multi-

family and commercial uses, less so for industrial uses. Much of the industrial

development in the case study communities is aging past 20 years, while the majority

commercial and multi-family residential development ranges between 30 and 60 years

old. The age and depreciation of non-residential buildings puts central cities like Pontiac

and Royal Oak, which rely heavily on a non-residential property tax base, at a distinct

disadvantage. Commercial and industrial depreciation rates adopted by the Michigan Tax

Commission shown in Table 12 suggest that much of the non-residential development in

the case study communities is reaching or is past its effective life.

25

Table 12. Commercial and Industrial Depreciation Rates

 

Annual Life
Building Characteristics Depreciation Expectangy
Heavy reinforced concrete or heavy steel
ﬁame with masonry 1.5% 60-65
Reinforced concrete, steel or wood frame with
masonry or wood walls 2.0% 50
Light steel or wood frame with masonry or
wood walls 2.5% 40
Inexpensive retail structures and low-cost
storage buildings 3.0% 30-35
Shedg low-cost hangars and utility build'ﬂs 4.0% 20-25

 

In the absence of new commercial and industrial development, case study cities
are forced to rehabilitate or redevelop physically obsolete buildings to combat a
collectively aging built environment and depreciating property tax base.
2.5 Land Value in Case Study Communities

In the analysis of TCV little has been said about the value of land and for good
reason. The Michigan State Assessors Board provides very little guidance on the
determination of land value. Land values are usually determined by comparable vacant
land sales and are inﬂuenced by several factors, such as location, economic conditions,
physical differences, and restrictions such as zoning and building codes (MSAB, 2002, p.
7-1). Two factors in the valuation of land maintain important ﬁscal implications to
Michigan communities. First, land does not physically depreciate therefore land values
usually remain stable or increase. This ﬁnding suggests that land preservation initiatives
can play an important role in maintaining ﬁscal stability as long as vacant land sales do
not signiﬁcantly decrease. Second and perhaps most germane to the planner, is the role of

highest and best use in the valuation of land.

26

Highest and Best Use (HBU) is a fundamental starting point in the assessment of
land value (MSAB, 2002, p. 3-17). HBU usually coincides with current use, which
supports equitable and legally cautious assessment practices. Appendix A. provides
detailed tables on vacant land values in all case study communities. An example of
vacant land values in two cities is included here in Table 13 to illustrate the concept of
land valuation and the ﬁscal implications of zoning practices.

Table 13. Vacant City Land Values Based on Highest and Best Use, 2007
Total Land Average Average Total Tax

 

 

 

 

 

CLASS Acres Value $/Acre ﬂax/Parcel Revenue
Pontiac
Commercial Improved 19.8 $2,641,221 $106,597 $242,971 $485,941
Commercial Vacant 197.7 $15,127,195 $147,170 $1,069 $256,529
Apartment Improved 0.9 $78,326 $87,119 $1 1,076 $11,076
Apartment Vacant 71.7 $4,339,012 $98,797 $405 $29,185
Industrial Vacant 251.9 $17,544,204 $107,679 $1,598 $201,394
Residential Improved 8.3 $436,000 $61,179 $1,701 $69,754
Residential Vacant 630.7 $39,867,674 $89,384 $207 $598,513
Lake Vacant 1.0 $12,000 $12,021 $315 $315
Condo Improved 2.6 $533,200 $216,235 $2,975 $35,698
Condo Vacant 107.0 $21,690,088 $310,969 $429 $214,847
TOTAL 1291.6 $102,268,920 $79,] 78 $490 8 1,903,255
Royal Oak

Corrnnercial Vacant 6.0 $2,402,675 $422,077 $5,441 $54,41 1
Apartment Vacant 0.1 $17,520 $174,235 $287 $287
Industrial Vacant 0.5 $90,833 $166,785 $965 $1,929
Utility Vacant 0.1 $15,318 $205,340 $386 $386
Residential Improved 3 .3 $930,515 $286,762 $2,792 $58,624
Residential Vacant 20.8 $5,619,254 $272,617 $1,498 $155,760
Condo Improved 2.2 $525,000 $319,338 $3,079 $21,554
Condo Vacant 2.3 $711,530 $306,859 $1,627 $17,899
TOTAL 35.2 $10,312,645 $292,704 $1,980 $310,851

 

The City of Pontiac contains over 1,291 acres of vacant land compared to only 35
acres in the City of Royal Oak. Classiﬁcations of vacant land identify properties with
vacant building improvements as well as those that do not include any improvements. In
the City of Royal Oak, a smaller supply of available land, coupled with an assumed

higher demand, create higher vacant land values for commercial, residential and condo

27

uses when compared to the City of Pontiac. Zoning regulations, which determine the

ﬁinction of property, clearly maintain ﬁscal consequences. On average, commercial and

industrial land values per acre are greater than residential land values. Higher density

residential zoning that creates HBUs of multi-family and condo uses are also associated

with higher land values than less dense residential land values in cities.

2.6 Composition of the Property Tax Base in Case Study Communities

In Michigan, property tax assessment is based on a measure of true cash value.

Table 14 and 15 present the three basic components of true cash value; improvement

values, which include buildings and yard improvements and land values for case study

cities and townships and villages.

Table 14. Basic Components of Property Tax Revenue in Case Study Cities, 2007

 

 

 

 

Wding Yard

Improvement Tax Total Land Improvement Tax Total Tax
Land Use Revenue Tax Revenue Revenue Revenue
Commercial $16,028,331 $6,559,527 $468,290 522.9833 63
Industrial $10,575,485 $2,145,481 $531,326 $13,252,293
Mobile Home Park $182,864 $48,047 $1.261 $182,864
Multi-Family $8,538,240 $1,305,702 $96,194 $9,678,102
Public/ Institutional $809,284 $137,972 $3,892 $951,148
Railroad $0 $13,351 $58 $13,408
Recreation/Conservation $100,967 $30,749 $1,549 513 3 .265
Single Family; More than one unit per $41,315 $125,719 S76 $167.11]
acre
Single Family: Less than 8,000 s.f. $45,567,828 $12,243,017 $76,969 $57,890,538
Single Family. 8.000 to 13,999 s.f. $13,514,955 $4,260,158 $39,858 $17,814.97]
Single Family; 14,000 to 43.559 s.f. $4,158,914 $1,317,181 $25,323 $5,501,418
Single Family, 1 to 2.5 acres $170,127 $53,097 $1,439 $224,663
Single Family; 2.5 to 4.9 acres $25,842 $14,151 $225 $40,218
Single Family. 5 to 9.9 acres $2,559 $1,515 556 $4,130
Transportation, Utility. $265,250 $203,043 $27,156 $504,403
Communication
Vacant $615,924 $1,581,346 $16,835 $2,214,105
Water $0 $53 $0 $53
TOTAL $100,597,885 $30,040,108 $1,290,507 $131,556,053

 

Table 15. Basic Components of Property Tax Revenue in Case Study Townships and Villages, 2007

 

 

 

 

Wilding Yard

Improvement Tax Total Land Improvement Tax Total Tax
Land Use Value Tax Revenue Revenue Revenue
Agricultural $7,506 $60,673 $55 $68,233
Commercial $6,451,314 $3,155,984 $307,954 $9,915,255
Extractive $1.764 $138,276 $71 $140,111
Industrial $6,759,545 $1,885,296 $271,097 $8,915,934
Mobile Home Park $1,037,799 $361,463 $495 $1,037,799
Multi-Farnily $842,884 $203,402 $8,449 $981 .365
Public/Institutional $226,471 $87,507 $3,294 $317,272
Recreation/Conservation $528,959 $961,436 $9,307 $1,499,702
Single Family: More than one unit per
parcel $52,567 $18,250 $77 $70,892
Single Family; Less than 8,000 s.f. $2,254,698 $1,730,235 $9,518 $3,994,455
Single Family. 8.000 to 13,999 s.f. $12,697,098 $7,431,058 $64,350 $20,192,516
Single Family. 14,000 to 43.559 s.f. $28,784,497 $12,427,968 $234,719 $41,460,350
Single Family; 1 to 2.5 acres $10,425,224 $4,311,990 $1 19,485 $14,856,701
Single Family, 2.5 to 4.9 acres $5,829,239 $2,731,091 $74,359 $8,634,694
Single Family, 5 to 9.9 acres $1,776,236 $1,090,058 $24,416 $2,890,707
Single Family; 10 acres or greater $754,146 $753,443 $10,162 $1,517,750
Transportation, Utility,
Communication $23,547 $125,333 $909 $149,789
Vacant $413,560 $5,581,034 $49,698 $6,044,291
Water $0 $4 $0 $4
TOTAL $78,867,054 $43,054,502 $1,188,415 $122,687,822

 

Building improvements account for a greater share of the property tax base in

cities (76%) than in townships (63%), demonstrating that density and build-out play an

important role in shaping the property tax base. Land values account for 10% more of the

property tax base in townships and villages when compared to case study cities. Land

values are not inﬂuenced by depreciation, therefore the larger share of the property tax

base accounted for by land values may help stabilize revenues in rural and developing

communities.

Cities maintain a more diversiﬁed tax base compared to townships which rely

primarily on residential land uses as a source of property tax revenue. Single family

residential land uses account for 76% of the total tax base in townships and villages

29

compared to 62% in cities. Commercial and industrial uses account for a combined
27.5% of the total property total tax base in cities compared to 15% in townships and
villages. Multi-family uses account for 7.4% of the cities property tax base but less than
one percent in townships and villages.

2.7 Fiscal Implications of Michigan’s Property Tax Equation

Michigan’s property tax equation highlights four driving factors that affect
property tax revenues produced by different land uses; construction quality, building age
building size and market conditions. The quality of construction serves as the foundation
for determining improvement value through the speciﬁcation of appropriate cost
schedules and depreciation rates. A high quality building stock will add greater value to a
community and will depreciate at a slower rate (in the case of non-residential uses) than a
low quality building stock.

Due to the reproduction cost framework that guides the cost approach to
assessment Michigan, building size maintains a positive relationship with assessed
manual values and property tax revenues. Communities that zone for larger buildings or
have the real estate market to support such developments, receive higher tax
contributions per building improvement, especially those built to superior construction
quality.

Calculations of cost new, less depreciation, create a distinct ﬁscal advantage for
newer communities when compared to older communities, holding all else constant. As
communities age, real estate market conditions expressed through the economic condition
factor can play a vital role in mitigating or compounding the negative ﬁscal impacts of

building depreciation. Within a built-out community, the value of real estate transactions

3O

must increase at a rate equal to or greater than the rate of depreciation to maintain or
grow the property tax base of the current building stock. In the absence of market
transactions, improvement values and associated tax revenues would continue to decrease
at the rate of depreciation. In a worst case scenario, real estate sales that are less than the
estimated manual value (resulting in an ECF of less than one) would lower the estimated
true cash value of other comparable properties, creating a ripple effect that could decrease
a community’s property tax base signiﬁcantly.

2.8 Informing Hedonic Pricing Analysis of Indirect Fiscal Impacts through
Applications of Michigan Property Tax Equation

Upon review of Michigan’s property tax equation, it is determined that the
hedonic pricing function can adequately measure the implicit contributions of land use
and natural amenity patterns to assessed residential and non—residential property values.
However, to translate implicit values derived ﬁ'om assessed property values to a measure
of ﬁscal impact, the hedonic pricing function must be supplemented with other
computational techniques. “Figure 2. Methodological Framework” presents a diagram of
the analytical model that is used to measure indirect ﬁscal impacts. Each stage in the
model is justiﬁed through discussion of Michigan’s General Property Tax Act and
established assessor practice on subsequent pages. This introduction to the
methodological framework aims to lay the foundation for a more detailed methodological

discussion presented in Chapter 3.

31

Figure 2. Methodological Framework

 

 

 

 

 

ASE-age One; Define Dependent arid Independent Variables t9 bekMpeasuredL . ”j

 

 

 

Linear Multiple Regression Analysis for Linear Multiple Regression
Building Value Analysis for Land Value
Dependent Variable Dependent Variablg
Building Value (Cost new, less Land Value
depreciation adjusted by ECF)
Indepgdpnt Variables Independent Vgiaples
Building Characteristic Variables Lot Size
Adjacent Land Use Variables Adjacent Land Use Variables
Buffer Land Use Variables Buffer Land Use Variables
Socioeconomic Variables Socioeconomic Variables
Stage Two: Interpret Signiﬁcant Determinants of ASsessed Building and Land 4:.
.- . . .Value. .. .. . l
Multiple Regression Analysis (MRA) Interpretation & Adjustments
Analyze and improve model

Identify statistically signiﬁcant determinants of building and land values
Coefﬁcients = Implicit Price for independent variables in linear MRA models

 

‘ StageThree i’C’alcu‘late “Implicit” Tax Revenues Associated with Building and 1
Land Values for each Parcel

 

x. .

Implicit Building Value Calculations Implicit Land Value Calculations

Calculate implicit prices for building values Calculate implicit prices for land values
[3 Vi IL Vi

 

Calculate implicit tax revenues for building Calculate implicit tax revenues land value

value coefﬁcients (B VITRi) for each parcel x coefﬁcients (L VITRi) for each parcel x
BVITRix =IBVix *PTVx *Mx LVITRL'x =ILVILx *PTVx *Mx

Where IB Vix is the implicit building value Where ILVDC is the implicit land value for
for independent variable 1’, property x, independent variable i, property x
multiplied by the PTV percent taxable value multiplied by the PTV percent taxable

and the local property tax millage rate M for value and the local property tax millage
each parcel x. rate M for parcel x.

 

'- VStage Four: Summation of “Implicit” Tax Revenues for Building and Land Value]
1 .e w .. Models. w.

Summation of Implicit Tax Revenues
ZlTRix = BVIIRix + LVITRix

 

Aggregate total implicit tax revenues (ITR) are calculated for signiﬁcant independent
variables (i) for building and land values for all parcels (x)

32

In stage one and two, the hedonic pricing function can be used to decompose the
assessed true cash value of real property given three key assumptions. First, the measure
of “true cash value” is assumed to be an accurate measurement of a property’s market
value. Given legislative mandates set forth by Michigan’s General Property Tax Act and
applications of Economic Condition Factors to capture market trends, assessed true cash
value meets this criteria while allowing for accurate modeling of ﬁscal impacts in stages
three and four. The second assumption is that the assessed true cash value of building
improvements and land should be modeled separately to be consistent with assessor
practice. Based on review of the Michigan Tax Assessors Training Manual it is clear that
a unique set of attributes are considered when determining the value of land versus the
value of a land improvement; selection of independent variables aim to capture these
factors using the best data available. The third assumption is that independent variables
included in the hedonic model maintain a statically signiﬁcantly causal relationship with
assessed true cash value at the 95 percent conﬁdence interval and that only those
signiﬁcant variables are associated with indirect ﬁscal impacts.

Stages three and four, transform modeling outcomes measured in assessed true
cash value to a measure of taxable value ﬁom which property tax revenues are derived
and ﬁscal impacts can be estimated. Michigan General Property Tax speciﬁes that local
property taxes are paid based the “taxable value” of real property, which represents the
lesser of “state-equalized value” and “capped value.” Taxable value is not to exceed 50

percent of true cash value for a property, as equalizedl (Sec. 27a.1). This is the state

 

1 The true cash value of property “as equalized” refers to the equalization process applied by County Board
of Commissioners (CBC) and State Tax Commission (STC) which is used to ensure equitable and accurate

33

equalized value or SEV. The Michigan General Pr0perty Tax also establishes that, “the
taxable value of each parcel of property adjusted for additions and losses shall not
increase by more than the increase in the immediately preceding year in the general price
level, or 5 %, whichever is less until ownership is transferred. When full or part
ownership of the parcel of property is transferred as deﬁned by law, the parcel shall be
assessed at the applicable proportions of current state equalized value” (See. 2(a)-2(b),
MSAB, 2002, p.8-1). The Michigan State Assessors Board emphasizes that, “the
constitutional limit or cap on annual increases in taxable value is not applicable to
increases in assessed values, (only to taxable value). Assessors are required to annually
estimate the true cash value of each parcel and establish an assessment based upon that
estimate.” (MSAB, 2002, 8-2).

As discussed earlier in stages one and two, the hedonic model assumes that
assessed true cash value is the most complete and accurate value from which implicit
prices for spatial variables can be derived. The reasoning is that taxable value may differ
between two exact properties depending on when the property is sold or the rate of the
previous year’s increase, which therefore can create different ratios of taxable value to
assessed true cash value. Due to the fact that taxable value, not the assessed true cash
value, determines the amount of property taxes paid, two additional steps must be added
to the hedonic model to capture the ﬁscal impact of model outcomes shown in steps three
and four. In step three, each coefﬁcient is multiplied by the ratio of taxable value to

assessed true cash value for each parcel affected. This step transforms the implicit prices

 

assessment of true cash value. When the CBA and STC check annual tax rolls. each body determines if
properties are less or greater than 50% of true cash value, and is required to “equalize” inadequate
assessments by adding to or deducting from the valuations in order capture 50% of true cash value which
represents the true cash value of that class of property.

34

of each variable from a measure of assessed true cash value to taxable value, which is
then multiplied by the applicable millage rate for each parcel to derive the “implicit tax
revenue” associated with a hedonic modeling coefﬁcient. Step four, simply aggregates
the implicit tax revenues for any given signiﬁcant value determinant to ascertain the

indirect ﬁscal impact of that value determinant on a community wide basis.

35

CHAPTER 3
MEASURING THE INDIRECT FISCAL IMPACTS OF THE BUILT AND
NATURAL ENVIRONMENT

3.1 Measuring Indirect Fiscal Impacts through Hedonic Pricing

Planners have traditionally analyzed the ﬁscal impacts of development based on a
direct contribution of costs or revenues to the community where the development is
taking place (Burchell, 1978). The failure to account for the indirect ﬁscal implications of
amenity value and value interactions between land use patterns is a major limitation of
traditional ﬁscal analysis techniques (Kotval and Mullins, 2007). To quantify the indirect
ﬁscal impacts of land use patterns and natural amenities, a modiﬁed hedonic pricing
method is applied to the property tax base in case study communities. The hedonic
pricing method uses multiple regression analysis to quantify statistically signiﬁcant
relationships between assessed real property values and surrounding land use. as well as
natural and public amenity patterns, from which the indirect ﬁscal impacts are obtained.

Hedonic pricing is an econometric tool commonly used in non-market valuation
studies. Hedonic pricing techniques have been used to assign an implicit monetary value
to non-market goods such as parks and recreational opportunities (Kluvankova, 1998;
Pendleton, L. & Mendelsohn, R. 2000), air quality (Zabel & Keil 2000; Chattopadhyay,
1999), water quality (Boyle et. a1, 1999; Legget & Bockstael, 2000), wetlands (Mahan,
2000), view (Benson, 1998), transit (Cervero & Duncan, 2001) school quality and crime
(Dubin & Goodman, 1982), zoning (Asabere & Huffman, 1997) and land use patterns
(Bockstael, 1996; Geoghegan et al., 1997).

“The hedonic pricing method uses market transactions to determine the value of

characteristics associated with a product. By observing the price differential between two

36

variations in the product, the monetary trade-offs individuals are willing to make with
respect to those characteristics are indirectly inferr ” (Taylor, 2002 p. 332). Taylor
(2002, p. 340-360) identiﬁes six fundamental steps in estimating the hedonic pricing
function which include:

1. Deﬁne the value to be estimated

2. Select the dependent variable

3'. Select independent variables

4. Determine sample Me

5. Specify the functional form of the hedonic price ﬁmction

6. Address spatial dependence and correlation.

These steps are used to frame mythological assumptions and techniques applied
within this study.

(1) Deﬁne the value to be estimated. The hedonic model will quantify the
“implicit value” of surrounding built and natural land use patterns by decomposing
monetary variations in assessed true cash value. Hedonic pricing ﬁrnctions are estimated
for single and multi-family residential, commercial and industrial land uses. Implicit
values for each hedonic pricing function must be translated to a measure of “implicit
property tax revenue.” To obtain implicit property tax revenues, statistically signiﬁcant
implicit value coefﬁcients are multiplied by the assessed to taxable value ratio and the
local millage rate for each parcel. Parcel records are then aggregated to ascertain net
implicit property tax revenue.

(2) Selecting the dependent variable. Assessed building and land values serve as

the dependent variable in the hedonic pricing function. Consistent with assessment

37

practices, the hedonic pricing functions for building improvements and land are estimated
separately and then combined to create a cumulative estimate of implicit value. Assessed
“true cas ” building and land values, which are the basis for property tax assessment in
Michigan, are used instead of market transactions. Assessed building and land values
allow for accurate quantiﬁcation of “implicit property tax revenues” associated with
independent variables.

The use of assessed values rather than market transactions represents a
fundamental departure from traditional hedonic pricing methods. Assessed values for all
parcel records are used, therefore eliminating sample selection biases, a ﬁequently cited
shortcoming of sales transaction data (Taylor, 2002, p. 343). The primary reasoning for
using sales transactions instead of assessed value is founded on assessor bias. Assessed
value may not reﬂect the value that a given property owner places on the value of a
property’s characteristics. This limitation is addressed in two ways.

First, the examination of assessor methods detailed in Chapter Two identify
methodological issues and assumptions applied in the calculation of assessed TCV which
inform accurate hedonic pricing techniques using assessed true cash building and land
values. Secondly, a rich set of assessor data is supplemented using other key
observational characteristics, which allow for a more robust determination of factors that
may inﬂuence assessed building and land values.

(3) Selecting independent variables. Independent variables included in this
hedonic pricing study include two types of independent variables: property characteristics
and location characteristics. “Table 16. Description of Property Variables Used in

Regression Analysis,” identiﬁes independent variables that describe property

38

characteristics. Property variables for case study communities were derived using

Oakland County assessment data provided by the Oakland County Equalization

Department.

 

Table 16. Description of Property Variables Used in lggression Analysis

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Variable Variable Description
Basic Building Characteristics
ACRES" Lot size
AGE* Effective age of a building
FLRAREA" Square feet of ﬂoor area
Property class based on quality of construction
CLASS“ 1 ifD; 1.5 ifCD; 2 ifC; 2.5 ifBC; 3 ifB; 3.5 ifAB; 4 ifA
NOBLD Number of residential buildings per lot
RESFULLB Number of full baths
RESHLFB Number of half baths
RESNOBED Number of bedrooms
RESGRGAREA Square feet of garage area
RESGARTYP Number of car garages
Building Exterior
RESEXTl Aluminum,vinyl,aluminum,vinyl; 1 if yes, 0 if no
RESEXTZ asbestos; 1 if yes, 0 if no
RESEXT3 Block; 1 if yes, 0 if no
RESEXT4 Brick,brick/siding,brickcrete; 1 if yes, 0 if no
RESEXTS Composition; 1 if yes, 0 if no
RESEXT6 Logpine/cedar, wood; 1 if yes, 0 if no
RESEXT7 Stone, stone/siding; 1 if yes, 0 if no
RESEXTS Stucco; 1 ﬁles, 0 if no
RESEXT9 Asphalt, dryvit, lap siding. masonite, metal; 1 if yes, 0 if no
BuildingSgrle
RESSTYLEI Bi-level; 1 if yes, 0 if no
RESSTYLE2 Bungalow; 1 if yes, 0 if no
RESSTYLE3 Cape cod; 1 if yes, 0 if no
RESSTYLE4 Colonial/2; 1 if yes, 0 if no
RESSTYLES Contemporary; 1 if yes, 0 if no
RESSTYLE6 Log 1 if yes. 0 if no
RESSTYLE7 Multi-tenant; 1 if yes, 0 if no
RESSTYLES Other; 1 if yes, 0 if no
RESSTYLE9 Raised ranch; 1 if yes, 0 if no
RESSTYLEIO Ranch; 1 if yes, 0 if no
RESSTYLEll Single family; 1 if yes, 0 if no
RESSTYLEIZ TriLevel/qudor; 1 if yes, 0 if no

 

 

39

 

 

Table 16 cont. Description of Property Variables Used in Regression Analysis

 

RESSTYLE13 Tudor, 1 if yes, 0 if no

 

RESSTYLE14 Townhouse/duplex; 1 if yes, 0 if no

 

 

RESSTYLEIS Mobile/module; 1 1 if yes, 0 if no

 

 

 

Three types of spatial variables are incorporated in the regression model to
document the “advantages and disadvantages of location.” The three typologies of
location-based variables include: socio-economic characteristics, adjacent land use
development patterns and surrounding land use development patterns. The ﬁrst set of
characteristics examines the impact of socioeconomic characteristics on assessed value.
The vast majority of existing hedonic studies include information on race, income, and in
some cases education. Examples from existing literature indicates that increases in
median income maintain a positive relationship with non-residential property values
(Bollinger et. a1, 1998), educational attainment increases residential property values
(Asabere & Huffman, 1997) and that racial diversity may reduce the selling price of
residential homes (Cervero & Duncan, 2004).

Table 17. Social Characteristics of Case Study Communities, 2000

 

 

 

Race Educational Attainment
Percent Non- High Percent Percent Master's
Community White White School Bachelor's or Greater
City of Pontiac 39% 61% 33% 7% 3%
City of Royal Oak 95% 5% 23% 26% 4%
Charter Township of Commerce 97% 3% 24% 23% 6%
Village of Wolverine Lake 98% 2% 30% 17% 6%
Charter Township of Milford 97% 3% 24% 20% l 1%
Village of Milford 97% 3% 27% 22% 9%
Charter Township of Oakland 94% 6% 18% 29% 19%

 

Table 17 shows that case study communities are relatively homogenous with
respect to race, with the exception of the City of Pontiac. Educational attainment is

highest in Milford and Oakland Townships, with over 10 percent of the population

40

maintain masters or professional degrees and over 20 percent of the population four-year
bachelor’s degrees. While educational attainment is evenly represented in case study
townships and villages, the City of Royal Oak has a population with a much higher
educational attainment than that of Pontiac. Table 18 demonstrates that wealth is much
higher on average in case study townships and villages than in cities. The City of Pontiac
has the lowest median household income of just over $31,000 which is over $20,000 less
than the City of Royal Oak which has the second lowest median household income.

Table 18. Housing and Income Characteristics in Case Study Communities, 2000
Housing and Income Characteristics

 

 

Housing Percent Percent Median
Community Units Own Rent Household Income
City of Pontiac 24,234 92% 8% $31,207
City of Royal Oak 28,880 96% 4% $52,252
Charter Township of Cormnerce 12,361 96% 4% $72,702
Village of Wolverine Lake 1,683 97% 3% $65,682
Charter Township of Milford 5,470 97% 3% $67,672
Village of Milford 2,435 97% 3% $59,688
Charter Township of Oakland 4,341 96% 4% $102,034

 

Socioeconomic characteristics included in the regression analysis are shown in
Table 19. Socio-economic characteristics were derived using 2000 Census block group
SF 1 and SF3 datasets. Census block group boundaries were overlaid with parcel
boundaries using a GIS system to assign individual socioeconomic values to individual

parcels.

41

 

 

 

 

 

 

 

 

 

 

 

Table 19. Description of Socio-Economic Variables Used in Regression Analysis
Variable Variable Description

PERWHITE Percent of total pqiulation that is white exclusively
PERNWHITE“ Percent of total population that is not white exclusively
PERHS Percent of total Jopulation with high school attainment

Percent of total population with bachelors, masters or PhD

PERBSMSPHD attainment
MEDTNCOME Median household income

PEROWN Percent of owner occgiied housing units

PERRENT'” Percent renter occupied housing units

 

 

“Removed from the multiple regression analysis to avoid the dummy variable trap.

The second set of location-based independent variables focus on the role that
adjacent uses play in the determination of assessed value, shown in Table 192. Spatial
information on land use, natural amenities, roads, and public facilities were provided in a
GIS format by the Oakland County One-Stop Shop while classiﬁcations for spatial
variables were provided by the Department of Community and Economic Development.

Adjacent variables describe land uses, natural amenities, public facilities and
capital investments including roads that are physically adjacent to the land use of interest.
Using GIS, each category of land use to be modeled was selected, and then adjacent land
uses were classified using the “select by location, share a line segment, use selected
features” criteria and were coded accordingly. Adjacent variables represent true
adjacency. Parcels that were located across the road from each other do not share parcel
boundaries therefore are not deﬁned as adj acent.3 It should also be noted that for each
regression model, adjacent uses that match the land use being modeled were removed

from the analysis (i.e. adjacent industrial parcels were removed from industrial models).

 

2 Missing numbers in variable codes represent removal of an independent variable early in model design
3This may be too rigid of a test to determine the monetary effect of mixed use zoning or incompatible land
uses. For future analyses a 100 or 200 meter buffer may provide more telling results.

42

 

Table 20. Description of Adjacent Land Use Variables Used in Regression Analysis

 

 

 

Variable Variable Descriition
Ajjacent Land Uses
A002 Parcels adjacent to multi-family; 1 if adjacent, 0 if otherwise

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A003 Parcels adjacent to commercial or ofﬁce; 1 if adjacent, 0 if otherwise
A005 Parcels adjacent to industrial; 1 if adjacent, 0 if otherwise
A006 Parcels adjacent to mobile home park; 1 if adjacent, 0 if otherwise
A007 Parcels adjacent to public and institutional; 1 if adjacent, 0 if otherwise

Parcels adjacent to single family parcels with more than one unit; 1 if adjacent,
A008 0 if otherwise

Parcels adjacent to single family with lot size 1 - 2.5 acres; 1 if adjacent, 0 if
A009 otherwise

Parcels adjacent to single family 14000 to 43,559 sq ft; 1 if adjacent, 0 if
A010 otherwise

Parcels adjacent to single family with lot size 2.5 - 5 acres; 1 if adjacent, 0 if
A011 otherwise ‘

Parcels adjacent to single family with lot size 5 - 10 acres; 1 if adjacent, 0 if
A012 otherwise

Parcels adjacent to single family 8000 - 13999 sq ﬁ; 1 if adjacent, 0 if
A013 otherwise

Parcels adjacent to single family with lot size greater than 10 acres; 1 if
A014 adjacent, 0 if otherwise

Parcels adjacent to single family less than 8000 sq ft; 1 if adjacent, 0 if
A015 otherwise

Parcels adjacent to utility and transportation land uses; 1 if adjacent, 0 if
A016 otherwise
A017 Parcels adjacent to vacant parcels; 1 if ad'Lacent, 0 if otherwise

 

Adjacent Natural Amenities

 

Parcels adjacent to lakes or ponds less than 19 acres; 1 if adjacent, 0 if

 

 

 

 

 

 

 

 

A019 otherwise
Parcels adjacent to lakes or ponds between 19 and 49 acres; 1 if adjacent, 0 if
A020 otherwise
Parcels adjacent to lakes or ponds less than 49 and 100 acres; 1 if adjacent, 0 if
A021 otherwise
Parcels adjacent to lakes or ponds greater than 100 acres; 1 if adjacent, 0 if
A022 otherwise
A023 Parcels adjacent to river or stream'Ll if aciacent, 0 if otherwise
A024 Parcels adjacent to wetland (any size); 1 if adjacent, 0 if otherwise
A027
* parcels adjacent to natural areas with priority one; 1 if adjacent, 0 if otherwise
A028
* parcels adjacent to natural areas with priority two; 1 if @acent, 0 if otherwise
A029 parcels adjacent to natural areas with priority three; 1 if adjacent, 0 if

*

 

otherwise

 

 

*Natural conservation priorities are established in Oakland County Conservation Plan

 

43

 

 

Table 20 cont. Description of Adjacent Land Use Variables Used in Regression

Analysis

Adjacent to Public Infrastructure

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A030 Parcels adjacent to recreation area (of any type); 1 if adjacent, 0 if otherwise
Parcels adjacent to municipally owned park or recreation area; 1 if adjacent, 0
A031 if otherwise
A034 Parcels adjacent to school; 1 if adjacent, 0 if otherwise
A040 Parcels adjacent to urban interstate; 1 if adjacent, 0 if otherwise
A041 Parcels adjacent to urban principal arterial; 1 if adjacent, 0 if otherwise
A042 Parcels adjacent to urban non-freeway; 1 if adjacent, 0 if otherwise
A043 Parcels adjacent to county, local, urban local road; 1 if adjacent, 0 if otherwise
A044 Parcels adjacent to county, local, rural local road; 1 if adjacent, 0 if otherwise
Parcels adjacent to county, local, urban collector road; 1 if adjacent, 0 if
A045 otherwise
Parcels adjacent to county, primary, rural minor arterial road; 1 if adjacent, 0
A046 if otherwise
Parcels adjacent to county, primary, rural major collector road; 1 if adjacent, 0
A047 if otherwise
Parcels adjacent to county, primary, urban non-freeway; 1 if adjacent, 0 if
A048 otherwise
Parcels adjacent to county, primary, urban minor arterial road; 1 if adjacent, O
A049 if otherwise
Parcels adjacent to cormty, primary, urban collector road; 1 if adjacent, 0 if
A050 otherwise
A051 Parcels adjacent to city, major, urban local road; 1 if adjacent, 0 if otherwise
Parcels adjacent to city, major, urban collector road; 1 if adjacent, 0 if
A052 otherwise
Parcels adjacent to city, major, urban minor arterial road; 1 if adjacent, 0 if
A053 otherwise .
A054 Parcels adjacent to city, major, urban non-freeway; 1 if adjacent, 0 if otherwise
A055 Parcels adjacent to cig, minor, urban local road; 1 if agacent, 0 if otherwise

 

 

The ﬁnal group of location-based independent variables focuses on the

 

conuibutions of surrounding land uses and natural amenities to assessed building and

land values. “Proximity variables,” shown in Table 21 define land uses located within a

44

quarter mile of a given property“. Quarter mile buffers are a frequently cited distance
incorporated within many hedonic studies and represent a commonly cited walking
distance in the planning profession. To create proximity variables a GIS buffer frmction
was used Quarter mile buffers were created for each land use using the “select by

location, intersect function” to identify and code proximity variables.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 21. Description of Proximity Variables Used in Regression Analysis

Variable Code Variable Description

BOO] Within .25 miles of agriculture; 1 if yes, 0 if no

B002 Within .25 miles of multi-family; 1 if yes, 0 if no

B003 Within .25 miles of commercial and ofﬁce; 1 if yes, 0 if no

BOOS Within .25 miles of industrial; 1 if yes, 0 if no

B006 Within .25 miles of mobile home park; 1 if yes, 0 if no

BOO7 Within .25 miles of public or institutional; 1 if yes, 0 if no

BOO8 Within .25 miles of railroad; 1 if yes, 0 if no

8009 Within .25 miles of recreation area (of any type); 1 if yes, 0 if no

B013 Within .25 miles of utility and transLorta 'on; 1 if yes, 0 if no

3014 Within .25 miles of lake or pond, between 0 and 19 acres; 1 if
yes, 0 if no

3015 Within .25 miles of lake or pond, between 19 and 49 acres; 1 if
yes, 0 if no

B016 Within .25 miles of lake or pond, between 49 and 100 acres; 1 if
yes, 0 if no

B017 Within .25 miles of lake or pond greater than 100 acres; 1 if yes,
Oifno

B018 Within .25 miles of stream or river; 1 if Les, 0 if no

B021 Within .25 miles of natural areas, priority one; 1 if yes, 0 if no

B022 Within .25 miles of natural areas, priority two; 1 if yes, 0 if no

B023 Within .25 miles of natural area, priority three; 1 if yes, 0 if no

B024 Within .25 miles of schools; 1 ifyes, 0 if no

B032 Within .25 miles of government facility 1 if yes, 0 if no

B033 Within .25 miles of ﬁre station; 1 if yes 0 if no

B034 Within .25 miles of police station; 1 if yes 0 if no

3037 Within .25 miles of wetland; 1 ifyes 0 if no

 

 

 

 

Few studies have examined the impact of land use patterns on property values.

Bockstael (1996) and Geoghegan et. a1 (1997) emphasize that the value of a residential

 

4 Missing numbers or ill-logical sequence of numbers in variable codes represents removal or manipulation
of an independent variable early in model design.

45

parcel is affected by the pattern of surrounding land uses and landscape, not just the
speciﬁc features of point locations. Geoghegan et a1. (1997) and Acharya and Bennet use
indices to address the spatial differences inland use patterns and the affects on property
values. While these indices present some indication of the surrounding number of land
uses, they fail to indicate the land uses that compose that mix. For this reason, emphasis
of this study relies on dummy variables to model the implicit values of proximity
variables.

(4) Sample frame. The hedonic price function is an equilibrium function that
describes a speciﬁc market (Taylor, 2002, p. 350). In Oakland County, separate real
estate markets are assumed to exist for single family residential uses in cities versus
townships and villages, therefore two single-family hedonic pricing functions are
estimated Commercial, industrial and multi-family are assumed to operate in one real
estate market in Oakland County, therefore hedonic pricing functions include parcel
records from all case study communities. It should be noted that the vast majority of
records in non-residential and multi-family models are found in the City of Pontiac and
City of Royal Oak.

(5) Functional form of the hedonic price function Linear multiple regression
analysis is used to estimated the hedonic price ﬁmction for assessed building and land
values. While linear multiple regression analysis is usually not the preferred econometric
technique (Taylor, 2002, p. 339) it offers several beneﬁts given the orientation of this
study. Linear regression techniques enable straight forward implicit price interpretation
and calculations. Linear regressions also meet the research goal of identifying statistically

signiﬁcant relationships in which land use patterns and natural amenities inﬂuence

46

assessed property values. It should be noted that linear functional forms will likely distort
the implicit value of square footage and lot size which maintain non-linear relationships
with property value. Box-cox regression forms that allow for linear and non-linear
speciﬁcations were tested (see limitations for more information).

(6) Address Spatial Dependence and Correlation. Appendix B. explores potential
affects of spatial bias on statistically signiﬁcant regression coefﬁcients. Appendix B
breaks down hedonic model samples, identifying statistically signiﬁcant independent
variables and the number and percentage of cases in the sample and associated average
building and land values. Given the strong spatial component of independent variables
spatial autocorrelation could exist, which is especially true for adjacent variables which
usually demonstrate a very small percentage of the sample. Simple logic tests founded on
the economic principles of scarcity and utility are applied when interpreting spatial
biases. For example, lakefront properties compose less than ﬁve percent of the total
overall residential sample. In this case, scarcity and utility of private waterfront
ownership increase demand, resulting in higher building and land values, therefore the
issue of spatial autocorrelation diminishes. Spatial biases are recognized when the utility
of the property is not clearly inﬂuenced by an adjacent land use or natural feature, and
location merely coincides with extreme observations in reliable value determinants such
as age, class and square footage.
3.2 Empirical Estimates and Regression Results

Linear multiple regression analysis is used to estimate the hedonic pricing

function for assessed “true cash” building value and land value for residential,

47

commercial, industrial and multi-family land uses. Table 55 and 56 presents the

summaries of MRA model outcomes.

Table 22. Building

Value Model Summaries

 

 

 

 

 

R Adjusted Std. Error of
Land Use R Square R Square the Estimate df F Sig;
City Single Family Residential 0.935 0.874 0.874 25,217.87 87 3,054.66 0.000
Township Single Family Residential 0.929 0.863 0.862 43,837.61 96 1,401.25 0.000
Commercial 0.898 0.806 0.800 541,526.02 55 144.79 0.000
Industrial 0.954 0.910 0.900 2,479,714.58 48 97.46 0.000
Multi-Family 0.956 0.913 0.906 401,287.69 57 118.82 0.000

Table 23. Land Value Model Summaries

R Adjusted Std. Error of
Land Use R Square R Square the Estimate df F Sig;
City Single Family Residential 0.829 0.688 0.687 13,001.17 60 1,403.02 0.000
Township Single Family Residential 0.813 0.661 0.660 37,144.31 70 596.24 0.000
Commercial 0.825 0.681 0.673 283,826.72 51 80.57 0.000
Industrial 0.618 0.381 0.323 587,397.96 44 6.56 0.000
Multi-Family 0.598 0.358 0.310 145,808.93 48 7.56 0.000

 

Building value and land value models produced statistically signiﬁcant results

demonstrated by R-squared values that explain between 35.8% to 68.8% of the variance

in land values and 80.6% to 91.3% of the variance in assessed building values. Building

value models demonstrate a high degree of explanatory power and relative consistency in

comparison to land value models. Building value models explain approximately 87% of

the variance in assessed value for city single family residential buildings, 86% for

township and village residential buildings, 80% commercial buildings, and 91% for

industrial and multi-family buildings. Land value models explain 83% and 81% of the

variance inland values for single family residential, 68% for commercial, 38% for

industrial and 35% for multi-family residential.

Constant terms were signiﬁcant for city single family residential building value

48

(-82,552) models, township and village single family residential building value (126,271)
and land value models (326,066), and commercial building value (468,755) models.5
Constant terms were insigniﬁcant for city single family land values, industrial and multi-
family models and commercial land value models. Constant terms, or the starting point
(y-intersect) for projecting the best-ﬁt regression line, play an important role in
interpretation of regression coefﬁcients. Of greatest consequence, the negative constant
term for city residential building values indicates that several positive relationships exist
that raise residential values up to positive values ﬁ'om a negative starting point.

Regression coefﬁcients are considered statistically signiﬁcant if they demonstrate
a signiﬁcance level of less than 5 percent. Regression coefficients represent the implicit
values assigned to independent variables. In the case of dummy variables, the coefficient
equals the implicit price of the associated variable. In the cases of independent variables
expressed as a scale or percentage, implicit prices must be transformed to show the
incremental change in monetary value attributed to an incremental change in percentage
or scalar value.
3.2.1. Basic Property Attributes

“Table 24. Linear Regression Coefﬁcient Estimates for Residential Building
Characteristics,” demonstrates that basic property attributes including class, age, square
footage, lot size and the number of structures were statistically signiﬁcant in almost all
cases for residential and non-residential building value models, in line with Michigan

assessor practice.

 

5 The unit of measurement for the constant term is dollars, just as the unit of measurement for implicit
regression coefﬁcients. The constant term represents the starting value from which the best-ﬁt regression
line is projected.

49

Table 24. Linear Regression Coeﬁicient Estimates for Basic

 

 

 

 

 

 

Property Attributes
Land Use Coefﬁcient Std. Error T Sig_._
City Single Family Residential
Class 17,114.06 710.68 24.08 .000
Age -l,207.94 15.95 -75.72 .000
Floor Area 63.60 0.49 130.07 .000
Lot Size (BV) 6,377.24 1,375.20 4.64 .000
Lot Size (LV) 80,765.45 550.85 146.62 .000
Number of Buildings 48,563.28 763.75 63.59 .000
Township Single Family Residential
Class 33,846.08 1,249.15 27.10 .000
Age -l,257.21 41.37 -30.39 .000
Floor Area 70.37 0.76 92.71 .000
Lot Size (BV) -87.57 127.03 -.69 .491
Lot Size (LV) 11,084.12 106.46 104.11 .000
Number of Buildings -37,186.06 4,289.36 -8.67 .000
Commercial
Class -25,603.43 35,465.73 -.72 .470
Age -5,737.43 555.12 -10.34 .000
Floor Area 55.33 1.28 43.18 .000
Lot Size (BV) 14,455.45 12,490.79 1.16 .247
Lot Size (LV) 194,264.04 4,159.73 46.70 .000
Number of Buildings -8,975.12 14,653.13 -.61 .540
Industrial
Class 1,114,488.29 545,361.33 2.04 .042
Age -29,512.95 7,057.14 -4.18 .000
Floor Area 18.36 0.46 39.95 .000
Lot Size (BV) 43,644.73 4,234.67 10.31 .000
Lot Size (LV) 9,830.25 755.52 13.01 .000
Number of Buildings 745,183.30 48,979.63 15.21 .000
Multi-Family
Class 165,572.62 44,360.42 3.73 .000
Age -355.21 950.42 -.37 .709
Floor Area 33.25 0.67 49.68 .000
Lot Size (BV) 5,002.63 5,503.07 .91 .364
Lot Size (LV) 14,674.95 - 1,848.86 7.94 .000
Number of Buildings -4,834.07 4,142.23 -1.17 .244

 

50

3.2.1.1 Construction Class

Construction class is a statistically signiﬁcant determinant of building value for single
family, industrial and multi-family land uses, conﬁrming that a positive relationship
exists between quality of construction and building value. Construction class is modeled
using a scale of one to four with one (typically referred to as Class D) representing the
lowest construction quality and four (Class A) the highest quality. The implicit price of
an incremental change in residential construction class accounts for $33,846 of building
value in townships and villages and $17,114 in cities. Construction quality can account
for a potential difference of $101,538 in building value when comparing class A and D
residential properties in township and village residential markets and $51,342 in city
markets. Construction class accounts for $1.11 million incremental change in building
value for industrial properties, however it should be noted that over 80% of the industrial
sample is designated as class C (2) construction. For multi-family uses, construction
quality adds an incremental value of $165,500 to building value.

Construction class was found to be an insigniﬁcant determinant of commercial
building value, indicating that other factors related to building or location characteristics
may outweigh the role that quality plays in the determination of commercial building
value.
3.2.1.2 Age

Age demonstrates a statistically signiﬁcant negative relationship with single
family, commercial and industrial building values. An incremental increase in age for
non-residential properties is associated with a $25,603 decrease in commercial properties

and $29,512 for industrial properties, which adequately represent standard depreciation

51

schedules adopted by the State Tax Commission. Age has a similar impact in case study
residential markets representing, a $1,207 incremental decrease of building value in city
residential structures and $1,257 for township and village structures.
3.2.1.3 Square Footage

In line with a priori expectations, square footage is a statistically signiﬁcant
determinant of building values in all cases. Square footage coefﬁcients are the highest for
township and village single family residential buildings at $70.37 per square foot and
lowest for industrial buildings at $18.36 per square foot. Single family residential
building values (city $63.60 and township-village $70.37) are higher than non-residential
(commercial $55.33, industrial $18.36) and multi-family ($33.25) values due to
economies of scale realized from larger building footprints.
3.2.1.4 Lot Size

Lot size was modeled for both building value and land value models. While lot
size is not a building attribute, it is assumed to correlate with market preference for built
structures, which is inline with application of the ECF. Lot size is positively correlated
with land values in all cases. Lot size contributes $194,264 per acre to land value for
commercial land uses, $80,765 for city single family residential, $14,675 for multi-family
uses, $11,084 for township and village uses and $9,830 for industrial lands. Lot size
variables for land value models clearly demonstrate the value implications of density. For
high density uses such as single family residential, commercial and multi-family uses, the
value of land increases, but for low density or large lot sizes land values are lower.

Lot size shows both signiﬁcant and insigniﬁcant relationships with building

value. Lot size is a signiﬁcant determinant of building value for city residential ($6,377

52

per acre) properties and industrial properties ($43,644). In both of these cases, market
preference for larger lots likely drives the signiﬁcant positive relationship with building
value, while in the case of commercial, township residential and multi-family uses lot
size does not determine increased market demand for buildings.

3.3.1.5 Number of Structures

The number of structures, while not clearly interpreted by assessors, represents
“secondary structures,” which can be signiﬁcant for industrial properties in particular.
The number of buildings is positive and signiﬁcant for city residential properties
($48,563) and negative and signiﬁcant for township and village single family uses (-
$37,186). Each additional industrial structure, adds $745,183 in building value for
industrial properties. It should be noted that the number of buildings, as coded and
interpreted by assessors requires additional study to be accurately modeled.

3.3.2 Residential Building Attributes.

Residential building attribute, shown in “Table 25: Linear Regression Coefﬁcient
Estimates for Residential Building Characteristics,” including the number of bathrooms -
and garage size contribute positively to building value in line with modeling
expectations. The number of bedrooms however demonstrates signiﬁcant negative
relationships with building value in both city and township and village models. One
possible explanation for the negative coefﬁcient associated with bedrooms is “double
counting” of ﬂoor area. The number of bedrooms and, bathrooms is already accounted for
in square footage, therefore the value of square footage and bathroom coefﬁcients may be
artiﬁcially high while the negative value of bedrooms may demonstrate a necessary

reduction in building area values. When modeling the value contributions of a garage, a

53

similar double counting scenario appeared when both garage area and garage type

(number of cars) were modeled together. Exclusion of bedrooms and bathrooms would

likely improve the efﬁciency of models and more accurately measure the value impacts

of ﬂoor area or vice versa.

Table 25. Linear Regression Coemcient Estimates for Residential Building Characteristics

 

 

 

 

City Model Township and Village Model
Coefﬁcient Std. Error T Sig._ Coefﬁcient Std. Error T Sig._
Bedroom -3,288.90 204.15 -16.11 .000 -6,779.77 413.57 -16.39 .000
FullBath 15,504.38 309.16 50.15 .000 24,483.87 630.17 38.85 .000
Half Bath 9,642.36 326.55 29.53 .000 12,553.68 677.68 18.52 .000
Garage Area 24.72 0.64 38.52 .000 36.84 1.49 24.65 .000
Garage Type 0.00 0.01 -.18 .855 -6,073.56 608.64 -9.98 .000
Residential Exteriors*
Asbestos -l,212.68 547.75 -2.21 .027 11,687.86 3,321.17 3.52 .000
Block -5,541.91 1,920.34 -2.89 .004 -2,473.29 4,964.57 -.50 .618
Brick 14,702.11 341.63 43.04 .000 2,227.06 759.60 29.26 .000
Composition -1,629.94 1,206.10 -1.35 .177 13,335.27 9,491.91 1.40 .160
Wood 572.96 1,179.54 .49 .627 8,559.83 3,630.38 2.36 .018
Stone 9,542.60 3.345.91 2.85 .004 60,058.08 8,651.71 6.94 .000
Stucco 5,949.01 1,406.81 4.23 .000 31,199.89 11,778.11 2.65 .008
Other -4,669.68 3,853.77 -1.21 .226 -5,367.44 5,023.22 -1.07 .285
Residential Style“
Bi-Level -22,846.53 2,604.36 -8.77 .000 -33,249.74 6,624.74 -5.02 .000
Bungalow 13,735.98 378.05 36.33 .000 21,609.48 3,815.83 5.66 .000
Cape Cod -4,439.44 1,888.26 -2.35 .019 2,539.84 5,561.94 .46 .648
Colonial 2,821.52 530.99 5.31 .000 -4,569.26 1,829.56 -2.50 .013
Contemporary -713.22 4,053.71 -.18 .860 33,402.40 2,618.90 12.75 .000
Other -21,290.76 838.03 -25.41 .000 -l3,151.59 3,426.78 -3.84 .000
Single Family 16,169.85 401.13 40.31 .000 -518.42 2,193.45 -.24 .813
Tri—Level/Qudor -34,503.17 1,389.28 -24.84 .000 -11,476.05 21,977.44 -.52 .602
Tudor -26,251.95 3,205.58 -8.19 .000 66,885.02 7,710.93 8.67 .000
Townhouse/Duplex -9,996.63 3,611.25 -2.77 .006

 

*Aluminum, aluminum and vinyl, vinyl were removed ﬁ‘om the analysis to avoid the dummy variable trap,
coeﬁicients represent a deviation in value when compared to aluminum and vinyl exteriors.

"Ranch building styles were removed from the analysis to avoid the dummy variable trap, coefﬁcients

represent a variation in value when compared to ranch housing styles

Wood represent log, pine and cedar and wood building materials
Other represents asphalt, dryvit, lap siding, masonite, metal and other building materials

3.2.2.1 Residential Exteriors

Residential exteriors are associated with both positive and negative contributions

to building value in case study communities. High quality building materials such as

54

brick ($14,700 city; $22,220 township), stone ($9,540 city; $60,000 township) and stucco
($6,000 city; $32,000 township) add signiﬁcant value to residential buildings. Low
quality materials, especially those found in older city residential properties, such as
asbestos (-$1,200), block (-$5,500) and composition (-$1,630) detract ﬁ'om building
value.
3.2.2.2 Residential Building Style

Residential building styles contribute to a high degree of variation in value
between city and township residential markets. For example, contemporary ($33,400) and
tudor ($66,900) building styles add signiﬁcant value to residential buildings in townships
and villages but subtract value from city residential properties (contemporary -$700 and
tudor, ~$26,250). Similarly, colonial building styles contribute $2,800 to building value in
cities but subtract over $4,500 in townships and villages. Bi-level, tri-level and “other”
building styles diminish building value in both residential markets while bungalow
building styles add value. These results help conﬁrm that fundamentally different real
estate markets exist between case study communities.
3.2.3 Socio-Economic Characteristics

Building and land value models reveal that social and economic conditions play
an inﬂuential role in determining residential property values and to a lesser extent non-
residential property values. “Table 26. Linear Regression Coefﬁcient Estimates for
Socioeconomic Attributes” demonstrates that race, educational attainment and household
income variables maintain a positive relationship with residential building and land
values in cities and a negative relationship in townships and villages. While outcomes for

socioeconomic variables fell in line with expectations for city models, unexpected

55

relationships in the townships and village could be explained by large census block
conﬁgurations with relatively consistent socioeconomic attributes causing a spatial
statistical bias.

Table 26. Linear Regression Coefﬁcient Estimates for Socioeconomic Attributes

 

 

 

 

 

 

 

 

Buildinglalue Models Land Value Models
Land Use Coefﬁcient Std. Error T Sig._ Coemcient Std. Error T Sig_._
City Single Family Residential
White (%) 319.23 737.88 43.26 .000 104.53 361.78 28.89 .000
H.S.Education(%) 152.84 2,345.20 6.52 .000 -30.87 1,200.63 -2.57 .010
College Education (%) 915.42 2,042.72 44.81 .000 502.82 994.98 50.54 .000
Med Household Income 0.21 0.02 10.40 .000 0.20 0.01 19.22 .000
Owner-Occupied (%) ~82.10 1,088.92 -7.54 .000 -97.10 550.55 -17.64 .000
Township Single Family Residential
White (%) -699.05 18,261.06 -3.83 .000 -2,029.85 15,225.03 -l3.33 .000
H.S. Education (%) -613.28 10,330.89 -5.94 .000 -564.53 8,678.46 -6.50 .000
College Education (%) -147.80 6,623.42 -2.23 .026 95.92 5,427.28 1.77 .077
Med Household Income -0.26 0.04 ~6.57 .000 -0.05 0.03 -1.38 .167
Owner-Occupied (%) -475.83 6,406.41 -7.43 .000 -316.23 5,400.13 -5.86 .000
Commercial
White (%) -469.86 75,517.88 -.62 .534 1,293.91 39,352.28 3.29 .001
H.S.Education(%) 2,837.37 227,617.19 1.25 .213 1,955.99 118,884.34 1.65 .100
College Education (%) 3,345.28 167,977.95 1.99 .047 4,536.46 87,713.26 5.17 .000
Med Household Income 3.50 1.54 2.27 .024 -0.69 0.81 -.85 .394
Owner-OccupiedQ/o) -3,607.17 100,826.94 ~3.58 .000 -1,182.13 52,757.19 -2.24 .025
Industrial
White (%) -7,807.72 899,267.86 -.87 .386 -25.96 202,707.06 -.01 .990
H.S.Education(%) -53,581.04 2,985,471.96 -1.79 .073 8,171.23 674,789.78 1.21 .227
College Education (%) -26,661.59 1,740,583.90 -l.53 .126 9,194.39 393,952.56 2.33 .020
Med Household Income 18.88 15.78 1.20 .232 0.77 3.57 .21 .830
Owner-Occupied (%) -32,767.42 1,014,022.79 -3.23 .001 -9,502.98 229,504.84 —4.14 .000
Multi-Family
White (%) 574.84 123,182.75 .47 .641 11.83 41,472.88 .03 .977
H.S.Education(%) 1,953.57 274,964.18 .71 .478 1,535.11 97,840.76 1.57 .117
College Education (%) 2,300.12 233,418.36 .99 .325 1,574.84 82,859.90 1.90 .058
Med Household Income -l.15 2.78 -.42 .678 -0.27 0.97 -.27 .784
Owner-Occupied (%) -2,652.01 156,249.43 -1.70 .090 -1,268.14 54,870.66 -2.31 .021

 

Coefﬁcients are adjusted to show dollar increase in building and land value associated with percentage
increase in independent variables that are expressed as a percentage

Med. household income stands for median household income, H.S. stands for high school education, owner-
occupied stands for percentage of households that are owner—occupied

Racial diversity is a statistically signiﬁcant determinant of building and land value
for residential properties but not for multi-family and only commercial land value in the

case of non-residential properties. Regression results show that the percentage of the

56

population that is white maintains a negative relationship with townships and cities
residential property values, demonstrated by an incremental decrease in value of
approximately $700 in building value and over $2,000 in land value. In case study cities,
as the percentage of the white population increases, so to does residential building value
($319) and land value ($104).

College education maintains a positive relationship with city single family
building ($915) and land values ($502), commercial property (building, $3,345) and
industrial land values $4,536) and demonstrates the broadest and highest value impacts
across land use type of any statistically signiﬁcant socioeconomic variable.

3.2.4 Adjacent and Surrounding Land Use Patterns

Adjacent and buffer variables measure the dynamic value interactions between
different land uses and geographic land use patterns. For each land use model, coinciding
land use variables (i.e. adjacent and buffer commercial variables are removed from
commercial building and land models) are removed to better interpret value implications
of mixed use development. In the case of single family residential models, buildings with
a lot size less than 8,000 square feet, the highest density residential designation, are
removed to interpret the implications of surrounding residential densities on building and
land values. It should be noted that residential buffer variables are removed ﬁ'om the
analysis because they cover the vast majority of land area within the case study

communities.

57

Table 27. Impacts of Land Use Patterns on City Single Family Residential Property
Values

Linear Regression Coefficient Estimates for Adjacent Land Uses

 

 

 

 

 

 

 

 

 

BuildinLValue Models Land Value Models
Land Use Coefficient Std. Error T Sig.__ Coefﬁcient Std. Error T Sig.__
City Single Family Residential
Multi—Family -1, l 99.91 714.71 -1.68 .093 434.05 366.79 1.18 .237
Commercial -4.5 87.69 795.01 -5.77 .000 885.71 409.43 2.16 .031
Industrial -2.996.66 1,852.39 -1.62 .106 -3,071.69 953.25 -3.22 .001
Public-Institutional -2,283. 13 836.68 -2.73 .006 -2, 134.40 428.80 -4.98 .000
Utility-Transportation 475.54 731.31 .65 .516 -290.48 3 76.90 -.77 .441
Vacant 482.10 409.65 1.18 .239 -1,384.72 209.11 -6.62 .000
Single Family Residential
More than one unit per lot 13,771.19 4,236.64 3.25 .001 21,873.97 2,166.82 10.09 .000
Lot size; 8,000 - 13,999 s.f. -2,846.59 276.30 -10.30 .000 2,983.83 139.52 21.39 .000
Lot size: 14.000-25,000 5.11 4,427.49 459.15 9.64 .000 -154.80 228.99 -.68 .499
Lot size; 1-2.5 ac. 1,626.89 1,815.73 .90 .370 -38.548.59 906.42 -42.53 .000
Lot size; 2.5 - 5 ac. 248.00 3,401.56 .07 .942 -51,271.94 1,721.92 -29.78 .000
Lot size; 5 - 10 ac. 13,936.70 8,838.79 1.58 .115 -86.925.45 4.527.88 -19.20 .000
Linear Regression Coefficient Estimates for Land Uses within Quarter Mile Buffer
ﬂ Building Value Models Land Value Models
Land Use Coeﬂicient Std. Error T Sig; Coefﬁcient Std. Error T Sig_._
City Single Family Residential
Multi-Family -1,744.92 360.35 -4.84 .000 -2,404.91 250.35 -4.92 .056
Commercial 4,522.34 409.93 11.03 .000 -l,835.69 210.38 -8.73 .000
Industrial 243.69 339.68 .72 .473 -2,481.56 173.74 -14.28 .000
Mobile Home -7,672.21 1,098.47 -6.98 .000 -5,554.27 565.21 -9.83 .000
Public-Institutional -729.67 432.15 -1.69 .091 -1,308. 17 222.31 -5.88 .000
Railroad 1,488.58 402.44 3.70 .000 -136.07 205.66 -.66 .508
Utihty-Trangsportation -1,734.27 307.18 -5.65 .000 168.49 157.86 1.07 .286

 

Table 27 demonstrates that adjacent commercial land uses decrease city single
family residential building values by -$4,587 but increase building value by $4,522 when
within a quarter mile, suggesting that neighborhood commercial nodes may elevate
surrounding residential livability and property values if properly designed. Public and
institutional uses detract from city residential property values in the cases of adjacency (-
$4,417 combined building and land implicit values) and proximity (-$1,308 land value
only). Industrial land uses do not impact city residential building values, but do impact
land values to the magnitude of -$3,072 in the case of adjacency and -$2,481 in the case

of quarter mile proximity. Multi-family and mobile homes located within a quarter mile

58

coincide with decreases in city building value of -$1,744 and -$7,672 respectively,

however adjacent condo developments (residential uses with more than one unit per lot)

may increase building value signiﬁcantly, suggesting that quality or design of

development may mitigate the impacts higher residential densities on single family

property values. It is unclear whether city residential densities impact property values.

Adjacent residential lot sizes of 8,000 to 14,000 square feet decrease city building values

by -$2,846 but increase land values by $2,983, therefore having minimal overall impact.

As lot size decreases, adjacent land values fall in cities, however few samples exist to

conﬁrm this ﬁnding.

Table 28. Impacts of Land Use Patterns on Township and Village Single Family
Residential Property Values

Linear Regession Coefﬁcient Estimates for Adjacent Land Uses

 

 

 

 

 

 

 

 

 

Building Value Models Land Value Models
Land Use Coefficient Std. Error T Sig._ Coeﬁcient Std. Emr T Sig._
Township Single Family Residential
Multi-Family 6,058.62 2,891.22 -2. 10 .036 4,856.66 2,445.00 -1.99 .047
Commercial 7,002.13 2,403.50 2.91 .004 3,507.70 2,033.72 1.72 .085
Industrial 5,188.24 5,254.06 .99 .323 -12, 183.30 4,421.36 -2.76 .006
Public-Institutional 5,174.10 2,948 .58 1.75 .079 2.386.15 2,496.73 .96 .339
Utility-Transportation -8.247.15 3,833.86 -2.15 .031 6,040.10 3,242.61 1.86 .063
Vacant 7,639.46 753.70 10.14 .000 11,391.77 636.04 17.91 .000
Single-Family Residential
More than one unit per lot -20,229.08 4,554.60 -4.44 .000 -32.837.22 3,841.87 -8.55 .000
Lot size; 8,000 - 13.999 s.£ -5,351.37 825.65 -6.48 .000 -6,767.27 682.60 -9.91 .000
Lot size; 14,000-25,000 s.f. 2,044.71 720.49 2.84 .005 2,426.98 596.11 4.07 .000
Lot size; 1-2.5 ac. 884.61 861.67 1.03 .305 51.31 719.04 .07 .943
Lot size; 2.5 - 5 ac. 398.51 1,099.23 .36 .717 2,583.02 915.94 2.82 .005
Lot size; 5 - 10 ac. -1,500.87 1,305.76 -1.15 .250 5,035.91 1,101.76 4.57 .000
Lot size; greater than 10 ac. -5,193.15 1,597.22 -3.25 .001 3,002.52 1,350.23 2.22 .026
Linear Regression Coefficient Estimates for Land Uses within Quarter Mile Buffer
Building Value Models Land Value Models
Land Use Coefﬁcient Std. Error T Sig._ Coefﬁcient Std. Error T Sig;
Township Single Family Residential
Agricultural -12,112.80 1,577.24 -7.68 .000 6,593.11 1,325.85 4.97 .000
Multi-Family -378.73 967.83 -.39 .696 -3,881.03 815.98 -4.76 .000
Commercial 1,492.40 849.35 1.76 .079 -4,976.92 715.40 -6.96 .000
Industrial 1,299.60 1,339.53 .97 .332 -4,991.11 1,130.71 -4.41 .000
Mobile Home -1 ,268.23 2,742.64 -.46 .644 -14,545.88 2,311.95 -6.29 .000
Public-Institutional 1,464.49 796.86 1.84 .066 -6,721.56 672.87 -9.99 .000
Railroad -8,691.54 2,004.46 -4.34 .000 -1 1,518.91 1,679.63 -6.86 .000
Utility-Transportation -4,302.34 801.40 -5.37 .000 -1,784.34 674.86 -2.64 .008

 

59

Table 28 shows that residential land values in townships and villages are

signiﬁcantly affected by surrounding non-residential land uses and residential densities

but not building values. Single family land values are negatively impacted by all non-

residential uses located within a quarter mile proximity except for agricultural, with value

reductions ranging from -$14,545 for mobile homes to -$1,784 for utility and

transportation uses. Nearby commercial and industrial uses reduce single family

residential land values by nearly -$5,000 each. Adjacent residential densities of less than

14,000 square feet (-$12,1 18 combined building and land value) and condos (-$53,066

combined building and land value) reduce property values signiﬁcantly while lot sizes of

14,000 to 25,000 square feet ($4,470 combined building and land value), 2.5 to 5 acres

($2,583) and 5 to 10 ($5,035) increase property values.

Table 29. Impacts of Land Use Patterns on Multi-Family Property Values

Linear Rggression Coefﬁcient Estimates for Ag’Lacent Land Uses

 

 

 

 

 

 

 

 

Building Value Models Land Value Models
Land Use Coeﬂicient Std. Error T Sig. Coefﬁcient Std. Error T 5L
Multi-Family
Commercial 97,161.55 51,117.16 1.90 .058 81,372.57 18,066.10 4.50 .000
Industrial -204,129.23 114,382.65 -l.78 .075 -106,668.03 40,960.78 -2.60 .009
Public-Institutional -12,781.11 61,717.99 -.21 .836 35,474.43 21,970.85 1.61 .107
Utility-Transportation -303,855.11 157,416.62 -1.93 .054 126,272.66 55,666.94 2.27 .024
Vacant 3,320.72 48,489.64 .07 .945 5,611.52 17,326.39 .32 .746
Single Family Residential
Lot size; less than 8.000 s.f. -122,901.75 37,941.55 -3.24 .001 -13,086.74 13,428.56 -.97 .330
Lot size; 8,000 - 13.999 s.f. 31,731.27 38,498.04 .82 .410 30,613.97 13,790.43 2.22 .027
Lot size; 14,000-25,000 s-f. 57,480.40 57,152.86 1.01 .315 24,831.09 20,606.16 1.21 .229
Lot size; 1-2.5 ac. -38,048.28 143,816.27 -.26 .791 53,169.87 51,641.12 1.03 .304
Linear Regression Coefficient Estimates for Land Uses within Quarter Mile Buffer
Building Value Models Land Value Models
Land Use Coefﬁcient Std. Error T Sig._ Coefﬁcient Std. Error T Sig.
Multi-Family
Commercial 236,907.86 93,490.80 2.53 .012 14,316.76 32,903.97 .44 .664
Industrial -66,765.66 40,271.31 -1.66 .098 -7,872.28 14,345.08 -.55 .583
Mobile Home 517,624.42 191,420.15 2.70 .007 85,865.39 67,980.33 1.26 .207
Public-Institutional -l25,228.10 64,178.12 —1.95 .051 -75,902.36 23,166.84 -3.28 .001
Railroad -18,502.33 50,785.74 -.36 .716 18,670.43 18,049.67 1.03 .301
Utility-Transportation -25,348.94 43,685.40 -.58 .562 3,266.18 15,565.47 .21 .834

 

60

Table 29 demonstrates that adjacent and surrounding land uses impact the value

of multi-family properties in different ways. Commercial land uses add signiﬁcant value

to multi-family properties, with adjacent commercial properties adding $81,372 to land

values and commercial uses within a quarter mile adding $236,907 to building value.

Public-institutional and utility-transportation uses located within a quarter mile detract

from multi-family value, while the presence of mobile homes may increase multi-family

property values.

Table 30, Impacts of Land Use Patterns on Commercial Property Values

 

 

 

 

 

 

 

 

Linear R sion Coefﬁcient Estimates for Adjacent Laud Uses
BuildingValue Models Land Value Models
Land Use Coeﬂ'rcient Std. Error T Sig. Coefﬁcient Std. Error T Sig.__
Commercial
Mimi-Family 24,536.60 55,062.68 .45 .656 87,542.09 28,817.55 3.04 .002
Industrial -194,305.85 57,665.17 -3.37 .001 -107.556.11 30,115.50 -3.57 .000
Public-Institutional -66,903.20 51,098.83 -1.31 .191 32,620.20 26,772.36 1.22 .223
Utility-Transportation -226,711.67 80,011.53 -2.83 .005 -129.1 1 1.30 41,850.66 -3.09 .002
Vacant 28,206.13 34,219.72 .82 .410 -25.l85.33 17,905.79 -1.41 .160
Single Family Residential
Lot size; less than 8,000 s.f. -25,323.45 32,770.66 -.77 .440 -24,895.79 17,128.45 -1.45 .146
Lot size; 8,000 - 13.999 s.£ 37,661.44 40,134.69 .94 .348 6.549.24 20,978.65 .31 .755
Lot size; 14,000-25.000 s.f. 8,030.67 54,960.57 .15 .884 -17,671.47 28,631.34 -.62 .537
Lot size; 1-2.5 ac. -121,7l4.18 134,981.65 -.90 .367 -120.311.38 70,720.23 -1.70 .089
Lot size; 2.5 - 5 ac. 567,080.44 199,584.43 2.84 .005 -777,696.10 102,202.38 -7.61 .000
Lot size; 5 - 10 ac. 220,017.96 219,072.29 1.00 .315 -788,537.45 114,050.14 -6.91 .000
Linear Regression Coefﬁcient Estimates for Land Uses within Quarter Mile Buﬁ'er
Building Value Models Land Value Models
Land Use Coefﬁcient Std. Error T 81L Coefﬁcient Std. Error T Sig._
Commercial
Multi-Family 76,496.50 40,704.15 1.88 .060 49,156.92 21,268.22 2.31 .021
Industrial -41,733.46 32,838.93 -1.27 .204 -34,460.57 17,142.28 -2.01 .045
Mobile Home -175,969.96 83,848.39 -2.10 .036 -238,432.54 43,904.03 -5.43 .000
Public-Institutional -82,554.67 49,797.53 -1.66 .098 -91,950.45 26,041.03 -3.53 .000
Railroad 43,047.06 35,266.43 1.22 .222 -l6,371.02 18,320.30 -.89 .372
Utility-Transportation -10,812.95 31,804.75 -.34 .734 45,828.70 16,627.99 2.76 .006

 

Table 30 shows that commercial building and land values appear to be affected by
fewer adjacent and surrounding land uses when compared to residential properties,

however the magnitude of value impacts is much higher when signiﬁcant relationships

61

exist. Regression ﬁndings suggest that certain mixes of land uses are ﬁscally beneﬁcial

while others are not. For example, adjacent multi-family land uses increase commercial

building values by $87,542, while adjacent high-intensity land uses such as industrial

facilities (-$301,861) and utility-transportation ('-$251,896) uses detract from commercial

property values. Multi-family land uses located within a quarter mile increase

commercial land values by $49,156 while industrial properties within a quarter mile

coincide with decreases of -$34,460.

Table 31. Impacts of Land Use Patterns on Industrial Property Values

Linear Regression Coefﬁcient Estimates for Adjacent Land Uses

 

 

 

 

 

 

 

 

Building Value Models Land Value Models
Land Use Coefﬁcient Std. Error T Sig. Coefﬁcient Std. Error T Sig.__
Industrial
Multi-Family 78,239.53 539,471.79 .15 .885 -72,024.61 122,312.66 -.59 .556
Commercial —146,7l7.50 276,296.48 -.53 .596 59,704.50 61,921.43 .96 .335
Public-Institutional -300_.492.46 614,328.89 -.49 .625 -150,861.65 136,264.95 -1.11 .269
Utility-Transportation 500,870.62 426,141.29 1.18 .240 42,854.35 96,346.89 .44 .657
Vacant 482,849.78 269,071.02 -1.79 .073 73,677.14 60,631.62 1.22 .225
Single Family Residential
Lot size: less than 8.000 s.f. -509,481.26 $54,360.50 -.92 .359 44,414.48 125,166.82 -.35 .723
Lot size; 8,000- 13,999 s.f. 334,703.46 525,338.57 .64 .524 46,728.16 118,313.81 -.39 .693
Lot size; 14,000-25.00051. 213,569.90 822,078.78 .26 .795 -348,953.32 185,836.96 -1.88 .061
Lot size; 1-2.5 ac. 992,379.80 796,766.66 1.25 .214 98,071.61 180,421.16 .54 .587
Linear Legression Coefﬁcient Estimates for Land Uses within Quarter Mile Buffer
Building Value Models Land Value Models
Land Use Coefﬁcient Std. Error T Sig._ Coefﬁcient Std. Error T Sig._
Industrial
Multi-Family -984,917.59 384,186.07 -2.56 .011 -277,221.22 84,880.60 -3.27 .001
Corrmrercial 561,106.38 422,985.07 1.33 .185 -365,780.60 94,583.50 -3.87 .000
Mobile Home 406,084.15 646,229.44 .63 .530 41,775.70 144,515.26 -.29 .773
Public-Institutional 141,123.23 281,438.38 .50 .616 15,470.56 63,652.01 .24 .808
Raih'oad 317,110.73 291,343.35 1.09 .277 203,290.35 65,883.89 3.09 .002
Utility-Transportation -129,836.71 318,866.48 -.41 .684 -165,890.58 71,341.25 -2.33 .020

 

Table 31 shows that surrounding land use patterns signiﬁcantly inﬂuence

industrial property values while adjacent land uses do not. Regression coefﬁcients

demonstrate that multi-family uses located within a quarter mile, can detract over -$1.2

million dollars (building and land coefficients combined) from industrial property values

62

while nearby commercial facilities detract -$365,780 in industrial land value. Railroads,
which often serve to support industrial facilities, increase industrial land values by
$203,290.
3.2.5 Adjacent and Surrounding Natural Amenities

“Table 32. Impacts of Natural Amenities on Single Family Residential Property
Values in Cities” reveals that natural amenities contribute positively and negatively to
property values in cities Lakefront properties add signiﬁcant value to city single family
residential properties. Small lakes, less than 19 acres do not signiﬁcantly impact building
values but increase adjacent land values by $50,850. Lakes between 19 and 49 acres in
size contribute $19,621 to building values but subtract over $15,000 in land value. Large
lakes, over 100 acres in size, add over $187,000 to adjacent property values. While
lakefront locations add value to buildings and land, properties within a quarter mile of
lakes of all sizes coincide with statistically signiﬁcant decreases in building and land
value in most cases, suggesting that distance to natural amenities may have both negative
and positive implications to property values. River and wetlands are associated with
negative contributions to city residential land values and are statistically signiﬁcant in all
cases except the adjacent riverside building value, suggesting that these areas should be
buffered and preserved. Natural areas are associated with decreases to adjacent residential
land values, however properties within a quarter mile of natural areas coincide with

$2,450 increases in residential land value.

63

Table 32. Impacts of Natural Amenities on Single Family Residential Property

Values in Cities

Linear Regression Coefﬁcient Estimates for Adjacent Natural Amenities

 

 

 

 

 

 

 

 

 

IBuildinLValue Models Land Value Models
Land Use Coefﬁcient Std. Error T §g_ Coefﬁcient Std. Error T 8'5;
City Single Family Residential
lake; 0-19 ac. 5,792.67 7,482.04 .77 .439 50,849.94 3,815.56 13.33 .000
Lake; 1949 ac. 19,621.46 4,446.30 4.41 .000 -15,013.57 2,288.28 -6.56 .000
Lake; 100+ ac. 95,824.07 9,016.60 10.63 .000 91,440.04 4,627.82 19.76 .000
River -1,598.10 3,492.97 -.46 .647 -38,320.62 1,786.11 -21.45 .000
Wetland -18,542.98 3,481.20 -5.33 .000 -24,695.33 1,782.00 -13.86 .000
Natural Area; 4,833.94 2,983.54 -1.62 .105 -15,269.86 1,532.73 -9.96 .000
priority three
Linear Rgegre_ssion Coefﬁcient Estimates for Natural Amenities within Quarter Mile Buffer
Build'ﬂValue Models Land Value Models
Land Use Coefﬁcient Std. Error T Sig._ Coefﬁcient Std. Error T Sig._
City Single Family Residential
lake; 0-19 ac. 42.67 356.99 -.12 .905 -894.56 181.99 4.92 .000
Lake; 1949 ac. -17,589.89 1,056.86 -16.64 .000 -2,612.07 540.61 -4.83 .000
Lake; 100+ ac. -2,824.00 977.26 -2.89 .004 2,791.44 496.04 5.63 .000
River -5,497.61 577.05 -9.53 .000 -3,720.10 296.92 -12.53 .000
Wetland -9,305.82 731.95 -12.71 .000 -1,982.07 371.06 -5.34 .000
Natural Area; 49,694.16 25,274.47 -1.97 .049 -24,673.84 13,023.62 -1.89 .058
priority two
Natural Area; 516.64 705.29 .73 .464 2,453.29 361.08 6.79 .000
priority three

 

Table 33 demonstrates that natural amenities contribute positively and negatively

to property values in townships and villages more so than in case study cities. Lakes of

all sizes contribute signiﬁcantly and positively to adjacent residential building and land

values, with value contributions ranging from approximately $5,000 (lake 0-19 acres) to

nearly $175,000 (lakes over 100 acres). The size of the lake plays a key role in

determining contribution to value, with signiﬁcant variations demonstrated in both

building and land value models. Unlike city residential properties, township and village

residential properties beneﬁt from quarter mile proximity to lakes, with statistically

signiﬁcant results demonstrated in most cases. Similar to city residential properties,

wetlands and rivers decrease property values for adjacent and proximate township and

village residential properties. This ﬁnding suggests that preservation of sensitive

64

environmental resources, and buffering those resources from development, may create a
win-win scenario for environmental protection and ﬁscal planning objectives. Natural
areas contribute signiﬁcantly to residential building and land values in some cases, with
adjacent and buffer properties experiencing decreases in land values and increases in
building value.

Table 33. Impacts of Natural Amenities on Single Family Residential Property
Values in Townships and Villages

Linear Regression Coefﬁcient Estimates for Adjacent Natural Amenities

 

 

 

 

 

 

 

 

 

Building Value Models Land Value Models
Land Use Coefﬁcient Std. Error T Sig._ Coefﬁcient Std. Error T Sig._
Township Single Family Residential
Lake; 0—19 ac. 2,731.60 1,602.58 1.70 .088 5,004.94 1,355.54 3.69 .000
Lake; 1949 ac. 19,644.35 4,412.70 4.45 .000 58,435.87 3,732.68 15.66 .000
Lake; 49-100 ac. 11,973.92 3,202.98 3.74 .000 11,796.40 2,708.54 4.36 .000
Lake; 100+ ac. 35,926.12 1,421.96 25.27 .000 139,041.43 1,183.92 117.44 .000
River 827.36 2,675.13 .31 .757 12,156.41 2,257.15 5.39 .000
Wetland -5,264.95 1,609.35 -3.27 .001 -1,600.35 1,359.47 -1.18 .239
Natural Area; 4,529.91 2,758.66 1.64 .101 -12,789.11 2,317.31 -5.52 .000
priority one
Natural Area; -3,350.11 2,035.43 -1.65 .100 -3,406.31 1,719.27 -1.98 .048
priority two
Natural Area; 7,116.20 1,774.50 4.01 .000 -164.44 1,499.19 -.11 .913
priority three
Linear Regression Coefﬁcient Estimates for Natural Amenities within Quarter Mile Buffer
4 Buildinﬂalue Models Land Value Models
Land Use Coefﬁcient Std. Error T Srg' . Coefﬁcient Std. Error T Sig._
Township Single Family Residential
Lake; 0-19 ac. -2,034.92 1,093.46 -1.86 .063 9,278.34 919.19 10.09 .000
Lake; 1949 ac. 6,764.65 1,261.74 5.36 .000 906.20 1,064.93 .85 .395
Lake; 49-100 ac. 11,599.85 1,657.60 7.00 .000 -1 1,322.13 1,379.00 -8.21 .000
Lake; 100+ ac. 13,817.19 941.10 14.68 .000 9,281.73 777.68 11.94 .000
River -1,496.60 840.24 -1.78 .075 -2,512.02 708.66 -3.54 .000
Wetland -3,636.49 762.33 -4.77 .000 -6,667.28 643.86 -10.36 0
Natural Area; -846.65 1,298.68 -.65 .514 -11,460.44 1,081.08 -10.60 .000
priority one
Natural Area; 5,652.21 924.96 6.11 .000 -251.88 778.82 -.32 .746
priority two
Natural Area; 6,980.95 726.83 9.60 .000 4,147.45 613.35 -6.76 .000
priority three

 

65

For multi-family and non-residential land uses, natural amenities rarely contribute
to property values. Table 34 demonstrates statistically signiﬁcant cases for multi-family
uses that include lakefront properties (-$129,862) and properties within a quarter mile of
rivers (-$48,600). Both commercial ($79,861) and industrial ($332,019) property values
beneﬁt from lakeside ownership. Industrial properties adjacent and within a quarter mile
of wetlands, coincide with decreases in building value.

Table 34. Impacts of Natural Amenities on Non-Residential and Multi-

 

 

 

 

 

 

 

 

 

Family Property Values
Linear Regression Coefﬁcient Estimates for Adjacent Natural Amenities
-Building Value Models Land Value Models

Land Use Coefﬁcient Std. Error T Sig._ Coefﬁcient Std. Error T Sig._
Multi-Family

Lake;0-19 ac. -37,495.73 166,658.88 -.225 .822 -129,863.00 59,263.51 -2.191 .029
Wetland -203,071.47 165,099.64 -1.230 .219 -110,520.69 58,795.96 -1.880 .061
Commercial

Lake; 0—19 ac. 219,476.21 110,451.83 1.987 .047 -l39,615.62 57,369.90 -2.434 .015
River 27,934.19 142,176.17 .196 .844 -94,833.76 74,386.25 -1.275 .203
Wetland 29,619.80 156,545.24 .189 .850 -227,460.39 80,975.18 -2.809 .005
Industrial

Lake; 0-19 ac. 496,230.67 418,284.08 1.186 .236 332,019.76 93,850.32 3.538 .000
Wetland 4,540,143.87 715,685.57 -2.152 .032 51,414.42 161,147.21 .319 .750

Linear Regression Coefﬁcient Estimates for Natural Amenities within Quarter Mile Buffer
Building Value Models Land Value Models

Land Use Coefﬁcient Std. Error T Sig._ Coemcieut Std. Error T Sig;
Multi-Family

River 60,475.50 69,601.84 .87 .385 48,607.33 24,619.53 -1.97 .049
Wetland -64,647.04 71,827.60 -.90 .368 40,380.50 24,873.81 -l.62 .105
Commercial

River 38,893.72 42,291.10 .92 .358 —36,142.94 22,027.05 -1.64 .101
Wetland 16,100.75 46,338.40 .35 .728 67,747.82 24,256.48 2.79 .005
Industrial

River -674,879.47 480,539.87 -1.40 .161 -15,018.53 108,724.76 -.14 .890
Wetland -785,535.12 301,939.14 -2.60 .010 53,167.66 68,425.20 .78 .438

 

3.2.6 Adjacent and Surrounding Capital Facilities

“Table 35. Impacts of Capital Facilities on Single Farrrily Residential Property
Values” shows the impacts of adjacent and surrounding parks, recreation areas, schools,

government facilities, and police and ﬁre stations on residential property values in cities.

66

Recreation and parks contribute signiﬁcantly to adjacent residential buildings in both
cities and townships. Recreation uses contribute positively to adjacent city residential
building values ($6,076) and township residential land values located within a quarter
mile ($3,294) but detract from adjacent township building values (-$7,238) and city
property values within a quarter mile (-$2,222). Schools signiﬁcantly inﬂuence city
residential property values when located within a quarter mile distance, however impacts
are marginal. Government uses located within a quarter mile distance detract value from
residential building values in cities (-$4, 1 97) but add value to township and Villages
($9,378). Fire stations located within a quarter mile increase city residential building
values, but do not signiﬁcantly contribute to property values in townships and Villages.
Police stations located within a quarter mile increase city residential property values

($3,456), but detract from township and village residential building values (-$l6,724).

67

Table 35. Impacts of Capital Facilities on Single Family Residential Property Values

Linear Reggession Coefﬁcient Estimates for Adjacent Capital Facilities

 

 

 

 

 

 

Building Value Models Land Value Models
Land Use Coefﬁcient Std. Error T Sig._ Coefﬁcient Std. Error T Sig;
City Single Family Residential
Recreation 6,076.15 1,821.00 3.34 .001 1,342.52 937.47 1.43 .152
Park -13,58l.27 2,053.91 -6.61 .000 -216.33 1,056.39 -.20 .838
School -2,340.98 2,539.59 -.92 .357 -2,818.00 1,307.40 -2.16 .031
Government -1,401.95 7,672.09 -.18 .855 5,897.01 3,953.62 1.49 .136
Township and Village Single Family Residential
Recreation -7,283.65 1,484.35 -4.91 .000 -l,908.63 1,253.68 -l.52 .128
Park 4,034.46 3,575.06 —1.13 .259 3,037.46 3,024.85 1.00 .315
School -6.479.65 4,692.25 -1.38 .167 419.75 3,972.82 .11 .916
Government 3,552.57 5,600.75 .63 .526 3,919.95 4,742.05 .83 .408
Linear Regression Coefﬁcient Estimates for C3pital Facilities within Quarter Mile Buffer
Building Value Models Land Value Models

 

Land Use Coeﬁicient Std. Error T Sig; Coefﬁcient Std. Error T Sig._
City Single Family Residential

 

 

Recreation -1,153.19 545.72 -2.11 .035 -1,069.94 279.91 -3.82 .000
Park 3,135.15 359.42 8.72 .000 -288.27 183.39 -1.57 .116
School 813.38 317.45 2.562 .010 -791.65 163.06 4.855 .000
Goverrmrent 4,197.20 812.89 -5.163 .000 698.47 414.16 1.686 .092
Fire 4,440.98 953.37 4.658 .000 928.81 486.59 1.909 .056
Police 5,805.79 1,971.56 2.945 .003 -2,349.89 1,014.10 -2.317 .020
Township and Village Single Family Residential

Recreation 51.75 750.18 .07 .945 3,294.10 633.36 5.20 .000
Park 2,304.59 1,014.16 2.27 .023 645.65 855.08 .76 .450
School 1,125.70 1,236.28 .911 .363 -l,326.3l ’ 1,044.47 -l.270 .204
Government 9,378.51 3,031.94 3.093 .002 -1,992.92 2,560.94 -.778 .436
Fire 4,641.64 3,248.61 -1.429 .153 4,498.92 2,746.40 1.638 .101
Police -l6,724.87 3,607.37 4.636 .000 -2,650.53 3,048.23 -.870 .385

 

Table 36 demonstrates that in most cases, the location and type of capital facilities
do not impact multi-family and non-residential property values. Police stations located
within a quarter mile contribute positively to multi-family property values ($323,995) and
commercial building values. Industrial property values may beneﬁt greatly from quarter

mile proximity to govermnent uses ($1.5 million) and recreation uses ($242,506).

68

Table 36, Impacts of Capital Facilities on Non-Residential and Multi-Family

 

 

 

 

 

 

 

 

 

Property Values
Linear Regression Coefﬁcient Estimates for Adjacent Capital Facilities
Building Value Models Land Value Models

Land Use Coefﬁcient Std. Error T Sig; Coefﬁcient Std. Error T Sig._
Commercial

Recreation -79,660.87 103,802.54 -.767 .443 -127,501.84 54,329.25 -2.347 .019
Park 38,372.81 142,581.30 .269 .788 15,751.15 74,698.75 .211 .833
School -66,312.59 131,167.63 -.51 .613 79,423.59 68,444.81 1.16 .246
Government -323,329.01 140,943.36 -2.29 .022 45,973.77 72,238.90 -.64 .525
Industrial

Recreation 245,035.46 627,609.27 .390 .696 41,431.47 142,149.78 -.291 .771
Park -1,716,731.28 1,048,135.55 -l.638 .102 -193,623.87 236,468.65 -.819 .413
Multi-Family

Recreation -30,243.19 172,502.30 -.175 .861 149,210.81 61,860.92 2.412 .016
Park 588,377.34 222,216.89 2.648 .008 -58,720.06 80,276.76 -.731 .465
School 381,162.29 229,263.27 1.66 .097 -106,873.06 82,833.46 -1.29 .197

Linear Regression Coefﬁcient Estimates for Capital Facilities within Quarter Mile Buffer
Building Value Models Land Value Models

Land Use Coefﬁcient Std. Error T Sig._ Coefﬁcient Std. Error T SiL
Commercial

Recreation -35,015.88 39,987.56 -.88 .381 -10,187.11 20,859.12 -.49 .625
School 449.08 31,048.33 -.01 .988 4,291.39 16,245.16 .26 .792
Govermnent 17,482.51 45,065.25 .39 .698 38,930.23 23,533.15 1.65 .098
Fire -71,161.93 55,211.18 -1.29 .198 -3,509.06 28,822.08 -.12 .903
Police 140,958.09 60,601.47 2.33 .020 -27,078.04 31,655.68 -.86 .392
Industrial

Recreation 252,919.40 “328,979.10 .77 .442 242,506.64 74,138.54 3.27 .001
School -308,137.17 362,148.41 -.85 .395 48,944.29 81,878.76 -.60 .550
Govermnent 1,215,231.31 443,888.78 2.74 .006 343,059.78 100,464.14 3.41 .001
Multi-Family

Recreation 79,593.75 69,549.82 1.14 .253 5,828.05 24,351.77 .24 .811
Park 39,412.33 45,960.83 .86 .391 27,216.34 16,162.08 1.68 .093
School -22,538.39 42,114.00 -.54 .593 -3.368.92 15,209.03 -.22 .825
Government 23,598.19 56,592.10 .42 .677 -14,069.60 20,086.70 -.70 .484
Fire -100,292.28 82,535.61 -1.22 .225 -23,547.36 29,445.19 -.80 .424
Police 261,859.27 87,020.89 3.01 .003 62,096.69 30,472.25 2.04 .042

 

3.2.7 Transportation Infrastructure and Access

Transportation access and trafﬁc patterns were hypothesized to have signiﬁcant

impacts on assessed residential and non-residential property values, particularly land

values. The application of road classiﬁcations identiﬁes impact of location along

69

 

interstate, ﬁeeway, arterial, collector, or local road classes managed by county and cities

as deﬁned by the US Department of Transportation.

Table 37. Impacts of Transportation Infrastructure on City Residential Property

 

 

 

 

 

Values
Linear Regression Coefﬁcient Estimates for Road Classiﬁcations
‘ Building Value Models Land Value Models
Land Use Coefﬁcient Std. Error T Sig._ Coefficient Std. Error T Sig._
City Single Family Residential
Road not classiﬁedinPA -8,833.07 1,013.03 -8.72 .000 12,345.12 509.32 24.24 .000
51
Urban Interstate 4,380.94 2,724.28 -1.61 .108 -7,021.01 1,402.73 -5.01 .000
Urban Non-Freeway -5,506.62 1,172.93 4.69.000 -1,868.05 598.03 -3.12 .002
CountyLocal, Urban -14,911.11 4,909.56 -3.04 .002 -7,301.14 2,528.72 -2.89 .004
Local
CountyPrimary, Urban -22,796.41 3,921.95 -5.81 .000 -3,689.39 2,020.54 -l.83 .068
Non-Freeway
CountyPrimary, Urban -13,544.27 3,262.76 4.15 .000 -264.17 1,680.24 -.16 .875
Minor Arterial
County Primary, Urban -2,382.49 4,068.52 -.59 .558 1,123.95 2,096.81 .54 .592
CollectorPrimary
City Major, Urban Local 4,848.41 512.69 9.46.000 -263.17 262.52 -1.00 .316
City Major, Urban 3,397.41 653.13 5.20.000 -2,617.69 335.42 -7.80 .000
Collector
CityMajor, Minor -11,108.37 948.40 -11.71 .000 -1,955.97 486.77 4.02 .000
Arterial
CityMajor, Urban Non— -7,697.75 748.93 -10.28 .000 -6,508.41 384.80 -16.91 .000
Freeway
City Minor, Urban Local 3,164.05 532.15 5.95 .000 329.71 273.16 1.21 .227

Road

 

Table 37 shows that road functions coincide with statistically signiﬁcant impacts

on single family residential building and land values in cities. Hi gh-impact, high traveled

roads detract from adjacent residential property values. Urban interstates (with and

without ﬁ'eeway status) subtract -$7,021 and -$7,3 74 from residential property values,

county primary non-freeway roads subtract over -$26,000 and arterial roads detract

between -$13,808 and -$13,063 respectively. Increases in residential property value

coincide with location along urban local and collector roads ranging from a high of

$4,848 for urban major local roads to $780 for urban collector roads.

70

Table 38. Impacts of Transportation Infrastructure on Residential Property Values

in Townships and Villages

Linear Regression Coefﬁcient Estimates for Road Classiﬁcations
Building Value Models

Land Value Models

 

 

 

Land Use Coefﬁcient Std. Error T Sig._ Coefﬁcient Std. Error T Sig._
Township and Village Single Family Residential

Road not classiﬁed in PA 1,408.35 936.45 1.50 .133 1,756.84 789.07 2.23 .026

51

Urban Interstate 6,889.12 45,203.18 .15 .879 212,251.28 38,146.69 5.56 .000
Urban Non-Freeway -12,395.62 18,112.57 -.68 .494 5,099.50 15,333.90 .33 .739

County Local, Urban -3,162.04 949.71 -3.33 .001 -7,l62.17 801.78 -8.93 .000

Local

CountyLocal, Rural 12,533.40 2,378.43 5.27 .000 6,750.61 2,002.87 3.37 .001

Local

CountyLocal, Urban -l6,891.13 2,457.98 -6.87 .000 -6,325.85 2,071.85 -3.05 .002

Collector Local

County Primary, Rural -20,582.48 4,831.32 4.26 .000 16,579.30 4,071.46 4.07 .000

Minor Arterial

Cormty Prirmry, Rural -11,066.32 3,723.92 -2.97 .003 21,869.50 3,121.77 7.01 .000
Major Collector

County Primary, Urban 4,393.31 3,951.78 -1.11 .266 4,399.09 3,343.58 -1.32 .188

Non-Freeway

County Primary, Urban -1,533.37 1,226.68 -1.25 .211 -6,694.03 1,029.98 -6.50 .000
Minor Arterial

Cormty Primary, Urban 4,676.29 2,131.09 -2.19 .028 -9,584.57 1,797.78 -5.33 .000
Collector Primary

City Major, Urban Local -1,579.25 2,709.24 -.58 .560 -6,674.44 2,288.41 -2.92 .004
City Major, Urban 19,743.66 5,042.43 3.92 .000 4,491.68 4,254.15 1.06 .291

Collector

City Major, Minor 11,713.06 3,041.97 3.85 .000 6,154.22 2,566.67 2.40 .017
Arterial

City Minor, Urban Local 2,234.54 1,384.37 1.61 .107 1,159.56 1,155.89 1.00 .316
Road

 

Table 38 demonstrates that unlike cities, township and Village residential land
Values located along interstates coincide with value increases of $212,251. This ﬁnding
may suggest that township and village residents value efﬁcient access to regional
destinations which supports their rural or suburban living preference. Arterial roads
impact township, Village and city residential prOperty values in similar ways. County
rural primary (-$4,003) and urban arterials (-$6,694), along with county urban collector (—
$23,216) all detract from residential property values however collector roads in Villages

add $19,743 and arterials add $11,713 to property values. Rural local roads add $39,026

71

to property values while urban local roads subtract between $-6,674 and $10,324
respectively. Overall, road functions and class seem to have a larger impact on residential
property values in townships and villages than cities, however it should be noted that
spatial autocorrelation may bias regression coefﬁcients.

Table 39. Impacts of Transportation Infrastructure on Multi-Family Property

 

 

 

 

Values
Linear R ion Coelﬁcient Estimates for Road Classiﬁcations
Building Value Models Land Value Models
Land Use Coefﬁcient Std. Error T ﬁg. Coefﬁcient Std. Error T Sig._
Multi-Family
Road not c'lassiﬁedinPA 24,573.37 63,403.29 .39 .698 68,614.25 21,701.20 3.16 .002
51
UrbanNon-Freeway -252,818.31 74,118.23 -3.41 .001 -7,311.87 26,563.27 -.28 .783
CountyLocal, Urban -317,586.95 189,791.61 -1.67 .095 -172,960.61 67,167.22 -2.58 .010
Local
CountyPrirnary, Urban 66,495.39 144,324.58 .46 .645 29,348.59 51,951.16 .56 .572
Non-Freeway
CountyPrimary, Urban -135,l91.42 90,837.63 -1.49 .137 -20,242.69 32,791.58 -.62 .537
Minor Arterial
CityMajor, Urban Local -97,063.50 50,934.06 -1.91 .057 -8,387.18 18,367.16 -.46 .648
City Major, Urban 36,709.13 49,909.47 .74 .462 12,753.14 17,942.51 .71 .477
Collector
City Major, Minor -348,201.70 86,663.16 -4.02 .000 29,060.11 30,483.77 .95 .341
Arterial
City Major, Urban Non- 144,014.36 46,810.06 3.08 .002 80,596.92 16,618.80 4.85 .000
Freeway
City Minor, Urban Local -20,666.13 42,179.26 -.49 .624 46,517.42 14,843.56 3.13 .002
Road

 

Unlike, single family residential and non-residential land uses, regression

coefﬁcients for multi-family uses shown in Table 39 do not demonstrate a clear

relationship with road classiﬁcation. Multi—family locations along city non-freeway

interstates coincide with a $224,610 increase in property value, however, state non-

freeway interstates subtract -$252,818 from multi-family building values. Locations

along minor local roads result in an increase of $46,5 1 7 in multi-family land values,

however county urban local roads coincide with a decrease of -$172,960 in land values.

72

Table 40. Impacts of Transportation Infrastructure on Commercial Property

 

 

 

 

Values
Linear ion Coeicient Estimates for Road Classiﬁcations
Building Value Models Land Value Models
Land Use Coefﬁcient Std. Error T Sih Coefﬁcient Std. Error T Sig._
Commercial
Road not classiﬁedinPA -60,490.83 46,962.79 -1.29 .198 45,343.48 24,548.05 1.85 .065
51
Urban Interstate -30,064.67 38,890.25 -.77 .440 147,893.97 20,270.87 7.30 .000
CountyLocal, Urban 160,281.39 73,742.37 2.17 .030 89,635.71 38,646.36 2.32 .020
Local
CountyPrimary, Urban 78,284.69 68,009.25 1.15 .250 99,009.51 35,525.36 2.79 .005
Non-Freeway
CountyPrimary, Urban -59,789.93 48,745.77 -1.23 .220 -63,783.77 25,399.41 -2.51 .012
Minor Arterial
CountyPrirnary, Urban -83,312.98 133,250.71 -.63 .532 47,961.38 69,734.67 -.69 .492
CollectorPrimary
CityMajor, Urban Local 30,134.32 39,372.46 -.77 .444 1,901.79 20,571.91 .09 .926
CityMajor, Urban 39,386.71 41,050.52 .96 .337 -18,702.04 21,503.97 -.87 .385
Collector
City Major, Minor -44,135.51 41,501.82 -1.06 .288 55,780.09 21,689.10 2.57 .010
Arterial
CityMajor, Urban Non- 48,313.43 36,449.38 -1.33 .185 28,080.45 18,925.20 1.48 .138
' Freeway
CityMinor, Urban Local -50,633.73 27,767.31 -1.82 .068 -17,319.19 14,534.38 -1.19 .234
Road

 

Table 40 shows that commercial land values are signiﬁcantly inﬂuenced by road

functional classiﬁcations, however building values are not. Commercial locations along

highly traveled urban interstates ($147,893), urban non-freeway ($99,009) and city

arterials ($55,780) increase land values while locations along minor county arterials (-

$63,783) detract from property values.

73

Table 41. Impacts of Transportation Infrastructure on Industrial Property Values

Linear Regression Coefﬁcient Estimates for Road Classiﬁcations
- Building Value Modeb Land Value Models
Land Use Coefﬁcient Std. Error T SiL Coefﬁcient Std. Error T Sig._
Industrial
Road not classiﬁed in PA -536,704.82 277,754.64 -1.93 .054 -19,194.53 62,963.87 -.30 .761
51

 

 

Urban Interstate -l,168,102.65 669,949.46 -1.74 .082 -62,905.34 151,146.15 -.42 .677
County Local, Urban -837,911.86 359,109.78 -2.33 .020 414,944.94 81,089.36 -.55 .580
Local

CountyPrimary, Urban 1,303,341.30 739,550.35 1.76 .079 432,081.75 166,485.62 2.60 .010
Non-Freeway

CountyPrimary, Urban 142,479.99 390,852.51 .36 .716 -67,164.30 88,398.57 -.76 .448
Minor Arterial

CountyPrimary, Urban -578,804.03 980,553.92 -.59 .555 -14,645.59 221,365.32 -.07 .947
CollectorPrirnary

City Major, UrbanLocal 228,313.75 384,978.19 .59 .553 313,689.56 86,614.45 3.62 .000
CityMajor, Urban 213,377.80 379,313.13 .56 .574 292,063.81 82,833.25 3.53 .000
Collector

City Major, Minor -580,258.82 538,657.21 -1.08 .282 418,349.26 119,939.22 3.49 .001
Arterial

CityMajor,UrbanNon- -1,896,920.86 580,206.22 -3.27 .001 442,581.37 125,248.15 3.53 .000
Freeway

City Minor, Urban Local -560,424.52 325,216.42 -1.72 .086 78,921.74 72,265.97 1.09 .275
Road

Table 41 demonstrates a positive relationship between travel intensity and
industrial land values. Industrial location along city local ($313,689), collector
($292,063), minor arterials ($418,349) and non-freeway ($442,581) road classes along
with county non-freeway ($432,081) classiﬁcations increase land value.

3.1 Quantiﬁcation of Implicit Property Tax Revenues Attributed to Surrounding
Land Use Patterns and Natural Amenities

Through the application of multiple regression analysis techniques and hedonic
pricing theory, statistically signiﬁcant indirect relationships between assessed property
values and surrounding land use and natural amenity patterns were discovered The value
dynamics between property values and surrounding land use patterns and natural
amenities translate into implicit property tax revenues for the case study communities

which are presented in Tables 42 through 47.

74

Table 42. City Total Implicit Single Family Property Tax Revenues
Attributed to Surrounding Land Use Patterns

 

 

Land Use City of Pontiac City of Royal Oak City Total
Multi—Family -$321,684 -s424,773 -$746,458
Commercial $543,944 $643,678 $1,187,622
Industrial -$305,048 -$lO9,637 -$414,684
Mobile Home -$66,830 $0 -$66,830
Public-Institutional -$320,809 -$335,910 -$656,718
Utility-Transportation -$ 195,042 -$ 144,079 -$339,121
Vacant -$106,947 -$7,292 -81 14,240

 

Table 42 reveals that within case study cities, commercial properties adjacent to
and within a quarter mile of single family residential properties contribute over $543,000
to the property tax base in the City of Pontiac and over $673,000 to the property tax base
of the City of Royal Oak. In cities, adjacent commercial land uses signiﬁcantly detract
from assessed building values but signiﬁcantly add to adjacent single family land values.
Overall adjacent commercial properties decrease single family property tax revenues in
the case study communities by nearly -$54,000. Commercial uses located within a quarter
mile proximity increase property tax revenues attributed to building value by over $2.03
million while subtracting from property taxes from residential land values by nearly
$796,000. Overall commercial lands add nearly $1.2 million dollars in property tax
revenue through value-added relationships with single family residential lands in case
study cities.

Other adjacent and surrounding land uses, detract from single family property
values and ultimately single family property tax revenues. The biggest indirect ﬁscal
impact on city residential single family property taxes are associated with multi-family

land uses (-$746,45 8), public-institutional uses (-$656,718) and industrial uses

75

(-$414,684).

Table 43. Township and Village Total Implicit Single Family Property Tax Revenues Attributed to
Surrounding Land Use Patterns
Commerce Wolverine Milford Milford Oakland Township and

 

Community TWp. Lake Village TWp. Village Np. Village Total
Agricultural $45,238 $0 $1 5,207 $0 $20,04 $80,486
Mimi-Family -$107,4 14 -$51,404 -$9,032 -$1 19,455 -$13,349 -$300.653
Commercial -$196,042 -$55,549 -$2 1 ,3 73 -$1 15,356 -$42,3 63 -$430,683
Industrial -$4.188 $0 -$5,213 -$2,757 $0 -$12.158
Mobile Home -$33,188 $0 -$17,086 $0 -$3.071 -$53,345
Public-Institutional -$31 1,510 -$100,587 -$54,205 -$162,394 -$92,280 -$720,975
Utility-Transportation -Sl 94,441 -SS .455 -$27,728 -$79,315 -$89,217 -$396, 157
Vacant $504,679 $72,668 $255,587 $71,671 $248,235 $1,] 52,840

 

Table 43 demonstrates that in case study townships and villages, surrounding
agricultural uses and vacant land add value to residential properties and ultimately the
residential property tax base, suggesting that “rural amenities” indirectly beneﬁt the
residential property tax base in case study townships. When located within a quarter mile
radius, agricultural uses increase residential property tax revenues via land values by over
$80,486 while vacant parcels, when located adjacent to residential properties, increase
single family property tax revenues by over $1.15 million in townships and villages.
Surrounding “urban” non-residential and multi-family uses detract from the residential
property tax base, conﬁrming that “rural character” is a source of value to both residents
and government in rural communities. Unlike cities, adjacent and surrounding
commercial uses detract from residential building and land values resulting in an overall
decrease of -$430,683 from the residential property tax base. Other uses such as public-
institutional and utility-transportation and multi-family uses detract from single family

property values and property tax contributions.

76

Table 44. Township and Village Total Implicit Single Family Property Tax Revenues Attributed to
Surrounding Water-based Natural Amentites

Lake, 19-49 Lake, 49-100 Lake, 100+

 

 

Community Lake, 0-19 acres acres acres acres River Wetland
Commerce Twp. $1,132,842 $186,520 $33,358 $3,050,171 $12,904 —$296,232
Milford Twp. $306,550 $21,280 $14,202 $144,997 $3,964 $112,719
Milford Village $167,876 $0 $7,805 $15,948 $1,343 -$65,858
Oakland Twp. $492,731 $6,619 $17,619 $0 $27,527 -$214,634
Wolverine Lake Vill. $210,649 $16,182 $0 $1,220,969 $0 -$38,909
mship and

Village Total $2,310,648 $230,601 $72,984 $4,432,085 $45,737 -$728.3 53

 

Modeling the relationship between single family residential property taxes and
surrounding natural amenities conﬁrms that “rural” natural amenities, especially lakes
and natural conservation areas, are a major source of public value in rural and suburban
townships and villages. Table 44 demonstrates that lakes of all sizes, contribute to the
property tax base indirectly by adding value to single family residences. Small lakes, 0 —
19 acres, add over $2.3 million in property tax revenues. Small lakes add over $56,000 in
property tax revenue through adjacent lakefront properties and add over $2.25 million to
township and village revenues by increasing surrounding residential land values located
within a quarter mile. Large lakes, over 100 acres in size, contribute a total of $4.4
million in property tax revenues indirectly and over $3 million in property tax revenue to
Commerce Charter Township alone. Large lakes increase waterfront property tax
revenues by $2.6 million and increase residential property tax revenues located within a
quarter mile by $1.8 million. While lakes add value to waterfront and surrounding home
values, not all water-based natural amenities contribute positively to the property tax
base. Locations next to wetlands decrease property taxes from residential building value
by over -$64,000 while quarter mile proximity decreases property taxes by over -

$663,000.

77

Table 45. Township and Village Total Implicit Single Family Property Tax
Revenues Attributed to Surrounding Land-based Natural Amentites

 

 

 

Natural Area, Natural Area, Natural Area,
Community priority one priority two priority three
Commerce Twp. -$9,073 $81,004 $369,781
Milford Twp. -$31,312 $44,815 $140,322
Milford Village -$919 $20,867 $24,976
Oakland Twp. -$10,985 $105,681 $168,248
Wolverine Lake Vill. $0 $0 $24,080
Township and
Village Total -$52,289 $252,367 $727,407

 

Modeling outcomes shown in Table 45 reveal that natural conservation initiatives
add value to the residential property tax base indirectly by increasing surrounding
residential home values. While natural areas with the highest level of conservation
priority detract from adjacent residential land values, quarter mile proximity to priority
two and three natural areas increase property taxes derived from building values by
$280,000 and $656,000 respectively. Modeling results suggest that preserving high
priority natural areas such as wetlands while enabling residential access to high quality
lakes and natural areas can provide win-win scenarios for ﬁscal and environmental
planning objectives.

Table 46. City Total Implicit Single Family Property Tax Revenues
Attributed to Surrounding Natural Amentites

 

 

Community City of Pontiac City of Royal Oak City Total
Lake, 0-19 acres -$56,292 -$4l,544 -$97,836
Lake, 19-49 acres -$40,655 $0 -$40,655
Lake, 100+ acres $43,026 $0 $43,026
River -$348,098 $0 -$348,098
Wetland -$267,873 $0 -$267,873
Natural Area, priority two -$269,349 $0 -$269,349
Natural Area, priority three $37,569 $10,498 $48,067

 

78

Contrary to ﬁndings in townships and villages, Table 46 shows that city
residential properties do not beneﬁt monetarily from surrounding water-based natural
amenities. While lakes only marginally increase city waterfront land and building values,
they detract from surrounding neighborhood property values as a whole, suggesting that
water quality or access may play a major role in determining indirect value contributions
of natural amenities to a city’s property tax base. Priority three natural conservation areas
add value indirectly to property tax revenues by increasing residential land values located
within a quarter mile, suggesting that open space may amplify the positive ﬁscal impacts

of higher housing densities.

Table 47. Total Implicit Non-Residential and Multi-Family Property Tax
Revenues Attributed to Surrounding Land Use Patterns
Surrounding Land Uses

 

 

Land Use Analyzed Multi-Family Commercial Industrial
Multi-Family - $3,615,585 -$37,227
Commercial $1,958,748 - -$1,524,416 .
Industrial -$8,271,685 -$4,176,748 -

 

Dynamic ﬁscal relationships exist between surrounding land use patterns and non-
residential and multi-family property values that create positive and negative indirect
sources of property tax revenue within the case study communities. “Table 47. Total
Implicit Non-Residential and Multi-Family Property Tax Revenues Attributed to
Surrounding Land Use Patterns,” demonstrates how surrounding land uses (shown in
columns) impact the property tax revenues of multi-family, commercial and industrial
land uses (in rows). When compared to single family residential land uses, comparatively

fewer statistically signiﬁcant relationships exist between surrounding land uses patterns

79

 

and non-residential property values, however, the magnitude of these monetary
relationships can be great.

A mutually beneﬁcial ﬁscal relationship exists between multi-family and
commercial land uses. When located within a quarter mile, commercial land uses increase
multi-family building values, indirectly increasing property tax revenues by over $3.45
million in the case study communities. Multi-family land uses located within a quarter
mile of commercial properties, increase commercial land values, resulting in property tax
increases of $1.7 million. Commercial and multi-family land uses also positively beneﬁt
from being located adjacent to one another, with land value increases contributing over
$386,000 to the bottom line.

In case study cities, regression results reveal that mixed-use zoning policies have
contributed positively and negatively to private property values and public property tax
revenues. Findings suggest that mixed-use policies that encourage a mix of residential,
multi-family and commercial development result in net increases in property values and
property tax revenues within case study cities. In these mixed-use development scenarios,
integration of commercial uses appear to mitigate the negative ﬁscal relationships
between surrounding single family and multi-family property values through increased
neighborhood livability.

Mixed-use policies that encourage a mix of industrial, multi-family and
commercial development result in a net decrease in property values and property tax
revenues within case study communities. Industrial properties located adjacent to multi-
family and commercial uses decrease commercial and multi-family property tax

contributions by over -$7 97,000 combined. Industrial uses located within a quarter mile

80

of commercial properties, detract approximately -$764,000 in property tax revenue
derived from commercial land. While it comes as no surprise that high intensity industrial
uses negatively impact neighboring land uses, regression results reveal that multi-family
and commercial land uses have a much greater negative impact on industrial land values
than vice versa. Multi-family and commercial land uses located within a quarter mile of
industrial uses, subtract over $12 million in industrial property tax contributions through
decreases in building and land value. Findings suggest that clustering non-residential land
uses while buffering industrial uses can maximize industrial property tax contributions

while mitigating nuisance effects such as trafﬁc, noise, air pollution, and aesthetics.

81

CHAPTER 4.
SUMMARY OF FINDINGS, LIMITATIONS, SUGGESTIONS FOR FUTURE
RESEARCH AND APPLICATIONS TO THE PLANNING PROFESSION
4.1 Summary of Findings
The spatial distribution of land use patterns and natural amenities are found to
have signiﬁcant impacts on assessed building and land values, and consequentially,
property tax revenues. Applications of a hedonic pricing framework within property tax
analysis demonstrates that the total value of a property tax base represents not only direct
contributions associated with the structural characteristics and function of land
development, but also a set of location speciﬁc lcharacteristics which determine the value
and utility of real property. The attributes of location that inﬂuence property value are
deﬁned through land use planning, regulation and development. Results conﬁrm that the
value of “location, location, location,” is deﬁned differently by different users of the land.
“Figure 3. Summary of Indirect Fiscal Impacts Attributable to Surrounding Land
Use Patterns in Case Study Communities,” presents a summary of the indirect ﬁscal
impacts of land use planning and development patterns in the case study communities.
Findings demonstrate that there are key ﬁscal tradeoffs associated with certain mixes of

land uses.

82

Figure 3. Summary of Indirect Fiscal Impacts Attributable to Surrounding
Land Use Patterns in Case Study Communities

 

Land Use Pr0perty Values
Read down to evaluate impacts of surrounding land uses in rows
on indirect property tax revenues in columns

 

Residential Commercial Industrial Multi-Family

 

 

 

 

 

Residential ” ~ h I

City Total

TwPNillage Total ' - - $738,475
Commercial 1 I ' I 1’
CM”! 7121872622 - + $4,176,748 +$3,615,585

 

Twp/Village Total -$430,683

 

in...“ 1 l W 1

 

Indirect Fiscal Impacts of
Surrounding Land Uses

City Total -$4 14,684
TWp/Village Total -$12,158

 

-$1,524,416 - ‘ $37,227

 

Multi-Family l t l

 

City Total -$746,458

 

+$l,958,748 -$8,271,685 -‘

 

 

 

 

TWP/Village Total $300,653

 

 

 

 

Key Decrease in Increase in Insigniﬁcant
Property Value Property Value Relationship with
Property Value
Change in Value
I 1 ‘->

 

 

 

 

In case study cities, single family residential, commercial and multi-family land
uses located within a quarter mile proximity of each other create “value-added”
relationships that account for $6.7 million in property tax revenue. Increased livability
associated with residential locations within walking distance of commercial uses,

accounts for $1.2 million (2%) of the city single family property tax base and $3.6

83

 

million (33%) of the multi-family tax base. These ﬁndings suggest that mixed-use
policies that encourage high density nodes of commercial and multi-family development,
surrounded by walkable single family neighborhoods, create both public and private
value in cities.

The property values of non-residential uses are also impacted by surrounding land
use patterns. A negative ﬁscal relationship exists between industrial, multi-family and
commercial uses in the case study communities. Industrial properties located adjacent to
multi-family and commercial uses decrease commercial and multi-family property tax
contributions by less than 2% combined. However, multi-family and commercial land
uses located within a quarter mile of industrial uses, subtract over $12 million in
industrial property tax contributions, accounting for decrease in value equal to
approximately 55% of the total industrial property tax base.

Consistent with a preference for rural character, single family residential land uses
in townships and villages do not value mixed use land use patterns, but rather value close
proximity to natural amenities. Figure 4. presents a summary of indirect ﬁscal impacts
attributable to surrounding natural amenities in case study communities. Proximity to
water and land based natural amenities account for nearly $7.3 million in tax revenue and
nearly eight percent of the total single family residential property tax base in case study
townships and villages.

Surrounding parks and recreation areas indirectly increase property tax revenues
from multi-family uses. This ﬁnding suggests that negative ﬁscal impacts associated with
neighboring single and multi-family land uses can be mitigated through green

infrasﬁucture planning.

84

Figure 4. Summary of Indirect Fiscal Impacts Attributable to Surrounding
Natural Amenities in Case Study Communities

 

 

 

 

 

 

 

 

 

 

 

 

Land Use Property Values
Read down to evaluate impacts of surrounding natural amenities in rows on property values in
columns
City Twp/Village
Residential Residential Multi-Family
Lakes 1 t I
Total -$95,456 +$7,046,318 -$41,762
Wetlands l I h
Total -$267,873 -$728,353 -
.3 Rivers 1 t h
'E
.5 g Total $348,098 +$45,737 -
g “g Natural Areas I I I h
a.
E a
g E, Total $221,283 +$927,486 —
'3 :5 Recreation &
>. 3 parks ' <—>
.9: .
{2 Total +3 82,5 1 5

 

 

 

 

 

Decrease in Increase in Insigniﬁcant
Property Value Property Relationship
Value with

Property
Value

I l «—

While surrounding land use patterns and natural amenities do maintain

 

 

 

 

 

 

 

 

statistically signiﬁcant relationships with assessed building and land values, they account
for a relatively small portion of the variance in assessed value and ultimately the property

tax base. Building characteristics, namely construction quality, age and size play a major

85

role in determining the composition and resiliency of the property tax base. Residential
structures depreciate at a linear rate of one percent each year until they reach 55 years of
age at which time depreciation detracts a constant 45% of the reproduction value. Non-
residential structures can depreciate anywhere from 1.5 to four percent depending on the
quality of construction and maintain an effective life of anywhere from 20 to 65 years.

The negative ﬁscal consequences of depreciation are mitigated through a healthy
real estate market, in fact, only actual and comparable sales transactions greater than the
manual value (cost new, less depreciation) result in property tax revenue increases due to
the application of the ECF. Only when a property is sold on the market can its maximum
taxable value be achieved. When real estate market conditions are strong, legislative
limits imposed by theMichigan General Property Tax Act cap the annual rate of property
tax increase by less than ﬁve percent or the rate of inﬂation which ever is less. However,
when real estate market conditions are weak, as is the case in Michigan currently, sales
transactions less than the manual value will systematically drive down assessed property
values.
4.2 Limitations and Suggestions for Future Research

The hedonic pricing model adopted in this study represents a theoretical starting
point from which a ﬁscal analyst can document and measure the positive and negative
value contributions of land use planning, natural amenities and development
characteristics to property values and property tax revenues. Improvements to this ﬁscal
analysis method can be made in three primary areas ( 1) functional form of the hedonic
model (2) design of independent variables and (3) the temporal dimension of property tax

revenues .

86

 

4.2.1 Functional Form of the Hedonic Model

The functional form of the multiple regression models employed in this thesis can
be improved by accounting for linear and non-linear relationships associated with
development characteristics. A double-log linear box-cox transformation may present an
ideal model speciﬁcation from which the non-linear relationship of variables such as
square footage, age and lot size can be more adequately modeled while still incorporating
the linear shift in value that adjacent and proximity variables were proven to possess. The
double log transformation is desirable due to the comparative ease of calculating implicit
prices of independent variables which requires the analyst to measure implicit prices
within the framework of the difference in value from a mean property value. The double
log transformation requires a much higher degree of mathematical aptitude than most
planners or ﬁscal analysts are likely to posses and may be better suited for the
experienced economist.

It should be noted that the analyst did attempt three different box-cox
transformations (right-hand log, left-hand log and double log) for residential building and
land values models using the same data set used in the linear regression models. The
number of dummy variables resulted in model interference amongst all transformations
except the right-handed log model, demonstrated by an r-squared value of one. Removal
of dummy variables resulted in models free of interference, but less explanatory power,
especially amongst land use adjacent and buffer variables. For this reason, the box-cox
transformation may be better suited for hedonic studies that focus on the value impacts of

a single set of public goods or attributes.

87

4.2.2 Design of Independent Variables

The explanatory power of dummy variables as designed in this study, especially
land use proximity variables, is limited. For example, a dummy variable that states “the
property is located within 0.25 miles of a commercial use” does not identify how many
commercial uses are present within the quarter mile area. Similarly, dummy variables do
not account for the percentage of land area within the quarter mile buffer that consists of
any given land use. The degree to which different land use patterns impact property
values is likely to be inﬂuenced by the density and the percentage of land area maintained
by different uses.

Two alternatives were tested to address the limitations identiﬁed in buffer
variable design. First, the analyst created quarter mile buffers around each parcel and
applied the clip GIS function to dissolve parcel buffers over land use layers therefore
capturing the land use composition of each buffer. Using Microsoft Access, the analyst
was able to group common land use elements and count the'number of common land uses
as well as sum the areas of land uses within each parcel buffer. This method was tested
on a very small sample using Commerce Township parcels and water features, which
delivered adequate results but longer than desirable processing times. When the method
was extended to all case study communities using land use layers in MSU’s GIS lab,
processing time ran for over eight hours and was terminated before completion.
Exploration of alternative GIS methods that can more accurately and efﬁciently capture
the proximate land use composition of very large parcel layers merits future attention and

is likely to advance the hedonic modeling of land use planning and other geographic

88

variables. Vector-raster conversions, density applications, and VBA coding offer
potential solutions.

The second method adopted to improve explanatory power of land use proximity
variables made use of indices as speciﬁed by Acharya and Bennett (2001). The design of
mix indices speciﬁes the number of land uses within a distance of any given parcel. For
example, a mix index of ﬁve demonstrates that there are ﬁve different land uses within a
quarter mile of a parcel, but does not identify what those ﬁve land uses are. Similar to
buffer variables, the mix index does not demonstrate the number of common land uses or
the area that any one land use maintains within any given parcel buffer. The
methodological design of Acharya and Bennett indices warrants further exploration.

Application of hedonic pricing techniques within a ﬁscal framework presents a
promising new direction for research in the ﬁelds of non-market valuation. Given the
methodological similarities between hedonic pricing techniques and property assessment,
future research efforts Should begin with a comprehensive inventory of assessment
practices established through state legislation, emphasizing similarities and differences as
they relate to the identiﬁcation of market and non-market factors that determine value,
application of assessment and MRA methods, and implications of assessment methods to
the bottom line in local communities.

The International Association of Assessment Ofﬁcers encourages the use of
hedonic pricing techniques as tools to support mass appraisal and sales comparison
assessment methods (IAAO). Applications of multiple regression analysis however, were

not uncovered through the case study analysis.

89

4.2.3 Temporal Dimensions

Finally, several potential applications exist in the hedonic modeling of time-series
data. Identifying and measuring changes in property values and property tax revenues
requires assessor’s data for two points in time. The logistics of accounting for new
construction, parcel divisions, sales, and a host of other potential issues will greatly
increase the complexity of already complex data management and integrity. Possibilities
do exist to measure the impacts of social, environmental and economic change on current
values by designing certain independent variables to account for change.

Using a time series approach may also further support the use of assessor’s data
over the traditional use of sales data. Hedonic time series studies requires properties be
sold twice, greatly limiting and most likely biasing samples. The use of time series
assessor’s data may allow for a more adequate measurement of factors that inﬂuence
market conditions and value as they relate to public property appraisal. Extrapolation of
property tax revenues at future points in time would require critical assumptions to
account for changes in millage rate, predictions on inﬂation and the nuances of capped
vs. equalized value, future sale prices, and a host of other issues.

4.3. Applications to Flaming Policy and Practice

Integrating knowledge of property tax assessment within the constructs of
planning enables the capacity to pursue ﬁscal sustainability through informed long-range
planning efforts. The methodological framework tested in this thesis helps support
comprehensive land use planning and regulation, by assigning an actual fiscal value to

development patterns built under planning guidelines.

90

In many ways, the ﬁndings of this thesis conﬁrm what planners have always
thought to be valuable planning, ﬁscal or otherwise. For example, mixed-use policies
often aim to achieve objectives in urban livability, walkability, and local commerce.
Modeling outcomes demonstrate that mixed-use nodes of commercial and multi-farnily
uses in urban single family neighborhoods are ﬁscally beneﬁcial, therefore conﬁrming
this win-win scenario for planning policy. Mixed use policies that concentrate high-
intensity uses can detract value from these land uses in urban areas, especially over time
as these buildings depreciate. During a Visioning session for the 2008 City of Pontiac
Master Plan, local residents cited the mix of land uses as something that they didn’t like
about their community. Residents described the mix of high-intensity land uses as a
“concrete jungle” that creates nuisances for local residents.

While the methods adopted in this thesis effectively document and assign a
monetary value to the dynamic relationships between land uses, the time, resources and
knowledge required for completing such an analysis will greatly limit its application in
planning practice. This method most likely won’t tell planners anything they don’t
already know, but it will provide ﬁscal statistics to ﬁrrther support decision making.
Application of this method has little relevance for a local jurisdiction, but is best applied
at a regional level. Only in comrties or metropolitan areas, and especially those engaged
in grth management and inter-jurisdictional planning, should this technique be
considered a worth while endeavor. In order to make this method and its outcomes
applicable to local jurisdictions, much more work needs to be done. This methodological
framework must be replicated in regions across America to create a critical mass of work

that can conclusively document common signiﬁcant indirect ﬁscal relationships and their

91

value. A comprehensive collection of common ﬁndings, including indirect ﬁscal
“multipliers” like the cost multipliers presented in Burchell’s seminal Fiscal Impact
Guidebook, would allow for more efficient and holistic local ﬁscal evaluation and
planning efforts.

Detailed examination of the nexus between land use policy and property tax
revenues should continue to go beyond the measure of direct ﬁscal impacts. While this
study was successful in quantifying the indirect ﬁscal impacts of built and natural land
use patterns, ﬁndings are limited to a single point in time. Changes in the physical and
socioeconomic conditions, which are of primary interest to community planners, drive
changes in property value and property tax revenue over the long-term Integrating a
temporal dimension to this analysis would greatly support pursuit of ﬁscal sustainability
and examination. A regional scope of analysis would be highly favored in a property tax
hedonic modeling that incorporates a time series framework. In this regard, adopting a
comparative community approach would likely provide more robust ﬁndings that could
help explain and predict local real estate market dynamics while putting ﬁndings in a

more understandable perspective.

92

APPENDICES

93

Appendix A. Highest and Best Use Analysis

Table A-1. Vacant City Land Values Based on Highest and Best Use, 2007

 

 

 

 

Total Land Average Average Total Tax
CLASS Acres Value $1 Acre STax/Parcel Revenue
Pontiac
Commercial Improved 19.8 $2,641,221 $106,597 $242,971 $485,941
Commercial Vacant 197.7 $15,127,195 $147,170 $1,069 $256,529
Apartment Improved 0.9 $78,326 $87,] 19 $1 1,076 $1 1,076
Apartment Vacant 71.7 $4,339,012 $98,797 $405 $29,185
Industrial Vacant 251.9 $17,544,204 $107,679 $1,598 $201,394
Residential Improved 8.3 $436,000 $61 ,1 79 $1,701 $69,754
Residential Vacant 630.7 $39,867,674 $89,384 $207 $598,513
Lake Vacant 1.0 $12,000 $12,021 $315 $315
Condo Improved 2.6 $533,200 $216,235 $2,975 $35,698
Condo Vacant 107.0 $21,690,088 $310,969 $429 $214,847
TOTAL 1291.6 $102,268,920 $79,178 $490 $1,903,255
Royal Oak

Commercial Vacant 6.0 $2,402,675 $422,077 $5,441 $54,41 1
Apartment Vacant 0. 1 $ 17,520 $174,235 $287 $287
Industrial Vacant 0.5 $90,833 $166,785 $965 $1,929
Utility Vacant 0.1 $15,318 $205,340 $386 $386
Residential Improved 3.3 $930,5 15 $286,762 $2,792 $58,624
Residential Vacant 20.8 $5,619,254 $272,617 $1,498 $155,760
Condo Improved 2.2 $525,000 $319,338 $3,079 $21,554
Condo Vacant 2.3 $711,530 $306,859 $1,627 $17,899
TOTAL 35.2 $10,312,645 $292,704 $1,980 $310,8ﬁ

 

94

Table A-2.Vacant Township and Village Land Values Based on Highest and Best Use

 

 

 

 

Total Land Average Average Total Tax
CLASS Acres Value $/Acre STax/Pareel Revenue
Commerce Towmhip
Commercial Improved 5.9 $730,950 $132,610 $34,894 $104,681
Commercial Vacant 254.6 $27,822,243 $246,678 $8,728 $366,580
Apartment Vacant 37.9 $4,251,516 $116,753 $4,168 $37,516
Industrial Improved 8.8 $281,615 $32,052 $16,536 $16,536
Industrial Vacant 164.2 $23,1 12,153 $178,939 $2,632 $178,964
Residential Improved 5.4 $1,092,565 $245,105 $4,488 $58,345
Residential Vacant 968.9 $54,342,282 $249,807 $1,097 $61 1,204
Lake Improved 0.6 $581,390 $1,185,771 $1 1,377 $22,754
Lake Vacant 128.3 $24,009,177 $476,002 $1,026 $216,441
Condo Improved 10.7 $2,286,847 $259,473 $4,592 $96,438
Condo Vacant 203.5 $40,985,845 $265,331 $580 $261,456
TOTAL 1788.8 $179,496,583 $1,970,915
Wolverine Lake Village

Commercial Vacant 15.8 $2,303,091 $139,223 $5,747 $22,990
Residential Improved 0.3 $79,296 $265,968 $3,846 $3,846
Residential Vacant 28.5 $4,523,264 $239,267 $905 $67,850
Lake Vacant 4.3 $2,710,058 $751,620 $1,464 $29,284
TOTAL 48.9 $9,615,709 $123,970

 

95

Table A-3.Vacant Township and Village Land Values Based on Highest and Best Use

 

 

 

 

 

 

 

Total Land Average Average otal ax
CLASS Acres Value $/Acre $Tax/Parcel Revenue
Milford Township
Farm Vacant 25.0 $763,976 $30,618 $2,558 $5,116
Commercial Vacant 101.8 $6,145,302 $65,758 $4,731 $52,041
Apartment Vacant 8.2 $742,154 $91,418 $1,299 $3,897
Industrial Improved 2.6 $201,508 $77,900 $8,183 $8,183
Industrial Vacant 632.6 $14,004,51 1 $53,389 $2,747 $148,359
Utility Vacant 7.6 $139,820 $15,245 $890 $2,669
Residential Improved 12.] $589,124 $48,303 $7,582 $30,329
Residential Vacant 1673.2 $41,538,350 $64,418 $1,575 $522,919
Lake Improved 1.2 $192,500 $165,818 $521 $521
Lake Vacant 103.8 $3,256,026 $185,199 $965 $39,578
Condo Improved 22.1 $2,042,000 $151,734 $6,106 $79,372
Condo Vacant 252.0 $21,767,470 $171,650 $1,149 $236,681
TOTAL 2842.] $91,382,741 $1 , 129,663
Milford Village
Commercial Improved 3.4 $1,627,619 $489,581 $52,351 $104,701
Commercial Vacant 4.2 $710,186 $269,395 $1,972 $9,861
Apartment Vacant 26.4 $1,199,620 $43,776 $2,321 $13,928
Industrial Vacant 15.8 $1,397,753 $104,354 $706 $5,650
Utility Vacant 7.5 $111,744 $14,884 $2,022 $2,022
Residential Improved 0.2 $54,550 $316,884 $1.710 $1,710
Residential Vacant 66.0 $3,547,184 $173,236 $964 $53,019
Lake Vacant 7.7 $409,344 $63,860 $447 $5,807
Condo Vacant 26.6 $5,158,650 $373,545 $507 $31,417
TOTAL 157.7 $14,216,650 $228,] 17
Oakland Township
Commercial Vacant 13.8 $488,071 $99,652 $1,209 $7,255
Industrial Improved 137.7 $1,420,760 $10,321 $22,458 $22,458
Industrial Vacant 214.4 $2,814,543 $57,606 $4,258 $29,803
Residential Improved 22.4 $1,179,598 $1 16,876 $6,358 $44,504
Residential Vacant 4427.7 $107,757,482 $123,113 $2,512 $1,431,750
Lake Vacant 26.6 $1,555,862 $95,833 $1,136 $20,439
Condo Improved 10.] $2,098,595 $298,974 $7,216 $101,021
Condo Vacant 320.6 $78,221,422 $319,592 $1,843 $934,396
TOTAL 5173.4 $195,536,333 $2,591,626

 

96

Appendix B: Exploration of Spatial Autocorrelation

Table B-1. Exploration of Spatial Autocorrelation in Residential Building Values

 

 

Average Residential
Property Average ECF Average BV Total No. Cases
RES 1.45 $145,915 59752

 

Signiﬁcant Variable at .1 Average ECF Average BV No. of Cases % of Cases

 

A002 1.6 $133,864 1,607 3%
A003 1.47 $117,516 1,441 2%
A008 1.07 $248,441 135 0%
A009 1 .01 $222,743 5,226 9%
A012 0.96 $230,456 1,632 3%
A013 1.45 $136,079 24,729 41%
A014 0.94 $228,817 1,013 2%
A017 1.14 $140,949 11,237 19%
A019 0.98 $239,085 923 2%
A020 1.14 $219,704 144 0%
A02] 1 .22 $164,754 243 0%
A022 1.28 $245,186 1,31 l 2%
A024 0.95 $232,178 1,042 2%
A027 1.07 $189,872 341 1%
A029 1 .03 $246,775 845 1%
A030 1.26 $157,202 2,422 4%
A039 1.07 $212,018 6,140 10%
A042 1.31 $95,231 521 1%
A043 1.02 $201,670 11,525 19%
A044 0. 95 $241,549 45 5 1%
A045 0.91 $192,872 356 1%
A046 0.96 $169,038 90 0%
A047 0.88 $176,021 167 0%
A048 1.34 $136,319 172 0%
A049 1.07 $172,136 1,665 3%
A050 1.14 $155,539 519 1%
A051 1.8 $131,375 4,248 7%
A052 1.96 $135,720 1,949 3%
A053 1.57 $113,284 1,031 2%
A055 1.7 $116,744 1,329 2%

 

97

Table B-1 cont. Exploration of Spatial Autocorrelation in Residential Building

 

 

Values
Average Residential
Property Average ECF Average BV Total No. Cases
RES ‘ 1.45 $145,915 59752
B001 0.95 $214,667 965 2%
B002 1.62 $120,215 35,481 59%
B003 1.56 $119,895 39,619 66%
B004 1.] $237,919 413 1%
3005 1.38 $100,309 14,277 24%
B007 1.54 $123,812 41,685 70%
8008 1.75 $129,501 6,923 12%
B009 1.53 $136,423 47,554 80%
B010 1.68 $127,292 31,251 52%
3013 1.42 $122,857 19,089 32%
B014 1.2 $172,273 27,899 47%
B015 1.] $169,740 2,389 4%
B016 1.17 $142,775 1,231 2%
B01 7 1.13 , $163,083 7,253 12%
BO] 8 1.1 $148,169 7,534 13%
3021 1.07 $175,961 1,898 3%
8022 1 $228,471 3,874 6%
B023 1.09 $194,979 9,401 16%
3024 1.65 $123,902 12,217 20%
BO32 1.44 $113,279 4,831 8%
B033 1.4 $106,217 3,208 5%
BO34 1.44 $140,333 537 1%
B037 1.04 $197,256 16,068 27%

 

98

Table B-2. Exploration of Spatial Autocorrelation in Residential Land Values

 

 

Average Residential
Property Average Acres Average LV Total No. Cases
RES 0.51 $59,071 59751

 

Signiﬁcant Variable at .1 Average Acres Average LV No. of Cases % of Cases

 

A002 0.31 $43,419 1607 3%
A003 0.33 $45,755 1441 2%
A005 0.72 $37,496 266 0%
A009 2.14 $111,055 5256 9%
A010 0.59 $89,276 16540 28%
A012 5 $150,001 1632 3%
A013 0.25 $58,519 24729 41%
A014 6.19 $164,779 1013 2%
A017 0.95 $66,033 1 1236 19%
A019 5.13 $154,276 923 2%
A020 0.78 $123,129 144 0%
A02 1 0. 81 $96,493 243 0%
A022 0.59 $237,847 1311 2%
A023 3.63 $130,226 379 1%
A024 4.72 $140,835 1042 2%
A027 4.94 $145,821 341 1%
A028 4.57 $146,753 645 1%
A036 1.46 $92,593 82 0%
A039 1.41 $107,002 6140 10%
A040 0.8 $38,521 123 0%
A042 0.19 $20,488 521 1%
A043 0.83 $100,909 1 1525 19%
A044 4.27 $145,115 455 1%
A046 5.88 $163,949 90 0%
A047 9 $21 1,503 167 0%
A048 0. 77 $102,212 172 0%
A049 1.9 $107,114 1665 3%
A050 1.61 $101,892 519 1%
A051 0.2 $41,081 4248 7%
A052 0.2 $37,668 1949 3%
A054 0.24 $31,359 1329 2%
A055 0.19 $37,518 36412 61%

 

Table B-2 cont. Exploration of Spatial Autocorrelation in Residential Land Values

 

 

Average Residential

Property Average Acres Average LV Total No. Cases
RES 0.51 $59,071 5975]

B001 1.2 $100,790 965 2%
B002 0.31 $43,419 1607 3%
B003 0.23 $41,813 39618 66%
BOO4 1.5 $118,091 413 1%
3005 0.28 $29,763 14277 24%
3006 0.84 $45,433 926 2%
B007 0.26 $42,439 41684 70%
B008 0.23 $40,532 6923 12%
B01 0 0.28 $42,381 31251 52%
BO] 3 0.29 $44,375 19089 32%
3014 0.81 $84,038 27899 47%
B016 0.71 $81,537 1231 2%
B017 0.43 $111,734 7253 12%
B018 0.76 $71,845 7534 13%
B021 2.18 $109,591 1898 3%
3022 1.86 $1 12,073 3874 6%
B023 0.91 $87,401 9401 16%
8024 0.23 $39,985 12217 20%
B034 0.39 $75,750 537 1%

 

100

Table B-3. Exploration of Spatial Autocorrelation in Multi-Family Building Values

 

 

 

 

Average Multi-Family

Property Average ECF Average 3V Total No. Cases
APT 1.17 $634,100 70]
Signiﬁcant Variable at .1 Average ECF Average 3V No. of Cases % of Cases
A003 1.2 $1,225,493 93 13%
A005 1.3 $535,990 16 2%
A015 1.21 $475,041 368 52%
A016 1.03 $1,832,877 8 1%
A031 1.15 $2,852,072 9 1%
A034 1 .12 $1,996,667 22 3%
A042 1 .21 $937,848 40 6%
A043 1.02 $641,026 12 2%
A051 1.2 $643,386 108 15%
A053 1.07 $894,016 32 5%
A054 1.18 $1,123,748 141 20%
3003 1 . 17 $600,568 671 96%
B005 1.21 $567,835 248 35%
3006 1.25 $4,047,188 6 1%
3007 1.16 $609,589 628 90%
3034 1.33 $1,074,412 43 6%

3038 1.14 $1,480,955 32 5%

101

Table 3-4. Exploration of Spatial Autocorrelation in Multi-Family Land Values

 

 

Average Multi-Family
Property Average Acres Average LV Total No. Cases
APT $634,100 701

 

Signiﬁcant Variable at .1 Average Acres Average LV No. of Cases % of Cases

 

A003 2.28 $176,713 93 13%
A005 3.31 $81,194 16 2%
A013 1.77 $126,486 190 27%
A016 8.7 $298,839 8 1%
A019 20.3 $219,707 10 1%
A024 15.2 $225,333 9 1%
A030 5.66 $272,471 34 5%
A039 3.73 $145,844 144 21%
A043 20.5 $167,299 9 1%
A054 1.85 $169,354 141 20%
A055 1.18 $103,521 457 65%
B007 1.27 $91,526 628 90%
B010 1.28 $118,476 428 61%
BO] 1 1.16 $39,267 13 2%
B018 2.31 $77,723 78 11%
3034 0.56 $150,405 43 6%
3038 5.07 $243,375 32 5%

 

Table B-5. Exploration of Spatial Autocorrelation in Commercial Building Values
Average Commercial
Property Average ECF Average 3V Total No. Cases

 

 

COMM l. 15 $578,554 1977

 

Signiﬁcant Variable at .1 Average ECF Average 3V No. of Cases % of Cases

 

A005 1.06 $1,005,777 118 6%
A011 1.07 $1,933,505 9 0%
A016 1.11 $1,338,481 54 3%
A019 1.09 $2,381,354 29 1%
A036 1.1 1 $4,540,431 20 1%
A043 , 1.14 $1,561,034 75 4%
3002 1.16 $529,944 1,629 82%
3006 1.07 $778,710 50 3%
3007 1.16 $507,693 1,765 89%
3034 1.31 $693,924 172 9%

 

102

Table B-6. Exploration of Spatial Autocorrelation in Commercial Land Values
Average Commercial
Property Average Acres Average LV Total No. Cases

 

COMM 0. 83 $244,906 1977

 

Signiﬁcant Variable at .1 Average Acres Averge LV No. of Cases % of Cases

 

A002 1.37 $462,488 109 6%
A005 2.09 $333,655 118 6%
A009 1.73 $245,670 18 1%
A01 1 7.45 $624,286 9 0%
A012 8.12 $711,893 7 0%
A016 2.46 $461,042 54 3%
A019 5.9 $990,938 29 1%
A024 7.8 $1,367,972 15 1%
A030 2.8 $519,113 77 4%
A039 2.4 $574,172 166 8%
A042 0.67 $305,189 510 26%
A043 3.19 $722,381 75 4%
A048 2.3 $605,819 104 5%
A049 1.14 $239,771 208 1 1%
A053 0.95 $260,355 286 14%
3002 0.62 $225,173 1,629 82%
3005 0.84 $179,306 ‘ 1,015 51%
3006 1.98 $238,248 50 3%
3007 0.64 $206,629 1,765 89%
3013 0.78 $232,703 1,116 56%
3032 0.78 $266,713 556 28%
3037 2.7 $531,355 300 15%

 

103

 

 

 

 

 

 

Table B-7. Exploration of Spatial Autocorrelation in Industrial Building Values
Average Industriall Property Average ECF Average BV Total No. Cases
IND 0.98 $1,644,324 513
Signiﬁcant Variable at .1 Average ECF Average BV No. of Cases % of Cases
A017 0.99 $2,519,846 150 29%
A042 0.95 $710,616 19 4%
A043 1.07 $534,975 68 13%
A054 0.92 $9,230,618 23 4%
A055 0.91 $1,964,201 185 36%
B002 0.92 $905,013 252 49%
B018 0.96 $973,417 41 8%
B032 0.95 $3,682,034 38 7%
B037 1.03 $1,144,252 190 37%

 

Table B—8. Exploration of Spatial Autocorrelation in Industrial Land Values
Total No. Cases

Average Industrial Property Average Acres Average LV

IND

5.5

$326,932

513

 

Signiﬁcant Variable at .1

A024
A048
A05 1
A052
A053
A054
A055
B002
BOO8
B032
BO33
B038

45.2
5
13.6
9.55
11.1
20.6
4.02
2.2
3.44
7
0.92
10.83

104

$906,829
$751,245
$787,359
$518,012
$1,015,514
$1,030,621
$264,552
$231,143
$323,473
$691,705
$153,516
$254,208

13
11
60
75
27
23
185
252
219
38
16
118

Average Acres Average LV No. of Cases

% of Cases

3%
2%
12%
15%
5%
4%
36%
49%
43%
7%
3%
23%

REFERENCES

105

References

Asabere, P. & Huffman, F. (1997). Hierarchical zoning, incompatible uses and price
discounts. Real Estate Economics. 25 (3) 439-451

Benson, E. (1998) Pricing residential amenities: The value of a view. Journal of Real
Estate Finance and Economics. 16 (1) 55-73

Bockstael, N. E. (1996). Economics and ecological modeling: The importance of a spatial
perspective. American Journal of Agricultural Economics. 78 (5) 1 168-1 180

Bollinger, C., Ihlanfedt, K. & Bowes, D. (1998) Spatial variation in ofﬁce rents within
the Atlanta region. Urban Studies. 35 (7) 1097-1118

Boyle, K., Poor, J. & Taylor, L. (1999). Estimating the demand for protecting freshwater
lakes from eutrophication. American Journal of Agricultural Economics 74 (4)
99 1- 1002

Burchell, R, & Listokin, D. (1978) The ﬁscal impact handbook. New Brunswick, NJ:
Rutgers University, Center for Urban Policy Research.

Cervero, R. & Duncan, M. (2004). Neighbourhood composition and residential land
prices: Does exclusion raise or lower values? Urban Studies.4l (2) 299-315

Cervero, R & Duncan, M. (2001) Transit ’s Value-Added: Eﬂects of Light and Commuter
Rail Services on Commercial Land Values. Department of City and Regional
Planning, University of California Berkley. Retrieved on October 4, 2007 from
ht_tp://www.apta.condresgch/info/brieﬁngs/documerusjcervero duncanpdf

Chattopadhyay, S. (1999). Estimating the demand for air quality: New evidence based on
the Chicago housing market Land Economics. 75 (1) 22-3 8.

Chesire, P. & Sheppard, S. (1995). On the price of land and the value of amenities.
Economica. 62. (246) 247-267 Retrieved November 1, 2007 from
http://linksjstor.org/sici?sici=0013-

0427%28199505"/0292%3A62%3A246%3C247%3AOTPQLA%3E2.O,CO%3B2-
ﬂ

Dekel, G. (1995). Housing density: A neglected dimension of ﬁscal impact analysis.
Urban Studies. 32, (6), 935-931.

Dubin, R. & Goodman, A. (1982). Valuation of education and crime neighborhood
characteristics through hedonic housing prices. Population and Environment. 5
(3) 166-181

Field, B. (2001). Natural resource economics. New York: Irwin/McGraw-Hill.

106

Geoghegan, J ., L. A.Wainger, and N. E. Bockstael. (1997). Spatial landscape indices in a
hedonic framework: An ecological economics analysis using GIS. Ecological
Economics 23 (3) 251-264.

Heikkila, E. & Davis, W. (1997). Rethinking ﬁscal impacts. Journal of Planning
Education and Research. 16, 202-211

Huddleston, J. (2005). The intersection between planning and the municipal budget.
Prepared for the Lincoln Institute of Land Policy workshop on Curriculum for
Graduate Planning Programs: Nuts and Bolts of Development Finance. Retrieved
October 15, 2007 from

hgp://www.lincolninst.edu/subcenters/TFDP/materials/huddleston-Dlan-
budgetpdf.

 

International Association of Assessment Ofﬁcers. Standard on Mass Appraisal of Real
Property Retrieved on September 1, 2008 ﬁ'om:

www.iaao.orgZuploads/StandardsonMassAppraisal.com

Kotval, Z. & Mullin, J. (2007). Fiscal impact analysis: Methods, cases and intellectual
debate. Prepared for the Lincoln Institute of Land Policy workshop on Curriculum

for Graduate Planning Programs: Nuts and Bolts of Development Finance.
Retrieved October 1, 2007 from

hgp ://www.lincolninstedu/pubs/PubDetail.aspx?pubid=l 252

 

Kluvankova, T. ( 1998). Valuation of national parks in transitional economies.
Eckonomicky Casopis. 46(5) 671-694

Legget, D. & Bockstael, N. (2000) Evidence of the effects of water quality on residential
land prices. Journal land prices. Journal of Environmental Economics and
Management 39 (2) p. 121-144

Mahan, B., Polasky, S., Adams, R. (2000). Valuing urban wetlands: A property price
approach. Land Economics. 76 (1) 100-113.

Michigan Compiled Laws § 21 1.1-21 1.15 7. Michigan general property tax act. Retrieved
on June, 15 2008 from http://www.legislature.mi.gov/docugnents/mcl/pdﬂmcl—a_c_:g
206-of-1893mdf

Michigan State Assessors Board. (2002). Assessor’s training manual. Lansing, MI:
Michigan State Assessors Board

Pendleton, L. & Mendelsohn, R. (2000) Estimating recreation preferences using hedonic

travel cost and random utility models. Environmental & Resources Economics. 17
(1) 89-108.

107

Taylor, L. (2003) A primer of nonmarket valuation (Champ, P., Boyle, K. & Brown, T.
Ed.) Norwell, MA: Kluwer Academic Publishers.

Zabel, J .E., & Kiel, K.A. (2000) Estimating the demand for air quality in four US cities.
Land Economics. 76 (2) 174-194.

108