The Socio-Economic Benefit of Home Ownership in Low and

Moderate Income Communities

Thomas P. FitzGibbon III

DISSERTATION.COM

Boca Raton

The Socio-Economic Benefit of Home Ownership in Low and Moderate Income Communities

Copyright © 2010 Thomas P. FitzGibbon III

All rights reserved. No part of this book may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information

storage and retrieval system, without written permission from the publisher.

Dissertation.com Boca Raton, Florida

USA • 2010

ISBN-10: 1-59942-362-6 ISBN-13: 978-1-59942-362-3

iii

ABSTRACT

The U.S. government spends billions of dollars and implements regulations involving

community lending initiatives to force banks to lend to low and moderate income

communities. However, little research has assessed the effectiveness of these moneys and

programs. The purpose of this study is to assess the relationship between low income

home ownership and community benefit, measured through several socio-economic

measurements within two Chicago community areas and two counties in Indiana.

Although welfare economic theory may support these investments and regulations, the

public also expects community improvement. The research design was quantitative using

existing data from Chicago Public Schools, Police Department and the U.S. government.

Analyses using regression and a one-tailed t test concluded that no significant differences

in crime rates, unemployment, high school graduation, and standardized test scores in a

community with higher housing growth versus a community without housing growth.

These results suggest that, if public investment in housing does not yield greater

community benefit, the financial support of low and moderate income housing should not

continue. Public funds may be more appropriately directed toward other efforts such as

education.

v

DEDICATION

I would like to dedicate this work to my maternal grandparents, Thomas W. and

Katherine M. Caven. They were both very hard working people who instilled a great

sense of responsibility in me. While my time with them was short, their personal ethics

and dedication to their family helped me understand how truly great they were.

vi

ACKNOWLEDGMENTS

There are several people I would like to acknowledge in their support in this

research effort. First and foremost, my committee chair, Dr. Mohammed Sharifzadeh and

secondly both Dr. Reza Hamzaee and Dr. Lilburn Hoehn who served on my dissertation

committee. It was with their constant support and guidance that I was able to complete a

research project that was both personally interesting and useful in the real world.

Additionally, I would like to acknowledge the significant support of my wife Jennifer.

She allowed me to spend a lot of my free time on this project while shouldering a lot of

the responsibilities at home.

vii

TABLE OF CONTENTS

LIST OF TABLES .................................................................................................................x CHAPTER 1: INTRODUCTION TO THE STUDY.............................................................1 Introduction to the Study ....................................................................................................1 Statement of the Problem ....................................................................................................3 Background of the Problem ................................................................................................4 Purpose of the Study ...........................................................................................................6 Theoretical Framework for the Study .................................................................................6 Assumptions ........................................................................................................................8 Scope and Delimitations .....................................................................................................8 Limitations ..........................................................................................................................9 Definitions of Terms ...........................................................................................................11 Nature of the Study .............................................................................................................12 Research Questions and Hypotheses ..................................................................................12 Significance of the Study ....................................................................................................13 Summary .............................................................................................................................14 CHAPTER 2: LITERATURE REVIEW ...............................................................................15 Introduction .........................................................................................................................15 Search Strategy ................................................................................................................16 Regulatory Actions of the United States Government ........................................................16 The Impact of Government Funded Programs ....................................................................27 Socio-Economic Impact of Government Funded Programs and Regulations ....................45 Gap in Research ..................................................................................................................54 Summary .............................................................................................................................56 CHAPTER 3: RESEARCH METHOD .................................................................................57 Introduction .........................................................................................................................57 Description of the Research Design ....................................................................................57 Target Population ................................................................................................................60 Sample and Sampling Methods ..........................................................................................61 Data Collection ...................................................................................................................62 Data Analysis ......................................................................................................................63 Validity and Reliability .......................................................................................................69 Measures for Participant Protection ....................................................................................73 Conclusion ..........................................................................................................................73 CHAPTER 4: RESULTS .......................................................................................................78 Introduction .........................................................................................................................78 Demographics of the Community Areas and Counties.......................................................78 Socio-Economic Performance Indicators ...........................................................................82 Crime Rates ......................................................................................................................83

viii

High School Selection......................................................................................................91 High School Test Score Performance ..............................................................................92 High School Graduation Performance .............................................................................94 Median Income Data for the Community Areas ..............................................................96 County Based Unemployment Data .................................................................................96 Summary ..........................................................................................................................97 Results of the Data Analysis ...............................................................................................98 Introduction ......................................................................................................................98 Test of Hypotheses ...........................................................................................................98 Hypothesis One .............................................................................................................98 Test of Hypothesis One: Murder Rate .......................................................................99 Test of Hypothesis One: Sexual Assault ....................................................................99 Test of Hypothesis One: Robbery ..............................................................................100 Test of Hypothesis One: Aggravated Assault and Battery ........................................100 Test of Hypothesis One: Burglary .............................................................................100 Test of Hypothesis One: Theft ...................................................................................101 Test of Hypothesis One: Motor Vehicle Theft ..........................................................101 Test of Hypothesis One: Arson ..................................................................................102 Test of Hypothesis One: Aggregate Crime ................................................................102 Hypothesis Two ............................................................................................................104 Test of Hypothesis Two .............................................................................................104 Hypothesis Three ..........................................................................................................105 Test of Hypothesis Three ...........................................................................................105 Hypothesis Four ............................................................................................................106 Test of Hypothesis Four .............................................................................................107 Correlation of Variables ................................................................................................108 Correlation of Crime Rate Measurements in New City .........................................109 Correlation of Crime Rate Measurements and Education in New City .................109 Correlation of Crime Rate Measurements in Austin ..............................................110 Correlation of Crime Rate Measurements and Education in Austin ......................110 Analysis of Autocorrelation .......................................................................................111 Summary ..........................................................................................................................114 CHAPTER 5: SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS .................115 Overview .............................................................................................................................115 Conclusions .........................................................................................................................115 Hypothesis One ................................................................................................................115 Hypothesis Two ...............................................................................................................116 Hypothesis Three .............................................................................................................117 Hypothesis Four ...............................................................................................................117 Implications.........................................................................................................................119 Recommendations for Action .............................................................................................122 Recommendations for Further Study ..................................................................................124 Implications for Social Change ...........................................................................................125

ix

Summary .............................................................................................................................126 REFERENCES ......................................................................................................................128

x

LIST OF TABLES

Table 1 Racial Composition for the Community Areas Under Review ................................79 Table 2 Percentage of Owners-Occupied Housing in the Community Areas .......................80 Table 3 Racial Composition of the Community Areas Under Review ..................................81 Table 4 Percentage of Owners-Occupied Housing in the Community Areas .......................82 Table 5 Murder Rates by Year in the Community Area ........................................................83 Table 6 Sexual Assault Rates by Year in the Community Area ............................................84 Table 7 Robbery Rates by Year in the Community Area ......................................................85 Table 8 Aggravated Assault Rates by Year in the Community Area ....................................86 Table 9 Burglary Rates by Year in the Community Area ......................................................87 Table 10 Theft Rates by Year in the Community Area .........................................................88 Table 11 Motor Vehicle Theft Rates by Year in the Community Area .................................89 Table 12 Arson Rates by Year in the Community Area ........................................................90 Table 13 Total Crime Rates by Year in the Community Area ..............................................91 Table 14 ACT Scores for the New City High Schools ..........................................................92 Table 15 ACT Scores for the Austin High Schools ...............................................................93 Table 16 High School Population Taking the ACT Examination .........................................93 Table 17 Weighted Average ACT Score Performance by Community Area ........................94 Table 18 Graduation Rates by High School by Graduation Year ..........................................94 Table 19 Potential Graduating Population .............................................................................95 Table 20 Weighted Average Graduation Rates by Community Area ....................................95 Table 21 Average Median Family Income for the Community Areas ..................................96 Table 22 Unemployment Data by County .............................................................................97 Table 23a Summary of Regression Analysis Data for Variables Predicting Crime ..............103 Table 23b Summary of Regression Analysis Data for Variables Predicting Crime ..............103 Table 24 Summary of Regression Analysis Data for Variables Predicting ACT Scores ......106 Table 25 Summary of Regression Analysis Data for Variables Predicting High School Graduation..............................................................................................................................108

1

CHAPTER 1:

INTRODUCTION TO THE STUDY

Introduction

Over the past thirty years, the United States Government has played an active role

in developing greater home ownership opportunities for low and moderate income

families. The government intended to support these opportunities by a combination of

regulatory changes and direct program funding to increase the options available to the

targeted groups. The government’s immediate goal was to improve the funding options

for underserved communities; its long-term goal was to improve the overall socio-

economic condition of the community.

Although research has examined the socio-economic benefits of home ownership

in general, less attention has been given to the benefits of changes in home ownership in

low and moderate income communities. Given this lack of attention, the intent of this

study was to identify the regulatory and programmatic interventions to understand

whether improvements in home ownership occurred, and to assess the socio-economic

benefits of those programs in a low and moderate income community.

Given the breadth of government regulations from the Community Reinvestment

Act to the Home Mortgage Disclosure Act, the government provided the encouragement

to lenders in providing mortgage options that address the needs of low and moderate

income communities that were not previously available in the market. Additionally, the

government has established several funded programs to provide a financial stimulus to

2

lenders, community groups, and individuals in the effort to support home ownership to

the low and moderate income markets.

An additional problem is the inconsistency in available data that would support a

relationship between home ownership and accepted socio-economic measurements within

low and moderate income communities. Secondly, related to the impact of government

regulations and funding efforts, it is proving to be quite difficult to assess the benefits of

one particular program or regulation since there is typically a combination of several

options that could be used at any point in time. For example, in instances where multiple

funding sources used on a single project it could prove difficult to determine whether one

source was any more effective than another.

To address these problems, the aim of this study was to study the socio-economic

benefit of home ownership in low and moderate income communities. Evidence suggests

that the public investment has not always yielded greater community benefit (Czerwinski,

2006). For example, programs such as HOPE VI have intended to spur home ownership

in low and moderate income homes, yet the effectiveness of such programs has not been

measured. Acts like the Community Reinvestment Act, also intended to encourage home

ownership, have also been limited in their effectiveness. Such acts have tended to relax

the underwriting standards of banks to get more customers, thereby increasing loan

delinquency and foreclosure (Dreier, 2003). Given the disconnect between government

initiatives and effectiveness in the community, alternate investment channels may yield

better results. For example, research has found a link between educational attainment

and general community benefit. These findings suggest that an investment in education

3

may be a more effective alternative to housing that could yield an equal or better benefit

for the community.

Statement of the Problem

The United States government has spent billions of dollars in direct financial

support and administrative expenses to improve community conditions (Wood, 2007).

However, little evidence exists to substantiate the effectiveness of government efforts on

home ownership in low and moderate income communities. An initial review of

literature revealed that the nature of the relationship between home ownership and socio-

economic benefit is unclear. Therefore, the problem is that, while there are benefits that

individual home owners may have as a result of these programs and investments, how the

change in home ownership has to greater community health remains unclear.

To determine the socio-economic impact of these government programs, I

examined the relationship between home ownership (i.e., the non-manipulated

independent variable) and crime rate, unemployment rate, high school graduation rate,

and high school test scores (i.e., the dependent variables). My intention was to assess

housing growth between two low and moderate income communities from the

perspective of the ratio of total owners-occupied housing within the total housing

available. This ratio identified both the community area with no growth and the

community area with high growth. Because the demographic characteristics of the

communities were similar, this process of indexing served to reduce any inconsistencies

in growth between the community areas.

4

This information provided an inference as to whether home ownership related to

other socio-economic measurements. This information then provided a basis to the larger

discussion on welfare economics with specific application to Pareto efficiency whereby

any improvement in economic conditions of one individual is not to the detriment of

another. In this context, there is not a shifting of wealth from one individual which then

allows another individual to purchase a home.

Background of the Problem

The underlying intent of government programs has been to develop greater

opportunities for home ownership in the low and moderate income communities. The

goal of these efforts was to improve the quality of communities through the increase in

home ownership. By increasing home ownership, it was believed, other socio-economic

factors would improve, such as a decrease in crime rate and increase high school

graduation rates.

Historically, lenders have accepted deposits from the communities they served,

only lending to individuals with the least risk of defaulting on a loan. The result of this

practice has been that few individuals with a low and moderate income have been able to

get loans or establish credit. The majority of loans have been given to individuals earning

an upper income. These individuals have also comprised the population that typically

owned their own home, thereby leaving the low and moderate income population to

either rent or live in less-than-adequate owned housing (Dreier, 2003).

To address this discrepancy in home ownership, the government began

implementing regulations to encourage lenders to serve a larger community. Among the

5

most popular regulations was the Community Reinvestment Act of 1977. The primary

focus of this act was an improvement to the oversight of lenders that would force lenders

to provide credit to a demographic representation of the communities served. From this

act, banks would be forced to lend to the same populations from which they had accepted

deposits. Banks were also forced to consider the penalties associated with non-

compliance with the regulations. For example, non-compliance to the requirements of the

Community Reinvestment Act would limit a bank’s ability to create and offer new

products or even open additional branches in the community, thus stopping any growth

efforts until the non-compliance issues were resolved.

Initially, the lenders fought the regulations, arguing that lending to people

previously considered non-qualifying may put their business at peril and result in poorer

performance to investors. However, the government’s regulations passed and were put

into law. As a result, the lenders were forced to create new strategies to address the needs

of the low and moderate income community.

Along with these new regulations, there was also a need for the government to

directly fund programs that would support home ownership in the low and moderate

income communities. The government considered the regulations to be an effective first

step in the process, the potential market still needed to provide financial incentives to

potential home owners, community groups, and lenders. Over time, the government

established several different funded efforts such as the HOPE VI program and Bank

Enterprise Award program to provide a financial stimulus to a greater population.

6

The programs did increase home ownership, but little research has measured the

community changes from such programs, particularly the socio-economic benefit to low

and moderate income communities. Additionally, limited attention has been given to the

management of government funds, and little guidance is available for local organizations

to report performance back to the government. In the absence of this guidance, some

community organizations may use government funding for efforts unrelated to the

program goals.

Purpose of the Study

The purpose of the study was to examine the socio-economic changes in home

ownership in low and moderate income communities. Historical data were examined,

relating to home ownership, crime rate, employment rate, high school graduation rate,

median income rates, and academic test scores in two low and moderate income

communities in Chicago and two counties in Indiana. Analysis of these variables

provided empirical evidence of the socio-economic impact of home ownership on low

and moderate income communities.

Theoretical Framework

Arrow (1983) and Sen’s (1997) welfare economics theory, and Keynes (1936) and

Friedman’s (2002) market economic theories were used to define areas of government

regulation and distribution of government funds and to study the impact of government

funding on community performance.

Welfare economics theory posits that no one individual should become better off at the

expense of another individual. Arrow (1983) and Sen’s (1997) theories of welfare economics

7

conceive of taxes as a means to equalizing wealth across the population. Taxes could equalize

other factors such as educational opportunities, which have been considered by some (Keynes,

1936) as one of the primary environmental factors for equal opportunity. This tenet is

commonly referred to as Pareto optimality.

However, a variety of theorists, including Friedman (2002), consider that a typical

market economy is not Pareto optimal. The market economy in general contests the idea of

Pareto optimality, as there will always be an unequal distribution of wealth. Keynes’ (1936)

theory on the multiplier effect of money conflicts with Friedman’s market theory. For

example, Keynes’ theory posited that greater funds in a system would correlate with greater

consumer spending, which would, in turn, spur improvements in overall income.

Friedman’s (2002) theories do not support active regulation or long-term programs to

support low and moderate income communities. On the other hand, Keynes’ (1936) theory

suggests that it would be more effective for the government and the market to invest in this

effort. Keynes held the perspective that with the investment the government makes, that

investment would result in improved economic conditions for the individuals receiving support

from the programs. Arrow (1983) and Sen (1997) considered these housing development

programs to be the goals of welfare economics. They saw these programs as providing a

greater opportunity for low and moderate income families to improve their financial condition.

However, they followed different reasoning compared to Keynes. Arrow and Sen considered

government programs to be one of many income distribution programs to provide greater

opportunity in the community. Both considered education to be the key vehicle to provide an

8

individual with the ability and the skill to be more professionally productive in their

professional lives.

All of these theories were brought into a greater context of the current state of the

housing market as well as the performance of programs and regulations developed to support

improvements in housing programs in low and moderate income communities. Both welfare

and market economics theories were used to determine whether the government efforts

resulted in Pareto optimization as well as a resulting socio-economic return on investment.

Assumptions

Given that the purpose was to assess the socio-economic benefits of home ownership in

low and moderate income communities, I assumed that the people in this study wanted to own

a home and did not want high residential turnover. I also assumed that they wanted to improve

themselves socio-economically.

Scope and Delimitations

The study included a socio-economic assessment of a low and moderate income

community in the city of Chicago. Two low and moderate income neighborhoods were

studied. Although these neighborhoods shared similar demographics, one neighborhood had a

history of growth in home ownership whereas the other had limited housing growth. The

study focused on two counties with a significant low and moderate income population in order

to assess the relationship between owner-occupied housing stock and unemployment.

Variables under study were limited to crime rates, high school graduation rates, rates of

home ownership as well as unemployment performance. This information was accessed both

9

from the Federal Government level as well as public information managed by the city of

Chicago. The delimitation of the study was that it was not feasible to examine every low and

moderate income community in the U.S. Instead, the focus was on the particular

neighborhoods or counties discussed above. While these areas were defined as low and

moderate income, the delimitation may have an impact on the ability to apply the specific

findings to other low and moderate income communities in the United States.

Limitations

The predominant limitation of the study related to data access. Although most of the

data were in the public domain, data on the historic high school academic performance were

not available. Additionally, the data was limited to two Chicago community areas and two

Indiana counties. For example, one of the data points reviewed was the American College

Testing Program test scores of high school students in the target neighborhoods. Although the

ACT test has been offered as an optional examination for over twenty years, the text was not

required in the state of Illinois until the 2000-2001 academic year. Thus, these data only

identified the academic performance of students starting high school in 1997 or later.

This study was also limited by the gentrification of neighborhoods over the period

under review. From this limitation, outcome measures could reflect the changing dynamics of

the neighborhood as well as the city. However, by comparing reasonably similar

neighborhoods within the city, the likelihood that similar dynamic changes would affect both

neighborhoods in the comparative study was lessened.

10

Another limitation was that the data related to crime rate only included those crimes

that were reported to government authorities. The true number of crimes may have been

higher than what was reported.

Furthermore, unemployment related data only reported on those not working and

actively seeking employment. This data did not report on those who are unemployed, but not

seeking work. This limitation may have resulted in an underreporting of unemployment

information. Since unemployment data were not tracked at the census tract level, it was

necessary to utilize employment data at the county level in order to provide an assessment of

the effect of changes in home ownership. Given that unemployment data were not available at

the community area level, I considered median income information at the census tract level as

a substitute for unemployment data within the overall community area analysis.

Data were also limited to public students, and did not include private or magnet school

students. With this limitation, data were excluded because it was difficult to assess the

performance of individual neighborhood residents when their performance was reported along

with other students who did not reside in the neighborhoods under study.

This study was also limited by other uncontrollable external factors that could have

impacted the performance of any neighborhood. For example, in a generally declining

economy, events such as drastic changes in unemployment will deteriorate in most

communities but can also have an adverse impact on low and moderate income communities at

a higher level which could result in increases in crime and decreases in educational

performance. Although these changes may have hindered the generalizability of these findings

11

to different communities, the selection of two community areas within the same city served to

mitigate some of these limitations.

Definition of Terms

The following terms will be used throughout the text of this study.

Academic Scorecard: An annual report provided for each public school within the

State of Illinois that lists the quantifiable information related to test score performance,

graduation rates and overall enrollment at the school, district and state level.

Bank Enterprise Award (BEA): A United States Government funded program that

provides financial incentives for banks to support low income housing programs.

Community Development Block Grant (CDBG): A United States Government funded

program that provides targeted funding assistance to low and moderate income communities.

Government Sponsored Enterprise (GSE): Financial services corporations established

by the United States Government to provide greater access to credit. Fannie Mae and Freddie

Mac are considered to be Government Sponsored Enterprises.

Low Income Housing Tax Credit (LIHTC): Provides a federal tax credit to private

investors who develop low and moderate income housing programs.

Prairie State Examination: A mandatory standardized test required of all Illinois Public

High school students that is completed at the end of the 11th grade. The ACT test is included

within this examination.

12

Nature of Study

This study employed a quantitative design, involving collection and analysis of the

existing data regarding the effects of home ownership on certain socio-economic

characteristics of low and moderate income communities. Because the existing data were used,

a quantitative design was the most appropriate method to use for this research. Secondly, as

noted above, while past research may not have focused on this community, the gathered data

as well as the analysis methods used were consistent with this research as well.

Research Questions and Hypotheses

For the purposes of this study, indexed changes in home ownership in the low and

moderate income community were considered as the non-manipulated independent variable.

With that, the following research questions were considered.

1. What is the relationship between home ownership in low and moderate income

communities and crime rate?

2. What is the relationship between home ownership in low and moderate income

communities and unemployment?

3. What is the relationship between home ownership in low and moderate income

communities and standardized test scores?

4. What is the relationship between home ownership in low and moderate income

communities and high school graduation?

Other variables within this analysis were considered, such as median income, to

address potential external factors impacting the collected data. The inclusion of these

13

variables provided a more ecologically valid understanding of the economic conditions of the

selected low and moderate income communities.

Significance of the Study

Although research has illustrated the benefits of government funding on home

ownership in general, little attention has been given to the impact of government funding on

low and moderate income communities. Thus, while research has supported further public

funding of these programs, little research has indicated the impact of those funds on low and

moderate income communities.

Many constituencies can benefit from this research. First, the government may use this

information to determine whether tax dollars should continue to be invested in improvements

to home ownership, and whether existing regulations for greater lending are benefiting banks

and communities. Second, the community members may use this information to redirect

available funds, identify interventions that could provide better community benefit or to

identify a more effective set of regulations that can improve community health.

This research could also benefit those who are seeking to implement community-wide

interventions for greater home ownership. This information may lead to the development of

different initiatives such mixed-income housing developments or commercial real estate

developments within the community to better address negative performance of some of the

important socio-economic performance measurements.

14

Summary

A conflict exists when discussing the perspectives of welfare economics and market

economics applied to low and moderate income home ownership programs. There is a desire

from the public to assist low and moderate income families in owning a home, it is unclear

which stakeholders should facilitate the effort. Market theorists like Friedman (2002) hold

that lenders should identify the market and address the demand, whereas welfare economic

theorists like Arrow (1983) argue that the government should spearhead the effort. In section

2, I consider a contemporary application of both market and regulatory actions in the low

income housing market, grounded in market and economic theories. In chapter 3, I discuss the

method used to collect and analyze my data. In chapter 4, I present the results to my analyses

for each research question. In chapter 5 I explain the applicability of the study results and

suggest recommendations for future research related to this topic.

15

CHAPTER 2:

LITERATURE REVIEW

Introduction

For the past thirty years, the United States Government has provided additional

opportunities for low and moderate income families to own their own home. The

government’s intent was to address the discriminatory practices by. These steps were

driven by the theory that home ownership would also improve other socio-economic

aspects of low and moderate income communities.

In this chapter, I will present some of the major regulatory and government-

funded programs that have been implemented to meet the needs of the low and moderate

income population. I will also discuss the market response from lenders and consumers as

well as the impact these efforts have in low and moderate income communities. I will

also detail the response from the lending industry both from a compliance perspective as

well as the identification of new product options designed specifically for potential low

and moderate income home owners. From there, the my focus will turn to reviewing

several examples of government funded programs developed to provide financial

incentives to lenders and communities to provide further investment in low and moderate

income communities throughout the United States. Finally, my review will continue with

an application of the general socio-economic community benefits of high home

ownership within the general context of welfare economic theory and its applicability to

low and moderate income housing programs.

16

Search Strategy

Using the EBSCO and ProQuest Libraries, I conducted key word searches within

peer reviewed journals and government publications that focused on welfare economics,

low and moderate income housing programs, crime rate, socio-economic benefits of

home ownership, and the effect of home ownership and educational performance.

Regulatory Actions of the United States Government

Since the Great Depression, lending institutions have been regulated by the

United States Government. These regulations range from requirements of safety and

soundness of an institution to the different products the bank is able to offer to the public.

However, it was not until the late 1960s that regulations related to the service of low and

moderate income communities came to the market. The need for greater regulation and

oversight was not driven by the government, but by local community members who

wanted to gain political influence and change the existing banking environment (Dreier,

2003).

From a theoretical perspective, the existing lending system was inefficient. That

is without proper service to low and moderate income communities, the needs of a

potential market were not being addressed. The government’s rationale was that if the

lenders would not provide the service on their own, the government would create

regulations that would enable the lenders to provide loan products and meet the needs of

more consumers.

Considering the application of general welfare economic theory to these

regulations, one could assume that the regulations would provide a greater Pareto optimal

17

market. That is with the new regulations, the needs of more people could be met without

any loss to those that were already able to qualify for existing loan products, taking what

could be considered a less Pareto optimal system prior to the regulations and allowing the

system to perform at greater optimization with the needs of more consumers being met.

The implementation of regulations may not be the only answer. Market

economists like Friedman would conclude that these regulations are not necessary in that

if there was an actual need, the market would address that need if a market intervention

was a fiscally sound option. From the government’s perspective, the lenders concluded

that low and moderate income borrowers had a high risk of loan default and had no desire

to enter a risky market. Without the government’s regulatory intervention, the lenders

would likely not consider entry into the low and moderate housing market.

With the evidence indicating the need to address the lack of financing

opportunities for low and moderate income families, it was necessary for community

groups to band together to create the message that the status quo needed to change. The

first step in that process was that the victims of discrimination had to be identified

(Dreier, 2003). This identification would allow the stakeholders to visualize the people

that did not have access to home ownership.

Secondly, the community organizations needed to identify a set of solutions to the

problem (Dreier, 2003). The solution was not simply to require that the banks lend to

anyone that applied. What was necessary was the development of solutions for lenders,

developers, governments and individuals that would align the needs of the individual

stakeholders as well as the overall goals of the effort. This may involve new loan

18

products, services and counseling with the end results being financing options that would

meet the needs of all stakeholders.

Third, the groups needed to align themselves with politicians and public

organizations that would be the long term partners after any program or regulatory

implementation (Dreier, 2003). The need to partner with these groups will continue to be

necessary to keep the needs of low and moderate income families in the spotlight. Not

only do politicians and public organizations have community influence, but they can also

influence any government funding or support as well. The ability of the community

groups to leverage relationships with public organizations and politicians can have

tangible benefits with potential government funding and regulations, but can also

establish a higher level of credibility within the community.

Finally, the structure of the local organizations should not only to be aligned with

the overall goals of improving home ownership, but also should also provide a

mechanism to learn from one another (Dreier, 2003). There is no single solution that

would meet the needs of all low and moderate income families, but the need for a

structure that can share these best practices will result in more effectiveness for all local

organizations can benefit a wider set of community groups. Organizations such as

Neighborhood Reinvestment Corporation and Neighborhood Housing Services are non-

governmental organizations that focus on supporting local groups in addressing housing

development programs. Programs such as these provide counseling and operational

guidance that can leverage the best practices of a variety of successful local

organizations.

19

Starting with the Fair Housing Act, the Home Mortgage Disclosure Act, and the

Community Reinvestment Act, banks were required to offer their products and services

to a wider customer base. As a result, banks were required to not only accept deposits

from their community, but would also have to lend to that same community (Dreier,

2003). These regulatory changes were designed to address historic issues where banks

were willing to accept deposits from any customer, but would only provide credit to those

customers who the bank felt were the least risky. The result of the regulatory

implementation was that individuals considered either low or moderate income had

improved access to credit.

The regulations were implemented because of what was defined as discriminatory

practices (Freeman & Hamilton, 2004). However, there is some dispute as to whether the

alleged discrimination actually occurred. From a fundamental economics perspective, it

has been speculated that it would not be in a bank’s best interest to simply ignore a

population of potential customers without examining the financial quality of those

customers (Newman & Wyly, 2004). Doing so would diminish the potential revenue of

the bank by ignoring an entire segment of customers. Assuming that the discrimination

did occur, banks were considering characteristics such as race as a risk factor rather than

a simple analysis of the credit quality of the loan applicant.

Kaersvang (2006) argued that the Federal Housing Act may be focused more on

providing financing options for inner city residents that would give those residents a

choice to live someplace else other than their present location. As such, it may not be a

matter of providing financing options that would eventually improve the conditions of

20

inner city neighborhoods, but to give people an outlet to depart the inner city all together.

The result was that the Federal Housing Act was not necessarily a vehicle for urban

redevelopment, but simply an additional financial option that would give residents a

greater opportunity of choice.

With the implementation of the Community Reinvestment Act, banks would be

measured on their ability to fairly service all potential customers in their service area. In

the event of a satisfactory rating from the federal auditors, a bank would be able to

operate without further intervention. If a bank review was unsatisfactory, a bank would

not be allowed to open a new branch, offer a new product, or install a new automatic

teller machine. Thus, a bank could not expand in any operational area until the identified

compliance issues were resolved. Given the growth limitations provided under the

regulatory framework, compliance with the Community Reinvestment Act was critical to

the future success of the bank.

To address the potential compliance issues, banks were required to consider both

their product offerings as well as their underwriting requirements for credit (Fennell,

2008). However, before creating the new products and services, the banks needed to

develop an understanding of what options this new market needed. One potential area for

consideration were the requirements involving minimum down payment options for low

and moderate income home buyers. Traditionally, banks would require at least 10% of

the purchase price as a down payment from the borrower. With the down payment, the

borrower would assume a limited level of investment in the property by providing the

funds for the down payment while the bank would provide the balance of the funds

21

needed to purchase the property through a traditional mortgage. Additionally, the down

payment would also provide a financial cushion in the event of a downturn in the housing

market which would belay the risk associated with a mortgage balance that was in excess

of the current property value (Wray, 2006). Some may consider a 10% down payment to

be a minimal investment for the purchaser, but this amount would force a low or

moderate income family into either substandard housing or to simply not consider home

ownership as the required down payment was not affordable (Freeman & Hamilton,

2004).

Government Sponsored Enterprises such as Fannie Mae and Freddie Mac

developed loan products that would require down payments of less than three percent for

qualified low and moderate income families (McDonald, 2005). These organizations did

not lend directly to the public, the partner banks could now offer these more attractive

loans knowing that either Freddie Mac or Fannie Mae would eventually purchase the

loans from the banks. The Federal Housing Administration provided default insurance

for qualified loans (McDonald, 2005) or in some circumstances, Fannie Mae or Freddie

Mac would provide the default guarantees for loans not eligible for FHA insurance

(Jaffee & Quigley, 2008). With the insurance, banks would consider these loans to be of

lower risk as they would not be maintained by the bank after transfer to the secondary

lender.

With the lower risk, the Government Sponsored Enterprise loans also proved to

be very successful in several large markets in exceeding the goals of available affordable

housing (McClure, 2005). The participation of the Government Sponsored Enterprises

22

served to not only mitigate the perceived risk of the banks, but build upon that to provide

better products and investment in the targeted communities. However, it can not be said

that the success was consistent across all major cities, there was a clear indication that the

available mortgage products along with the secondary market acquisition of the

originated mortgages that there was a greater incentive to meet the needs of the low

income residents. However, Frame and White (2005) conclude that while the

Government Sponsored Enterprises did create and offer these more flexible loan products

to the market, the data that would indicate that the existence of these products is

inconclusive in relation to any significant improvement in home ownership (McDonald,

2005).

As a result, the banks viewed these new products as low risk along with meeting

new compliance requirements and the potential generation of fee revenue in the

origination process. Given the opportunity to access a new group of customers along

with the guarantees provided by the Government Sponsored Enterprises, lenders now

considered low and moderate income families to be a worthy investment. The result of

the willingness of lenders to provide products to these customers was that low and

moderate income families were now able to realize the opportunity of owning a home as

it was now a more affordable alternative with the lower down payment requirements.

In addition to the down payment requirements, banks consider a customer’s credit

rating to be an indication of their ability to pay the loan back once the funds are

disbursed. The credit rating requirements also proved to be challenging for low and

moderate income families as in many cases this population either had a very limited or

23

low quality credit history (Ibarra & Rodriguez, 2006). In order to qualify for most

traditional mortgages, a typical borrower must have an established and reliable history of

paying previous debts in a timely matter. With an established credit history, the bank

could assume that if a customer has paid their other debts on time they would likely also

pay their mortgage on time as well.

As a result of a lack of credit history, banks were forced to consider alternatives to

credit history that would also serve to provide evidence of consistent payment of

financial responsibilities. Banks would need to identify other financial responsibilities

that required routine timely payments but did not commonly appear on a credit report.

For example, banks could consider a customer’s payment history on rent expenses and

utilities as an indication of payment history. These expenses would not appear on a

typical credit report or have an impact on a credit score. However, an objective report of

payment history could be reviewed by the bank which would then provide an alternative

to the standard credit rating process.

After resolving the credit history and down payment concerns, the last remaining

challenge for this community were the requirements surrounding gross income

requirements. For most traditional mortgages, a customer cannot have a monthly

payment that would exceed 35% of an individual’s monthly gross income. Again, the

intent of this requirement is that the applicant is not accessing more debt than what the

applicant can afford to repay. For many low and moderate income families, this

maximum payment requirement would either result in the need for a significant down

payment in advance of the mortgage or the purchase of a home that was of limited value.

24

In order to provide some flexibility in the maximum payment requirement, banks

first reviewed a borrower’s rent history. It is common that low and moderate income

families pay a very high percentage of their income towards rent, in some examples, rent

payments were nearly 50% of an individual’s gross income per month (Mueller &

Schwartz, 2008). In reviewing this information as well as the consistency of timely rent

payment of the customer, the bank could then consider a higher monthly payment. The

lender could make the assumption then that if the customer had consistently made their

high rent payment, it was likely that the consistency of a mortgage payment of a similar

amount would result. Banks would likely not allow a mortgage payment to be at 50% of

gross income, they would likely consider something higher than the current maximum

which would allow the customer to purchase a more valuable property.

The result of these actions by the banks was the development of new mortgage

options for low and moderate income families that would provide a greater opportunity to

finance and purchase a home. Banks considered this to not only be a social benefit to the

community, but it would also provide a financial benefit by accessing a new customer

segment. The social benefit provides a greater amount of goodwill in the community

paired with the most important factors of risk, profit, and regulatory compliance. These

new products were able to effectively address these needs.

While banks have loosened their credit standards and processes, any future

adjustments to underwriting requirements may conflict with the existing risk assessment

processes of the bank (McClure, 2005). The result of this risk aversion is that there is

still a customer segment with unmet needs. With the existing unmet demand, there was a

25

market for an additional lending industry outside of traditional banks, the subprime

lender (Newman & Wyly, 2004). Subprime lenders would operate in a similar manner to

banks in the function of providing loans with the exception of not accepting deposits

from customers. The subprime lender would generate new loans and then immediately

sell those loans on the secondary market to organizations like Fannie Mae and Freddie

Mac. The business focus of subprime lenders was low and moderate income borrowers

that may be of higher risk with normal underwriting standards (Shlay, 2006).

Subprime lenders offered more creative loan products that were not commonly

offered by traditional banks. For example, subprime lenders would offer options such as

interest-only mortgages, loans that would finance more than 100% of the property value

or adjustable rate mortgages which would offer a low initial monthly payment with the

risk that the payment may change in the future terms of the mortgage. Most traditional

banks were unwilling to offer similar loan products offered through subprime lenders as

the interpretation was that the default risk was much higher when compared to existing

mortgage options.

With the collapse of the mortgage market over the past few years, the perception

of risk for subprime loans appeared to be correct. As Jaffee and Quigley (2008) note,

nearly 9% of subprime mortgages were already in foreclosure. However, these defaults

and foreclosures were not solely related to low and moderate income borrowing. In

addition to low and moderate income families, more affluent individuals also entered into

default and foreclosure. The new customers were different. These affluent borrowers

simply assumed more mortgage debt than what could be paid over the terms of the

26

mortgage. The resulting collapse in the mortgage market that we are witnessing now is

that not only did many subprime lenders fail, but those organizations that purchased the

loans also failed.

From a compliance perspective, the banks could demonstrate to the regulators that

they were providing additional options for mortgages that would serve a wider portion of

the community. From a risk perspective, the banks were still following an underwriting

process that addressed the need to assess the customer’s ability to pay the mortgage back.

Along with the underwriting perspective, there was a readily available secondary market

with Fannie Mae and Freddie Mac that were willing to purchase these loans shortly after

origination resulting in the default risk moving from the bank to the new purchaser

(Freeman & Hamilton, 2004). At a minimum, the banks could also see a short term profit

in the generation of origination fees associated with the loans. The banks were able to

identify a new customer channel for their products that was relatively low risk and would

generate a steady income (Shlay, 2006).

Outside of lenders, there was a history of discrimination on the part of property

insurers. It was uncommon for insurance companies to discriminate in terms of race, but

several insurers would simply ignore entire areas of a city due to the perception of a risk

of loss. This lack of available options for property insurance proved to be detrimental to

families wishing to purchase a home as without property insurance, there was no

potential for a mortgage. Without property insurance, lenders would not finance the

property against damage or loss as it would be detrimental to the property value and the

underlying mortgage. Changes in available insurance would also result in existing

27

residents choosing to leave a community due to continually rising insurance rates

(Kaersvang, 2006).

From the insurers’ perspective, the perception was that areas with high crime and

urban blight were too risky without any improvements in the socio-economic factors or

the general condition of the insured properties (Kaersvang, 2006). The insurance

companies could simply charge higher premiums for those in the community, leave the

community all together or charge higher premiums to those in lower risk communities.

Just like lenders, insurance companies are businesses and are measured by their ability to

be profitable. The amount of profit for an insurance company is measured by the amount

of claims paid against the amount of premiums received. Where insurance companies see

a net loss, they are forced to consider other options of how to reduce that loss.

The government did further regulate insurance companies to provide more options

to particular communities, but regulations alone did not solve the problem. There was a

need to develop partnerships between community groups, individuals and insurance

companies that would provide a mechanism where the insurance companies could see

that the investment by lenders, individuals and governments would result in socio-

economic improvements within communities. Those improvements would then result in

a lower risk for the insurers which would then result in lower losses or increase profits for

the insurance companies.

The Impact of Government Funded Programs

Beyond simply implementing several regulations that would force banks to

provide better products and services to low and moderate income communities, the U.S.

28

Government took steps to create several government funded programs that would provide

direct funding to banks and communities to support local efforts to strengthen home

ownership in underserved communities. These government programs range from grant

funding to targeted communities to grants provided to banks that were complying with

the new regulations.

With the government providing direct funding into the market, the underlying

intent was that a small investment would encourage other outside investors to also fund

efforts in the targeted communities. There is a clear basis for this theory within Keynes’

multiplier theory. The underlying intent from the government was that the minimal

federal investment would then stimulate other investments that could then result in socio-

economic improvements such as reductions in crime rate and unemployment within the

community (Hannsgen, 2007). Keynes would surmise that the initial public investment

and potential gains in employment would also result in more local spending by those

living in the community. Improved employment conditions would encourage those

newly employed individuals to spend more money, resulting in further increases in

employment. Keynes considers that the investment multiplies as the funds circulate

through the economy. Even though Keynes’ approach was based on accepted economic

theory, the actual existence of the multiplier effect in the community could be questioned.

For example, one could consider the investment of funds into an endeavor which directly

creates new jobs. Those gaining employment would spend their income on goods and

services which would generate further investment. However, Keynes’ theory may be

contradictory to the results of housing programs. Regardless of whether a project focuses

29

on developing new housing or rehabilitating existing housing, any resulting increases in

employment would be temporary and end at the conclusion of a specific project.

There are several challenges with assessing the effectiveness of the government

funded programs. First, with the various programs in existence, it is difficult to prove the

effectiveness of any single program in the market (Erickson, 2006) due to several

instances of program overlaps (Staudt, 2006). Secondly, many of the funded programs

require that the government funding is not the only source used on a particular project.

As discussed above, the intent of the government programs are that the public funds are

used to leverage private partnership and funding to support an overall project. However,

what is the common practice is that the seed funding from the government is simply used

to leverage other public funds from a federal level (Shear, 2007). This is also supported

by Basolo (2006) where it was not only clear that the vast majority of funds came from

federal grants, but in “over half of the cities spent no local dollars on housing programs”

(p. 107) which provides further evidence on the overdependence on federal support for

program funding. In contrast, Super (2005) notes that the lack of local investment could

also be the result of local governments waiting for the federal government to spend their

funds first rather than having the funds originate from local budgets.

One could consider the impact of the HOPE VI program in support of home

ownership. The HOPE VI program was designed to create more mixed-income

communities that would support both low and moderate income home ownership. The

intent of the HOPE VI program was that with a greater mix of people from different

backgrounds, all residents of the community would benefit (Jois, 2008). However, the

30

implementation of the program did not encourage middle-income families to enter the

specific community (Varady, Raffel, et al, 2005). Like those with low or moderate

income, the expectation of safe neighborhoods, good schools and a strong community

identity would be necessary prerequisites in order to enable prospective residents to

consider moving into a HOPE VI project (Hanngsen, 2007).

When considering the impact of school quality and neighborhood selection, one

could consider the city of Chicago to be an example of where school quality is a

significant factor in neighborhood selection. The Chicago Public School system is

composed by both locally assigned schools where residents of a particular area are

assigned to a specific school or magnet schools where students have an option to attend

out of neighborhood schools based on previous academic performance. Along with the

neighborhood and magnet schools, there are several charter school programs throughout

the city that commenced operations within the last five years.

There is significant diversity of location for the Chicago Public School System,

there is also a wide range of quality at the high school level. With the current school

funding model is based on property taxes in the local neighborhoods, one would

generally find that the schools of higher quality are located in higher income areas of the

city. In contrast, those schools with historically weak performance records are typically

in low income communities in the city. In the Chicago market, programs like HOPE VI

may prove difficult to promote until there was a supporting improvement in school

quality within the existing low income communities.

31

Varady, Raffel, et al. (2005) reviewed the performance of the HOPE VI program

in Cincinnati, Ohio. In this HOPE VI implementation, the local officials charged with

promoting the new housing program supported by the HOPE VI program focused on

developing housing in low income neighborhoods that would be attractive to market-rate

middle-income home owners. However, the officials “ignored the issues of schools and

middle-income families (p. 155).” In this circumstance, the goal of the project was to

provide a more diverse income community, but there was a failure to attract higher

income residents due to the perceived weakness of the local schools.

To make matters worse, Jois (2008) concluded that there were several examples

of HOPE VI programs that actually resulted in a net loss of affordable housing in

comparison to the environment prior to the HOPE VI project. What was missing was an

expectation that the resulting project should at least offer a break even in affordable

housing units. Unfortunately, while that should be an obvious expectation, the growth in

available housing units expectation is not currently built into the overall requirements of

the HOPE VI program. HOPE VI is not alone in this result of a net loss of affordable

housing stock. In fact, between the implementation of the Housing Act of 1949 and the

creation of the Department of Housing and Urban Development in 1965, there was a

propensity for a net loss of new housing stock throughout the period (Erickson, 2006).

The rationale for the Housing Act of 1949 was to build more affordable housing, the

result was that the while new housing was developed, there was less available to not only

the existing population in the community, but to new residents as well.

32

According to Wood (2007) of the Government Accountability Office of the U.S.

Government, there were also several concerns related to the performance of the program

as well as the Department of Housing and Urban Development’s oversight of the

program. Within the requirements of HOPE VI, the government is not to be the only

funding source for individual projects. The government funds should be leveraged to

access private funding sources for the project. In actual practice, nearly 79% of the

funding for HOPE VI projects came directly from government funds. Some of this lack

of leveraging may be due to inconsistent application processes for other sources of public

and private funding there is a need for a process that better supports the leveraging goals

and expectations (Shear, 2007).

Even though evidence of community improvement in the areas surrounding new

HOPE VI projects did exist, the Government Accountability Office was unable to

attribute the improvement to the HOPE VI project or other factors in the community.

The Government Accountability Office concluded that the Department of Housing and

Urban Development’s operational oversight was lacking. As a specific example, “HUD

did not have an official enforcement policy to deal with grantees that missed project

deadlines” (Wood, p. 9). This is problematic in that it is indicative of an inefficient use

of government funds provided in the construction grant.

However, in their same study, Varady, Raffel, et al. (2005) did identify one

community where the previous high crime area was transitioned into a mixed-income

community with significantly lower crime rate and higher levels of community

involvement. This community was the Park DuValle community in Louisville,

33

Kentucky. The planning for this program was significantly different than the planning

for the community in Cincinnati, Ohio discussed above. At the inception of the planning

process, the Park DuValle community planners focused on identifying methods not only

to attract low and moderate income first time home buyers, but also developed a plan to

promote the community to middle-income families as well. This HOPE VI

implementation was significantly more in line with the goals of the project and the HOPE

VI program.

The Park DuValle also established a local advisory council for the community and

worked with the City of Louisville to establish a new school that would better support the

expectations of middle-income families with children. The school effort as well as the

establishment of a community center and playground was attractive to both low income

and middle-income families. When comparing the changes in crime rate within the

Cincinnati and Louisville HOPE VI projects, the Park DuValle project saw a very

significant reduction in crime in comparison to the environment in the community prior

to the project launch. The only failure of this project, as the researchers note, was that the

goals of racial integration were not met as the resulting community continued to be

predominantly African-American (Varady, Raffel, et al, 2005).

There are examples of both successful and less than successful programs in these

efforts, Squires and Kubrin (2005) theorize that the motivation for programs such as

HOPE VI and the Community Reinvestment Act tend to focus more on issues of location

rather than supporting individuals within a community. They conclude that the focus

tends to be on improving a particular area or development that may have a high

34

concentration of a specific population. This focus on location rather should not be the

sole focus of a project. As they note, there is a greater need to “reduce the concentration

of poverty and segregation (p. 60)” rather than simply improving the housing and not the

overall demographics of the community.

As discussed above in the HOPE VI programs in Cincinnati and Louisville, the

program in Cincinnati failed in its attempt to encourage middle-income families with

children to enter the community. This could be defined as a focus on place rather than

people. This is in contrast to the Park DuValle project which focused on not only

providing better housing for the existing community, but also provided an incentive for

residents from outside of the community to consider entering Park DuValle. Given the

increase in community diversity of the Park DuValle project, this project would be

considered to be the result of a need to focus not only on the location and the property,

but to also focus on lowering the concentration of poverty in the area as well.

There must be a fine balance between both place and people. For example, if the

focus is too highly placed on the needs to decentralize poverty, there is a risk where a

majority of the population may simply leave the area and move to either other

neighborhoods or the suburbs rather than stay in a central city. The focus on place based

programs have a limited ability to show a positive benefit to the community (Tranel &

Handlin, 2006) with the result that those remaining in the central city population further

concentrate both poverty and in many cases, race as well (Bayoh, Irwin & Habb, 2006).

Where government programs are used, the need exists that the resulting newly developed

property should not only be attractive to new members, but also should be affordable to

35

the existing community as well (Fennell, 2008). If the suburbs are more attractive from a

cost and benefit perspective, there is less incentive for someone to remain in the

community when there are better alternatives in other areas.

HOPE VI is not the only example of a government program to support the

development of low and moderate income housing programs. Staudt (2006) concludes

that the multitude of programs might actually be causing more harm that good. The issue

with these multiple programs is that there is a limited level of mutual exclusivity between

individual government programs. The result is that the multiple programs overlap in

attempting address the same community need. This overlap results in two challenges.

First, there are additional costs associated with the redundancy and secondly, as discussed

above, it is difficult to determine if any one program is successful. In discussing the

theories of Milton Friedman, Staudt (2006) surmises that this lack of “coordination and

potential incompatibility of the programs can prove to be troublesome when assessed

separately but, but when investigated together, they border on the absurd (p. 1209).” This

conclusion further supports the idea that the apparent redundancy in these programs may

actually result in the effort being less efficient and more expensive when compared to the

potential outcomes of a single program and oversight body responsible for meeting the

market need.

This overlap occurs at the jurisdictional oversight level as well (Staudt, 2006).

When reviewing the processes related to federal government oversight of subsidized

housing programs, there are three different oversight bodies responsible for these

programs. In the House of Representatives, the Finance Committee oversees all

36

subsidies, the Banking Housing and Urban Affairs Committee in the Senate. The

Department of Housing and Urban Development are responsible at the Executive level of

government. One could conclude that with these overlapping oversight functions, there is

an inherent risk that there will be a direct conflict between the overall goals of the

program and the methods to achieve those goals.

In addition to HOPE VI, another government funded program is the Community

Development Block Grant program. The Community Development Block Grant program

was designed to provide funding to targeted communities where the existing housing

stock was falling into ill repair. The funding would then be transferred to state or city

government administrators for distribution to the targeted community (Super, 2005).

While the mission of Community Development Block Grant was clear as an investment

in low and moderate income communities, what was lacking was a realistic method of

assessing whether or not a specific community was eligible to receive Community

Development Block Grant funds.

Posner (2005) noted in his testimony to the United States House of

Representatives that the current funding eligibility model does not appropriately consider

either population size or poverty status when determining funding to an eligible

community. The result of the failure of the funding model is that there is a propensity for

cities to receive large grants where the actual needs in other cities receiving lower funds

is greater. The funding is getting out into the market, but the challenge is that with the

existing funding model, there is a risk that the funds are not getting to the communities

with the highest need Additionally, as Czerwinski (2006) noted that in addition to the

37

issues of population size and poverty status, the Community Development Block Grant

funding formula also gives additional weight to the amount of pre-1940s housing in a

specific community. It is common that older properties tend to need more repairs both in

frequency and expense, but there are two flaws with this additional age-based weighting.

First, it is likely more expensive to improve property that is over sixty years old and

secondly, that this method may ignore the needs of communities with newer housing

stock that may also have a significant need. The result of the age-based weighting is that

more funds may be spent to improve fewer older housing rather than trying to benefit the

needs of more home owners who may need lower cost repairs. The result is that fewer

residents are assisted at the expense of improving a smaller number of older properties.

Super (2005) considered that programs like Community Development Block

Grant where the program administration is transferred from the federal to the local level

is misguided. The transfer in administration results in a misalignment of programs goals

between the federal and state administrators. This misalignment is another example

where the federal government is providing the funds without the necessary guidance or

accountability to ensure that those funds are used in a manner that meets the program

requirements. A case could be made that the local governments have a higher awareness

of a particular neighborhood problem, there is a greater need for oversight by a joint

federal and local effort.

Along with the government funding provided to housing projects, the government

also provides funding to local community groups who are supporting local housing

programs. The government funds are administered through Technical Assistance Grants

38

that are provided to the local organizations. McCool (2002) from the Government

Accountability Office notes that there are problems with this program as well. While the

Technical Assistance Grants are designed to provide training support to local

organizations, the General Accountability Office often found that the funds were not used

for that purpose. Examples where grant funding was used for purposes such as training

grant writers to developers on more effective ways of navigating government funding

programs were cited as not being aligned with the requirements of the Technical

Assistance Program.

This failure is not as much a fault of the local organizations, but it is a failure of

the Department of Housing and Urban Development’s oversight and guidance as to how

the funds should be used (McCool, 2002). With the lack of oversight, the local groups

simply receive the funds and spend them on what they consider to be the highest

operational need at the time of receipt. Similar to the problems with Community

Development Block Grant and HOPE VI, the Department of Housing and Urban

Development does not measure the effectiveness of the Technical Assistance Grant

program within the local communities receiving the funds. Since there is limited

oversight on the actual use, it would be quite difficult to determine whether the program

in general has been effective without a thorough understanding of exactly what the funds

were used to support. Clearly, the General Accountability Office concerns about the lack

of measurement were justified since the U.S. Government sets aside between $100

million to $200 million in funds on an annual basis for the grant program.

39

Finally, one could also consider the effectiveness of incentive programs that are

provided to banks participating in loan programs for low and moderate income

communities. A prime example of a bank incentive program is the Bank Enterprise

Award program. The Bank Enterprise Award program was designed to give cash awards

to banks that were supporting community reinvestment activities in low and moderate

income communities (Scott, 2006). However, the Government Accountability Office

noted that there were several problems with the current administration of the Bank

Enterprise Award program. First, as with the other programs discussed above, there was

no objective measurement to determine whether the program was successful in spurring

new investment. In fact, there were several examples noted by the Government

Accountability Office that would indicate that not only had the actual impact been

overstated, but that specific research conducted by the Government Accountability Office

indicated that the impact of the program was of limited significance. Secondly, there was

no requirement that the banks spend the award on any specific area, it simply went into

the bank's balance sheet (Scott, 2006).

There were also several examples where the calculations used to determine what

the award amount should be were so poorly structured that there was a risk of

overpayment to a bank. Even when appropriate award amounts were provided, those

amounts were so small that the value of the award from the perspective of the bank was

insignificant when compared against the overall cash position of the bank. Finally, there

was a limited ability to determine any difference between incentives already provided

under the Community Reinvestment Act and those incentives paid through the Bank

40

Enterprise Award program. Instead, it was more common that banks focus on the

requirements under the Community Reinvestment Act rather than any cash incentive that

would be provided to them under the Bank Enterprise Award program (Scott, 2006).

Beyond all of the programs and financial incentives to the various stakeholders

supporting low and moderate income housing development, there was also a need to

develop a plan to encourage prospective residents to consider that home ownership is a

possibility. With limited information, the vast majority of low and moderate income

families will simply conclude that home ownership is not possible. Home ownership is

not a possibility for everyone, but it is necessary for the stakeholders to have a process in

place that can identify those individuals who have the highest potential to purchase a

home and provide an environment where the property acquisition process is both efficient

and customer focused.

Even though society is transitioning into an ownership society (Wray, 2006) one

should also consider that for many years, that same society has conditioned low income

individuals to become dependent on many programs such as Section 8 and welfare

(Grinstein-Weiss, Irish, et al, 2007) rather than home ownership. As such, when working

with people in these groups, there may be resistance to losing the benefits they have had

with the purchase of a home. It is often necessary to not only offer the available

programs to this population, but to also discuss how home ownership will create more

financial stability and improvement for the individual (Di, 2007). The idea being that

home ownership is promoted as an economic benefit, but in the eventuality, the new

owner must be financially self-sufficient (Wray, 2006). Along with the personal

41

economic benefit, these programs, when compared to entitlement programs like Section 8

and welfare can improve resident tenure in a community which results in a more cost-

effective use of federal funds (Hoff & Sen, 2005).

The communication process should provide support prior to and after the home is

purchased. This could involve programs such as matching savings programs that will

provide financial support for saving funds for a down payment (Grinstein-Weiss, Irish, et

al, 2007), programs to establish a reliable credit history as well as after purchase support

programs on personal budgeting and home maintenance. With the end result being that

the potential home owner is prepared to financially support the purchase.

Programs such as these could be supported within the theories of welfare

economics. At the heart of welfare economic theory is that wealth is redistributed to

those who need to have greater access to the funds. For home ownership programs, a

portion of paid taxes are redistributed to those individuals who would meet the programs

qualifications. In the example discussed above, public funds would be used to provide

either direct down payment assistance or matching of saved funds. In contrast to other

government funded efforts discussed here, these are examples of funds that would go

directly to the consumer rather than an intermediary such as a bank or local organization.

Another example of a government funded program, albeit at the state and local

level, is the Low Income Housing Tax Credit program. Low Income Housing Tax Credit

programs allow a local government to use a tax incentive to a property developer when

that developer is providing greater housing access to low and moderate income families

(Mueller & Schwartz, 2008). As a result of a Low Income Housing Tax Credit, the

42

developer is able to use the tax credit against future taxes owed to the issuer of the credit.

In addition, Low Income Housing Tax Credits can be traded or sold in a similar manner

to other assets owned by a developer with the new owner having the ability to apply the

tax credit in the same manner that the original owner possessed.

Low Income Housing Tax Credits can prove to be a very effective method in

facilitating new development of low income housing (Jaffee & Quigley, 2008). The

developer receives a financial incentive to develop housing to meet the needs of this

population and will often pass some of the savings to new home owners with price

reductions when the property is available for sale. However, with Low Income Housing

Tax Credits, there is no funding that actually changes hands. Since this is a tax credit, the

local government is not providing direct funding to the developer, the government is

simply allowing the developer to apply a credit towards future taxes owed.

Nevertheless, while there is a benefit to the utilization to Low Income Housing

Tax Credits, there has also been a failure to appropriately assess the long term financial

impact on government budgets for these expenditures (Erickson, 2006). Unlike a

program that provides direct funding for a particular program or project that is normally

budgeted, Low Income Housing Tax Credits can be used at any point after the tax credit

is awarded. Given the flexibility of use, it may be difficult for local governments to

predict how these credits will impact future tax revenues.

Fennell (2008) considers that there could also be a financial incentive for a new

resident to enter a community where that incentive works in a similar manner to a Low

Income Housing Tax Credit for a developer. In this option, the buyer and developer

43

could receive a tax credit. After the property acquisition, the buyer could either get a

credit towards property taxes owed or could have a structured reduction of their property

taxes over a determined period of time that would provide the additional financial

incentive (Conley & Gifford, 2006). Dreier (2003) also concluded that these tax credits

could act in a similar manner to the Earned Income Tax Credit currently available to

qualified applicants.

Even with the financial incentives for developers to offer low income housing,

that does not necessarily imply that the Low Income Housing Tax Credit program has

been successful. As Mueller and Schwartz (2008) discuss, there is limited data that

would indicate that the tax credit related to the targeted development of low income

housing actually satisfies the actual market demand. The conclusion being that the

financial incentive needs to be sufficient for the developer to provide the housing. While

there is a social benefit to meet the needs of the population, if the developer cannot offer

the project at a sufficient profit in comparison to non low income housing projects, there

would be little likelihood that the developer would implement a project where the

potential of profit maximization does not exist.

Hendrikson (2006) concluded that the focus of civic leaders cannot simply be of a

financial nature. There is also a necessity for governments to provide more favorable

terms that not only balance the financial risks associated with developments in the inner

city. This balancing should include the financial incentives, but should also include

easing requirements on permits and other steps necessary to efficiently develop the

property. One should also consider the impact on regulations such as zoning on the

44

community (Staley & Claeys, 2005). While zoning areas as single-family or owners-

occupied would support home ownership over renting, those same practices could also

exclude renters from residing in a particular neighborhood (Nelson, Dawkins & Sanchez,

2004). The result is that the government is not simply being a funding source, but they

are also engaging in a partnership with the developer as well. Thereby balancing the

needs of home ownership, businesses and renters to support effective community

development programs (Jois, 2008).

Even with the varied programs discussed above there is limited data that actually

assesses an individual program’s effectiveness in the community. By U.S. government

requirements, program performance must be assessed by objective performance

measurements. This is not simply based on the amount of money invested, but the

outcomes that resulted as this would be a better measurement of the return on the

government’s investment (Staudt, 2006). There are many examples of an investment

yielding an increase in housing stock, but in most cases those improvements were the

result of several programs rather than one particular program under review.

Conley and Gifford (2006) conclude that there is little connection between

programs that provide opportunities to provide economic parity and home ownership.

From their research, programs that offer additional financial assistance or incentives do

not necessarily result in any significant increase in home ownership within a community.

In fact, as they note, “home ownership tends to prevail where state commitments to social

insurance programs are smallest” (Conley & Gifford, p. 75). There is a clear limit to the

information that would justify not only why government programs and regulations may

45

not necessarily generate improvements in home ownership. As discussed below, there is

cause to question that there are any socio-economic benefits for the community in the

event that home ownership does increase based on the investment.

Socio-Economic Impact of Government Programs and Regulations

The long standing justification for the regulatory changes and funding from the

U.S. government was that funded programs and regulations would result in improved

housing conditions in low and moderate income communities. As a result of those

changes, other socio-economic benefits of home ownership would also occur (Frame &

White, 2005). These benefits were to include improvements in employment, crime rate

and educational performance within the community.

When specifically examining the impacts of home ownership and general health

of home owners, Laaksonen, Rahkonen, et al (2005) indicated that while there was a

relationship between general health and home ownership, over 50% of good health was

explained by other socio-economic factors rather than home ownership as a primary

factor. There was an indication that home owners were healthier than those that did not

own homes, but the notion that there was relationship between the two variables could

not be completely supported.

When considering the relationship between crime rate and home ownership, there

was limited evidence to prove a relationship that would indicate that those communities

with higher home ownership rates have lower crime rates. Other factors such as

“poverty, racial composition and instability” (Tita, Petrus & Greenbaum, p. 310) also

have a role in crime rate. There are communities that have a higher than average ratio of

46

home ownership that still have significant issues of crime due mainly to other socio-

economic factors that also impact the condition of a community.

Research does support that crime can actually lead to the decline of a community

which could then result in declines in home ownership (Tita, Petrus & Greenbaum,

2006). One should not assume that the reverse is true. While declines in home

ownership do lead to increases in crime, it does not necessarily hold that increases in

home ownership would lead to decreases in crime. One could surmise that this decline

could very well be the result of individuals leaving the community to other areas

considered to be safer, but it could also be the result of declines in existing property

values resulting from the impact of negative changes in factors such as crime rate

(Hanngsen, 2007 and Gibbons, 2004). Lynch and Rasmussen (2004) conclude that the

data supporting a relationship between crime rates and home values is inconsistent. They

conclude that it is a combination of the immediate surroundings of a property as well as

the general environment in the city as a whole. Individuals may interpret higher crime

areas as risky from a housing investment perspective, the data to support that conclusion

may not be generally applicable to all communities.

The opposite is true when there are improvements in crime rate. If the housing

situation it is likely that property values will increase. This is a positive result for the

home owner, but it is also necessary to consider the impact that increases in property

values have on the remaining renters. As property values increase, property taxes will

also increase. The result is that those property tax increases are commonly passed from

landlord to tenant in the form of rent increases (Hoff & Sen, 2005). With the improved

47

property values, there is a corresponding increase in net worth of home owners along

with a similar decrease in net worth of renters. To make matters worse, if a renter does

elect to purchase a home in the future, they would end up paying a higher price for the

home when compared to the price if they entered the market with the initial offering

(Jeske, 2005).

Increases in crime rate can lead to a downward spiral for the entire neighborhood.

As more people take the option to leave to other areas and fewer owners enter the

community, existing property would likely be neglected and result in more residents

departing and conditions further deteriorating (Gibbons, 2004). In the eventuality, the

result is that the remaining community members are those with no choice but to stay in

the community while simply attempting to survive the poor conditions. While factors

such as crime rate do have a negative impact on most communities, they tend to have a

much more significant impact on disadvantaged neighborhoods (Tita, Petrus &

Greenbaum, 2006).

When considering the impact of home ownership on employment, Munch,

Rosholm and Svarer (2006) theorized that home ownership may be a barrier to

employment. Their perspective was that people who owned homes were less mobile for

employment opportunities in comparison to renters within the same locations. The idea

being that people who owned homes were less likely to consider relocation for

employment as it would likely require selling their home. The lack of mobility could

then result in a loss of investment value or equity in a slow housing market. The result

being that the home owner would have to balance the potential gain with the new

48

employment opportunity against a potential loss that could result from the sale of their

existing home. There was a potential of improved employment in the event of

improvements in home ownership, but the research indicates that home ownership could

stifle future employment opportunities for those same home owners.

In the same study, Munch, Rosholm and Svarer (2006) also concluded that when

home owners became unemployed, their transition to regaining employment was much

shorter in comparison to renters. This quicker transition may be the result of other socio-

economic factors such as average level of education between renters and home owners

which would not be a reflection on whether or not an individual owned a home. While

this study did not focus solely on low and moderate income communities, the issues of

employment tend to be significantly higher in these communities when compared to the

general population. Unemployment is typically much higher in low and moderate

income communities when compared to the general population. While the study does

indicate that home owners tend to have a lower transition period from one job to another,

the results did not indicate that improvements in home ownership would necessarily lead

to improvements in employment, in the general population or in low and moderate

income communities. Based on the existing research one cannot objectively assess

whether or not a relationship between home ownership and employment exists.

In general, there is a strong link between education and socio-economic status.

Those with higher levels of education tend to be better off from a socio-economic

perspective which then supports improved employment opportunities (Tolnay &

Eichenlaub, 2007). Additionally, Zhan (2006) considered that education of single

49

mothers has a significantly greater influence on socio-economic mobility when compared

to home ownership. In their research, Caner and Wolff’s (2004) concluded that

individuals who did not complete college or university “were twice as likely” (p. 500) to

be asset-poor in comparison to those who completed college. One could assume that if

home owners have less flexibility in a job search in comparison to renters, Munch,

Rosholm and Svarer (2006) surmise that the shorter transition time is likely due to a

combination of other factors exclusive of home ownership.

There is also a disparity in race and education. Segal (2007) notes, that while

education is an “indicator of future economic success” (p. 69), there is a significantly

higher likelihood that a white person will complete a college education in comparison to

African-American or Latino-American individuals. One could conclude that the issues of

affordable housing are not simply the result of having affordable financial options

available to loan applicants. The lack of a college education may also be a barrier to

affordability. Without a significant level of education, individuals may not have

sufficient financial success that would make a home more affordable. This link between

education and economic success might serve to justify an increased government

investment in educational programs.

The link between education and economic strength is another example where a

relation may exist, but the order of the events is reversed. It is education that leads to

home ownership rather than an assessment of whether home ownership results in

improved educational performance. The Segal (2007) study concludes that those with

higher levels of education tend to have higher levels of income making home ownership a

50

more realistic possibility. The resulting change in home ownership is due more to the

resulting changes in wealth from educational attainment. While this study did not focus

on a particular community, it also did not conclude that improvements in home

ownership would lead to improvements in educational attainment or performance.

Home ownership could also be interpreted as an investment opportunity for the

home owner to build a greater sense of financial security when compared to those who

rent (Hogan, Solheim, et al., 2004 and Zhan, 2006). This financial stability is supported

by the idea that home owners can build equity by owning a home where renters do not

have the opportunity to build equity. That equity then can be accessed in the events of

future financial hardship or in the event of a future sale of the property (Caner & Wolff,

2004). Owners also gain significant tax advantages by ownership with the tax

deductibility of property taxes and mortgage interest that is not available to renters

(Fennell, 2008 and Beland, 2007).

In addition to the investment benefit, there is also a benefit that home ownership

can bring to any neighborhood, stabilization (Galster, Marcotte, et al, 2007). Typically in

neighborhoods with higher ratios of home ownership, community members tend to stay

in their properties for a longer period of time when compared to renters. The result is that

with more consistency, individuals are able to establish longer relationships with their

neighbors as well as take a greater amount of pride in their communities (Guest, Cover, et

al, 2006).

The link between low resident turnover and wealth building could then support

the desire of individual residents to establish a stronger community to support that

51

investment (Guest, Cover, et al, 2006). However, as discussed above, in order for the

investment value to grow, the general socio-economic performance of the community

must at least be maintained or improve over time. A home can be considered to be a

worthy investment, but other factors such as crime rate are also considered by a

perspective home owner in their assessment of the risk of that investment as well as

supporting a decline in existing home values (Tita, Petrus & Greenbaum, 2006).

In a similar study, Galster, Marcotte, et al. (2007) also concluded that the socio-

economic conditions of a neighborhood would have a significant influence on whether or

not a purchaser would select a specific neighborhood to enter. These additional socio-

economic conditions would have an impact on both affluent and low income populations

in their selection process. More affluent buyers would be much less likely to select a

neighborhood if there was a perception of existing socio-economic problems.

Conversely, those low income families considering the purchase of a home may be

limited in their selection process due to prices in these same, socio-economically

challenged neighborhoods.

Racial homogeneousness can also be a selection factor for potential home buyers

within the community. In the event that a community has a significant majority of one

particular ethnic group, that issue may discourage members of other ethnic groups from

considering the purchase of a home in the community (Tita, Petrus & Greenbaum, 2006).

For example, if a community had a majority of African-American residents, someone of

Hispanic or Western European origin would be less likely to consider that community

52

and more likely to consider other communities where their ethnic group was either in the

majority or of a significantly larger portion of the existing population.

Research does not show that individuals will choose a neighborhood based solely

on a social context. When an individual is considering a particular neighborhood to live

in, they do not typically select that neighborhood based on any existing social network

(Guest, Cover, et al, 2006). One could conclude that since this network is not ‘known’

to the perspective resident, it would be difficult for the resident to assess this

characteristic as a motivation to purchase property and live in a particular neighborhood.

Fong, Bowles and Gintis (2005) built upon welfare economics theory in their

conclusion that from a behavioral perspective, people are generally more than willing to

provide financial assistance to the poor, either through private donations or by tax payer

support. However, if there is a public perception that those being assisted are taking

advantage of the financial support, there will be significant resistance to providing

continued support for future needs. This perception is similar to the view of programs

like welfare and Section 8 where the public considers that these programs were designed

to address a temporary hardship of the recipient not for long term use. Where individuals

may receive support from these welfare related programs, the challenge was the related to

the expectation that people would eventually no longer need the benefit. Even though

recent legislation has placed limits on the length of time spent on welfare or receiving

Section 8 funds, there is still a perception that individuals are dependent on the funding.

There is a desire to assist people in need, but if those receiving the funds are not using

those funds in a manner that will eventually improve their socio-economic standing, the

53

funding will not see continued support The public perception of housing assistance

programs may not be as much about giving people an opportunity to improve their

financial condition, but more aligned with the perception that these programs are simply a

different version of existing welfare programs. Rather than assisting low and moderate

income individuals with a one-time investment of tax funds, society may view these

programs as another method of income redistribution without a measureable benefit.

The benefit also needs to be balanced against “egalitarian policies that reward

people independently whether or not they contribute to society (Fong, Bowles & Gintis,

p. 292).” It is often necessary to position the support in a manner that would not only

benefit the individual receiving the support but the greater society that is providing the

support. In considering the example of programs supporting low and moderate income

home ownership, those responsible for program administration will commonly discuss

factors such as appreciating property values and lower crime rates that can result from

these programs. The wider benefit can then be used as a justification for the financial

support as the larger society, by assumption, could stand to benefit as well.

Basolo (2007) notes that a majority of survey respondents indicated that housing

for moderate income families should be driven by market forces and not government

support. In reference to low income, there was greater support that those programs

should be driven by a public and private partnership. This may indicate that there may be

a desire for support, but that support should be needs-tested at different levels rather than

attempting to create subsidies that address the needs of both low and moderate income

families (Shlay, 2006).

54

From a political perspective, as populations grow in the suburbs and shrink in the

cities, the political influence of the cities will diminish (Hendrikson, 2004). The result of

the reduced political influence is that while there may be evidence of a need to invest in

inner city housing and general infrastructure, there is less political support to do so

resulting from the increases in representation that could result in the suburbs (Jois, 2008).

The perspective would be that the more affluent suburbs would expect that a significant

percentage of the taxes collected would be spent locally rather than redistributed to the

inner city (Segal, 2007). Programs that encourage a shift to the suburbs could further

serve to perpetuate the inner city problems that already exist (Staudt, 2006).

Along with the shift in voters to suburban districts, there are also political limits

that tend to arise if a program or regulation remains unchanged over a period of time. It

is often necessary to evaluate the relevance of a program as the environment that program

was designed to support also changes (Beland & Wadden, 2007). In order for the success

to continue, the program must evolve as the community evolves. A program cannot be

developed that will consistently meet the market needs over a significant period of time,

but what tends to occur in the present environment is that a program or regulation is

implemented and future updates are rarely implemented.

Gap in Research

There was significant research related to understanding the socio-economic

benefits of home ownership, but the research was limited to studying general populations

rather than low and moderate income populations. The gap in the research lies within the

lack of existing research related to the objective benefits of home ownership within low

55

and moderate income communities. As discussed in detail within this chapter, the

existing research does support an objective benefit to home ownership in general, but

there was limited literature to support or reject the theory that there is a greater

community benefit to home ownership within low and moderate income communities.

Secondly, an underlying gap existed in that there was limited research or literature related

to the any improvements in the dependent variables under consideration in this study.

This was especially important given the amount of Federal, State, and private investment

in the development of low and moderate income home ownership programs where the

actual benefit of those programs to the target community is relatively unknown. There

appeared to be a moral rationale for giving low and moderate income families a greater

opportunity to own a home, there was a question as to whether a redistribution of wealth

by means of tax dollars was the appropriate manner to address the socio-economic

conditions of a particular community. One must consider whether a financial incentive to

this community to purchase a home was any different than a payment made by means of

welfare or other entitlement programs. What was unknown and worthy of further study

is whether or not a change in home ownership and the related monetary investments to

support it yielded a greater benefit to the community beyond the individual home owner.

In conclusion, the only applicable research not reviewed or included in this

chapter were additional examples of the benefits of home ownership within general

populations. As I have already included several examples of this research, the addition of

more studies related to this topic were redundant with the existing information. Finally,

while these were valid examples of the benefits of home ownership, these studies did not

56

provide any greater insight into the needs of low and moderate income communities

which, as stated above, was the research gap that this research desired to fill.

Summary

Over the past thirty years, the U.S. government has implemented programs and

services in order to establish greater opportunities for home ownership for low and

moderate income families. During this period, the government regulations have focused

on holding a greater level of accountability for lenders to better serve their communities

while funding programs that would provide a financial incentive for all stakeholders to

participate in the effort. The regulations were implemented and the money was spent, but

there were limited examples where specific regulations and programs succeeded in

meeting their individual objectives. There were thousands of examples where low and

moderate income families were able to purchase their own homes, but there was limited

information about the long term socio-economic benefits of home ownership in these

communities. As such, it was necessary to assess whether those long term goals are

within the influence of these regulations and government funded programs.

Continuing in chapter 3, I discussed the specific research methodologies that were

implemented in order to effectively assess the impact of home ownership for low and

moderate income communities. In addition to an overview of the methodologies, a

summary of both the independent variable, home ownership, and the dependent variables

of crime rate, unemployment rate, high school graduation rate and high school test scores

were discussed as well as any resulting effect that occurred in higher home ownership

communities.

57

CHAPTER 3:

RESEARCH METHOD

Introduction

In this chapter, I will focus the discussion on the three major components of the

research method: the research design, population of the study and the definitions of the

collected data. Within those major components, I will also discuss how the collected data

relate to the research questions and hypotheses discussed in chapter 1. At the conclusion

of this discussion and leading into chapter 4, I will discuss the collected data and how

that data support or reject the hypotheses discussed in chapter 1.

Description of the Research Design

My focus of the research design was the analysis of existing historical data

defining both the independent and dependent variables. Given that the focus of the my

work was to determine the socio-economic effects of low and moderate income home

ownership, the most applicable design was to analyze the existing data for both the

independent and dependent variables in order to assess whether any there was any effect

of home ownership on any of the socio-economic variables. In considering the socio-

economic measurements discussed earlier, I can use this information to serve as a gauge

for measuring overall community health, thereby formulating a conclusion as to the

impact that home ownership has on community health within low and moderate income

communities. Quantitative analysis of the existing data can also serve to support my goal

of establishing findings that may be generalizable to a larger population outside of the

sample population.

58

The first of the socio-economic indicators I considered related to crime rates in

the community areas. The overall crime rate can serve as an indicator of community

health, but it is also necessary to consider the major categories of crime within the data

collection and analysis as well. By considering both aggregate crime and the individual

crime categories, I will be able to assess the relationship that home ownership has on

crime in general as well as the specific categories. While a qualitative analysis could be

used to gauge the perception of crime within a community, those results would not be

generalizable to communities outside of the survey population. Since the generalizability

of the information is a goal of my research, a quantitative analysis of the data is

appropriate for this study.

When considering unemployment data, a quantitative analysis would also yield

useful results. As is the case with crime rate information, unemployment information is

generally available in a numerical format. A qualitative study could certainly identify the

number of people unemployed or not regardless of home ownership; it would not have

supported an assessment of a relationship between the two variables. Qualitative data

would indicate the percentage of participants who were or were not home owners as well

as those who were or were not employed. Similarly to the discussion of crime data, the

findings would not be generalizable to the larger community.

Finally, when considering the variables related to high school academic

performance, I intended to see whether home ownership was related to improvements in

academic performance. For the two considered variables, high school graduation rate and

ACT test score performance, a qualitative study was not applicable. Given that both of

59

these variables were considered to be quantifiable data, a qualitative study was not

appropriate. The community perception of improvements in educational performance

may be helpful, it was the objective performance measurements that would define

whether a relationship existed or not.

There are likely examples of qualitative analysis that could apply to the context of

this study. However, my goal of this research was not to consider subjective factors such

as individual stories, perceptions or the dynamics of specific groups within the

communities. Instead, the goal was to collect objective and accepted data for the

variables that could result in findings that would be generalizable to other low and

moderate income communities. Given the goal of generalizability of the findings, my

conclusion was that a quantitative analysis was the most applicable method.

I used regression analysis to determine the relationships between the individual

dependent variables and whether those shared relationships had any impact on other

socio-economic measurements. First, I needed to ensure that my independent variables

were not highly correlated. To address the risk of multicollinearity, I ran regression

analyses between the major variables in the study: crime rate, home ownership, ACT

score and high school graduation. I also included data related to median income within

the regression analysis to address potential external economic factors. The regression

analysis also examined the extent to which home ownership, on its own, and in

conjunction with other independent variables, explained the variance in community

health (Singleton & Straits, 2005).

60

I also used one-tailed t test comparing two population means as it would be

possible to infer whether changes in an independent variable were related to any changes

in the dependent variables under consideration. With this information, I could make an

inference as to whether a change in home ownership also resulted in a change in any of

the dependent variables. For example, if the t test results indicate that an increase in

home ownership also resulted in an increase in ACT test scores, I could infer that a

relationship between home ownership and ACT test scores existed. In this example, it

would be a positive effect, but the t test could also identify a negative relationship as

well.

Once a relationship was identified between the independent variable and any of

the dependent variables, the next step in my research design involved the assessment of

the significance of that effect. The level of significance indicated the extent to which the

two variables were related to the exclusion of other variables. Thus, when the statistical

significance was higher it was less likely that factors other than the independent variable

caused the resulting changes in any of the dependent variables (Singleton and Straits,

2005).

Target Population

The target population for the study was two low and moderate income community

areas in Chicago, New City and Austin. The Austin community had a preceding history

of a 3.6% improvement in home ownership and New City had a preceding history of a

4% decrease in home ownership. This population was not surveyed or interviewed, but

the historical data related to both the independent variable and the dependent variables

61

were analyzed to either reject or fail to reject the hypotheses. These community areas

were selected based on their higher proportion of low and moderate income residents as

well as more recent changes in available owners-occupied housing within the

communities. The racial makeup of each area was different; where Austin had an 87.5%

African-American population and New City had a 32.8% Hispanic-American population.

Both communities maintained populations that were a majority of non-white residents.

While the ethnic composition of the population was different, the similarities were

supportive of an effective comparison.

In reference to data related to unemployment, the target population was Lake and

Marion Counties in the State of Indiana. Because unemployment data were not collected

at the community area level, it was necessary to consider county-based information.

Both Lake and Marion Counties met the needs for this study because they had a higher

than average proportion of low and moderate income families. In contrast to the Chicago

community areas, both Lake and Marion counties were composed of a majority of white

residents.

Sample and Sampling Method

Sampling criteria was based on an assessment of the historical changes in home

ownership within the community areas and counties. Because the purpose was to assess

whether the independent variable has any relationship to the dependent variables, the two

neighborhoods and counties were selected based on their relative changes in home

ownership ratios in comparison to each other. Other than the differences in total

population, the demographics in relation to age and minority composition in the

62

community areas were similar. It was necessary, then, to identify two community areas

that met the definition of low and moderate income.

The sample for this study was based on locating the highest concentration of low

and moderate income residents within the City of Chicago. In both the selected

community areas, there were a higher proportion of low and moderate income residents

when compared to the general composition of the city. The selection was based on

understanding the performance of the studied variables within this population. Given the

gap in the research when reviewing home ownership in this specific population, it was

necessary to identify specific areas where the low and moderate income population was

at a much higher level.

Data Collection

With the data for analysis in the public domain, the first step was to determine the

particular community areas being considered. As a basis for this analysis, I accessed

information via the Federal Financial Institutions Examination Council. Federal

Financial Institutions Examination Council data summarized information related to

housing data and median income level by census tract within the City of Chicago. For

the purpose of this research, the Department of Housing and Urban Development

definition of low and moderate income was considered to be those populations that have

an income at or less than 80% of the median income for the City of Chicago. Once the

general locations of low and moderate income areas were determined, the final step in the

neighborhood selection was to identify contiguous census tracts within the specific

community areas in the study. The same process was applied to the counties. I accessed

63

the data from the Federal Financial Institutions Examination Council at the county rather

than the census tract level to identify the best suited counties for the study. The resulting

information yielded the level of owners-occupied housing growth, as a percentage of total

available housing over the same data collection period.

Data Analysis

Given that the selected community areas and counties were identified, my data

analysis focused on assessing the effect of the observed changes in the owners-occupied

housing population against the other socio-economic variables discussed above. The data

related to each socio-economic variable was studied for each community area. In the

detailed summary below, I will provide further information on each of the variables in the

study.

Using the information provided by the Chicago Police Annual Reports, the crime

rate information was categorized into eight classifications covering all types of crimes

that occur within the city. The first of those categories was murder. Murder was

considered to be an act of an individual that results in the death of another individual.

There are several categories or degrees of crime, this classification included all incidents

where the criminal activity resulted in the intentional death of the victim. The number of

reported incidents was considered to be those that occurred within the specific

community area.

The second category is criminal sexual assault. Criminal sexual assault was

defined as a criminal activity where the event was sexual in nature. It was not assumed

that this category would only include incidents such as rape where a form or sexual

64

penetration can occur. As such, it was also necessary to include incidents such as

inappropriate touching and other acts that could also be defined as sexual in nature.

The third category of crime was robbery. Robbery was considered to be a crime

where the offender takes property by force or intimidation. As was the case with other

crimes, there were subcategories within robbery. As an example, an event that the

offender used a weapon in commission of the crime, was considered to be an armed

robbery. Actions such as a mugging where no weapon was used was also considered a

robbery.

The fourth category of crime was aggravated assault and battery. Aggravated

assault and battery crimes were defined as violent crimes committed against a victim

where the result of that crime is an injury to the victim that did not result in death. These

crimes were not commonly the result of any sexual activity or sexual touching as was the

case with sexual assault. These crimes did not include events where any sort of theft

resulted.

The fifth category was burglary. Burglary crimes were defined as crimes that

were the result of thefts that that took place within the residence of the victim. These

crimes would often result in the theft of personal property contained within the property

but outside of a personal residence. There were incidents where the residence is forcibly

entered and no theft has taken place. This was still considered a burglary as the residence

was entered without the owner’s consent.

The sixth crime category was theft. In this category, the crime was committed

when the personal property of the victim was taken without consent outside of the

65

victim’s residence. Theft crimes did not necessarily require that the victim be an

individual. For example, when considering the crime of shoplifting, the victim was not

an individual person. Instead the victim was the business itself.

The seventh crime category was motor vehicle theft. As was the case with theft,

motor vehicle theft also involved non-consensual seizure of property. The Chicago Police

Department considered that this form of crime should be tracked separately. With motor

vehicle theft, the stolen property was specifically related to the motor vehicle owned by

the victim. This form of crime included not only the theft of the motor vehicle itself, but

also included the theft of any installed equipment such as tires or radios in the vehicle.

The eighth crime category was arson. Arson crimes related to the intentional

damage to the victim’s residence or personal property by means of fire. Arson crimes

resulted by not only the intentional damage to property of the victim’s residence, but it

would also include the intentional damage to property, such as a motor vehicle, that was

not contained within the residence. It was necessary to consider all crimes where fire

causes damage to be categorized as arson.

It was necessary to consider the historical crime data for each category for both

community areas. In order to provide a sufficient baseline for the review, the reported

volume of occurrences were applied as ratios of the frequency against the total population

of the community area as identified in the 2000 United States Census. By using ratios, I

was able to address issues related to the different populations in each community area.

With this information, I was able to make an assessment as to the effect, if any, between

66

each category of crime, the overall crime rate, and the ratio of owners-occupied housing

within the specific community area.

As discussed above, unemployment information was not tracked at the census

tract or community area level. It was necessary to use a larger area where the

unemployment data is reliably tracked. In this circumstance, the data analysis assessed

the relationship, between the annual unemployment rate, reported by the Bureau of Labor

Statistics against any changes in housing trends within the counties since 1997.

In reference to the academic performance information, I examined the

performance at Manley and Marshall High Schools in the Austin Community area and

Tilden and Gage Park High Schools in the New City Community area. As discussed

above, the ACT score information as well as graduation information from 2001 – 2008

was available through each school’s Academic Scorecard Report. Those students who

take the ACT test as well as graduate in 2001 were applicable to those students who

entered the ninth grade of high school in 1997.

For the ACT score information, I focused on analyzing the weighted average

ACT Composite Score for each community area. By weighting the scores I was able to

address the different populations in each high school. With this information, I could

assess the relationship between home ownership and the ACT score performance. In

reference to high school graduation rates, I applied the same weighting process to

determine the weighted average graduation rate for the high schools in each community

area. As with the ACT information, I assessed the graduation rates against housing to

determine if any relationship exists between the two factors.

67

To answer the above research questions, I collected annual data on home ownership as

percentage of households, crime rate, unemployment rate, graduation rate, and test scores for

both community areas since 1997. I then analyzed the data to test the following research

hypotheses.

Hypothesis 1: Relationship between homeownership and crime rate

CRLCRH

CRLCRH

H

H

:

:

1

0

Where CRH was the average annual crime rate in the high homeownership community area

over the study period and CRL was the average annual crime rate in the low homeownership

community area over the study period. Crime rate was defined as the percentage of crimes

committed among the population of the neighborhood. The crimes were limited to those

reported to a government authority that could result in criminal punishment in the event of

prosecution of the offender.

I used a one tail t test for comparing two population means to test hypothesis 1.

Rejection of the null hypothesis implied that a relationship existed between home ownership

and crime rate in the low income community area with high homeownership. I also performed

a regression analysis to further study the relationship between crime rate and housing. In the

regression analysis, crime rate was considered to be the dependent variable. The independent

variables in the analysis were housing and ACT test score performance.

Hypothesis 2: Relationship between homeownership and unemployment rate.

68

UELUEH

UELUEH

H

H

:

:

1

0

Where UEH was the average unemployment rate in the high homeownership county over the

study period and UEL was the average annual unemployment rate in the low homeownership

county over the study period. Unemployment rate was defined as the percentage of the county

population that was not working but was actively seeking employment.

I used a one tail t test for comparing two population means to test hypothesis 2.

Rejection of the null hypothesis implied that a relationship existed between home ownership

and unemployment rate in the low income community with high homeownership.

Hypothesis 3: Relationship between homeownership and standardized test scores

TSLTSH

TSLTSH

H

H

:

:

1

0

Where TSH was the average ACT test score in the high homeownership community area over

the study period and TSL was the average ACT test score in the low homeownership

community area over the study period. The ACT test score was defined as the reported ACT

composite score resulting from students completing the mandated ACT test at the end of their

junior year of high school.

I used a one tail t test for comparing two population means to test hypothesis 3.

Rejection of the null hypothesis implied that a relationship existed between home ownership

and the average test score in the low income community area with high homeownership. I also

used a regression analysis to further study the relationship between ACT score performance

69

and housing. In the regression analysis, the ACT score performance was considered to be the

dependent variable. The independent variables in the analysis were housing and median

income.

Hypothesis 4: Relationship between homeownership and high school graduation

GRLGRH

GRLGRH

H

H

:

:

1

0

Where GRH was the average high school graduation rate in the high homeownership

community area over the study period and GRL was the average high school graduation rate in

the low homeownership community area over the study period. The high school graduation

rate was defined by the percentage of students entering public high school that completed the

graduation requirements within five years of entry.

I used a one tail t test for comparing two population means to test hypothesis 4.

Rejection of the null hypothesis implied that a relationship existed between home ownership

and high school graduation rate in the low income community area with high homeownership.

I also used a regression analysis to further study the relationship between high school

graduation rate and housing. In the regression analysis, the high school graduation rate was

considered to be the dependent variable. The independent variables in the analysis were

housing and ACT test score performance.

Validity and Reliability

Since all of the data in the study existed and was in the public domain, the data

was considered both valid and reliable. As discussed above, in some cases, it was

70

necessary to apply ratios where actual quantities are reported from the outside sources.

For example, the data related to housing from the Federal Financial Institutions

Examination Council was recorded as a total quantity rather than a ratio of available

housing to the population. This also applied to information on crime rates within the

community areas. As reported by the Chicago Police Annual reports, the incidents in

each community area were not reported as a ratio, but instead as the frequency of a

particular category of crime. To have an equal baseline of the data between community

areas, I considered crime incidents as a ratio against the total population of the

community area. Finally, in reference to the academic performance information, the

Chicago Public Schools reported information as actual incidents of graduation and

average scores of the ACT at the high school level. In order to address the different

populations of the high schools, I applied a weighting process to determine the average

graduation and test score data to account for the differences in population at each high

school.

Other than the need to convert the data to ratios, the data measurements were

considered to be reliable as the measurements measured the intended variables. Home

ownership, while considered a measured data set, simply quantified the amount of

owners-occupied homes that existed within a specific census tract. The census tract on

its own is of no significance. The quantity of owners-occupied homes was consistently

measured throughout the market. When considering the dependent variables, I

considered these to be reliable measurements as well. In reference to crime rates, since

the data for each category was the result of a specific reported crime, the location of that

71

crime has no impact. The definition for terms such as murder, arson or motor vehicle

theft were consistent both at the local and national level. There may have been some

variance related to the definition of assault or aggravated assault, but any inconsistencies

in reporting are likely insignificant between reporting agencies. Finally, when

considering the measurements of high school academic performance, the first

measurement, the ACT score results was assumed to be reliable. Since the ACT is an

examination that is administered throughout the U.S., I assumed that the data was reliable

regardless of testing location. This reliability also applied to high school graduation

rates. There may be certain state-based variations on high school graduation

requirements related to particular subjects, but the overall expectation of meeting a

required set of curriculum requirements was consistent regardless of location of the

student or high school. While the required classes may be slightly different, I considered

the overall graduation rates to be reliable.

There is still a risk that other unknown factors could be impacting any changes in

the socio-economic measurements, but the fact that these community areas are within the

general area of the city served to mitigate much of that risk. That is not meant to imply

that the risk is non-existent, but the varied common factors between both community

areas served to at least minimize the risk. However, there was a risk that the conclusions

may yield results that would be applicable to other low and moderate income

neighborhoods within the city of Chicago but may not be applicable to other low and

moderate income communities within the U.S.

72

There was a risk to the validity of the data and findings where it may not be

possible to generalize the results to a significantly different population. There might be

other factors within a larger city such as Chicago that might not be applicable to smaller

cities with low and moderate income communities. There was a potential risk that the

reliability will be limited to low and moderate income communities in larger cities within

the U.S. rather than attempting to generalize the results of this study to any low and

moderate income population existing in any city or country.

There may be a need in future studies that would take factors such as population

size and density into context so as to reduce the risks associated with validity. For

example, while the research in this study related to the Chicago Metropolitan Area, that

data may only apply to other areas of similar size to Chicago. There may also be a need

to apply the same methods utilized in this study to a smaller population. This secondary

research could either serve to provide a greater measurement of validity to the initial

study here or could perhaps identify additional factors that pertain to the smaller

population that may not apply to this study.

When considering both reliability and validity, my use of both regression analysis

and the one-tailed t test for comparing population means was the most appropriate data

analysis methods based on the goals of this study. With these analysis methods, I was

able to compare the measured statistics between the two community areas and counties

under review. Considering that the goal of the study was to compare the changes in the

dependent variables, my use of these analysis methods provided an ability to determine

73

whether the independent variable had any relationship with any of the dependent

variables.

The data gathered in this study was objective information available in the public

domain from generally acceptable sources such as the U.S. government, local

government authorities and local organizations responsible for assessing the economics

of low and moderate income communities. By using these acceptable sources, I used

data that was considered reliable by previous researchers. The definitions of particular

variables will follow the generally accepted definitions used in similar research.

Measures for Participant Protection

In this study, the research focused on data that was available in the public domain.

No information was sourced or reported that would indicate data related to an individual.

The participation of individual members of the community was not solicited, which

resulted in no need to create or implement any measures to protect the population.

Conclusion

As noted in Chapters 1 and 3, there was a need to further explore the relationship

between home ownership and socio-economic community benefit within low and

moderate income communities. In this study, my intention was to examine a community

with growth in home ownership against a community with no growth over the same

period. With that information, I then analyzed the data related to each dependent variable

in order to assess whether any relationship between housing and any of the dependent

variables exist.

74

As discussed above, the most appropriate statistical analysis options were both the

one-tail t test for population means and regression analysis. In these forms of analysis, I

tracked the performance of each dependent variable for both community areas and

counties in the sample. This allowed me to formulate an effective assessment as to

whether the dependent variables in the community area or county with housing growth

performed any differently when compared to the performance of the dependent variables

in the areas with no growth.

I considered that a simple improvement in any dependent variable did not

necessarily imply that a change in housing was the cause of the change. This is where

both the one-tail t test of population means and regression analysis were effective. In

these tests, I considered the impact of other outside variables that could have caused the

change. This was effective when considering that both the two community areas and the

two counties are within the same general area. With this test, I was simply comparing the

annually recorded information of each dependent variable between each community area

or county.

The one-tailed t test for comparing population means was also effective when

considering that the data related to each dependent variable for the areas under study was

not equal. When discussing variables such as high school graduation rates, the two

community areas did not have the exact same results. Specifically, one area may have a

better graduation rate while another did not. This was another example of where the one-

tail t test for comparing population means is an effective analytical test. With this test,

my intent was not to simply compare the graduation rate in one community area against

75

another community area. My intent was to assess whether a dependent variable

performed any differently as a result of the independent variable. What I was attempting

to gauge was whether the presence of housing growth within a specific community area

allowed any of the dependent variables to change in a manner that was different than the

area without growth.

The same goals applied to the use of regression analysis within the goals of this

study. As noted previously, regression analysis allowed me to study how a single

dependent variable can change as the result of changes in several independent variables

within the community area. This served to build upon the results of the one-tail t test by

considering the interaction of multiple independent variables on a single dependent

variable.

As noted in chapter 2, there was significant research related to the socio-economic

benefits of home ownership for a general population, but there was limited research

related specifically to the benefits in a low and moderate income community. Within the

context of socio-economic benefit, there was limited scholarly research related

specifically to the impact on crime within low and moderate income communities as well.

There was research that would infer that areas with low crime rates tend to have higher

home ownership, but there was limited research that would infer that higher home

ownership led to lower crime rates.

This gap also applied to educational performance. Again, as noted in chapter 2,

there was significant research related to the findings that those with higher educational

attainment tended to own homes rather than rent, but there was very limited research

76

related to whether home ownership led to improvements in educational attainment. The

gap lies in the fact that while there was research that would infer that those with higher

education own homes, but there was limited research that home ownership improvements

led to improved educational attainment.

Finally, one also needs to consider the gap in the existing research related to

unemployment. Again, as noted in chapter 2 of this study, there was significant research

that would infer that in areas with low unemployment there also tended to be higher

home ownership. There was limited research that would infer that higher home

ownership would lead to improvements in unemployment within the area.

I considered that there were two separate but relevant gaps in the literature.

Firstly, while there was evidence that home ownership and improved performance of the

dependent variables may co-exist, that did not necessarily imply that there was a

relationship between the variables. Secondly, the existing research was limited when

discussing the performance of low and moderate income communities. This research was

both necessary and relevant as there was a need to address both of these research gaps.

Finally, I also considered the overall social change impact of this information.

For over 30 years, the U.S. government has created both enhanced regulations and direct

public funding of programs to spur growth in low and moderate income home ownership.

The underlying goal of the programs and regulations was to provide more options to

potential home owners with a secondary goal of community benefit of this effort. The

hypothesis was that if more people own homes, the entire community will improve. The

basis for that hypothesis was the contemporary research discussed in chapter 2. As there

77

was limited current research that would support the hypothesis in the context of low and

moderate income communities, there was a risk that the funds and the regulations have

had no measureable benefit to the community.

The purpose of this study was to determine whether home ownership has a

relationship with the dependent variables. In the event that a relationship existed, there

would be a foundation to not only continue the funding and regulations, but to perhaps

enhance them further so as to have a greater long term benefit. If the relationship

between home ownership and the dependent variables cannot be supported, it will be

necessary to stop spending the public funds on this effort. As this was the conclusion of

this research, future research should focus on identifying other factors that may have a

greater impact on the dependent variables outside of housing.

78

CHAPTER 4:

RESULTS

Introduction

The purpose of this study was to examine the relationship between home

ownership in low and moderate income communities and community health, as measured

by crime rate, high school test score performance, high school graduation, and

unemployment. Guided by the theories of welfare economics, I sought to determine

whether public investment in improvements in home ownership results in improved

community health.

In this chapter, I review the demographics of the sample population and the

historical performance of each of the socio-economic measurements under consideration.

I also present the results of both the one-tail t test and regression analyses.

Demographics of the Community Areas and Counties

Crime rates, high school test performance, and high school graduation were

studied in two of 75 community areas within Chicago, Austin, and New City. Because

the focus of this study involved low and moderate income communities, communities

meeting the U.S. Census definition of low and moderate income were considered.

According to the 2000 United States Census, the Austin community area population was

117,527 and the New City community area population was 51,721. As I note in Table 1,

there were differences in ethnic composition for each community area, but both

community areas maintain a larger proportion of non-white ethnic groups when

compared to the city as a whole.

79

Table 1 Racial Composition for the Community Areas Under Review Community Area

Ethnic Group Austin New City

White 6.0% 24.0%

Black or African American 87.5% 23.7%

American Indian or Native Alaskan 0.1% 0.5%

Asian 0.5% 0.3%

Native Hawaiian or Pacific Islander 0.0% 0.1%

Hispanic 4.0% 32.8%

Other Races 1.8% 18.6% Source: 2000 United States Census

Beyond identifying two low and moderate income community areas, it was also

necessary to select the community areas based on any changes in owners-occupied home

ownership within the community area. It was necessary to identify one community area

that witnessed a growth in home ownership and another area that had either a reduction

or no growth in owners-occupied housing stock within the community area. The Federal

Financial Institutions Examination Council provides housing information at the census

tract level for the United States housing market. Table 2 notes the information related to

owners-occupied housing changes from 1998 to 2007 for both community areas.

80

Table 2 Changes in Owners-Occupied Housing in the Community Area

Community Area

Year Ending Austin

New City

1997 38.2% 32.5%

1998 38.2% 32.5%

1999 38.2% 32.5%

2000 38.2% 32.5%

2001 38.2% 32.5%

2002 38.2% 32.5%

2003 39.6% 31.2%

2004 39.6% 31.2%

2005 39.6% 31.2%

2006 39.6% 31.2%

2007 39.6% 31.2% Source: Federal Financial Institutions Examinations Council Census Reports (1997-2007)

Housing and unemployment performance data were examined at the county level.

Following similar identification processes noted above, I identified two low and moderate

income counties in the State of Indiana—Marion and Lake County—that met the

requirements of the study. According to the 2000 U.S. Census, both counties had a

higher proportion of residents at or below 80% of average median income level meeting

the definition of low and moderate income when compared to the average median income

for the State of Indiana: 11% of Marion County’s residents and 12.2% of Lake County’s

residents were at or below the poverty level; these percentages were higher than the state

average of statewide poverty level of 9.5%.

81

The ethnic distribution of Marion and Lake Counties differed in comparison to the

ethnic distribution in Austin and New City community areas. Marion and Lake Counties

are predominantly White, whereas the Chicago community areas were predominantly

non-white.

Table 3 Racial Composition for the Counties Under Review County

Ethnic Group Marion Lake

White 69.0% 60.5%

Black or African American 23.6% 23.0%

American Indian or Native Alaskan 0.2% 0.3%

Asian 1.4% 0.7%

Native Hawaiian or Pacific Islander 0.0% 0.0%

Hispanic 3.8% 11.1%

Other Race 1.9% 4.5% Source: 2000 United States Census

While there was a difference in the overall demographics between the community areas

and counties, the income demographics provided a stronger comparison as both the

community areas and counties had a higher than average proportion of low and moderate

income families. Continuing with Table 4 below, I have detailed the percentage of

owners-occupied housing in the counties.

82

Table 4 Changes in Owner-Occupied Housing in the Counties

County

Year Ending Marion Lake

1997 52.1% 65.0%

1998 52.1% 63.2%

1999 52.1% 63.2%

2000 52.1% 63.2%

2001 52.1% 63.2%

2002 52.1% 63.2%

2003 53.7% 64.3%

2004 53.8% 64.3%

2005 53.8% 64.3%

2006 53.7% 64.3%

2007 53.7% 64.3% Source: Federal Financial Institutions Examinations Council Census Reports (1997-2007)

With this information serving as a basis for the research, the next section of this chapter

reviews the detailed socio-economic data related to both the Chicago community areas as

well as the two counties in Indiana.

Socio-Economic Performance Indicators

Four socio-economic factors—crime rates, high school test performances, high

school graduation rates, and median income—were examined as indicators of community

health.

83

Crime Rates

The Chicago Police Department reported crime rates by eight major categories,

ranging from murder to arson. Crime rates are reported as the total number of incidents

within the community area as a percentage of the population of that community area.

Starting with Table 5 below, I have noted the data related to murder rates since 1997 for

both community areas.

Table 5 Murder Rates by Year in the Community Area Crime Rate

Year Ending Austin New City

1998 0.05% 0.06%

1999 0.04% 0.03%

2000 0.03% 0.05%

2001 0.04% 0.05%

2002 0.05% 0.04%

2003 0.04% 0.06%

2004 0.03% 0.02%

2005 0.03% 0.03%

2006 0.03% 0.03%

2007 0.03% 0.03% Source: Chicago Police Department Annual Reports (1998-2007)

At the start of the survey period, the murder rate in New City was higher when compared

to Austin. Over the duration of the survey period, both community areas witnessed a

reduction in the murder rate concluding with 1997 when the murder rate in both

community areas down to an equal level of 0.03%.

84

Continuing with Table 6 below, I have noted the criminal sexual assault rates

since 1997 for each community area.

Table 6 Sexual Assault Rates by Year in the Community Area Crime Rate

Year Ending Austin New City

1998 0.16% 0.12%

1999 0.12% 0.12%

2000 0.11% 0.07%

2001 0.13% 0.09%

2002 0.11% 0.08%

2003 0.11% 0.10%

2004 0.11% 0.07%

2005 0.10% 0.09%

2006 0.08% 0.08%

2007 0.09% 0.09% Source: Chicago Police Department Annual Reports (1998-2007)

The Austin Community Area did have a higher proportion of these crimes in 1998, both

communities were able to reduce sexual assault crimes to an equal level over the ten year

period.

In Table 7, I have detailed the historical data related to robbery rates for both

community areas. In 1998, the robbery rate in Austin was nearly twice as much as New

City. Since that time, the robbery rates for both community areas have decreased.

85

Table 7 Robbery Rates by Year in the Community Area Crime Rate

Year Ending Austin New City

1998 1.50% 0.83%

1999 1.16% 0.79%

2000 1.07% 0.91%

2001 0.99% 0.73%

2002 0.94% 0.88%

2003 1.04% 0.73%

2004 0.91% 0.65%

2005 0.93% 0.61%

2006 0.91% 0.70%

2007 0.86% 0.53% Source: Chicago Police Department Annual Reports (1998-2007)

Cases of aggravated assault have also decreased in both community areas. Austin

had a higher proportion of this crime in comparison to New City in 1997, but it is evident

that aggravated assault crimes in Austin not only decreased, but decreased at a faster pace

when compared to New City. When reviewing the historical data in Table 8 below,

Austin has consistently witnessed a reduction year-to-year, the annual trends in the New

City Community Area have not managed the same trend of reductions. Overall, there has

been a reduction, there have also been periods with short term increases in reported

incidents.

86

Table 8 Aggravated Assault Rates by Year in the Community Area Crime Rate

Year Ending Austin New City

1998 1.67% 1.56%

1999 1.37% 1.52%

2000 1.36% 1.44%

2001 1.27% 1.47%

2002 1.24% 1.60%

2003 1.03% 1.20%

2004 1.01% 1.07%

2005 0.92% 0.99%

2006 0.92% 1.20%

2007 0.92% 1.10% Source: Chicago Police Department Annual Reports (1998-2007)

When considering burglary rates, the trends are quite similar to changes in

aggravated assault. Both communities have made significant reductions since 1998, but

the reductions have been inconsistent. In 1998, Austin started with a much higher

proportion of reported burglaries when compared to New City. Concluding in 2007,

Austin had a much higher reduction in reported burglaries resulting in a reduction of

nearly 50% over the period. New City also had a significant reduction in reported

burglaries, but that reduction was much less than Austin over the survey period. As with

other measurements summarized above, the reduction in Austin was consistent over the

survey period while in New City, there was a significant increase in 2000 and then a

87

reduction going forward. In Table 9, I have detailed the historical trends in burglary rates

for both community areas.

Table 9 Burglary Rates by Year in the Community Area Crime Rate

Year Ending Austin New City

1998 1.59% 1.42%

1999 1.09% 1.36%

2000 1.02% 1.70%

2001 0.89% 1.19%

2002 0.85% 1.08%

2003 0.85% 1.11%

2004 0.93% 1.01%

2005 0.87% 1.01%

2006 0.72% 0.97%

2007 0.79% 0.95% Source: Chicago Police Department Annual Reports (1998-2007)

Theft related crimes were reduced in both New City and Austin over the survey

period. Contrary to other factors discussed above, New City started with higher

proportion of reported crimes in comparison to Austin. Neither community area

witnessed a consistent reduction in reported crimes since 1998. There was a reduction

from 1998, but there were also incidents where there were short term increases as well.

When examining the data since 2003 when both communities had nearly the same

proportion of reported incidents, Austin had a higher reduction in reported incidents in

comparison to the New City Community area. I have detailed the historical trends in

theft rates for each community area in Table 10 below.

88

Table 10 Theft Rates by Year in the CommunityArea Crime Rate

Year Ending Austin New City

1998 3.38% 3.73%

1999 3.06% 3.51%

2000 2.92% 3.59%

2001 2.79% 3.21%

2002 2.76% 2.92%

2003 3.14% 3.16%

2004 2.98% 3.42%

2005 2.77% 2.78%

2006 2.68% 2.81%

2007 2.66% 2.93%

Source: Chicago Police Department Annual Reports (1998-2007)

In reference to motor vehicle theft rates, again Austin started at a much higher

level compared to New City. As early as 2000, the proportion of reported incidents was

nearly equal between the two community areas. The improvement in Austin was short

lived as there was an increase in 2001 where the variance between communities was

high. Austin quickly recovered from the increase in 2001, but there is still a trend of

higher rates in reported motor vehicle theft. As was the case with other crimes discussed

above, the trend of improvement over the period was much higher in Austin. I have

detailed the historical trends in motor vehicle theft rates for both community areas in

Table 11 below.

89

Table 11 Motor Vehicle Theft Rates by Year in the Community Area Crime Rate

Year Ending Austin New City

1998 1.73% 1.15%

1999 1.44% 1.12%

2000 1.11% 1.15%

2001 1.30% 0.97%

2002 1.07% 0.91%

2003 0.99% 0.85%

2004 0.91% 0.86%

2005 0.97% 0.94%

2006 1.01% 0.87%

2007 0.85% 0.62% Source: Chicago Police Department Annual Reports (1998-2007)

In reference to arson rates, both Austin and New City started and ended the

survey period at nearly the same level. Arson related crimes have also lowered in both

community areas as well with incidents reduced by 50% or more over the survey period.

I have detailed the historical trends in motor vehicle theft rates for both community areas

in Table 12 below.

90

Table 12 Arson Rates by Year in the Community Area Crime Rate

Year Ending Austin New City

1998 0.09% 0.08%

1999 0.07% 0.07%

2000 0.06% 0.08%

2001 0.06% 0.07%

2002 0.05% 0.07%

2003 0.05% 0.05%

2004 0.05% 0.04%

2005 0.05% 0.03%

2006 0.03% 0.03%

2007 0.04% 0.04%

Source: Chicago Police Department Annual Reports (1998-2007)

Concluding with the overall crime rate for both community areas, Austin began

with a much higher crime rate compared to New City. Since that time, the crime rate for

both community areas has decreased to nearly the same level in 2007. I have detailed the

overall crime rate information for both community areas in Table 13 below.

91

Table 13 Total Crime Rates by Year in the Community Area Crime Rate

Year Ending Austin New City

1998 10.17% 8.95%

1999 8.35% 8.52%

2000 7.68% 8.99%

2001 7.48% 7.79%

2002 7.09% 7.58%

2003 7.24% 7.26%

2004 6.93% 7.16%

2005 6.63% 6.48%

2006 6.39% 6.69%

2007 6.25% 6.31%

Source: Chicago Police Department Annual Reports (1998-2007)

High School Selection

Chicago has third largest public school system in the U.S. At the high school

level, there were three types of high schools available for students, magnet schools,

academy schools and neighborhood schools. For the purposes of this study, the focus

was on the neighborhood schools that are zoned to specific community areas in the city.

Both the New City and Austin community areas had two neighborhood schools

assigned to the areas for prospective high school students who are not considering a

magnet or academy high school program. The schools assigned to the New City

community area are Manley and Marshall High Schools. The schools assigned to the

Austin community area are Tilden and Gage Park High Schools.

High School Test Score Performance

92

The State of Illinois Department of Education requires that each public high

school publicly report the academic performance of the enrolled students. Among the

standard measurements was the ACT test that was a required portion of the Prairie State

Examinations administered to high school students near the end of the 11th grade. The

ACT examination was required testing step starting in 2001. Along with the requirement

of the Chicago Public Schools, the ACT test was offered to students throughout the

United States and was often used by colleges and universities as an admissions criterion.

I reviewed the average ACT test score performance for each individual high

school in both community areas. The test score information below was reported by the

year the examination was taken and was an average test score for all test takers in each

school. Table 14 notes the average ACT information for New City schools and Table 15

noted the information for each of the Austin community area schools.

Table 14 ACT Scores for New City High Schools Test Year High School

2001 2002 2003 2004 2005 2006 2007 2008

Marshall 13.3 13.3 13.4 13.5 13.9 13.8 13.9 14.2

Manley 13.0 13.6 14.2 14.7 13.9 13.9 14.3 14.5 Source: Chicago Public Schools Scorecard Reports (2001-2008)

93

Table 15 ACT Scores for Austin High Schools Test Year High School

2001 2002 2003 2004 2005 2006 2007 2008

Tilden 13.7 13.7 14.6 14.5 14.4 14.1 14.2 14.0

Gage Park 14.1 14.1 14.2 14.2 14.5 14.2 15.1 14.8 Source: Chicago Public Schools Scorecard Reports (2001-2008)

It was also necessary to consider the total amount of test takers at each school for

each year of the examination. With this information, I could apply a weighted average

test score for each community area. The process of weighting accounted for variations in

school and community area size and provided an appropriate average measurement for

the combined schools within the community area. In Table 16, the number of test takers

for each year was noted.

Table 16 High School Population Taking the ACT Examination Test Year High School

2001 2002 2003 2004 2005 2006 2007 2008

Marshall 222 195 221 223 206 225 254 175

Manley 97 71 96 114 140 48 192 180

Tilden 123 131 92 125 137 161 191 203

Gage Park 255 246 234 240 316 309 210 205

Source: Chicago Public Schools Scorecard Reports (2001-2008)

With the population of test takers known, I then calculated the weighted average ACT

composite score for each community area. As noted in Table 17 below the Austin

schools started with a lower average ACT score when compared to New City. The 2008

94

ACT test results indicated that both Austin and New City were at the same level for the

overall average ACT score.

Table 17 Weighted Average ACT Score Performance by Community Area Test Year Community Area

2001 2002 2003 2004 2005 2006 2007 2008

Austin 13.2 13.4 13.6 13.9 13.9 13.8 14.1 14.4

New City 14.0 14.0 14.3 14.3 14.5 14.2 14.7 14.4

High School Graduation Performance

The next variable within the study was the graduation rate for the neighborhood

schools within the Austin and New City community areas. As noted in Chapter 3, the

graduation rate was calculated by the percentage of students who entered the ninth grade

and completed the requirements necessary to graduate within five years of starting

school. In order to begin the analysis, it was necessary to review the historical graduation

rate performance at each of the four high schools since 1999. I have detailed the

historical graduation rates for each high school in Table 18 below.

Table 18 Graduation Rates by High School and Graduation Year Percentage of Graduating Students High School

1999 2000 2001 2002 2003 2004 2005 2006 2007

Marshall 31.6% 38.5% 40.3% 36.0% 39.5% 47.1% 40.0% 48.1% 53.8%

Manley 31.8% 35.0% 36.4% 36.8% 42.8% 46.8% 43.2% 41.0% 39.5%

Tilden 28.2% 31.1% 29.5% 23.3% 31.7% 34.1% 27.4% 35.6% 38.1%

Gage Park 49.7% 51.1% 45.4% 46.0% 48.6% 50.7% 50.9% 47.3% 48.3%Source: Chicago Public Schools Scorecard Reports (2001-2008)

95

When comparing performance on the ACT Composite Score, it was necessary to

consider the size of each school to determine the weighted average graduation rate for the

community area. As such, in Table 19 below I have noted the cohort groups who were

eligible to graduate in each year in the event that the graduation requirements were met.

Table 19 Potential Graduating Population Number of Potential Graduating Students High School

1999 2000 2001 2002 2003 2004 2005 2006 2007

Marshall 248 272 235 238 138 226 183 216 242

Manley 665 453 465 386 310 408 365 389 471

Tilden 482 461 326 384 325 359 393 434 408

Gage Park 495 478 396 330 463 528 466 526 513 Source: Chicago Public Schools Scorecard Reports (2001-2008)

When calculating of the average ACT score for each community area, it was also

necessary to determine the weighted average graduation rate for the neighborhood high

schools within the each community area. As such, in Table 20 below I noted the high

school graduation rate performance for each community area.

Table 20 Weighted Average Graduation Rate by Community Area Graduation Year Community Area

1999 2000 2001 2002 2003 2004 2005 2006 2007

Austin 31.7% 36.3% 37.7% 36.5% 41.8% 46.9% 42.1% 43.5% 44.4%

New City 39.1% 41.3% 38.2% 33.8% 41.6% 44.0% 40.1% 42.0% 43.8%

96

Median Income Data for the Community Areas

In addition to the other socio-economic factors discussed above, I also found it

necessary to consider average median income information for both community areas. As

noted in Chapter 3, the median income information can shed light into the general

economic situation of a community area. In Table 21 below, I have noted the historical

median income information for both community areas.

Table 21 Average Median Family Income for the Community Areas Median Income Year Ending Austin New City

1997 $ 35,730 $ 29,657

1998 $ 38,099 $ 31,623

1999 $ 40,853 $ 33,909

2000 $ 43,478 $ 36,088

2001 $ 45,143 $ 37,469

2002 $ 48,281 $ 40,074

2003 $ 42,041 $ 33,288

2004 $ 42,191 $ 33,406

2005 $ 42,657 $ 33,775

2006 $ 44,867 $ 35,525

2007 $ 43,373 $ 34,342

Source: Federal Financial Institutions Examinations Council Census Data

County Based Unemployment Data

As noted above, unemployment data was considered with a different population

by using the historical data related to changes in unemployment in Marion and Lake

97

County, Indiana. In Table 22 below, I have noted the information related to historical

unemployment rates in each county.

Table 22 Unemployment Rate by County County Year Ending Lake Marion

1997 4.3% 3.0%

1998 3.9% 2.7%

1999 4.0% 2.6%

2000 3.6% 2.7%

2001 4.8% 3.7%

2002 6.4% 5.2%

2003 6.1% 5.4%

2004 6.3% 5.4%

2005 6.1% 5.5%

2006 5.7% 4.9%

2007 5.2% 4.5%

2008 6.2% 5.6% Source: United States Department of Labor, Bureau of Labor Statistics

Summary

Given the detailed demographic data I have provided above, I will continue with

an analysis of the data as applied to the hypotheses discussed in chapters 1-3. In the next

section, I reviewed each of the stated hypotheses and applied the statistical tests to each

in order to determine the relationship between owners-occupied housing and the major

socio-economic variables in the study.

98

Results of the Data Analysis

Introduction

As noted above as well as in chapter 1 of this study, the research questions

revolved around developing an understanding of the relationship, if any, in improvements

in home ownership in low and moderate income communities and the varied socio-

economic measurements discussed above. I used a one-tailed t test for two population

means for each variable in order to determine whether a relationship existed and the

extent of the relationship between the independent and dependent variables.

Furthermore, each t test will have a significance level α of 5%.

Test of Hypotheses

Hypothesis One

Hypothesis one involved the relationship between home ownership and crime rate in

the community areas.

CRLCRH

CRLCRH

H

H

:

:

1

0

Where CRH was the average annual crime rate in the high homeownership community area

over the study period and CRL was the average annual crime rate in the low homeownership

community area over the study period. The crime rate was defined as a percentage of the

reported crimes, by category within the total population of the community area. The crimes

were limited to those reported to a government authority that could result in criminal

punishment in the event of prosecution of the offender.

99

Test of Hypothesis One: Murder Rates

As discussed above, I considered the performance of each category of crime as

well as the aggregate crime rate for the community areas with growth and no growth in

owners-occupied housing. To conduct the t test, I tabulated crime rates for each

community area from 1998 through 2007 as noted in Table 5 of this study. Using

Microsoft Excel, I completed a one-tail t test of population means in order to assess the

relationship between home ownership and murder rates.

When comparing the historical information for both community areas since 1998,

the resulting analysis indicated a p-value = 0.5599 which one would accept 0H . The

conclusion was that no relationship existed between the murder rate and increases in

owners-occupied housing within a low and moderate income community.

Test of Hypothesis One: Sexual Assault

To conduct the t test, I tabulated crime rates for each community area from 1998

through 2007 as noted in Table 6 of this study. Using Microsoft Excel, I completed a

one-tail t test of population means in order to assess the relationship between home

ownership and sexual assault rates.

When comparing the historical information for both community areas since 1998,

the resulting analysis indicated a p-value = 0.0155 which one would reject 0H and accept

1H . The conclusion was that a relationship existed between the criminal sexual assault

rate and increases in owners-occupied housing within a low and moderate income

community.

100

Test of Hypothesis One: Robbery

To conduct the t test, I tabulated crime rates for each community area from 1998

through 2007 as noted in Table 7 of this study. Using Microsoft Excel, I completed a

one-tail t test of population means in order to assess the relationship between home

ownership and robbery rates.

When comparing the historical information for both community areas since 1998,

the resulting analysis indicated a p-value = 0.0006 which one would reject 0H and accept

1H . The conclusion was that a relationship existed between the robbery rate and

increases in owners-occupied housing within a low and moderate income community.

Test of Hypothesis One: Aggravated Assault and Battery

To conduct the t test, I tabulated crime rates for each community area from 1998

through 2007 as noted in Table 8 of this study. Using Microsoft Excel, I completed a

one-tail t test of population means in order to assess the relationship between home

ownership and aggravated assault rates.

When comparing the historical information for both community areas since 1998,

the resulting analysis indicated a p-value = 0.1914 which one would accept 0H . The

conclusion was that no relationship existed between the aggravated assault and battery

rate and increases in owners-occupied housing within a low and moderate income

community.

Test of Hypothesis One: Burglary

To conduct the t test, I tabulated crime rates for each community area from 1998

through 2007 as noted in Table 9 of this study. Using Microsoft Excel, I completed a

101

one-tail t test of population means in order to assess the relationship between home

ownership and burglary rates.

When comparing the historical information for both community areas since 1998,

the resulting analysis indicated a p-value = 0.0602 which one would not reject 0H . Since

the p-value is quite close to α, it would not be appropriate to assume that 0H was true.

Test of Hypothesis One: Theft

To conduct the t test, I tabulated crime rates for each community area from 1998

through 2007 as noted in Table 10 of this study. Using Microsoft Excel, I completed a

one-tail t test of population means in order to assess the relationship between home

ownership and theft rates.

When comparing the historical information for both community areas since 1998,

the resulting analysis indicated a p-value = 0.0375 which one would reject 0H and accept

1H . The conclusion was that a relationship existed between the theft rate and increases in

owners-occupied housing within a low and moderate income community.

Test of Hypothesis One: Motor Vehicle Theft

To conduct the t test, I tabulated crime rates from 1998 through 2007 as noted in Table 11

of this study. Using Microsoft Excel, this I a one-tail t test of population means in order

to assess the relationship between home ownership and motor vehicle theft rates.

When comparing the historical information for both community areas since 1998,

the resulting analysis indicated a p-value = 0.0720 which one would not reject 0H . Since

the p-value is quite close to α, it would not be appropriate to assume that 0H was true.

102

Test of Hypothesis One: Arson

To conduct the t test, I tabulated crime rates from 1998 through 2007 as noted in

Table 12 of this study. Using Microsoft Excel, I completed a one-tail t test of population

means in order to assess the relationship between home ownership and arson rates.

When comparing the historical information for both community areas since 1998,

the resulting analysis indicated a p-value = 0.7150 which one would accept 0H . The

conclusion was that no relationship existed between the arson rate and increases in

owners-occupied housing within a low and moderate income community.

Test of Hypothesis One: Aggregate Crime

To conduct the t test, I tabulated crime rates from 1998 through 2007 as noted in

Table 13 of this study. The summary data for aggregate crime rates is noted below in

Table 31. Using Microsoft Excel, I completed a one-tail t test of population means in

order to assess the relationship between home ownership and the aggregate crime rate.

When comparing the historical information for both community areas since 1998,

the resulting analysis indicated a p-value = 0.7551 which one would accept 0H . The

conclusion was that no relationship existed between the aggregate crime rate and

increases in owners-occupied housing within a low and moderate income community.

In addition to the one-tail t test of population means, I also conducted a regression

analysis to assess the relationship between aggregate crime rates and home ownership,

ACT scores, median income and high school graduate rates. I have detailed the results of

the regression analysis to predict aggregate crime in Table 23 below.

103

Table 23a Summary of Regression Analysis Data for Variables Predicting Aggregate Crime (N = 14)

Variable Β SE Β t p

High School Graduation -0.469 0.053 -0.880 0.401

ACT Scores -0.013 0.004 -2.794 0.020

Homes -0.112 -0.047 -2.391 0.040

Median Income -0.001 0.000 -1.210 0.257

However, when considering the four independent variables, the p-values for both

high school graduation (0.4014) and median income (0.2570) were above the 5%

significance level. It was necessary to remove those variables from the regression

analysis. I conducted a regression analysis to assess the relationship between aggregate

crime rates, home ownership and the ACT test score performance. I have detailed the

results of the regression analysis to predict aggregate crime in Table 24 below.

Table 23b Summary of Regression Analysis Data for Variables Predicting Aggregate Crime (N = 14)

Variable Β SE Β t p

ACT Scores -0.013 0.052 -4.444 0.001

Homes -0.124 0.031 -3.889 0.002

As I have noted in the table above, the p-values for all regression coefficients are

very small, implying that regression coefficients are significantly different from zero.

Given these results, the regression equation for aggregate crime rate would be the

following:

104

Aggregate Crime = 0.307 - (0.0139 x ACT Composite Score) - (0.124 x Home Ownership Rate)

Additionally, with a Variable Inflation Factor (VIF) that is less than 5.00, this indicated

that the risk of variable collinearity is low. Given the low VIF and p-values, it would be

appropriate to conclude that both ACT score performance and home ownership have an

impact on crime rates.

Hypothesis Two

Hypothesis two involved the relationship between homeownership and unemployment

rate.

UELUEH

UELUEH

H

H

:

:

1

0

Where UEH was the average unemployment rate in the high homeownership county over the

study period and UEL was the average annual unemployment rate in the low homeownership

county over the study period. Unemployment rate was defined as the percentage of the

neighborhood population that is not working but is actively seeking employment.

Test of Hypothesis Two

To conduct the t test, I tabulated unemployment rates from 1997 through 2008 as

noted in Table 22 above. Using Microsoft Excel, I completed a one-tail t test of

population means in order to assess the relationship between home ownership and

unemployment rates.

When comparing the historical unemployment information for Lake and Marion

Counties since 1997, the resulting analysis indicated a p-value = 0.0752 which one would

105

not reject 0H . Since the p-value is quite close to α, it would not be appropriate to assume

that 0H was true.

Hypothesis Three

Hypothesis three involved the relationship between high school standardized test

performance and home ownership.

TSLTSH

TSLTSH

H

H

:

:

1

0

Where TSH was the average ACT test score in the high homeownership community area

over the study period and TSL was the average ACT test score in the low homeownership

community area over the study period. The ACT test score was defined as the reported

ACT composite score resulting from students completing the mandated ACT test at the

end of their junior year of high school.

Test of Hypothesis Three

To conduct the t test, I tabulated the weighted average ACT Composite Scores

from 2001 through 2008 noted in Table 17 above. Using Microsoft Excel, I completed a

one-tail t test of population means in order to assess the relationship between home

ownership and high school standardized test performance.

When comparing the historical information for the Austin and New City

community areas since 1998, the resulting analysis indicated a p-value = 0.0063 which

one would reject 0H and accept 1H . The conclusion was that a relationship existed

between the ACT Composite Score and increases in owners-occupied housing within a

low and moderate income community.

106

In addition to the one-tail t test of population means, I also conducted a regression

analysis to assess the relationship between ACT test score performance, home ownership,

and median income. I have detailed the results of the regression analysis to predict the

ACT test scores in Table 24 below.

Table 24 Summary of Regression Analysis Data for Variables Predicting The ACT Score (N = 20)

Variable Β SE Β t p

Homes -7.325 1.744 -4.200 0.001

Median Income -0.001 0.001 -2.430 0.033

As I have noted in the table above, the p-values for all regression coefficients are

very small, implying that regression coefficients are significantly different from zero.

Given these results, the regression equation for ACT Composite Score would be the

following:

ACT Composite Score = 20.1 - (7.33 x Home Ownership Rate) - (0.000080 x Median Income)

The Variable Inflation Factor (VIF) for this regression is 1.00. This would indicate that

the risk of variable collinearity is low. Given the low VIF and p-values, it would be

appropriate to conclude that both median income and home ownership have an impact on

crime rates.

Hypothesis Four

Hypothesis four involved the relationship between high school graduation rates

and home ownership.

107

GRLGRH

GRLGRH

H

H

:

:

1

0

Where GRH was the average high school graduation rate in the high homeownership

community area over the study period and GRL was the average high school graduation rate in

the low homeownership community area over the study period. The high school graduation

rate was defined by the percentage of students entering public high school that complete the

graduation requirements within five years of entry.

Test of Hypothesis Four

To conduct the t test, I tabulated the weighted high school graduation rates from

1999 through 2007 as noted in Table 20 above. I completed a one-tail t test of population

means in order to assess the relationship between home ownership and high school

graduation rates.

When comparing the historical information for both community areas since 1999,

the resulting analysis indicated a p-value = 0.8646 which one would accept 0H . The

conclusion was that no relationship existed between the high school graduation rate and

increases in owners-occupied housing within a low and moderate income community.

In addition to the one-tail t test of population means, I also conducted a regression

analysis to assess the relationship between high school graduation rate, home ownership

and ACT Composite Scores.

108

Table 25 Summary of Regression Analysis Data for Variables Predicting High School Graduation Rates (N = 14)

Variable Β SE Β t p

Homes 0.838 0.248 3.377 0.006

ACT Scores 0.090 0.024 3.722 0.003

As I have noted in the table above, the p-values for all regression coefficients are

very small, implying that regression coefficients are significantly different from zero.

Given these results, the regression equation for the High School Graduation Rate would

be the following:

High School Graduation Rate = - 1.15 + (0.0903 x ACT Composite Score) + (0.839 Home Ownership Rate)

Additionally, with a Variable Inflation Factor (VIF) that is less than 5.00, the indication

was that the risk of variable collinearity is low.

Correlation of Variables

As discussed in Chapter 3 of this study, it was also necessary to assess whether

the individual dependent variables may have correlation between each other. As there

may have been a possibility that an interrelationship between variables resulted in

changes over the survey period that were not directly related to any changes in the

dependent variable. As such, I reviewed the correlation coefficients for each dependent

variable measured by the separate community areas.

109

Correlation in Crime Rate Measurements in New City

Starting with the murder rate, there were strong relationships with criminal assault

(0.68), burglary (0.70) and arson (0.62). Beyond the high correlation with the murder

rate, criminal assault was not highly correlated with any other dependent variable. In

reference to robbery, there was a high correlation with aggravated assault (0.84), burglary

(0.60), vehicle theft (0.61) and arson (0.66). Considering aggravated assault, along with

the correlations noted above, there was a strong correlation with burglary (0.67) and

arson (0.88). Furthermore, burglary also had a high correlation with vehicle theft (0.60)

and arson (0.85). Theft did not have a high correlation with any of the other dependent

variables.

Correlation in Crime Rate Measurements and Education in New City

Considering the two dependent measurements of educational performance, high

school graduation rate and ACT test score performance, there was a negative correlation

with many of the variables related to crime. This negative correlation existed between

educational performance measurements and robbery (-0.82 and -0.90 respectively),

aggravated assault (-0.84 and -0.82 respectively), burglary (-0.57 and -0.71 respectively),

vehicle theft (-0.57 and -0.74 respectively), and arson (-0.63 and -0.69 respectively).

When there was an increase in either high school graduation or ACT score performance,

there was a relative decrease in the rates of crime for the specific categories noted above.

Given the frequent correlations between the socio-economic variables within New

City, it was also necessary to consider that similar correlations could occur within the

socio-economic variables in Austin.

110

Correlation in Crime Rate Measurements in Austin As was the case with correlations in crime rate in New City, the data indicated

that similar correlations existed between several dependent variables in Austin as well.

The level of correlation in Austin was more extensive in comparison to New City.

Considering the correlation between the murder rate and other dependent variables, with

the exception of burglary (0.26) and theft (0.15), the murder rate was highly correlated

with every measurement. When reviewing the data associated with criminal assault,

there was a high correlation with every measurement with the exception of theft (0.43).

In reference to robbery, along with the other factors discussed above, robbery was also

highly correlated with theft (0.69), arson (0.65) and vehicle theft (0.56). In reference to

aggravated assault, other than the variables discussed above, this crime was highly

correlated with both vehicle theft (0.82) and arson (0.81). When considering the

historical performance of burglary, the data indicated a high correlation with theft (0.51)

and arson (0.76) along with the other measurements discussed above. Other than the high

correlations discussed above, theft was not highly correlated with any of the other

variables. Finally, when considering vehicle theft, arson (0.64) was the only other

variable not previously discussed where a significant correlation exists.

Correlation in Crime Rate Measurements and Education in Austin

Similarly to New City, the data in related to high school graduation and test score

performance was also highly correlated with nearly every dependent variable of crime in

the community area. The correlation between educational performance and crime was

negative. As educational performance improves, nearly every measurement of crime is

111

reduced. The correlation was much more frequent in Austin. Where both community

areas shared a high correlation with educational performance and robbery (-0.49 and -

0.71 respectively), aggravated assault (-0.82 and -0.92 respectively), vehicle theft (-0.77

and -0.92 respectively) and arson (-0.62 and -0.75 respectively), Austin also had a high

correlation between educational performance and murder (-0.88 and -0.83 respectively)

and criminal assault (-0.51 and -0.75 respectively). Both educational performance

measurements had a relatively low correlation between educational performance and theft

(0.15 and -0.17 respectively).

Analysis of Autocorrelation

Given that the regression analyses in tests for Hypotheses 1, 3, and 4 involved

time-series data, it was necessary to check for existence of autocorrelation.

Autocorrelation occurs when the regression error term in one period is linearly related to

the error terms in the previous periods. I applied a Durbin-Watson test as a first-order

review for autocorrelation. When considering Hypothesis 1, the Durbin-Watson result

was 1.10 which would indicate that there was no first order autocorrelation. When

considering Hypothesis 3, the Durbin-Watson result was 1.38 which would indicate that

there was no first order autocorrelation. Finally, when considering Hypothesis 4, the

Durbin-Watson result was 2.53 which would also indicate that there was no first order

autocorrelation.

It was also necessary to check for the existence of a high-order autocorrelation.

Using Minitab, I applied the Ljung-Box Q analysis to determine the existence of any

112

autocorrelation. The results of that analysis, for each variable, are noted below:

4321

1.0

0.8

0.6

0.4

0.2

0.0

-0.2

-0.4

-0.6

-0.8

-1.0

Lag

Aut

ocor

rela

tion

Aggregate Crime Rate

From the results, there was no high-order autocorrelation for the data associated with the

aggregate crime rate.

4321

1.0

0.8

0.6

0.4

0.2

0.0

-0.2

-0.4

-0.6

-0.8

-1.0

Lag

Aut

ocor

rela

tion

High School Graduation Rate

113

From the results, there was no high-order autocorrelation for the data associated with

high school graduation rates.

4321

1.0

0.8

0.6

0.4

0.2

0.0

-0.2

-0.4

-0.6

-0.8

-1.0

Lag

Aut

ocor

rela

tion

ACT Composite Score Performance

From the results, there was no high-order autocorrelation for the data associated with the

ACT composite score performance.

Summary

In summary, there were some inconsistencies when considering the relationship

that home ownership may have to several socio-economic measurements within the

context of this study. When considering incidents of crime, there were some categories

that appeared to have a relationship with changes in owners-occupied housing, while

there were several other factors that did not hold such a relationship. There were also

several incidents where housing alone may not have an impact on socio-economic

performance improvements, but instead the combination of housing with other variables

could support a conclusion of greater community improvement.

114

Thus, continuing with Chapter 5 of this study, I will summarize the key findings

of the study and how those relate to the research questions discussed in previous chapters.

In addition to the interpretation of the findings, the implications of the findings will be

discussed in the broader context of the issues of low and moderate income housing.

Finally, I will also discuss the additional research opportunities within the subject that

could be built upon the findings of this study.

115

CHAPTER 5:

SUMMARY, CONCLUSION, AND RECOMMENDATIONS

Overview

The goal of this study was to assess the relationship between housing and

community health, as measured through changes in crime rate, high school graduation

rates, standardized test score performance, and unemployment. Data were analyzed

through multiple regression and the one-tailed t test of population means. Results were

mixed regarding the effect of home ownership on community health: Some aspects of

home ownership improved, while others had no effect on community health.

Conclusions

In this chapter, I will present the results of the analyses, as well as the

interpretation of the findings. The significance level for results was set at 0.05.

Hypothesis One

The first hypothesis posited that home ownership would decrease community crime

rate. The crime rate was defined as a percentage of the reported crimes, by category, within the

total population of the community area. The crimes were limited to those reported to a

government authority that could result in criminal punishment in the event of prosecution of

the offender.

I applied a one-tail t test of population means to determine whether a relationship

existed between home ownership and each major crime category. No relationship was

found between changes in housing and rates of murder, aggravated assault, burglary,

arson, and aggregate crime. Results did support a relationship between home ownership

116

and reports of sexual assault, robbery, and theft. With these results, the conclusion is that

the data does not yield a consistent relationship between changes in home ownership and

crime rates.

I also applied a regression analysis for the aggregate crime rate. In the regression

analysis, the aggregate crime rate was considered the dependent variable with the ACT

score and home ownership as independent variables. Results indicated a linear

relationship between the variables, suggesting that improvements in both ACT test score

performance and home ownership led to improvements in the aggregate crime rate. This

finding is in contrast to the one-tail t test results, wherein no relationship existed. In the

regression analysis, the inclusion of both independent variables in the model resulted in a

potential improvement in crime rates.

Hypothesis Two

The second hypothesis posited that home ownership would decrease the unemployment

rate. The unemployment rate was defined as the percentage of the county population that is not

working but is actively seeking employment. The unemployment rate is limited to those

individuals who had indicated, by survey, that they were not employed but looking for work.

I applied a one-tail t test of population means to determine whether a relationship

existed between home ownership and unemployment. No relationship was found

between changes in housing and unemployment rate. It should be noted though that as a

result of the t test, the resulting p-value was close to the significance level. As such, the

results were inconclusive indicating that a relationship, if any, is not significant.

117

Hypothesis Three

The third hypothesis posited that home ownership would increase the standardized test

scores. The ACT test score was defined as the reported ACT composite score resulting from

students completing the mandated ACT test at the end of their junior year of high school. As

discussed in chapter 4, my focus for the ACT score measurement was on the ACT composite

score.

I applied a one-tail t test of population means to determine whether a relationship

existed between home ownership and ACT test score performance. A relationship was

found between changes in housing and ACT test score performance. Indicating that as

home ownership improved, the ACT scores also improved. The conclusion being that

improvements in home ownership lead to improvements in ACT test score performance.

I also applied a regression analysis for the aggregate crime rate. In the regression

analysis, the ACT test score was considered the dependent variable with median income

and home ownership as independent variables. Results indicated a linear relationship

between the variables, suggesting that improvements in both median income and home

ownership led to improvements in the ACT test score performance. In the regression

analysis, the inclusion of both independent variables in the model resulted in a potential

improvement in ACT test scores.

Hypothesis Four

The fourth hypothesis posited that home ownership would increase high school

graduation rates. The high school graduation rate was defined by the percentage of students

entering public high school that complete the graduation requirements within five years of

118

entry. High school graduation data was limited to those students attending a neighborhood

high school within the community area.

I applied a one-tail t test of population means to determine whether a relationship

existed between home ownership and high school graduation rates. No relationship was

found between changes in housing and high school graduation rates. This result indicates

that there is no impact that home ownership has in improving high school graduation

rates.

I also applied a regression analysis for the high school graduation rate. In the

regression analysis, the high school graduation rate was considered the dependent

variable with the ACT score and home ownership as independent variables. Results

indicated a linear relationship between the variables, suggesting that improvements in

both ACT test score performance and home ownership led to improvements in the high

school graduation rate. This finding is in contrast to the one-tail t test results, wherein no

relationship existed. In the regression analysis, the inclusion of both independent

variables in the model resulted in a potential improvement in crime rates.

Results suggested a relationship between improvements in home ownership and

three of the eight crime categories: criminal assault, robbery and theft, while there was no

relationship between positive changes in housing and three other crime categories:

murder, aggravated assault and arson. The remaining two categories, burglary and motor

vehicle theft, did not yield sufficient results to indicate whether a relationship with home

ownership growth existed. Results suggested no relationship between improvements in

home ownership and the aggregate crime rate within the community.

119

Results on the relationship between home ownership and high school academic

performance suggested that, while students may test better in areas with housing growth,

there was no evidence that students from high home ownership communities graduate

from high school at any better rate than students from low home ownership communities.

Results on home ownership and unemployment were not significant.

Implications

Considering the contemporary research related to the relationship between home

ownership and community improvement, a case could be made for the significant public

investment as well as the regulatory changes to force banks and other lenders to provide

greater financing opportunities for low and moderate income home buyers. As I

discussed in chapter 2, there were several studies that would indicate that in a general

population communities do improve as a result of increases in owner occupied housing.

As I have indicated in this study, the same cannot be said of the low and moderate

income communities in this study.

Certain measurements of crime did appear to improve as a result of increase in

owner occupied housing, but there were other categories of crime that did not improve as

a result of the increase in home ownership. There was no consideration to give one

particular crime category any subjective weighting based on an interpretation of severity,

but the results were clear that when considering the overall crime rate of the

communities, the theory that a higher ratio of ownership would result in reductions in

crime could not be consistently supported. There may have been improvements in some

areas, but the improvement was not consistent against crime in general.

120

The same inconsistency can be said of high school academic performance. As I

discussed earlier, while test scores did appear to improve when an increase in housing

was considered, the graduation rate did not. As I discussed in chapter 4, improvements in

test results should be applauded as a positive result, but the lack of a relationship to high

school graduation is troubling.

Based on the varied regression analyses I completed in this study, it is apparent

that the ACT test score performance does appear to have an impact on other socio-

economic variables. As I discussed earlier in this chapter, there were several examples

where the ACT test score, when combined with home ownership did provide strong

evidence of a linear relationship to improvements in other socio-economic measurements.

Housing alone would not act as a catalyst for future community improvement. Instead, it

is the combination of academic performance improvements and housing improvements

that would lead to a greater community benefit.

Test score performance alone will not enable individuals to be in a better

economic position to own a home. Without an emphasis on educational improvements,

the government may be creating another generation of individuals who are no more able

to own a home than the present population. This theory was supported by the

contemporary research discussed in chapter 2 in reference to the strong relationship

between level of educational attainment and home ownership. Those with higher levels

of education were much more likely to own their homes without the assistance of outside

entities.

121

When applying the results to context of welfare economics, one can see where the

theory would hold that the redistribution of wealth by means of government expenditures

should be temporary to address a short term need. Instead, we see situations similar to

those identified in this study where home ownership does increase, but the community

improvement is negligible. Certainly, the individual home owner does receive a benefit,

but the tax payers should and do expect a greater societal improvement, which is not the

case here.

In reference to unemployment, the lack of clarity of the results make an effective

conclusion challenging. As I noted in chapter 4, we cannot reject the hypothesis that a

relationship does not exist, but one can also not accept the hypothesis that a relationship

between home ownership changes and unemployment existed either. The lack of a

relationship on its own would suggest that the community benefit, if any, is minimal.

As argument can be made that with the limited objective evidence that would

indicate that a community benefit exists, why does the government continue to provide

the financial support? The idea of increases in home ownership is an admirable task, but

if that increase does not make people within the community better off, there is little

value. There seems to be little justification for the continued spending and regulations.

Given the current financial crisis, one should also question the idea that home

ownership creates wealth for the home owner. This question of wealth benefit is

especially true given some of the more creative loan products that banks offered to low

and moderate income families. With the loosening requirements for down payment as

well as the allowance of financing 100% or more of the purchase price, many of these

122

home owners not only have not had the opportunity to build wealth, but their debt

balance was higher than the current value of their properties. The result was that while

they may own their home now, that ownership may be short lived as the resident may

decide that their money could be better spent elsewhere.

Perhaps there is a need to return to the theories of Milton Friedman in reference to

Market Economics. Simply put, it may have been better to allow the market to determine

whether or not home ownership in these communities was a worthwhile investment.

Clearly, the results of this study indicate that the government investment in public funds

and regulation has not yielded the greater community benefit that was a desired outcome.

The programs and services have certainly benefited individual home owners, but the

community benefit was negligible at best. It may have been more appropriate for the

government to not enter the housing market with these subsidies as there does not appear

to be any tangible return on investment, crime has not been reduced, high school

graduation has not improved, nor has there been any relief on unemployment.

Recommendations for Action

The evidence indicated that continued investment in public funded programs for

home ownership in low and moderate income communities was not consistently yielding

the anticipated socio-economic benefits to the community. It is necessary to re-think the

current processes and programs. As I discussed in chapter 2, part of this issue may be the

result of competitive government programs as well as the overreliance on federally

funded programs to not only spur home ownership, but to also solve all of the community

ills as well. Secondly, given the apparent lack of measurable outcomes from these

123

government programs, one should question whether or not these efforts are an effective

use of public funds.

If the government wishes to continue to support housing programs in low and

moderate income communities, the first step would be to revamp the programs currently

in place. These programs should be streamlined into a single effort that meets the market

need rather than several programs that seem to either be competing against each other or

are redundant. Along with this streamlining process, objective performance

measurements need to be established and consistently measured to determine whether a

program or expenditure is actually adding value to the community. Both of these steps

are absent from the current system.

Secondly, one should also consider whether there is a benefit from this market

intervention from the government at all. As evidenced in this study, the relationship

between housing improvements and general societal benefit was tenable at best. The

desire to create and fund programs used for home ownership should be reconsidered.

Perhaps, as noted above, those funds could be used on other resources such as

improvements to education in the community as this would likely yield a better long term

benefit in comparison to assisting a smaller population in their effort to purchase a home.

Finally, before making any changes to the current system, the government should

further study the particular dynamics of low and moderate income communities. The

rationale for these programs and regulations appear to be driven more by the general

theories of welfare economics or even the market interventionist theories of John

Meynard Keynes rather than developing a thorough understanding of the needs of this

124

particular market. The conclusion being that the idea was to simply spend public funds

where the need was perceived rather than developing an understanding of the specific

dynamics of this community to target those funds to areas with the highest need. The

general research supported the idea that home ownership and greater community benefit

are related considering the overall population. The data in this study implicated that this

relationship does not exist when considering low and moderate income communities.

Recommendations for Further Study

The most evident area for further research would be to expand the study to a

larger area of low and moderate income communities. It would also be beneficial to also

consider a sample of areas that were in several cities in the United States. In either

circumstance, expanding the group may yield additional results that could either support

or contradict the findings in this study.

Secondly, it would also be helpful to conduct further research within these

community areas and counties to determine what programs would result in greater

community benefit. It may be better to solicit feedback from members of the community

rather than rely on the government’s perception that homeownership alone can improve

community health. It is likely that the respondents may have a wide variety of ideas on

how to better address areas of crime, education and employment, being a member of the

community under study may serve to provide a greater understanding of the need.

In any event, the intent should be to develop a greater understanding of what

interventions at the federal or local level could better address the ills of a particular

community. The intention of increasing home ownership was certainly admirable and in

125

many areas successful, but one should attempt to understand what other areas of

community development may yield stronger benefits. With this information, one can

better utilize the limited public funds available to have a wider community benefit.

Implications for Social Change

For the individual, owning a home is certainly beneficial and I would have

expected that if more people owned homes in a community, the community would

naturally improve. In these communities, while improvements may have occurred, that

improvement does not appear to have any relationship with changes in home ownership.

One needs to consider whether there may be alternative options for the budgeted public

funds that may have a relationship with community improvement. One clear example, as

noted in the findings in chapter 4, would be further investment in education.

As discussed I in chapter 2, the local funding model for the public school system

was based on property taxes collected within the community surrounding the school. The

result was that more affluent areas received more funding in comparison to the poorer

areas of the city. Unless there is an alternate source of additional funding, there is little

potential that schools in low and moderate income areas will improve. It may be

appropriate to consider redirecting the funds spent on housing subsidies, funded programs

and the infrastructure to monitor regulatory compliance to be spent on areas that would

improve factors such as high school graduation.

As I discussed above, having a high school degree will give an individual a better

chance to earn more and purchase a home if they desire rather than depending on a

subsidy to assist in the financing and property acquisition. In these community areas, less

126

than 50% of students attending schools in Austin or New City graduated within 5 years of

entry. The communities are not creating an environment that values educational

attainment. That environment needs to change.

It seems that redirecting public funds in areas such as education would be more

appropriate as there is an opportunity to effect a greater population within the

community. The redirection of funding alone is not a short term solution to the ills of a

particular community. Instead, it should be considered to be a long term investment. I

believe that Milton Friedman would consider this as a worthwhile expenditure of public

funds. While it may be expensive, the potential benefits of a higher educated society

would likely result in higher incomes, increased tax revenues and less need for

entitlements in the future.

Summary

There continues to be a need to build a mechanism that will provide a greater

sense of well-being in low and moderate income communities. The current process of

redundant programs and aggressive regulation of lenders does not appear to be the cure.

We should not blame the financial weakness on low and moderate income communities

alone, we need to consider some of the loose credit requirements and financing options

that lenders offered to this population as part of the problem. With the easing of

payment, credit rating and equity financing requirements, individuals took on more

financial responsibility than they should have. As property values dropped, wealth, in

terms of equity, disappeared.

127

As a result of the financial meltdown, the credit markets now provides very

limited opportunities to gain access to credit. This lack of credit access was a penalty for

anyone who owned or was interested in owning a home. The current government

considered this to be an overreaction to the situation, the lenders could no longer accept

the level of risk associated with these more exotic loan products. There are now two

challenges facing these communities. First, as this research indicates, there appears to be

virtually no objective benefit of homeownership in the community, and secondly, the

near collapse of the housing finance market which can be partially attributed to loan

products designed for low and moderate income customers. There was not only no return

on the investment in terms of socio-economic benefit alone, but the public will now have

to pay a significant cost to rescue the banks as a result of these bad loans.

In summary, the government should review its involvement in the housing

business and possibly needs to get out of it. Although government regulations should be

in place to protect the individual, the role of the government as a market participant has

failed. Granted, home ownership has increased, but when one considers the expense

coupled with the lack of community benefits, failure is the only viable conclusion to the

performance results.

128

REFERENCES

Arrow, K. J. (1983). Collected papers of Kenneth J. Arrow, volume 1: Social choice and justice. Boston: Harvard University Press.

Basolo, V. (2007, January). Explaining the Support for Homeownership Policy in US Cities: A Political Economy Perspective. Housing Studies, 22(1), 99-119.

Bayoh, I., Irwin, E., & Haab, T. (2006, February). Determinants of Residential Location Choice: How Important Are Local Public Goods in Attracting Homeowners to Central City Locations?. Journal of Regional Science, 46(1), 97-120.

Béland, D. (2007, June). NEO-LIBERALISM AND SOCIAL POLICY. Policy Studies, 28(2), 91-107.

Béland, D., & Waddan, A. (2007, December 15). Conservative Ideas and Social Policy in the United States. Social Policy & Administration, 41(7), 768-786.

Caner, A., & Wolff, E. (2004, December). ASSET POVERTY IN THE UNITED STATES, 1984–99: EVIDENCE FROM THE PANEL STUDY OF INCOME DYNAMICS. Review of Income & Wealth, 50(4), 493-518.

Cheek, J., Onslow, M., & Cream, A. (2004, September). Beyond the divide: comparing and contrasting aspects of qualitative and quantitative research approaches. Advances in Speech Language Pathology, 6(3), 147-152.

Chicago Police Department. (1998). Annual Report 1998 - Chicago Police Department.

Chicago, IL: Chicago Police Department. Chicago Police Department. (1999). Annual Report 1999 - Chicago Police Department.

Chicago, IL: Chicago Police Department. Chicago Police Department. (2000). Annual Report 2000 - Chicago Police Department.

Chicago, IL: Chicago Police Department. Chicago Police Department. (2001). Annual Report 2001 - Chicago Police Department.

Chicago, IL: Chicago Police Department. Chicago Police Department. (2002). Annual Report 2002 - Chicago Police Department.

Chicago, IL: Chicago Police Department. Chicago Police Department. (2003). Annual Report 2003 - Chicago Police Department.

Chicago, IL: Chicago Police Department.

129

Chicago Police Department. (2004). Annual Report 2004 - Chicago Police Department.

Chicago, IL: Chicago Police Department. Chicago Police Department. (2005). Annual Report 2005 - Chicago Police Department.

Chicago, IL: Chicago Police Department. Chicago Police Department. (2006). Annual Report 2006 - Chicago Police Department.

Chicago, IL: Chicago Police Department. Chicago Police Department. (2007). Annual Report 2007 - Chicago Police Department.

Chicago, IL: Chicago Police Department.

Conley, D., & Gifford, B. (2006, March). Home Ownership, Social Insurance, and the Welfare State. Sociological Forum, 21(1), 55-82.

Coughlan, M., Cronin, P., & Ryan, F. (2007, June 14). Step-by-step guide to critiquing research. Part 1: quantitative research. British Journal of Nursing (BJN), 16(11), 658-663.

Czerwinski, S. (2006, June 27). Community development block grant formula: options for

improving the targeting of funds. GAO Reports.

Di, Z. (2007, July). Do Homeowners Have Higher Future Household Income?. Housing Studies, 22(4), 459-472.

Dreier, P. (2003). The future of community reinvestment. Journal of the American Planning Association, 69(4), 341-354.

Dropout and Graduation Rates, 1999-2007 [Data File]. Chicago, IL: Chicago Public

Schools-Office of Research, Evaluation and Accountability.

Erickson, D. (2006, April). Community Capitalism: How Housing Advocates, the Private Sector, and Government Forged New Low-Income Housing Policy, 1968-1996. Journal of Policy History, 18(2), 167-204.

Federal Financial Institutions Examination Council-1997 FFIEC Census Report-Summary Census Housing Information MSA: 1600-Chicago, IL [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-1998 FFIEC Census Report-

Summary Census Housing Information MSA: 1600-Chicago, IL [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

130

Federal Financial Institutions Examination Council-1999 FFIEC Census Report-

Summary Census Housing Information MSA: 1600-Chicago, IL [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2000 FFIEC Census Report-

Summary Census Housing Information MSA: 1600-Chicago, IL [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2001 FFIEC Census Report-

Summary Census Housing Information MSA: 1600-Chicago, IL [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2002 FFIEC Census Report-

Summary Census Housing Information MSA: 1600-Chicago, IL [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2003 FFIEC Census Report-

Summary Census Housing Information MSA: 1600-Chicago, IL [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2004 FFIEC Census Report-

Summary Census Housing Information MSA: 1600-Chicago, IL [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2005 FFIEC Census Report-

Summary Census Housing Information MSA: 1600-Chicago, IL [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2006 FFIEC Census Report-

Summary Census Housing Information MSA: 1600-Chicago, IL [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2007 FFIEC Census Report-

Summary Census Housing Information MSA: 1600-Chicago, IL [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2008 FFIEC Census Report-

Summary Census Housing Information MSA: 1600-Chicago, IL [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-1999 FFIEC Census Report-

Summary Census Median Income Information MSA: 1600-Chicago, IL [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

131

Federal Financial Institutions Examination Council-2000 FFIEC Census Report-

Summary Census Median Income Information MSA: 1600-Chicago, IL [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2001 FFIEC Census Report-

Summary Census Median Income Information MSA: 1600-Chicago, IL [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2002 FFIEC Census Report-

Summary Census Median Income Information MSA: 1600-Chicago, IL [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2003 FFIEC Census Report-

Summary Census Median Income Information MSA: 1600-Chicago, IL [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2004 FFIEC Census Report-

Summary Census Median Income Information MSA: 1600-Chicago, IL [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2005 FFIEC Census Report-

Summary Census Median Income Information MSA: 1600-Chicago, IL [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2006 FFIEC Census Report-

Summary Census Median Income Information MSA: 1600-Chicago, IL [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2007 FFIEC Census Report-

Summary Census Median Income Information MSA: 1600-Chicago, IL [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2008 FFIEC Census Report-

Summary Census Median Income Information MSA: 1600-Chicago, IL [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-1997 FFIEC Census Report-

Summary Census Housing Information State: 18-Indiana County: 89-Lake [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-1998 FFIEC Census Report-

Summary Census Housing Information State: 18-Indiana County: 89-Lake [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

132

Federal Financial Institutions Examination Council-1999 FFIEC Census Report-

Summary Census Housing Information State: 18-Indiana County: 89-Lake [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2000 FFIEC Census Report-

Summary Census Housing Information State: 18-Indiana County: 89-Lake [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2001 FFIEC Census Report-

Summary Census Housing Information State: 18-Indiana County: 89-Lake [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2002 FFIEC Census Report-

Summary Census Housing Information State: 18-Indiana County: 89-Lake [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2003 FFIEC Census Report-

Summary Census Housing Information State: 18-Indiana County: 89-Lake [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2004 FFIEC Census Report-

Summary Census Housing Information State: 18-Indiana County: 89-Lake [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2005 FFIEC Census Report-

Summary Census Housing Information State: 18-Indiana County: 89-Lake [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2006 FFIEC Census Report-

Summary Census Housing Information State: 18-Indiana County: 89-Lake [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2007 FFIEC Census Report-

Summary Census Housing Information State: 18-Indiana County: 89-Lake [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2008 FFIEC Census Report-

Summary Census Housing Information State: 18-Indiana County: 89-Lake [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-1997 FFIEC Census Report-

Summary Census Housing Information State: 18-Indiana County: 97-Marion [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

133

Federal Financial Institutions Examination Council-1998 FFIEC Census Report-

Summary Census Housing Information State: 18-Indiana County: 97-Marion [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-1999 FFIEC Census Report-

Summary Census Housing Information State: 18-Indiana County: 97-Marion [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2000 FFIEC Census Report-

Summary Census Housing Information State: 18-Indiana County: 97-Marion [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2001 FFIEC Census Report-

Summary Census Housing Information State: 18-Indiana County: 97-Marion [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2002 FFIEC Census Report-

Summary Census Housing Information State: 18-Indiana County: 97-Marion [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2003 FFIEC Census Report-

Summary Census Housing Information State: 18-Indiana County: 97-Marion [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2004 FFIEC Census Report-

Summary Census Housing Information State: 18-Indiana County: 97-Marion [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2005 FFIEC Census Report-

Summary Census Housing Information State: 18-Indiana County: 97-Marion [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2006 FFIEC Census Report-

Summary Census Housing Information State: 18-Indiana County: 97-Marion [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2007 FFIEC Census Report-

Summary Census Housing Information State: 18-Indiana County: 97-Marion [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

Federal Financial Institutions Examination Council-2008 FFIEC Census Report-

Summary Census Housing Information State: 18-Indiana County: 97-Marion [Data File]. Washington, DC: Federal Financial Institutions Examination Council.

134

Fennell, L. (2008, Summer). HOMEOWNERSHIP 2.0. Northwestern University Law

Review, 102(3), 1047-1118.

Frame, W., & White, L. (2005, Spring). Fussing and Fuming over Fannie and Freddie: How Much Smoke, How Much Fire?. Journal of Economic Perspectives, 19(2), 159-184.

Friedman, M. (2002). Capitalism and freedom: fortieth anniversary edition. Chicago: University of Chicago Press.

Fong, C., Bowles, S., & Gintis, H. (2005, September). Behavioural Motives for Income Redistribution. Australian Economic Review, 38(3), 285-297.

Freeman, L., & Hamilton, D. (2004, May). The changing determinants of inter-racial home ownership disparities: New York City in the 1990s. Housing Studies, 19(3), 301-324.

Galster, G., Marcotte, D., Mandell, M., Wolman, H., & Augustine, N. (2007, September). The Influence of Neighborhood Poverty During Childhood on Fertility, Education, and Earnings Outcomes. Housing Studies, 22(5), 723-751.

Gibbons, S. (2004, November). The Costs of Urban Property Crime. Economic Journal, 114(499), F441-F463.

Guest, A., Cover, J., Matsueda, R., & Kubrin, C. (2006, December). Neighborhood Context and Neighboring Ties. City & Community, 5(4), 363-385.

Hannsgen, G. (2007, January). A Random Walk Down Maple Lane? A Critique of Neoclassical Consumption Theory with Reference to Housing Wealth. Review of Political Economy, 19(1), 1-20.

Hendrickson, K. (2004, August). Bush and the Cities. Policy Review.

Hoff, K., & Sen, A. (2005, September). Homeownership, Community Interactions, and Segregation. American Economic Review, 95(4), 1167-1189.

HUD management: impact measurement needed for technical assistance. (2002, 17 September). GAO Reports.

Ibarra, B., & Rodriguez, E. (2006, Summer). Closing the Wealth Gap: Eliminating Structural Barriers to Building Assets in the Latino Community. Harvard Journal of African American Public Policy, 12, 25-38.

135

Jaffee, D., & Quigley, J. (2008, Fall). Mortgage Guarantee Programs and the Subprime Crisis. California Management Review, 51(1), 117-143.

Jeske, K. (2005, 4th Quarter). Macroeconomic Models with Heterogeneous Agents and Housing. Economic Review (07321813), 90(4), 39-56.

Jois, G. (2008, January). Affordable Housing and Civic Participation: Two Sides of the Same Coin. (Cover story). BYU Journal of Public Law, 22(1), 1-41.

Kaersvang, D. (2006, August). The Fair Housing Act and Disparate Impact in Homeowners Insurance. Michigan Law Review, 104(8), 1993-2018.

Keynes, J. M. (1936). The general theory of employment, interest and money. New York:

Macmillan Cambridge University Press.

Laaksonen, M., Rahkonen, O., Martikainen, P., & Lahelma, E. (2005, August). Socioeconomic Position and Self-Rated Health: The Contribution of Childhood Socioeconomic Circumstances, Adult Socioeconomic Status, and Material Resources. American Journal of Public Health, 95(8), 1403-1409.

Local Area Unemployment Statistics-Lake County, IN: 1997-2008 [Data File]. Washington, DC: United States Department of Labor, Bureau of Labor Statistics.

Local Area Unemployment Statistics-Marion County, IN: 1997-2008 [Data File].

Washington, DC: United States Department of Labor, Bureau of Labor Statistics. Lynch, A., & Rasmussen, D. (2004, February). Proximity, Neighbourhood and the

Efficiency of Exclusion. Urban Studies, 41(2), 285-298.

McClure, K. (2005, July). Deconcentrating Poverty Through Homebuyer Finance Programs. Journal of Urban Affairs, 27(3), 211-233.

McDonald, H. (2005, July). Policy Review. Housing Studies, 20(4), 665-678.

McDonald, J., & McMillen, D. (2004, February). Determinants of Suburban Development Controls: A Fischel Expedition. Urban Studies, 41(2), 341-361.

Mueller, E., & Schwartz, A. (2008, Winter). Reversing the Tide: Will State and Local Governments House the Poor as Federal Direct Subsidies Decline?. Journal of the American Planning Association, 74(1), 122-135.

Munch, J., Rosholm, M., & Svarer, M. (2006, October). Are Homeowners Really More Unemployed?. Economic Journal, 116(514), 991-1013.

136

Nelson, A., Dawkins, C., & Sanchez, T. (2004, February). Urban Containment and Residential Segregation: A Preliminary Investigation. Urban Studies, 41(2), 423-439.

Newman, K., & Wyly, E. (2004, January). Geographies of Mortgage Market Segmentation: The Case of Essex County, New Jersey. Housing Studies, 19(1), 53-83.

Posner, P. (2005, April 26). Community development block grant formula: targeting assistance to high-need communities could be enhanced. GAO Reports.

Prairie State Achievement Exam (PSAE): 2001-2007 [Data File]. Chicago, IL: Chicago

Public Schools-Office of Research, Evaluation and Accountability. Scott, G. (2006, July 31). Treasury's bank enterprise award program: impact on

investments in distressed communities is difficult to determine, but likely not significant. GAO Reports.

Segal, E. (2007, September). Social Empathy: A New Paradigm to Address Poverty. Journal of Poverty, 11(3), 65-81.

Sen, A. K. (1997). On economic inequality. New York: Oxford University Press. Shear, W. (2007, June 25). Leveraging federal funds for housing, community, and

economic development. GAO Reports. Shlay, A. (2006, March). Low-income homeownership: American dream or delusion?.

Urban Studies, 43(3), 511-531. Singleton, R.A., & Straits, B.C. (2004). Approaches to Social Research. New York:

Oxford University Press.

Squires, G., & Kubrin, C. (2005, January). Privileged places: race, uneven development and the geography of opportunity in urban America. Urban Studies, 42(1), 47-68.

Staley, S., & Claeys, E. (2005, December). Is the Future of Development Regulation Based in the Past? toward a Market-Oriented, Innovation Friendly Framework. Journal of Urban Planning & Development, 131(4), 202-213.

Staudt, N. (2006, Spring). REDUNDANT TAX AND SPENDING PROGRAMS. Northwestern University Law Review, 100(3), 1197-1250.

Super, D. (2005, June). RETHINKING FISCAL FEDERALISM. Harvard Law Review, 118(8), 2546-2652.

137

Tita, G., Petras, T., & Greenbaum, R. (2006, December). Crime and Residential Choice: A Neighborhood Level Analysis of the Impact of Crime on Housing Prices. Journal of Quantitative Criminology, 22(4), 299-317.

Tolnay, S., & Eichenlaub, S. (2007, Winter). Inequality in the West. Social Science History, 31(4), 471-507.

Using Individual Development Accounts to Save for a Home: Are There Differences by Race?. (2007, December). Social Service Review.

Varady, D., Raffel, J., Sweeney, S., & Denson, L. (2005, June). Attracting Middle-Income Families in the Hope VI Public Housing Revitalization Program. Journal of Urban Affairs, 27(2), 149-164.

Wood, D. (2007, June 20). Public housing: information on the financing, oversight, and effects of the HOPE VI program. GAO Reports.

Wray, L. (2006, January). Neocons and the Ownership Society. Challenge (05775132), 49(1), 44-73.

Zhan, M. (2006, December). Economic Mobility of Single Mothers: The Role of Assets and Human Capital Development. Journal of Sociology & Social Welfare, 33(4), 127-150.