dissertation - banking regulations in li communities
TRANSCRIPT
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.