'w‘wfi

  

M

w
M
J:

 
     
       
   

C
V
H

[WW

1

1’ W

HI

7

HM

WWI/l:

J

—|(o_, ‘
I010.) l

This is to certify that the
thesis entitled

ANALYSIS OF SECTION 8 VOUCHER RECEIPIENTS IN
PONTIAC. OAKLAND COUNTY, MICHIGAN

presented by

ARUN NEUPANE

has been accepted towards fulfillment
of the requirements for the

MS. degree in RESOURCE
DEVELOPMENT/URBAN
AFFAIRS

 

 

W7v flaw

- Major Professor’s Signature

4. I710?

 

Date

MSU is an Affirmative Action/Equal Opportunity Institution

----.-.—o-o-n-.-n-n-o-n-a--o-.-.-o-.---n—n----n-v-o_—v-.—-—.--o---2---.—.-.—IL-._.-.-.-.-.-._.-.-.-.-.-.-.-.-._

 

 

LIBRARY

Michigan State

University

 

 

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

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/ClRC/DateDue.p65-p.15

 

ANALYSIS OF SECTION 8 VOUCHER RECEIPIENTS IN PONTIAC, OAKLAND
COUNTY, MICHIGAN

By

ARUN NEUPANE

A THESIS

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

MASTER OF SCIENCE

DEPARTMENT OF RESOURCE DEVELOPMENT
& URBAN AFFAIRS PROGRAM

2003

ABSTRACT

ANALYSIS OF SECTION 8 VOUCHER RECEIPIENTS IN PONTIAC,
OAKLAND COUNTY, MICHIGAN

B

ARUN NEUPANE
Section 8 Housing Choice Vouchers are a popular form of housing subsidy for
low-income renters. One of the alleged benefits is their potential in providing
residential mobility and thus enabling the poor to escape from poverty areas to
areas that are more mixed-income and more racially diverse. The research
analyses the relocation outcome for a group of voucher recipients in Pontiac
Michigan, based on their pre-move and post-move census tract characteristics.
Simple spatial analysis involving overlays of address location of vouchers on
distribution various social indicators at census tract level is used to assess success
in dispersal. The difference between pre-move and post-move census tracts in
measures of poverty level, racial composition, household income and employment
are statistically significant. The comparison suggested that the voucher holders
did move to areas that are dissimilar to the ones they left, registering marginal
gains in terms of decreased poverty and increased diversity in destination tracts.
However, the post-move tracts, in absolute terms, are still the poorest tracts in
Oakland County and have a very high proportion of minority. Voucher recipients
also concentrate on tracts that are most distressed for indicators of distress like
reliance on public assistance, adult illiteracy, and proportion of female-headed

households.

ACKNOWLEDGEMENT

I would like to sincerely thank Dr. George Rowan for his genuine support and
guidance. I would also like to thank Dr. Rex LaMore for his constant, valuable
guidance and Dr. John Schweitzer for his valuable feedback early on during this
research. Finally, I am very grateful to Dr. Joe T. Darden at the Center for Urban
Affairs who gave me the initial inspiration to conduct this research and also made

the data from his own project available for this study.

iii

TABLE OF CONTENTS

 

 

 

 

LIST OF FIGURES ................................................................................. v
INTRODUCTION.........-. ..... ............ - 1
OVERVIEW ....................................................................................................................... 1
PURPOSE OF THE STUDY ................................................................................................... 2
SCOPE OF THE STUDY ....................................................................................................... 3
NEIGHBORHOOD EFFECTS AND RATIONALE FOR HOUSING MOBILITY PROGRAMS .............. 4
SECTION 8 VOUCHERS: HISTORY AND REVIEW OF SUCCESS _ - - 8
SECTION 8 TENANT-BASED VOUCHERS AND ASSISTED HOUSING POLICY ........................... 8
SECTION 8 VOUCHERS: REVIEW OF SUCCESS ................................................................... 10
DATA AND METHODOLOGY .......................... -- - - 13
DESCRIPTION OF DATA ................................................................................................... 13
VARIABLES OF INTEREST ................................................................................................ 15
QUANTITATIVE COMPARISONS: OBJECTIVES ................................................................. l6
QUALITATIVE COMPARISONS AND SPATIAL ANALYSIS .................................................... 17
SPATIAL DISTRIBUTION OF VOUCHERS IN OAKLAND COUNTY: AN EARLY ANALYSIS ...... l8
ANALYSIS AND RESULTS .......................................................................................... 24
POVERTY ........................................................................................................................ 24
RACE .............................................................................................................................. 26
INCOME .......................................................................................................................... 29
EMPLOYMENT ................................................................................................................ 31
SOCIAL CONTEXT INDICATORS ....................................................................................... 34
SUMMARY OF RESULTS ................................................................................................... 40
CONCLUSION -- -- - - - - 46
BIBLIOGRAPHY ........................................................................................................... 49

iv

LIST OF FIGURES
FIGURE 1 VOUCHER LOCATIONS IN OAKLAND COUNTY. ................................................... 19

FIGURE 2: ORIGIN AND DESTINATION LOCATION OF VOUCHER RECIPIENTS WITHIN THE

EXTENT OF PONTIAC ................................................................................................... 20
FIGURE 3: PREFERENCE FOR CENSUS T RACTS WITHIN CITY OF PONTIAC ............................ 22
FIGURE 4: PERCENTAGE HOUSEHOLDS BELOW POVERTY IN OAKLAND COUNTY .............. 25
FIGURE 5: PERCENTAGE OF HOUSEHOLDS THAT ARE MINORITY, OAKLAND COUNTY ........ 28
FIGURE 6: OAKLAND COUNTY: MEDIAN HOUSEHOLD INCOMES ........................................ 30
FIGURE 7: UNEMPLOYMENT IN OAKLAND COUNTY ........................................................... 32

FIGURE 8: PROPORTION OF HOUSEHOLDS THAT ARE FEMALE-HEADED WITH CHILDREN
UNDER THE AGE OF 18 ................................................................................................ 37

FIGURE 9: PROPORTION OF HOUSEHOLDS RECEIVING INCOME FROM PUBLIC ASSISTANCE 38

FIGURE 10: PROPORTION OF POPULATION 25 YRS AND OVER WITH LESS THAN HIGH
SCHOOL EDUCATION .................................................................................................. 39

FIGURE 1 1: MEDIAN GROSS RENT IN OAKLAND COUNTY, MICHIGAN ............................... 42

FIGURE 12: MEDIAN GROSS RENT UP TO AND ABOVE THE 1999 FMR IN OAKLAND
COUNTY ..................................................................................................................... 44

Introduction

Overview

In the United States, concentration of poverty in the inner cities is a well-known
challenge for urban policy maker. The inner city residents face an array Of disadvantage
like high crimes, poor education, joblessness, and poor quality housing, compared to the
residents Of the suburbs (Turner, 1998). Moreover, the problem is perceived to be more
serious since the effect Of concentration is cumulative—poverty seems tO yield more
poverty and distress (Orefield, I997; Galster and Killen, 1995). This has given the
rationale for anti-poverty strategies that aim to deconcentrate poverty, in order to
alleviate it. Hughes (1995) classifies anti-poverty strategies into three general
categories—development, dispersal and mobility. The development strategies include in-
place approaches to deal with poverty in the inner cities. Mobility strategies attempt to
connect the disadvantaged poor in the central cities tO the Opportunities by way Of

commutes and networking.

Dispersal strategies aim tO move the poor themselves, and try tO change their geography
or opportunity. The goal Of the dispersal strategy is to help the families move from

disadvantaged neighborhoods to neighborhoods that are richer and more racially diverse.

Housing Choice Vouchersl are rent subsidies given by the department Of Housing and
Urban Development (HUD) to qualifying low-income families to find a decent rental unit
in the private market. The intended outcome Of the dispersal is that those families also
achieve residential integration into mixed race and mixed class neighborhoods, away
from mostly segregated, poor neighborhoods. Housing vouchers are supposed tO be a
relief for the low-income families compelled to live in poverty of the central cities due to
financial constrain (Turner, 1998). The rent subsidy makes better housing in better
neighborhood more affordable for the poor. The vouchers work like this. The tenant pays
at least 30% of her income towards the rent and the rest is covered by the voucher, but
only up to the Fair Market Rent (F MR) value. The FMR, set by the HUD, for a
metropolitan area or a city sets the upper limit for rent that can be paid using the voucher.
It is calculated based on distribution of rents in the private market in the designated rental

marketz.

Purpose of the study

Overall purpose of the research is tO explore effectiveness of Section 8 Housing Choice
Vouchers in achieving the desired dispersal, or deconcentration Of poverty in Pontiac,
Michigan. The Objective is to assess whether there is any difference between tract
characteristics of the origin and destination location of voucher recipients. If the vouchers

were working, in their function as a deconcentration tool, one would find that on average,

 

' Housing Choice Vouchers earlier existed in form of either Certificates or Voucher, which are now
consolidated

2 The rent calculated by the HUD at 40'h percentile of distribution of rents in the relevant local/metropolitan
housing market (source: http://www.huduser.org/datasets/finr.htmI)

the new locations would exhibit lower poverty rate, higher employment and lower

proportion of minority in population.

One would also expect some evidence Of physical dispersal: the new locations would be
more Spread out than the original locations with appreciable average distance covered in

the move. This will also be investigated in the study.

Scope of the study

According to HUD (2000), Section 8 Housing Choice Vouchers have been very
successful in improving housing opportunities for the 1.4 million households, nationwide.
This study is only limited to the Oakland County and the city of Pontiac in particular,

therefore the conclusions will only apply to Oakland County, Michigan.

The nature Of data analysis attempted in this study is simplistic-it is not based on survey
data of individuals voucher recipients, as is the case in other such studies on effectiveness
of mobility programs. (See Katz et. al, 2001; Rosenbaum and Harris, 2001, for example).
All the analysis is based on Census Tract characteristics Of average Origin Tracts and the
average Destination Tracts of the voucher recipients. Therefore the study is limited to
comparison of neighborhoods, and the Census Tract characteristics are assumed to
approximate neighborhood characteristics. The data item Obtained for the study were

only the addresses Of the voucher recipients, which included those that received vouchers

distributed by the Pontiac Housing Commission, and who were also able to relocate to a

new address using those vouchers.

Neighborhood effects and rationale for housing mobility programs

There is a wide body of literature that point to the evidence that where individuals live
affects their opportunities and life outcomes (Galster and Killen, 1995; Rosenbaum,

1991 ,1995;Turner, 1998). The theory suggests that individual achievements (like
motivation to seek gainful employment or performance in school) are as much a function
of the neighborhood or the immediate social network, as it is a function of family
background. Galster and Killen (1995) modeled youth’s decision-making behavior and
concluded that the local social network has an important effect on youth's decisions
regarding education, fertility, work, and crime. Similarly, Rosenbaum (1991, 1995)
studied the participants of the Gautreaux Assisted Housing Program3—a landmark,
supreme-court mandated, dispersal strategy that required Chicago Public Housing
Authority to relocate low-income black families to mixed income and mixed race areas——

and found some encouraging results suggesting that the place matters.

He found that the participants that moved to the suburbs, compared to the ones that

moved close to the city (closer to their origin) benefited from residential integration in the

 

3 In Gautreaux, the courts found that the Chicago Housing Authority (CHA) and the Department of
Housing and Urban Development (HUD) had discriminated against the black tenants, concentrating them
in large-scale developments located in poor, black neighborhoods. The decision against CHA in 1969
called for the creation of new public housing at “scattered sites” in non-minority communities. The court
also ordered a relief in the form of 7,100 Section 8 certificates (now called Housing Choice Vouchers) to
current and former CHA residents to use in the neighborhoods that were less than 30 percent black.

several key areas. The suburban movers had higher adult employment rate, better
prospect Of youth’s education and employment, and the children among the suburban
movers were more likely to be in school, more likely to be in college-bound tracts in
school, in jobs and in jobs with benefits and better pay (Rosenbaum 1995: pp 264). In
addition to better availability Of jobs, suburbs also housed the ‘role models’ that were
achieving success, finding and keeping jobs, and getting their kids educated in good
school. He also cites the result that suburban move lead to better social integration of

blacks with whites (Rosenbaum, 1995).

In order to determine further whether residential mobility programs improve the life
chances Of the poor families through the alleged neighborhood effect, the HUD
implemented Moving to Opportunity (MTO) Demonstration Program for the period of
1994 to 1998. Some analysis of early results is encouraging. The experimental design in
MTO is an explicit random assignment of participants and this is thought to make the
results more valid than that of Gautreaux (Shroder, 2002). The MTO program was
conducted in Baltimore, Boston, Chicago, Los Angeles, and New York. Early results

from Boston and Chicago will be discussed here.

The MTO program Offered housing vouchers to very low-income families living in public
housing projects in tracts that had high poverty rate (greater than 40 percent), and
randomly assigned these families into 3 groups. The first experimental or the MTO group
received housing counseling and were also required to locate to census tracts with

population poverty rate less than 10%. The second group—regular Section 8

Comparison—was given the housing vouchers tOO, but they did not have any locational
constraints. They could move anywhere, even in areas of high poverty if they wished.
The third group was the control group, which just remained in the Inner City Public
Housing establishments. They could move if they wished, but on their own—they did not
get the vouchers. Rosenbaum and Harris (2000) and Katz et. al (2001) have studied the

early results Of the MTO experiment for Chicago and Boston, respectively.

Rosenbaum and Harris (2001) based their finding on post-move telephone surveys data of

individual householder supplemented by pre-move baseline survey data to assess social
gains Of the move. For both the MTO and the Section 8 comparison group, they find that
the proportion employed and the proportion in the labor force increased. Employment for
Section 8 Group increased from 29.3 % to 42.9%, and the same for the MTO group
increased from 24.5% to 46.3%. The survey data also Showed Significant improvement in
housing satisfaction in their new location, for all the movers. (Rosenbaum and Harris,
2001). Although both the groups moved to areas that were very different from their

origins, MTO group was more successful in locating to suburban areas.

Katz et. al (2001) also evaluated short terms results Of MTO program in Boston. Their
results also provide some evidence that moving to opportunity did bring yield some
improvements for the movers. They also had the same benefit of MTO household-head
survey data for 540 households, taken before and after the move. Their Significant finding

is in the area of children’s mental and physical health. Compared to the control group

they find that the MTO and Section 8 groups experienced gains in terms fewer behavior

problems among boys.

These results from the MTO demonstration program, coupled with experience from
Gautreaux program, do Offer some evidence that place matters: that people from poor,
segregated areas improve their lives if relocated to wealthier and diverse neighborhoods.
However, whether all the mobility programs manage to relocate the poor to the right
areas is a separate issue. In the following chapter, success in residential integration using

the regular Section 8 vouchers will be discussed in more detail.

Section 8 Vouchers: History and review of success

Section 8 tenant-based vouchers and assisted housing policy

Public housing policies Of the past have had significant impact on concentration Of

poverty as well as Of minorities in the inner cities (Turner, 1998). The above-mentioned

Gautreaux program is a result Of such a revelation. In 1965, Black residents Of public
housing (represented by Gautreaux) filed a civil rights suit alleging that public housing in
Chicago was segregated (Rosenbaum, 2000). The ruling called for a reversal Of that trend
Of segregation and demanded that the Chicago Housing Authority actively seek new
developments at scattered Sites and that the residents get housing vouchers to relocate to

non-minority areas in the suburbs (Rubinowitz and Rosenbaum 2000)

In the housing policy environment, tenant-based rental subsidy like Section 8 Housing
Choice vouchers gained increasing popularity, especially in recent years. While there
were only 0.6 vouchers and certificates for each unit of public housing produced in
19703, there are 4.75 voucher per a unit Of existing public housing now. Cost-
effectiveness of vouchers is one Obvious reason for the Shift (Shroder and Reiger, 2000).
The other important reason, as hinted by Hartung and Henig (1997), is the perception that
vouchers help deconcentrate poverty, and enable the low-income families move out to
better opportunities. Especially, given the opportunity for deconcentration where 59% of

all public housing residents live in high poverty areas (Turner, 1998). The success Of

residential mobility program like the Gautreaux program has further propelled the shift in

favor Section 8 vouchers.

Along with deconcentration, Section 8 vouchers were also seen as a tool of
desegregation. If given the choice, it is assumed that poor blacks for example, would

disperse and integrate into areas that are predominantly white.

Housing and Community Development Act (1974) Stated the following to be the goal of
Section 8 Program:
“..reduction Of the isolation of income groups within communities and geo-
graphical areas and the promotion Of an increase in the diversity and Viability of
neighborhoods through the spatial deconcentration of housing Opportunities for
persons Of lower income.”

[Housing Act, 1974]

The act implies that diversity, or a racial integration, is an explicit and desired goal of the
Section 8 program. Section 8 vouchers therefore emerge as the low-income housing
subsidy with two-pronged goal—deconcentration of poverty and desegregation. Another
intended outcome of Section 8 vouchers seems to be spatial dispersal. The degree of
dispersal can be measured using spatial tools, in order to gaze the success Section 8

program and this is attempted in the study.

While there have been elements Of success in residential mobility programs, like in that
Of the MTO demonstration or the Gautreaux program, the success of regular Section 8

Vouchers, in terms Of achieving these the same Objectives seems to be mixed.

Section 8 vouchers: review of success

In the MTO studies for Chicago by Rosenbaum and Harris (2000), the average
destination tract for Section 8 group in the demonstration still had high minority
dominance (90% Non-Hispanic Blacks). Also, only 1.9% of the Section 8 Group landed
in suburban tracts, while same rate for MTO group was 31.0%. In the case Of MTO,
better success can be attributed to the better treatment received by the MTO group who
were better informed about the metropolitan housing opportunity through locational

counseling.

HUD (2000) claims that families with Section 8 assistance are much less concentrated
than those in public housing, and the families who move to the suburban areas do the
best. This is shown to be true for most of the studies on Section: that these tenant-based
program have been successful relative to project-bases subsidies to public housing in

achieving the goal of economic integration and mobility.

Turner (1998) finds that vouchers and certificate holders typically live in neighborhoods

that have poverty rate lower than those in and around public housing projects. However,

10

 

the full potential residential mobility offered by the Section 8 vouchers do not seem to
have been realized.

Hartung and Henig (1997) analyzed census tract locations Of over 1000 certificate and
voucher households in the Washington DC. and its suburbs, and compared those with
public housing and other project-based subsidies. They also find evidence that Section 8
vouchers may be succeeding in dispersing affordable housing Opportunities beyond the
central city boundaries. They mention that compared to the disproportionate distribution

of public and project-based housing—300% and 175% respectively Of the fair Share if it

 

were distributed evenly throughout the census tracts— certificates and vouchers are much
more evenly distributed, with average tract having 9.2 certificate households and 3.2

voucher households.

However the authors also find some evidence Of re-concentration within the suburbs: the
vouchers in the suburban jurisdictions were found to be concentrated in hi gh-minority,

low-income tracts (Hartung and Henig, 1997).

Turner (1998) also finds that vouchers and certificates are at least better dispersed than
public housing. She finds that nationally, voucher households are far less likely than
public housing residents tO live in high-poverty neighborhoods. Only 14.8 percent Of
voucher recipients live in hi gh-poverty neighborhoods (greater than 30 percent poor)
compared to 53.6% for public housing residents (Turner, 1998). She also warns

however, that there are differences within race: the share of Blacks and Hispanic

11

certificate holders living in high-poverty neighborhoods is much higher than the share Of

whites. (Turner, 1998)

Varady and Walker (2000) analyze the mobility outcome Of a group Of public housing
residents that were given housing vouchers tO move out from their public housing
establishment that were demolished. They find that participants moved only Short
distances and the majority continued to live in racially segregated neighborhoods.

Measured at a census block group, they found that recipient’s post-move neighborhood

 

had only marginally lower percentage Of African-American population.

Pendall (2000) classified more than 44,000 metropolitan census tracts across the nation
for distress levels (mild or severe) and compared proportions Of Section 8 renters, poor
renters and all other renters in the metropolitan distressed tracts. He found that the
Section 8 tenants were better than other poor tenants in terms Of their location in non-
distressed tracts. However, compared to all other renter households in US metropolitan
tracts, Section 8 tenants still lived in disproportionately distressed neighborhoods. Pendall
(2000) also speculates on why the vouchers concentrate in distressed tracts and contends
that vouchers concentrate in distressed tracts because rental housing also concentrates

there.

12

Data and Methodology

Description of data

Most studies that have studied relocation success have followed the voucher recipients
from the origin to destination locations. In such studies, baseline survey and post-move
survey of the participants themselves is the method used to elicit indicators Of success

(see Rosenbaum, 1995; Katz et. al, 2001).

In this study, the data available is only the addresses of the voucher recipients. 500 pairs
Of addresses (1000 in total) Of Section 8 voucher recipients were made available from the
Pontiac Housing Commission. These addresses were linked to their corresponding census
tracts to determine ‘average’ neighborhood conditions at the origin and the destination.
All analyses assume that census tract characteristics Of a particular address reflect the

neighborhood conditions Of that location.

All of these participants were living in public housing maintained by the Pontiac Housing
Commission before they received the vouchers. Although the exact period of the
movement for these movement in unknown, it is known at least that they moved between

1995 and 2000 and that a Significant majority Of the recipients are African-Americans’).

 

4 Personal Communication with Dr. Darden, Urban Affairs Program, MSU

l3

 

 

Of the 500 addresses that included destination and origin address for each voucher
recipient, corresponding census tracts could be verified for only 459 pairs. Among these,
72 individual households (15.7%) moved within the same tract—the corresponding
census tract was the same for both the origin address and the destination address. These
pairs are excluded from the main analysis. Table 1 shows the breakdown for the original

data pool of 500 pairs of addresses. The final sample size is 387 pairs.

Table 1: Breakdown of 500 pairs of addresses

 

Number of lPercentage of Percentage out of the

 

 

 

Characteristics of Addresses address pairs Total Total Matched Pairs
Matched pairs 459 91.8%

Same tract movement 72 14.4% 15.7%
Inter-tract movement [The Study Sample] 387 77.4% 84.3%

 

Removed from original 500 due to

unconfirmed address for both origin and
destination 20 4.0%
Removed from original 500 due to

unconfirmed address for either origin or
destination 2] 4.2%

Total in the Study 500 100.0%

 

 

 

 

 

 

 

 

The total number Of census tracts contained in the sample Of 387 pairs Of addresses is 55.
Most Of the origin tracts are in Pontiac, Oakland County Of Michigan (26). For the rest
Of the origin tracts, 7 are from other counties within Michigan and 14 tracts are from
counties outside Michigan. Most of the destination tracts in the sample are in Oakland

County, except in one case where destination is Osceola County Florida.

Quantitative analyses include census tract data of all 55 tracts—even Of those outside

Oakland County Michigan. Qualitative analysis—Spatial displays/interpretations—are

however limited to movement within the Oakland County.

14

 

 

"fl
‘

Variables of interest

The objective of the study is to determine whether residential relocation Of the voucher
recipients have landed them in neighborhoods that is racially and economically mixed.
Therefore the primary census tract-level variables Of interest are poverty status,
proportion Of minority (Non-Hispanic Blacks, Hispanics), and median household income.
The residential relocation is also supposed increase employment prospects. Therefore,
employment variables are also analyzed. These include adult employment rate
(proportion of population 16 years and above who are employed) and also Specific
category Of employment like employment rate among high-school dropouts (proportion

Of children 16-19 who are not in high school and not employed).

Third categories of variables that are analyzed are variables that indicate neighborhood
distress (Kasardra, 1993). The rationale for analyzing neighborhood distress is that a
successful residential relocation would relocate the voucher recipients to neighborhoods
that are less distressed than their origin. The distress variables includes poverty status,
and in addition includes the following: Proportion of female-headed households with
children under 18, proportion Of households receiving public assistance income,
proportion of adult population 25 years and over with less than high school education,
high school drop out rate (proportion of population 16 tol9 that are not in high school
and that don’t have a high school diploma), proportion of high school dropout that are

unemployed

15

Variables that help interpret the results in terms Of housing market are also included in
the analysis. For example, residential relocation could be hindered by high cost Of rental

housing in areas surrounding the city.

Quantitative comparisons: Objectives

Quantitative analysis, using weighted means Of census tract characteristics is based on the
sample of 387 complete pairs of addresses. Each of 387 voucher recipients was assigned
an id, and census tract characteristics of the address were attached to the row Of data for
that id. Weighted means of relevant variables (like poverty rate, percentage minority) Of

origin addresses (n=3 87) and the destination addresses (n=3 87) are compared.

These means are compared to determine whether the move was a success. For example,
if the average origin tract poverty rate is 25% and average destination tract poverty is
22%, it would constitute an improvement. The statistical significance of these
differences is also tested. The statistical test chosen for this purpose is SPSS’S paired
sample t-test. The paired sample t-test is designed to test the significance Of difference
between two treatments. The two treatments, for the voucher recipients are:

a. Neighborhood condition (a chosen census tract variable) of their origin address in
1999, and

b. Neighborhood condition (a chosen census tract variable) Of their destination
address in 1999

The paired-sample t-test tests whether the average difference between the origin variables

and the destination variables is zero. The test takes an average of all the differences

l6

(n=3 87) between the origin and the destination variables for each case. If this mean Of the
differences is different from zero, then one can conclude that there iS a significant

difference between the two treatments (origin and destination).

The success Of residential relocation will be explored, one variable at a time, based on the
mean values for origin and destination tracts, and the significance of difference. For the
primary variables, a move to a neighborhood with lower poverty rate, lower proportion Of
minorities, higher median income, and with difference Significant at 95% confidence

interval, constitutes success.

Qualitative comparisons and spatial analysis

A geographic analysis is very relevant to the study Since the objective Of the research is to
assess dispersal and deconcentration, and the nature of data—addresses—also makes it a
very relevant exercise. All the addresses were geocoded using ESRI Arc View Software
in to a street level map. Tiger 2000 file, which has the street coverage for the whole of

United States, and which is available free from Census Bureau, was used for this purpose.

Geocoding process in Arcview involves Batch Match where the software gives a score
ranging from 75 to 100 for a perfect address match. It is usually the case that all the
addresses do not match and therefore do not produce a dot in the map due to spelling

errors and limitation of the reference street map. The final number of origin-destination

l7

pairs matched was 350. The households that failed to yield address match on either the

origin or the destination were excluded from spatial display and subsequent analysis.

618 also enables overlays of different themes in the same map. Once the address points
are geocoded and the points labeled origin and destination, these can form the frame Of
reference for other data that can be overplayed. The outputs maps of Oakland County,
showing the distribution Of vouchers with the distribution of other census tract level
attributes like (poverty status, unemployment rate) would give a visual picture of location
of vouchers relative to distribution of poverty and unemployment. Such analysis has
been attempted before. Varady et a1. (2000) analyzed dispersal Of vouchers using similar

method, and were able to Show graphically the extent Of dispersal.

Spatial distribution of vouchers in Oakland County: an early analysis

Figure 4 below shows the distribution of voucher recipients within the extent Of the
Oakland County. The Figure Shows a heavy concentration Of Section 8 Housing
Vouchers within the confines Of Pontiac. Figure 5 shows Pontiac detail, covering only

the census tracts that fall within Pontiac.

Both these figures set the tone for the rest Of the analysis in this paper: that the vouchers

recipients are concentrated within the confines of Pontiac and there are very few

vouchers recipients that have landed to other parts of Oakland County.

18

 

 

 

 

i

3:! ov

 

8 8823... >\

as 8822...
28285

2...:5 so: :6 82.3.. B

n82» 82.3 25:5 2....ch D
552589
595

89852

3550 9.2.30 5 message... Cocoa;

DEED EEEBO 5 £833?— uogo=e> Q 2:»:—

 

19

 

 

3096
8:258 .

 

c_m_._O . 2:: v a o a

88062
Emuczom 69H Ell /

 

 

“EOECOUCMQM um“, amt—m

 

 

 

 

 

 

. A

 

‘ _

«$5 :_ 33. .02.. 25:00
oazcom ho .chm 2: 55:5 3:233»: Logo—:5 he cozaczmoo new 595

 

 

ensue.” no Eouxm 2: £515 Saki—tom .5535 .3 5:88:— uozaucmoa can EMEO "N PSwE

20

In both the maps (Figure 1 and 2), light dots Show the pre-move addresses and the dark
dots Show the post-move addresses. It can be generally inferred from the distribution of
the dots that pre-move and post move locations are more-or less within the Pontiac. There
seems to be very few moves out of Pontiac to other parts of Oakland County. A quick
look also confirms what was discussed earlier in terms Of patterns Of these dots. There is
no discernible spatial pattern (grouping or clustering) for either the dark dots or the
lighter dots, separating areas chosen from the areas rejected. Except for few instances like
in census tract 1425 where a contiguous area is evacuated, the pattern is not clear. The
origin and the destination points are more or less interspersed within the boundary of
Pontiac, suggesting uniform spatial distribution. However, some tracts are more preferred
than others as indicated by different frequencies of addresses for these tracts. A simple
preference index—ratio for each Census Tract, Of the number that moves in to the
number that moves out calculated to rank the preference among the tracts. Figure 3 maps
this preference for tracts within the city Of Pontiac, giving an overview of preference

within these areas Of highest voucher concentration.

The Census Tract 1425 Of the Oakland County is the only tract that resembles a spatial
case of distinct rejection (see also Figure 2 inset). A very small number of voucher
recipients have moved into this tract compared to those that have moved out. One
probable reason for this could be that tract is the site of public housing establishment.
The preference index for Census Tract 1425 is 0.05, and it is therefore is the least

preferred tract. The most preferred Census Tract is 1411 with the preference index Of 35.

21

 

 

 

m 3
Z
w.«
ma; .
8 o . o D
G30 .33: .oz _ E .332 .on .89: 3:2. 2
Snafu—I
09:6 O
w=_I _
Eng—3.0

Seat. wzmcoOuflonssz :90 Son.
62.6.63 «ween: 356E...— xcam 23:52 =20 oco
camEoE 3:300 2.6.an 62< enscon— 5
28¢ msmcoo 55:5 00:222.“.

 

 

2355 .3 :30 £53 822,—. 25:00 .5.— oonouououm "m 2%:—

22

For each Of the voucher recipient the distance between their origin and the destination
addresses was calculated. This was done using the Arcview software’s network analyst
extension, which allows for accurate measurement of distance. The procedure calculates
the distance between two given points using the street map as the reference. The distance
calculated is the shortest route using the street network, from origin to destination, for
each pair of address. Street distance was deemed more Of an appropriate measure, since
driving or walking distance is a more realistic reference in neighborhood interactions.

The average distance moved by Pontiac voucher recipients was 2.53 miles.

23

Analysis and Results

Poverty

Poverty deconcentration is an explicit goal Of Section 8 Voucher programs. AS mentioned
in the Housing Act Of 1974, Section 8 families are supposed to move from high-poverty

areas tO low-poverty areas.

In the study sample, the average tract that the voucher recipients left behind had a
population poverty rate of 29.7 percent, and the ones that they moved to had 22.7 percent.
Therefore, the families did manage to move to an area of lower poverty, and that
indicates some success. However, the destination tract poverty level of 22.7 percent is
still classified a ‘poverty area’ by census bureau standards, which defines poverty areas
as those tracts having poverty rate up to 20 percent. The destination tracts would have
been disqualified as destinations for the MTO participants, who were only allowed tO
move to ‘low poverty’ areas with poverty rate less than 10%. The destination tracts
would also not be considered an ‘Opportunity area’ by Baltimore Housing Authority
Standards, for example, which considers Opportunity areas as ones below the poverty
level Of 20 percent (See Swope, 2000). Compared to the Oakland County as a whole,
which has a poverty level of 5.5 percent, the destination tracts, and Pontiac, which has a
poverty rate Of 20.9 percent, are clearly poor (See Table 2). Figure 4 below maps the
poverty rate for the whole of Oakland County, and shows that the poorest neighborhoods

are in and around Pontiac.

24

 

 

            

L. or

              
 

    

 

fififlmflwrt“ IIVLEIIJI

IgflfluI-BAman-lflfll-II-ml
B-hPAMI'I'E fillnl'li I
Efiwfifiuqu-EIIIII

BEBE-Emmnduflfl I”
nuns-um hfll‘flflmfl .-

o

in Inna rial
.Illurlllla SSHmpal-Ii

 

1i“... BARE-Bull!
«REFER-Ea.

D.

IVIQEWWWV‘ I
n. A I. I ,

IBM—”II

.. u A II
imam-p

Lo 3 Loco—40> “6 5:83 352330 265 20g =me

'4

89. E Egon >>o_om 365026: ”3560 3:2me

 

 

@550 can—GEO E 3:525 33mm @2383: away—593m ”v 0.5»:—

25

Concentration Of destination address dots in the same area illustrates the failure of
dispersal into low-poverty areas, given the apparent Opportunity to locate in low-poverty
tracts outside Pontiac. The tracts where the voucher holders have located are also the

poorest tracts in the whole Of Oakland County.

Table 2: Poverty Rates in Origin and Destination Tracts, Pontiac and Oakland
County

 

 

 

 

 

 

 

Percentage Below Poverty Paired Sample T-Test
5 Population Household Families for Origin-
Destination Pairs
Origin Tracts 29.7% 28.32% 25.4% Significant (p <0.01)
Destination Tracts 22.7% 21.6% 18.6%
Pontiac 22.1% 20.9% 20.3%
Oakland County 5.5% 5.4% 3.8%

 

 

 

 

 

 

Source: Summary File 3 Tables. Census 2000

Race

Like poverty-deconcentration, another goal of Section 8 Vouchers is racial
deconcentration, or more specifically, racial integration Of the voucher recipients into
neighborhoods that mixed race or even white dominated. The rationale for this, as
mentioned earlier, is derived mainly from experiments like Gautreaux where racial
integration that has worked in favor of the minority who made the move. The espoused
benefit mechanism is the neighborhood effect—better role models.

Segregation in housing is a pervasive problem (Farley and Frey, 1994; Crowder, 2001)

and Detroit area was the most segregated metropolis in 1990 among 47 metropolises with

 

5 Population for which poverty status was determined

26

a population Of a million or more (Farley and Frey, 1994: 9). The results Show that both
the origin and the destination neighborhoods were still had a high percentage Of Non-
Hispanic Blacks. The average origin neighborhood had 67.3 percent population that was
Non Hispanic Black and the same for the destination neighborhood had 54.4 percent
Non-Hispanic Blacks (See Table 3). The difference between these means is significant at
p=0.01 level. Therefore, it can be concluded that, although the destination neighborhood
is still very segregated with 54.4 percent Black, the move does constitute improvement

for individual voucher holders.

Table 3: Minority Percentage in Origin and Destination, Pontiac and Oakland
County

 

 

 

 

 

 

 

 

 

 

 

 

 

Percentage of Population Paired Sample T'TCSt
. . . . for OrIgIn-
Non- Non- Hrspanrc Mrnorrty Destination Pairs
Hispanic Hispanic (Blacks and
Blacks Whites Hispanic)
Origin Tracts 67.3% 7.5% 21.0% 88.3% Significant: (p
Destination Tracts 54.3% 1 1.0% 29.7% 84% <0-01)
Pontiac 47.2% 34.6% 12.7% 59.9%
Oakland County 9.9 % 81.4% 2.5% 12.4%

 

The destination tract proportion Of Hispanic has actually increased from 21 percent to
29.7 percent. The proportion of Whites has increased from 7.5 percent in the origin to 11
percent in the destination. Despite marginal, but statistically Significant gains in terms Of
move to non-minority tracts, the 84 percent of population in average destination tracts are
minority. Oakland County, as whole has only 9.9 percent Blacks and 12.4 percent
minority. This is clear evidence that vouchers are concentrated on highly minority

dominated areas, within the Oakland County.

27

 

 

$8 3 2 I
$3 8 8 g
$3 2 a a
$2 2 a D

 

002 E 3:052 omflcoouon ”3560 9:2me

 

 

 

@550 c.3250 $3.852 PE :2: mEeaogem he «wag-.09.?— "m charm

28

Figure 5 shows the proportion of minority household, confirming that the tracts where
Section 8 vouchers households are concentrated are also the tracts that have the highest

proportion of minority households.

As with poverty status, the desired threshold for racial composition also varies according
to the mobility program. In Gautreaux program, the implementers had a stringent
threshold: they did not send the participating tenants to areas with minority rate
exceeding 7% (Rosenbaum, 1995:257). Targets vary according to region. For instance,
for the Baltimore Housing Authority discussed earlier, a minority rate of less than 20%
was sufficient and constituted the ‘opportunity’ area. There seems to be ample
“opportunity” areas, with low minority proportions, within the Oakland County where the

voucher recipients can potentially relocate.

Income

Move to mixed-income neighborhood is another intended outcome for the Section 8
Vouchers. Through provision of rent subsidy that supplement the rent amount after the
tenant contributes 30% of her household income, the tenants are expected to afford better
housing in a more expensive, but decent neighborhoods. If this aspect of Section 8
Instrument were working, one would expect that the voucher recipients move to higher-
income tracts. Median Households Income is taken as a measure of neighborhood wealth
in this analysis. Table 4 shows Median Income (for Households, Families, and the

Population) for the Average Destination Tracts, Average Origin Tracts and for the whole

29

 

.. F or
945‘ "I.

IF 3:
g. ,n

92.2 5 2 «8.3;

  
 

 

 

8...:
«:63 3 v3.3.
3...: o. 3...:
:4: BED
8.3:. .552 2.. :1
his...
5.358.. .
03.0 u
.— O c0380.— Szmczmmo .>>or_w 3.00 =me
mmmr E OEOOE _mzcc< BOLOmJOI CEUOE ”3:300 3:3me
mun-aces; mica—023°: mauve—Z "ban—OD Ufla—u—ao "w ou—uwmh

3O

of Oakland County and the Pontiac, along with results of Paired Sample Test for the

Origin-Destination Pairs.

Table 4: Median Incomes, Pontiac and Oakland County

 

Median Income Paired Sample T-Test for Origin-
Destination Pairs

 

 

 

 

 

 

 

 

Households Families
Origin Tracts 27,239 31,077 Median Household Income,
Destination Tracts 31,159 36,549 Significant: (p <0'01)
Pontiac 31,207 36,391
Oakland County 61,907 75,540

 

 

 

Source: Census 2000 Summary File 3

 

There is a slight gain in average median household income from origin to destination
tracts. The average median household income for the origin households is $27,239 and
that for the destination households is $31,159. The difference is statistically significant.
However, in terms of the potential to relocate in higher-income areas throughout the
Oakland County, the concentration of vouchers is still in areas that have the lowest

Median Household Income range (See Figure 6).

Employment

Central cities typically have a very high unemployment rate (Orefield, 2000). High-
unemployment is one among the various attributes of ‘distress’ often used to characterize
inner city areas. The conditions leading to a systematic unemployment in the central
cities are varied. Disinvestments and White flight (relocation of middle class whites to
suburbs and relocation of business to follow them) are some of the conditions cited as

being responsible (Orefield, 2000)

31

 

 

 

2me Locoso> *0 c0380.. 8:25me m5 265 £00 :95

 

mm? 5 EoE>anocD ”3:300 9:2me

 

.3550 maniac E Eon—@3525 uh charm

32

Section 8 vouchers, as exemplified by the Gautreaux experiment, would be performing

well if they are moving the low-income tenants to areas where there are jobs.

It is debated whether the observed gain in employment of participants in a successful
housing mobility program is due to increased availability of jobs in the suburbs, or
whether it is simply due to increased motivation (Rosenbaum, 1995, Turner, 1998). A
simple assessment of job availability is attempted here for the sample participants. A
measure of unemployment in civilian labor force, based on the census tract data is
attempted. Exact measure used is 16 years and older male and female in civilian labor

force as a percentage of total labor force. Figure 7 and Table 5 show the result.

Unemployment does decrease from 15.6 percent in the origin tracts to 11.2 in the
destination tracts, and this decrease is significant. This is encouraging, but as in the case
of poverty and race, voucher holders are still concentrated in areas of highest

unemployment (as indicated by Figure 7).

Table 5: Unemployment in Origin and Destination Tracts, Pontiac and Oakland
County

 

Unemployment Paired Sample T-Test for Origin-Destination

 

 

 

Pairs
Origin Tracts 15.6% Proportion of Civilian Population in Labor
Destination Force, unemployed, Significant: (p <0.01)
1 1.2%
Tracts
Pontiac 10.3%

 

 

 

 

 

Oakland County 3.7%
Source: Census 2000 Summary File 3

 

33

It confirms the earlier patterns of concentration in distressed tracts. The areas where
vouchers are located are also the areas with highest unemployment in the whole of

Oakland County.

Social context indicators

Poverty, race, income and employment are census tract characteristics that a Section 8
Program can constantly monitor, in order to achieve better success. As is evident in the
above discussions, the distribution of vouchers in Oakland County is not optimal in these
parameters. There are additional, conventionally accepted measures of neighborhood
distress, which further supplements the analysis. Kasardra (1993) and Hartung and Henig
(1997) use the following social context indicators—poverty, joblessness, proportion of
female-headed households, proportion of population receiving public assistance and high

school dropout rate.

Among these, female-headed households, households on public assistance, and
proportion of population 25 years or over who are less than high school graduate is
analyzed in the following section, to supplement earlier analyses of poverty. Figures 8
through the 10 show the distribution of these variables relative to destination locations of

the vouchers, Table 6 summarizes the results.

34

Table 6: Indicators of Neighborhood Distress and Performance of Origin and

Destination Tracts

 

 

 

 

 

 

 

 

 

 

 

 

High School Unemployment Proportion Proportion Proportion
Dropout among High of Female- of of
Rate School Dropouts Headed Population Population
Households Receiving 25 years and
Public over with
Assistance less than
high school
education
Origin Tracts 18.1% 23.6% 20.7% 8.6% 34.06%
Destination 19.1% 9.9% 23.6% 10.1% 17.1 1%
Tracts
Paired Sample T- N.Ot . Significant: Significant: Significant: Significant:
Test for Origin- Significant (p <0 01) (p <0 0]) (p <0 01) (p <0 01)
Destination Pairs (p >0.05) ' ' ° '
Pontiac 19.8% 11.6% 14.8% 31.1 7.6
Oakland County 5.5% 18.4% 4.9% 10.7 1.8

 

Source: Census 2000 Summary File 3

 

There are mixed results in terms of differences between the origin and destination tracts
characteristics of these indicators of distress. The only difference not significant and
p=0.05 is high-school drop out rate. There is not much difference in high-school drop out
rate between the old neighborhoods and the new neighborhoods.

However, adult educational attainment—adults over 25 years with less than high school
education—has increased after the move to new neighborhoods. The gain from origin to
destination neighborhood is 34.1 percent to 17 percent. New neighborhoods therefore
have more educated individuals, therefore are better, on average, in terms of potential for
offering ‘role models’. Another variable that improved for the destination tracts is
unemployment among high school dropouts. In the origin tracts 23.6 percent of the high-

school dropouts were unemployed, in the new neighborhood this dropped to 10 percent.

35

However, among the distress indicators, single parenthood and dependence on public
assistance actually increased in the new neighborhoods. The origin neighborhood had
20.7 percent of households that were female-headed with children under 18. This
increased to 23.6 percent in the destination neighborhoods and this difference was
statistically significant. Given that single parent females form the largest majority of
Section 8 voucher recipients nationwide (HUD 2000), this result is interesting, and could
lead to further revelations upon more analysis. Similarly, proportion of population
dependent on income from public assistance also increased from 8.6 percent to 10.1
percent. The proportion is still less than the average for the city of Pontiac (31.1 percent),
therefore indicating by this measure, both the origin and the destination neighborhoods

are better than Pontiac.

The results of comparison of distress indicators of pre-move and post-move tracts are
therefore mixed. While there were encouraging gains in terms of adult educational
attainment and unemployment among high school dropouts, there were no gains in the
measure of proportion of female-headed households and dependence on public
assistance. Whatever the incremental gains or loss as a result of the move, relative to
Oakland County as a whole, vouchers are still concentrated on distressed tracts for most
indicators of distress. The following Figures (8 through 10) show the distribution of three
of the distressed indicators, it is clear that in all these measures the location of destination

vouchers falls in the highest range of distress.

36

 

 

 

 

 

 

$83 3 8.3 I
$3.3 8 2d I ,
$96 8 a: E
$§e 8 3.“ D
$8.“ 2% D
contoao;

3 83235

 

2 3523:. \<
353:9:
acocoao>

 

2 .595 5.220 5:5 “Sumo—+25 Eon 98 «“5 $39.33: be 53.23:“.

 

a no ow< 2: .535 5.5—EU .53 3533.03th 0.3 2:: 3:voch no SEE—ohm "m 0.59%

37

 

 

 

. fig... '7' . .
mfiMflfl- filfl'fi
zfifimfir§Ig

 

arm‘fl" Ilhll
< numb-mu.-

 

.. ”My...
«E... 2% D
I: hO cos-Dan:

8 033.52.

on 3333:. \<

933:9:
20:03.5 .

 

 

 

 

oocflflmfl‘ 23:; Eat 088:. 9:230”. 0.22.3.5: be .8anch /

 

 

 

3.5333. 3115 Bob 2:8:— wanoom 33:83: no 5.3.83.5 & 9:.me

38

 

 

 

 

viral

a   #de

 

$2.3 8$mmm~ I
$3.8 2 32 I
$3.2 3 89 g
$32 8 5.0 D
$86 2% B

 

someone

cm 8328:.
E. 3823:. >\
«>952:
neocoao>

 

cosmosum Begum :91 :55 mm»... 55 +2> mw 5:223... no coEonoE

 

:ouaoaam Begum Ami :2: W84 53 .52. can 9; mm Ecru—23m mo scheme; "A: 953m

39

Summary of results

The above analysis clearly shows that the vouchers in Pontiac, Michigan are still
concentrated in poverty tracts (>20% poverty rate), tracts with high proportion of non-
Hispanic blacks (54.3%), and tracts with lowest range of median household income in
Oakland County. The spatial distribution associated with each of the variables analyzed
clearly shows that dispersal, as a strategy is not working, especially given the availability
of opportunity to locate to areas that are low poverty, low minority and high income,

within the extent of Oakland County.

However, there are some encouraging gains at individual levels. For example, the
destination neighborhoods had higher proportion of high school educated adults, and
lower unemployment among high school dropouts. Most importantly, for each individual
family that moved, the neighborhoods they moved to were different from the ones they
left, however little the difference. Exception to this was high school dropout rate, where
the difference in origin and destination neighborhoods for the individual family was not
statistically significant. Overall, vouchers have concentrated, more or less in distressed
tracts, within highest poverty and highest minority concentration areas within Oakland

County

MTO demonstrations discussed above have provided the evidence that relocation
counseling helps the low-income tenants to disperse more. There was a significant
difference between the performance of Section 8 group and the MTO group and their

ability to relocate to suburbs. Katz et al. (2001) found the even though the both groups

40

came from a similar background, 31.3 percent, among the MTO group relocated to
suburbs, as opposed to only 1.9 percent for the regular Section 8 comparison. One of the
differences in the treatment was that MTO group received housing relocation counseling
that included items like landlord outreach. They were informed about the opportunities
and their choices more so than the Section 8 group. The Gautreaux program also had
relocation counseling and therefore had a good success rate for relocation to suburbs
(Rosenbaum, 1995). This aspect seems to suggest that constraint to relocation is only the

lack of information, counseling, or logistic help (car for apartment hunting, for example).

Another obvious constraint is the cost of rental housing in private market. Pendall (2000)
contends that it should be no surprise that vouchers are concentrated in the distressed
tracts because the rental housing is also concentrated there. Decent housing in an area out
of the present cluster of Section 8 vouchers, which are mostly within the Pontiac

boundary, may cost more than the subsidy.

Section 8 vouchers do pay part of the rent but there is a ceiling to it—The Fair Market
Rent value, anything above which the voucher recipients have to pay out of their own
pocket. In a tight housing market, this FMR set by the HUD, maybe much lower than the
market rate for decent housing in that area. Table 7 shows the FMR for Detroit MPSA,
inclusive of Oakland County, for the period 1995-2000. The natural-break distribution of
Gross Median Rent for a two-bedroom apartment in the Oakland County is displayed in

Figure 11.

41

 

 

 

88» 3 8m;
8m; 2 08;
8o; 3 Ba
8: 3 28 D
NS» 3% D
uco

 

>932:
co .3535
33.0 .

whom: Ecoso> .6 5:80.. ,5szme 39.5 200 =me

82. E “com waO 5282 ”3:300 3:3me

 

:«wEfl—Z .5550 unaEaO 5 «com 895 5:62 u: 95me

42

The distribution clearly shows that Gross Median Rent varies throughout the county and
also has a special pattern. Median Gross Rents for census tracts outside the areas of
vouchers concentration is distinctly higher. Most of the voucher locations fall within

areas with a median gross rent under 706 dollars (first two legends in the map legend).

Table 7 shows the trend in F MR for two-bedroom unit in Detroit MPSA. Further analysis
of F MR as a constraint can be attempted based on 1999 FMR— Figure 12 displays the
resulting map. It clearly shows that the concentration of voucher recipients falls squarely
in the un-shaded areas, where the Median Gross Rent is up to the Fair Market Rent of 634
dollars. The shaded area (almost all the tracts outside the Pontiac area) has a Gross
Median Rent that is higher than the FMR, and has very little voucher recipients that have

relocated there, as evidenced by the low density of dots that area of the map.

Table 7 Fair Market Rents for Two Bedroom Units in Detroit PMSA

 

Wear 2000 1999 1998 1997 1996 1995

FMR_2BD $650 $634 $618 $567 $559 $550
Source: HUD Datasets (http://www.huduser.org/datasets/fmr.html)

 

 

 

 

 

 

 

 

 

 

43

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

$223301 hose—:5 .o 5:38.. 532330 on: iosm 3.5.
3:300 “=8.an E m2”. ammr tam Eom anew c3305.

 

 

@550 "Ea—6E0 E mzrm 33 of 32?. an“ 8 mp 28% $20 5:52 "Nfi Sana

44

This analysis, although simple, does indicate that housing costs may be the constraint to
achieving the desired dispersal. If this is the case for Pontiac, Michigan, relocation
success is more a reflection of cost of housing in the private rental market rather than the
quality of information and counseling the low-income tenants get. This needs further

investigation.

45

 

Conclusion

The analysis of relocation success of Section 8 vouchers recipients in Pontiac, based on
census tract characteristics of tenants, seems to indicate that economic and racial
integration has not resulted to the desired extent, nor is spatial dispersal evident. Average

distance moved by each voucher holder was only 2.53 miles.

Firstly, 15.7 percent of the households (from a total of 459) moved within the same tract.
For the rest who moved at least to a different census tract, there are marginal gains
towards racial an economic integration, and these gains are statistically significant. The
program as a whole though, still seems to concentrate vouchers in high-minority areas
(84 percent minority) and in poverty areas (greater than 20 percent of households below
poverty line). Additional analysis of widely used measures of neighborhood distress also
indicated that the cluster of voucher households fall squarely on the most distressed tracts
for the most indicators of distress— proportion of adults without high school diploma,

high school drop out rate, and proportion of female-headed households.

There could be several reasons for this lack of success in spatial dispersal and racial and
economic integration to the desired extent. One of the most straightforward explanations
is that vouchers concentrate in distressed tracts because affordable rental housing is also
concentrated in these tracts (Pendall, 2000). A simple spatial analysis of location of the
vouchers relative to FMR and Gross Median Rent for Oakland County demonstrated this

constraint, to some extent, for the Oakland County.

46

 

Personal transportation could be another major constraint. One can not expect the inner
city poor without a car to move to suburbs when it is logical for them to remain in areas
with familiar bus lines (Varady and Walker, 2000). It could also be that minorities
(majority of voucher recipients) feel more comfortable living in areas with a decent
proportion of minority, even when given the choice to move to mixed, or white-
dominated areas (Farley and Frey, 1994; Hartung and Henig, 1997). Section 8 tenants
also face actual discrimination from landlords who do not wish to take the voucher-
holders because of the associated stigma with low-income tenants (Pendall, 2000;
Hartung and Henig, 1997) and this might have limited moves to the suburban

developments.

There is evidence that regular Section 8 Vouchers can be made more efficient as a tool
for dispersal, if supplemented with relocation counseling and landlord outreach. MTO
demonstration program, discussed earlier indicate that these improve success rate.
Studies on success of Chicago’s Gautreaux program also point to importance of
counseling and landlord outreach (Rosenbaum 1995). There has also been suggestion to
improve inter-jurisdictional cooperation in the administration of Section 8 program in

order to achieve better success (Katz and Turner, 2001).

It is evident from this research that a voucher-based strategy for dispersing housing
opportunity for the poor has its limitations. Except for the MTO demonstration program,

mobility is not imposed—regular Section 8 voucher holders also have the choice not to

47

 

move. There are other strategies that aim for betterment of housing opportunity through
dispersal. HUD’s scattered site public housing is one such policy. The policy originated
from the Gautreaux ruling in Chicago that mandated that all new public housing be built
at scattered sites, and not concentrates new development in inner city locations.
Enforcement of fair share housing is another such strategy. Fair share housing is based on
the notion that jurisdictions should establish specific criteria for including their fair share
affordable housing in any new development, not only public housing developments.
States and jurisdictions that have articulated such policy mandate that any new housing
development, wherever they are developed, should produce a fixed proportion of units
that are affordable for the low-income residents. In light of the fact that vouchers do not
seem to have achieved the desired dispersal in Pontiac, a holistic approach might include

a mix of all strategies, including fair share and scattered site public housing.

48

 

BIBLIOGRAPHY

(1974). Housing and Community Development Act.

Crowder, K. D. (2001). "Racial Stratification in the Actuation of Mobility Expectations:
Micro Level Impacts of Racially Restrictive Housing Markets." Social Forces 79(4).

Farley, R. and W. H. Frey (1994). "Changes in the Segregation of Whites from Blacks
during the 19803: Small Steps Toward a More Integrated Society." American
Sociological Review 59(1).

Galster, G. C. and S. P. Killen (1995). "The Geography of Metropolitan Opportunity: A
Reconnaissance and Conceptual Framework." Housing Policy Debate 6(1).

Hartung, J. M. and J. R. Henig (1997). "Housing Vouchers and Certificates as a Vehicle
for Deconcentrating the Poor: Evidence from the Washinton D. C., Metropolitan Area."
Urban Affairs Review 32(3).

HUD (2000). Section 8 Tenant-Based Housing Assistance: A Look Back After 30 Years,
U. S. Department of Housing and Urban Development.

Hughes, M. A. (1995). "A Mobility Strategy for Improving Opportunity." Housing Policy
Debate 6(1).

Kasardra, J. D. (1993). "Inner-City Concentrated Poverty and Neighborhood
Distress21970 to 1990." Housing Policy Debate 4(3).

Katz, L. F., J. R. Kling, et al. (2001). "Moving To Opportunity In Boston: Early Results
Of A Randomized Mobility Experiment." The Quarterly Journal of Economics 116(2).

Katz, L. F. and M. A. Turner (2001). "Who Should Run the Housing Voucher Program?
A Reform Proposal." Housing Policy Debate 12(2).

Orfield, M. (1996). Metropolitics: A Regional Agenda for Community and Stability.
Washington, DC, Brookings Institution Press.

Pendall, R. (2000). "Why Do Voucher and Certificate Users Live in Distressed
Neighborhoods.” Housing Policy Debate 11(4).

Rosenbaum, J. (1991). "Black Pioneers--Do Their Moves to the Suburbs Increase
Economic Opportunity for Mothers and Children?" Housing Policy Debate 2(4).

Rosenbaum, J. (1995). "Changing the Geography of Opportunity by Expanding
Residential Choice: Lessons from the Gautreaux Program." Housing Policy Debate 6(1).

49

 

Rosenbaum, E. and L. E. Harris (2001). "Residential Mobility and Opportunities: Early

Impacts of Moving to Opportunity Demonstration Program in Chicago." Housing Policy
Debate 12(2).

Rubinowitz, L. S. and J. E. Rosenbaum (2000). Crossing the Class and Color Lines, from
Public Housing to White Suburbia. Chicago, The University of Chicago Press.

Shroder, M. and A. Reiger (2000). "Vouchers Versus Production Revisited." Journal of
Housing Research 11(1).

Shroder, M. (2002). "Locational Constraint, Housing Counseling, and Successful Lease-
up in a Randomized Housing Voucher Experiment." Journal of Planning Literature 17(1).

Swope, C. (2002). Federal housing vouchers are supposed to break up concentrations of
poverty. Often they just create new ones. Governing Magazine (online edition). May
2002.

Turner, M. A. (1998). "Moving Out of Poverty: Expanding Mobility and Choice through
Tenant-Based Housing Assistance." Housing Policy Debate 9(2).

Varady, D. P. and C. C. Walker (2000). "Vouchering Out Distressed Subsidized
Development: Does Moving Lead to Improvements in Housing and Neighborhood
Conditions." Housing Policy Debate 11(1).

Varady, D. P., C. C. Walker, et al. (2001). "Voucher Recipient Achievement of Improved

Housing Conditions in the US: Do Moving Distance and Relocation Services Matter?"
Urban Studies 38(8).

50

 

M IIIIIIIIIIIIIIIIIIIIIIIIIIIIII
1

111111111!1111/111111111111111111r11 11,1111

1
1
3 1293 02493 077