Title: The Effects of Public LowIncome Housing Vouchers on Social and Labor Market Outcomes
1The Effects of Public Low-Income Housing Vouchers
on Social and Labor Market Outcomes
- Deven Carlson
- Robert Haveman
- Tom Kaplan
- Barbara Wolfe
- Institute for Research on Poverty
- University of Wisconsin-Madison
- November 21, 2008
2Presentation Overview
- Prior Research Results
- Research question and motivation
- What effect does the receipt of a low-income
housing voucher have on social and labor market
outcomes? - Theoretical link between voucher receipt and
outcomes - Data sources and methodology
- Major conclusions
- Positive effect on neighborhood quality in the
long term - Short-term changes in household composition, but
greater subsequent stability - Greater use of TANF and state-subsidized child
care - Short-term negative effect on earnings diminishes
over time - Mixed effects on work effort
- Results vary by demographic subgroup
3Previous Studies
- Experimental
- Mixed Results
- Gautreaux Program- Chicago
- Moving to Opportunity- 5 Cities
- Welfare to Work- 6 Cities
- Chicago Housing Authority Natural Experiment
- Minnesota Family Investment Plan
- Nonexperimental
- Also mixed results
- Bania, Coulton, and Leete (2003)
- Harkness and Newman (2003 2006)
4Our Research Approach
- Effect of a housing voucher on social and labor
market outcomes for low-income families in
Wisconsin - Includes both urban and rural areas
- Large sample sizes allow us to examine several
demographic subgroups - Pool multiple years and follow recipients over a
longer time period
5What is the Section 8 Program? How does it work?
- Primary objective of program is to enable very
low-income families to choose and lease safe,
decent, and affordable privately-owned rental
housing. - Section 8 vouchers currently serve about 1.9
million families nationally (more than 850,000
families with minor children). - Recipients must have income below 50 percent of
area median income. - If awarded a voucher, recipients choose available
private rental housing and, if they find it,
contribute 30 percent of their income toward
rent. - The program then pays the difference between the
contribution and actual rent (up to a locally
defined fair market rent).
6Outcomes of Interest
- Social Outcomes
- Neighborhood quality
- Four measures
- Household composition changes
- Six measures of household composition changes
- Public program participation
- State-subsidized child care- Wisconsin Shares
- TANF-Wisconsin Works
- Labor Market Outcomes
- Earnings
- Employment
7Theoretical link between voucher receipt and
outcomes
- Vouchers stimulate mobility
- Opportunity for re-evaluation
- Disruption
- Economic theory
- Income and substitution effects
8Theoretical link between voucher receipt and
outcomes
- Social Outcomes
- Move to a better neighborhood
- Change structure of household
- Ambiguous effect on public program participation
- Labor Market Outcomes
- Short-term disruption in employment
- Long-term moves to areas with better employment
opportunities
9Data and Estimation Sample
- Data
- Wisconsin administrative data supplemented with
U.S. Census data - Sample
- All cases applying for or receiving Food Stamps
between 2000 and 2003 - Identified two groups voucher recipients and
nonrecipients - Calendar year cohorts
- Pooled sample
10Estimation Method
- Propensity score matching
- Estimate probability of rental subsidy receipt
- Rich set of covariates
- Match voucher recipients to members in the
control group - Nearest neighbor matching method
- Matching procedure succeeds in eliminating bias
on all observed covariates - Labor market, Neighborhood, and Household
Composition-Mean comparison - Public Program Participation- Regression framework
11Results
12Neighborhood Characteristic Results
Where applicable, t-stat in parentheses below
point estimate
13Case Composition Results
Where applicable, t-stat in parentheses below
point estimate
14Child Care Results- Eligible Cases
Estimates in bold indicate significance at plt.05
level
15TANF/Wisconsin Works Results- Full Sample
Estimates in bold indicate significance at plt.05
level
16Results
- Labor Market Outcome Results
17Work Effort and Earnings Results- Full Sample
Where applicable, t-stat in parentheses below
point estimate
18The Annual Earnings Pattern Pooled Cohorts
19Results
20Subgroup Results
- Patterns we see in full sample are generally
present in subgroups as well - Case composition
- Adult loss in base year especially prevalent
among young, female, rural, single parents - Child care participation
- Effect of voucher greatest among females,
Hispanics, and Milwaukee residents - TANF participation
- Effect of voucher greatest among females, young,
urban areas, and families with children
21Earnings Results- Selected Subgroups
Where applicable, t-stat in parentheses below
point estimate
22Earnings Results- Selected Subgroups
Where applicable, t-stat in parentheses below
point estimate
23Conclusions
- Vouchers provide opportunity for re-evaluation
- Living situation
- Employment
- Public benefits
- Also can cause short-term disruptions
- Effects appear to vary by demographic subgroup
- Future Work
- Additional programs
- Additional estimation strategies