Phillips and Sullivan (2023)

The economics of Housing & Homelessness

With Reference to

Evans et al. (2016)

Overview

"The primary concern is that most individuals and families at risk of losing their housing do not end up homeless, even if they do not receive financial assistance"

Motivation

"A study in Chicago, for example, found that among eligible people who sought financial assistance to prevent homelessness but were denied, only 2 percent entered an emergency shelter within 6 months (Evans et al., 2016).

"To-date there is no direct experimental evidence of the impact of such financial assistance"

Geography: Santa Clara, CA

Design

Clients who are on the margin of being eligible for assistance and who cannot demonstrate an ability pay future rent in a county with high and rapidly rising rents.

Results

Emergency Shelter Usage

2.5pp

Evans et al. (2016)

Funding Rate

Non-veteran tenants seeking one-months rent ($300-$900) who don't have a rental subsidy and whose income is less than twice the poverty line and have a social security number

Baseline Differences

Treatment Effects

Dynamic Treatment Effects

Background

  • It might not provide adequate support to many people
  • It might provide financial assistance for people who are unlikely to become homeless without the support

Why might homeless prevention programs which provide financial assistance not be effective?

\$4442

Average Financial Assistance

Financial Assistance

  • Rent, Security Deposits, Utility Bills
  • Tenants who are housed, behind on rent, and at imminent risk of being removed
  • Target tenants who cannot demonstrate an ability to pay after the financial assistance ends

Sample Eligibility

0
8

Not Eligible 

for Support

Eligible 

for Support

Assessment Score

13

Entitled to Financial Assistance

  • Exclude tenants who are expected to receive support if assigned to the control group (Potential Ethical Concerns?)

Analysis

Which is the most interesting column-column comparison?

Y_i = \alpha _i + \beta Z_i + \phi _{am} + \varepsilon_i

Temporary Offer of Financial Assistance

Agency & Month Fixed Effects

Formally Homeless

Regression Specification

First Stage

LATE

Treatment Heterogeneity

Big Picture Takeaways

  • Evans et al. (2016) indicates that temporary financial assistance for those with documented income decreases the takeup of homelessness services
  • The interesting aspect of this paper is that it studies tenants without documented future income

"The client must be able to demonstrate self-sufficiency after they receive assistance; (ii) the client must have an eligible crisis (e.g., job loss or medical emergency) that has led to the need for assistance; (iii) the client must face imminent risk of homelessness or utility shutoff; and (iv) the current crisis must be solvable by the financial assistance."

Presentation of Phillips and Sullivan et al. (2023)

By Patrick Power

Presentation of Phillips and Sullivan et al. (2023)

  • 162