Caspi and Rafkin (2024) study the effect of providing attorneys to tenants facing eviction in Memphis, Tennessee, in partnership with a nonprofit that represents tenants facing eviction called The Works, Inc. (TWI). Tenants could apply for legal assistance online, and eligibility was based on whether they had an eviction filing, the court date, and whether they had already been assigned a lawyer through a different program. Eligible tenants were randomly assigned to treatment and control groups, where treated tenants were assigned to receive an attorney through TWI or reject assistance. Data was collected using administrative court data, baseline surveys, and end-line surveys. The outcomes analyzed included tenant eviction judgment rates 180 days after eviction filing and informal outcomes, including on landlord/tenant bargaining, obtained through the endline surveys.
MVPF = 2.7
The net cost of the program is the administrative cost of hiring an attorney plus the fiscal externality on other government spending. TWI provides tenants with both internal and external attorneys. External attorneys were paid $325 per case, while the estimated direct cost of an internal attorney was $250. The paper uses the average, $287.50, in the MVPF calculation.
The paper considers four fiscal externalities: the effects on the government budget of emergency shelter use, hospital visits, earnings, and writs. The median length of an emergency shelter visit is one month, and the average shelter visit cost $2,100 in 2006 dollars, according to Hao et al. (2022). Collinson et al. (2023) find that eviction leads to a 3.4 percentage point increase in the likelihood of using a shelter. Thus, an eviction costs the government $103.65 through shelter use. Scaled by the IV estimate from Caspi and Rafkin (2024) on the effects of counsel on eviction over 180 days, the provision of an attorney reduces government costs by $8.29.
For hospital visits, the average emergency department visit is $530 in 2017 dollars (Moore and Liang (2020)), and an eviction increases visits by 0.188 (Collinson et al. 2023). Scaling by the same IV estimate as for shelter use, this works out to a decrease in government costs of $9.52.
Collinson et al. (2023) also find that evictions reduce earnings over two years by $936. Using a 12.9% tax and transfer rate for low-income populations, the reduction in government costs due to not losing the associated tax revenue becomes $9.66.
Finally, King County (where comparable data was available) estimates that fulfilling a Writ where the sheriff evicts a tenant costs no more than $135. The paper finds that half of the eviction judgments result in writs and that attorney provision reduces the rate of writs by 0.01 percentage points. Thus, the government saves $0.07 per attorney provided.
The net cost to the government is $287.50 – $8.29 – $9.52 – $9.66 -$0.07 = $260.
The paper estimates willingness to pay by having respondents participate in an incentivized multiple price list exercise, choosing between attorney representation and different cash amounts. Some of the choices that participants made were randomly selected to be implemented in this exercise, making their responses consequential. The average respondent indicated they would be willing to pay $691 for attorney representation.
The paper validates this high willingness to pay in a few ways. Attention checks and valuations for placebo goods reject that participant inattention drives the estimated willingness to pay. Incentivized belief elicitations show that the valuation is not driven by optimistic predictions about the effectiveness of attorneys. In fact, tenants report valuing attorneys for their assistance during the eviction process by reducing stress and uncertainty, not because they think the attorneys will help them win in court.
Using the survey-elicited willingness to pay $691 and the net cost to the government of $260 yields an MVPF of $691/$260 = 2.7.
The paper also estimates an alternative MVPF of 0.4 using an alternative willingness to pay estimate calculated from earnings effects documented in Collinson et al. (2023). The higher MVPF using survey responses suggests that tenants value attorney assistance beyond its impact on earnings, perhaps due to reduced stress, better housing outcomes, or other non-monetary benefits.
Caspi, Aviv and Charlie Rafkin (2024). “Legal Assistance for Evictions: Impacts, Mechanisms, and Demand.” Working Paper. https://www.charlierafkin.com/docs/cr_rtc.pdf
Collinson, Robert, John Eric Humphries, Nicholas Mader, Devin Reed, Daniel Tannenbaum, and Winnie van Dijk (2023). “Eviction and Poverty in American Cities.” Quarterly Journal of Economics, 139(1):57-120. https://academic.oup.com/qje/article/139/1/57/7276608
Hao, Haijing, Monica Garfield, and Sandeep Purao (2022). “The Determinants of Length of Homeless Shelter Stays: Evidence-Based Regression Analyses.” International Journal of Public Health, 66. https://www.ssph-journal.org/journals/international-journal-of-public-health/articles/10.3389/ijph.2021.1604273/full
Moore, Brian J. and Lan Liang (2020). “Costs of Emergency Department Visits in the United States, 2017.” Healthcare Cost and Utilization Project (HCUP) Statistical Briefs. https://www.ncbi.nlm.nih.gov/books/NBK566654/