Family Homelessness Prevention: A Randomized Control Trial

Last registered on May 17, 2023


Trial Information

General Information

Family Homelessness Prevention: A Randomized Control Trial
Initial registration date
May 15, 2023

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
May 17, 2023, 3:01 PM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.


There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Primary Investigator

University of Notre Dame

Other Primary Investigator(s)

PI Affiliation
Oregon State University

Additional Trial Information

On going
Start date
End date
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Mary’s Place is a nonprofit organization in Seattle that provides services to women and families with current and previous experience of homelessness with the goal of ensuring that no child sleeps outside. They are launching a new homelessness prevention service informed by Santa Clara County’s homelessness prevention program. Because the past policy focus has primarily been on treatment, there is little evidence about which prevention programs work best and for whom. Our research project attempts to address this gap in the research by comparing the efficacy and cost effectiveness of prevention as a whole as well as comparing two different prevention strategies within the sample population. First, when services are oversubscribed due to budget and staffing limits, Mary’s Place will use random assignment to determine who is served first. Comparing people higher and lower on the waitlist will measure the effectiveness of receiving assistance. Second, we will use a lottery among program participants to measure the effectiveness of (1) intensive homelessness prevention services with a large cap on financial assistance and access to repeat assistance relative to (2) a traditional model of a smaller-capped, one-time financial assistance. Thus, the primary difference between the services provided to the two groups will be the intensity of financial assistance.

The results of this study will be informative to policymakers and service providers in other communities that are interested in the most effective means of homelessness prevention. Enrollment began in May of 2023 and will continue until we have enrolled approximately 3,708 households. Researchers had not received data on outcomes at the time of filing of this pre-analysis plan. Given the potentially wide-ranging effect of homelessness prevention, we identified three outcome domains for study: (a) receipt of services, (b) housing stability, and (c) family and child well-being.
External Link(s)

Registration Citation

Bollman, Katie and David Phillips. 2023. "Family Homelessness Prevention: A Randomized Control Trial." AEA RCT Registry. May 17.
Experimental Details


Mary’s Place is launching a new homelessness prevention service informed by Santa Clara County’s homelessness prevention program evaluated in a previous LEO study. In this evaluation, LEO and Mary’s Place plan to study whether providing families intensive homelessness prevention services with no cap on financial assistance results in different outcomes compared to a traditional model of capped, one-time financial assistance.
Intervention Start Date
Intervention End Date

Primary Outcomes

Primary Outcomes (end points)
Housing stability
Primary Outcomes (explanation)
Our primary outcome of interest is housing stability, which we will measure with an indicator for whether an applicant requested or received homelessness services. Specifically, this indicator takes the value of 1 if an applicant requests assistance from the Mary’s Place family shelter call center and indicates themselves as homeless, or if they receive services from the Emergency Shelter, Street Outreach, or Diversion programs. We will construct this measure using Mary’s Place internal data.

Secondary Outcomes

Secondary Outcomes (end points)
Exposure to violence: Any 911 call at their address related to domestic violence or any Medicaid claim for healthcare to treat injuries from violence
Child welfare: Any record of contact with child protective services or foster care
Evictions: Eviction filing, eviction judgment against
Use or request of homeless services from other agencies: Enrolling in Emergency Shelter, Street Outreach, Permanent Supportive Housing, Rapid Re-Housing, Transitional Housing, or Diversion; Completing Coordinated Entry assessment
Housing changes: Address changes, new address starts, existing address ends, neighborhood characteristics for most recent address
Employment: hours worked, total earnings, employed
Receipt of benefits: receipt of cash benefits; receipt of food benefits
Credit: credit score, total debt balance, delinquency, total inquiries
Healthcare use: total healthcare visits; visits by type (inpatient, outpatient, emergency); mental health treatment; mental health prescription fills
Overall criminal justice system contact: any arrest; number of arrests; arrests by class and type of charge
Schooling outcomes: attendance, disciplinary outcomes, and test scores
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
This study will have a two-staged random assignment. First, eligible applicants will be randomly assigned to receive or not receive assistance. Second, people selected for assistance will be randomly assigned to receive one-time or intensive assistance. We will conduct two sets of analysis. . First, we will measure treatment effects of receiving any assistance across program pathways, we will use data from all applicants who enter the waitlist who were deemed eligible at the time of random assignment. Second, we will measure the relative effectiveness of the two treatment pathways. The sample for comparisons between treatment pathways will be restricted to those applicants who are selected from the waitlist, determined to be eligible, and establish an appointment with a prevention specialist. When analyzing outcomes external to Mary’s Place, we will also limit the sample to those who provide consent for us to obtain and use their external data.
Experimental Design Details
Not available
Randomization Method
There will be two stages of randomization which are done using the Mary’s Place internal database Agency. The first stage will take place online in the services application form on the Mary’s Place website. Those identified as having worked with Mary’s Place in the past or having been referred will be added to a waitlist for the one-time or the intensive prevention services group. Every week, Mary’s Place will randomly select part of each group to be assessed for eligibility.
Randomization Unit
Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters
3708 households
Sample size: planned number of observations
3708 households
Sample size (or number of clusters) by treatment arms
Approximately 1236 households will be assigned to the waitlist for intensive homelessness prevention services and 2472 households to the waitlist for the traditional model of a smaller-capped, one-time financial assistance. After three years of enrollment, we anticipate that we will have about 1,350 households served with one-third receiving intensive prevention services and two-thirds receiving one-time assistance.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

Institutional Review Boards (IRBs)

IRB Name
The University of Notre Dame Institutional Review Board
IRB Approval Date
IRB Approval Number
Analysis Plan

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information