Evaluating California’s Guaranteed Income Pilot Program

This RCT focuses on two unique low-income populations facing a key life transition that puts them at higher risk for adverse outcomes: pregnant people and youth aging out of extended foster care. Participants in the GI treatment group will receive unconditional cash payments of $600-$1,200 per month for 12-18 months (duration and amount of payment vary by site). GI recipients also receive benefits waivers for some benefits, ensuring that the income they receive from GI does not impact their current benefits. Benefits counseling and other support services are offered to enrolled participants, but participation in these services is voluntary. All participants receive a link for a self-guided benefits assessment that helps the participant determine any additional benefits for which they may be eligible and any benefits that may be impacted by the receipt of GI. Some sites also follow up with in-person benefits counseling for those who express interest.

Low-income people, particularly those who are pregnant or leaving foster care, are at higher risk of experiencing a range of negative outcomes (e.g., housing instability, food insecurity, child maltreatment). Many experience financial instability, and existing benefits programs require a high administrative burden and do not adequately meet people’s needs – all of which lead to stress, material hardship, and difficulty accessing needed services. Therefore, providing low-income people with a guaranteed income and benefits waivers, should increase financial stability, reduce administrative burden, and ultimately reduce their risk of negative outcomes.

2023-11-21

2026-04-30

The primary outcomes of interest for the FFY population are:
• Healthcare utilization and access
o Delay or avoid healthcare
o Obtain routine healthcare
• Mental health
o Depression
o Anxiety
o Perceived stress
• Financial strain and well-being
o Financial stress
o Financial well-being
o Household income
• Housing security
o Episode(s) of homelessness
o Worry about losing housing
• Food insecurity
• Employment
o Employment status
o Employment conditions
• Education
o School enrollment
• Participation in benefits
o Lose any benefits
o Gain any benefits

The primary outcomes of interest for the pregnant population are:
• Healthcare utilization and access
o Delay or avoid healthcare
o Obtain prenatal healthcare
• Mental health
o Depression
o Anxiety
o Perceived stress
• Financial strain and well-being
o Financial stress
o Financial well-being
o Household income
• Housing security
o Episode(s) of homelessness
o Worry about losing housing
• Food insecurity
• Child welfare involvement
o Maltreatment referrals
o Investigations
o Substantiations
o Case openings
o Foster care entries
• Employment
o Employment conditions
• Participation in benefits
o Lose any benefits
o Gain any benefits

The primary outcomes for FFY will be measured as follows:
• Healthcare utilization and access
o Delay or avoid healthcare – This outcome will examine whether participants in the treatment group were less likely to delay or avoid healthcare due to cost or other barriers compared to participants in the control group. This outcome will be measured through participants’ responses to survey items drawn from the BRFSS (Pierannunzi, Hu, & Balluz, 2013). We will also measure change over time within individuals.
o Obtain routine healthcare – This outcome will examine whether participants in the treatment group were more likely to obtain routine healthcare compared to participants in the control group. This outcome will be measured through participants’ responses to survey items about obtaining healthcare and also through MediCal claims data. We will also measure change over time within individuals.
• Mental health
o Depression – This outcome will examine whether participants in the treatment group were less likely to experience depression compared to participants in the control group. This outcome will be measured through participants’ responses to the PHQ-8 (Kroenke et al., 2009) on the survey. We will also measure change over time within individuals.
o Anxiety – This outcome will examine whether participants in the treatment group were less likely to experience anxiety compared to participants in the control group. This outcome will be measured through participants’ responses to the GAD-7 (Spitzer et al., 2006) on the survey. We will also measure change over time within individuals.
o Perceived stress – This outcome will examine whether participants in the treatment group were less likely to experience stress compared to participants in the control group. This outcome will be measured through participants’ responses to the PSS-4 (Karam et al., 2012) on the survey. We will also measure change over time within individuals.
• Financial strain and well-being
o Financial stress – This outcome will examine whether participants in the treatment group were less likely to experience financial stress compared to participants in the control group. This outcome will be measured through participants’ responses to the PRAMS Financial Stress instrument (https://www.cdc.gov/prams/index.htm) on the survey, specifically the question about how often they were stressed about having enough money to pay their bills. We will also measure change over time within individuals.
o Financial well-being – This outcome will examine whether participants in the treatment group were more likely to experience financial well-being compared to participants in the control group. This outcome will be measured through participants’ responses to the CFPB Financial Wellbeing instrument (https://files.consumerfinance.gov/f/201512_cfpb_financial-well-being-user-guide-scale.pdf) on the survey. We will also measure change over time within individuals.
o Household income – This outcome will examine whether participants in the treatment group were more likely to increase their household income (beyond GI) compared to participants in the control group. This outcome will be measured through participants’ responses to income questions on the application and survey, as well as administrative data from the Employment Development Department (EDD). We will also measure change over time within individuals.
• Housing security
o Episode(s) of homelessness – This outcome will examine whether participants in the treatment group were less likely to experience homelessness compared to participants in the control group. This outcome will be measured through participants’ responses to survey items about episodes of homelessness. We will also measure change over time within individuals.
o Worry about losing housing – This outcome will examine whether participants in the treatment group were less likely to worry about losing their housing compared to participants in the control group. This outcome will be measured through participants’ responses to survey items about being worried about losing their housing. We will also measure change over time within individuals.
o Housing mobility – This outcome will examine whether participants in the treatment group moved more than participants in the control group. This outcome will be measured through participants’ responses to survey items about their living situation changes.
o Housing satisfaction – This outcome will examine whether participants in the treatment group had greater housing satisfaction than participants in the control group. This outcome will be measured through participants’ responses to survey items about their satisfaction with their living situation.
• Food insecurity – This outcome will examine whether participants in the treatment group were less likely to experience food insecurity compared to participants in the control group. This outcome will be measured through participants’ responses to the USDA short form (Blumberg et al., 1999) on the survey. We will also measure change over time within individuals.
• Employment
o Employment status – This outcome will examine whether participants in the treatment group were more likely to be employed (either full-time or part-time) compared to participants in the control group. This outcome will be measured through participants’ responses to survey items about their employment status, as well as EDD administrative data. We will also measure change over time within individuals.
o Employment conditions – This outcome will examine whether participants in the treatment group were more likely to improve their employment compared to participants in the control group. This outcome will be measured through participants’ responses to survey items about whether they increased their employment benefits, increased their income, and the number of jobs reported. We will also measure change over time within individuals.
o Employment mobility – This outcome will examine whether participants in the treatment group changed jobs more than participants in the control group. This outcome will be measured through participants’ responses to survey items about their job changes.
• Education
o School enrollment – This outcome will examine whether participants in the treatment group were more likely to be enrolled in school or a training program compared to participants in the control group. This outcome will be measured through participants’ responses to survey items about their school enrollment status, as well as administrative data from the National Student Clearinghouse. We will also measure change over time within individuals.
• Participation in benefits
o Lose any benefits – This outcome will examine whether participants in the treatment group were less likely to lose any benefits compared to participants in the control group. This outcome will be measured through administrative benefits data (CalWORKs, CalFresh, MediCal, etc.).
o Gain any benefits – This outcome will examine whether participants in the treatment group were more likely to gain any new benefits compared to participants in the control group. This outcome will be measured through administrative benefits data (CalWORKs, CalFresh, MediCal, etc.).

The primary outcomes of interest for the pregnant population will be measured as follows:
• Healthcare utilization and access
o Delay or avoid healthcare – same definition as FFY
o Obtain prenatal healthcare – This outcome will examine whether participants in the treatment group were more likely to obtain prenatal healthcare compared to participants in the control group. This outcome will be measured through participants’ responses to survey items about obtaining prenatal care, as well as through MediCal claims data and vital records data. We will also measure change over time within individuals.
• Mental health
o Depression – same definition as FFY
o Anxiety – same definition as FFY
o Perceived stress – same definition as FFY
• Financial strain and well-being
o Financial stress – same definition as FFY
o Financial well-being – same definition as FFY
o Household income – same definition as FFY
• Housing security
o Episode(s) of homelessness – same definition as FFY
o Worry about losing housing – same definition as FFY
o Housing mobility – same definition as FFY
o Housing satisfaction – same definition as FFY
• Food insecurity – same definition as FFY
• Child welfare involvement
o Maltreatment referrals – This outcome will examine whether participants in the treatment group were less likely to be referred (i.e., reported) to the child welfare system for alleged maltreatment compared to participants in the control group. This outcome will be measured through child welfare administrative data.
o Investigations – This outcome will examine whether participants in the treatment group were less likely to be investigated for alleged maltreatment compared to participants in the control group. This outcome will be measured through child welfare administrative data.
o Substantiations – This outcome will examine whether participants in the treatment group were less likely to have a substantiated (i.e., confirmed) maltreatment referral compared to participants in the control group. This outcome will be measured through child welfare administrative data.
o Case openings – This outcome will examine whether participants in the treatment group were less likely to have a child welfare case opened compared to participants in the control group. This outcome will be measured through child welfare administrative data.
• Employment
o Employment conditions – same definition as FFY
• Participation in benefits
o Lose any benefits – same definition as FFY
o Gain any benefits – same definition as FFY

The secondary outcomes of interest for the FFY population are:
• Health
o Healthcare utilization and access: Medical debt, Hospitalizations, Chronic health conditions
o Perceived health
o Health-related quality of life
o Physical activity
o Sleep quality
o Diet
• Well-being
o Well-being
o Hope
• Financial strain and well-being
o Debt
o Unexpected household shocks
• Housing security
o Satisfaction with living situation
o Improve living situation
• Transportation security
• Child welfare involvement (for children of FFY)
o Maltreatment referrals
o Investigations
o Substantiations
o Case openings
o Foster care entries
• Employment
o Satisfaction with employment
• Education
o Achieve educational goals
• Criminal justice involvement
o Arrests
o Convictions
o Incarcerations
The secondary outcomes of interest for the pregnant population are:
• Health
o Healthcare utilization and access: Obtain routine healthcare, Medical debt, Hospitalizations
o Perceived health
o Chronic health conditions
o Health-related quality of life
o Physical activity
o Sleep quality
o Diet
• Well-being
o Well-being
o Hope
• Financial strain and well-being
o Debt
o Unexpected household shocks
• Housing security
o Satisfaction with living situation
o Improve living situation
• Transportation security
• Employment
o Employment status
o Satisfaction with employment
o Employment mobility
• Education
o School enrollment
o Achieve educational goals
• Perinatal health conditions
o Gestational hypertension
o Gestational diabetes
• Birth outcomes
o Pre-term birth
o Birth weight
o NICU admission
o NICU length of stay

The secondary outcomes for the FFY population will be measured as follows:
• Health
o Healthcare utilization and access
 Medical debt – This outcome will examine whether participants in the treatment group were less likely to have medical debt compared to participants in the control group. This outcome will be measured through participants’ responses to survey items about their medical debt. We will also measure change over time within individuals.
 Hospitalizations – This outcome will examine whether participants in the treatment group were less likely to be hospitalized compared to participants in the control group. This outcome will be measured through administrative data from the Department of Healthcare Access and Information (HCAI). We will also measure change over time within individuals.
o Perceived health – This outcome will examine whether participants in the treatment group were more likely to perceive themselves to be in good health compared to participants in the control group. This outcome will be measured through participants’ responses to survey items drawn from the BRFSS (Pierannunzi, Hu, & Balluz, 2013). We will also measure change over time within individuals.
o Chronic health conditions – This outcome will examine whether participants in the treatment group were less likely to be diagnosed with a new chronic health condition compared to participants in the control group. This outcome will be measured through participants’ responses to the survey items about chronic health conditions and through MediCal claims.
o Health-related quality of life – This outcome will examine whether participants in the treatment group reported having a better health-related quality of life compared to participants in the control group. This outcome will be measured through participants’ responses to survey items about their health-related quality of life (Moriarty, Zack, & Kobau, 2003).
o Physical activity – This outcome will examine whether participants in the treatment group report more physical activity than participants in the control group. This outcome will be measured through participants’ responses to survey items about their physical activity adapted from the BRFSS (Pierannunzi, Hu, & Balluz, 2013).
o Sleep quality – This outcome will examine whether participants in the treatment group report better sleep quality than participants in the control group. This outcome will be measured through participants’ responses to survey items about their sleep quality.
o Diet – This outcome will examine whether treatment participants report a better diet compared to control participants. This outcome will be measured through participants’ responses to survey items about their diet.
• Well-being
o Well-being – This outcome will examine whether participants in the treatment group were more likely to report improved wellbeing compared to participants in the control group. This outcome will be measured through participants’ responses to the Cantril Wellbeing Scale (Levin & Currie, 2014) on the survey. We will also measure change over time within individuals.
o Hope – This outcome will examine whether participants in the treatment group report more hope than participants in the control group. This outcome will be measured through participants’ responses to the Herth Hope Index on the survey (Herth, 1992).
• Financial strain and well-being
o Debt – This outcome will examine whether participants in the treatment group were more likely to decrease their debt compared to participants in the control group. This outcome will be measured through participants’ responses to survey questions about debt.
o Unexpected household shocks – This outcome will examine whether participants in the treatment group are better able to financially handle an unexpected household shock than participants in the control group. This outcome will be measured through participants’ responses to survey questions about unexpected household shocks.
• Housing security
o Satisfaction with living situation – This outcome will examine whether participants in the treatment group were more likely to report satisfaction with their living situation compared to participants in the control group. This outcome will be measured through participants’ responses to survey items about their living situation. We will also measure change over time within individuals.
o Improve living situation – This outcome will examine whether participants in the treatment group were more likely to improve their living situation compared (e.g., safety, quality, satisfaction) to participants in the control group. This outcome will be measured through participants’ responses to survey items about their living situation. We will also measure change over time within individuals.
• Transportation security – This outcome will examine whether participants in the treatment group were more likely to have adequate transportation compared to participants in the control group. This outcome will be measured through participants’ responses to the Transportation Security Index Short Form (Murphy, Gould-Werth, & Griffin, 2021) on the survey. We will also measure change over time within individuals.
• Child welfare involvement (for children of FFY)
o Defined the same as for the Pregnant Population primary outcome.
• Employment
o Satisfaction with employment –This outcome will examine whether participants in the treatment group were more likely to be satisfied with their employment compared to participants in the control group. This outcome will be measured through participants’ responses to survey items about their employment satisfaction. We will also measure change over time within individuals.
• Education
o Achieve educational goals – This outcome will examine whether participants in the treatment group were more likely to achieve their educational goals compared to participants in the control group. This outcome will be measured through participants’ responses to survey items about their education goals and status.
• Criminal justice involvement
o Arrests – This outcome will examine whether participants in the treatment group were less likely to be arrested compared to participants in the control group. This outcome will be measured through Department of Justice (DOJ) administrative data.
o Convictions – This outcome will examine whether participants in the treatment group were less likely to be convicted of a crime compared to participants in the control group. This outcome will be measured through DOJ administrative data.
o Incarcerations – This outcome will examine whether participants in the treatment group were less likely to be incarcerated compared to participants in the control group. This outcome will be measured through DOJ administrative data.
The secondary outcomes of interest for the pregnant population will be measured as follows:
• Health
o Healthcare utilization and access
 Obtain routine healthcare – same definition as FFY, see primary outcomes explanation above
 Medical debt – same definition as FFY
 Hospitalizations – same definition as FFY
 Chronic health conditions – same definition as FFY
o Perceived health – same definition as FFY
o Health-related quality of life – same definition as FFY
o Physical activity – same definition as FFY
o Sleep quality – same definition as FFY
o Diet – same definition as FFY
• Well-being
o Well-being – same definition as FFY
o Hope – same definition as FFY
• Financial strain and well-being
o Debt – same definition as FFY
o Unexpected household shocks – same definition as FFY
• Housing security
o Satisfaction with living situation – same definition as FFY
o Improve living situation – same definition as FFY
• Transportation security – same definition as FFY
• Employment
o Employment status – same definition as FFY, see primary outcomes explanation above
o Satisfaction with employment – same definition as FFY
• Education
o School enrollment – same definition as FFY, see primary outcomes explanation above
o Achieve educational goals – same definition as FFY
• Perinatal health conditions
o Gestational hypertension – This outcome will examine whether participants in the treatment group were less likely to develop hypertension during pregnancy compared to participants in the control group. This outcome will be measured through participants’ responses to survey items about perinatal health conditions, as well as MediCal claims data.
o Gestational diabetes – This outcome will examine whether participants in the treatment group were less likely to develop gestational diabetes compared to participants in the control group. This outcome will be measured through participants’ responses to survey items about perinatal health conditions, as well as MediCal claims data.
• Birth outcomes
o Pre-term birth – This outcome will examine whether participants in the treatment group were less likely to have a pre-term birth (i.e., birth prior to 37 weeks of gestation) compared to participants in the control group. This outcome will be measured through participants’ responses to survey items about their birth experiences, as well as MediCal claims data.
o Birth weight – This outcome will examine whether participants in the treatment group were less likely to have a baby with low birth weight compared to participants in the control group. This outcome will be measured through MediCal claims data.
o NICU admission – This outcome will examine whether participants in the treatment group were less likely to have a baby with a NICU admission compared to participants in the control group. This outcome will be measured through MediCal claims data.
o NICU length of stay – This outcome will examine whether treatment participants whose baby had a NICU admission had a shorter length of stay compared to participants in the control group whose baby had a NICU admission. This outcome will be measured through MediCal claims data.

The GI evaluation employs experimental methods, i.e. a randomized controlled trial (RCT). The eligible population for GI includes youth aging out of extended foster care at age 21 during the study enrollment period and pregnant individuals who are within the first 27 weeks of pregnancy at the time of application. Each of the five pilot sites in the RCT identify, screen, and enroll participants. Eligible participants who choose to enroll are randomized to either the treatment group which receives unconditional cash payments of $600-$1,200 per month for 12-18 months (duration and amount of payment vary by site) or the control group which receives $20 per month. One site is not using control group payments. We anticipate enrolling a total of 3,021 participants in the study. This includes 774 FFY (395 treatment, 379 control) and 2,247 pregnant people (1,156 treatment and 1,091 control).

Not available

We use random assignment to determine which eligible applicants will receive GI. For all but one site, the randomization process is as follows:

Enrollee data required to conduct randomization is transferred from the online application platform to the Urban Institute on a regular basis. Urban then conducts randomization using STATA. All randomization is stratified first by pilot grantee and then by target population (for pilots serving both the FFY and pregnant populations). Some sites are also stratified by county.

Each batch is randomized until the point is reached that there are more applicants than remaining slots. Each site has specified its own treatment to control ratio based on funding and eligible universe. Urban then transfers randomization results back to pilot sites, and the pilot site staff contacts applicants to inform them of their group assignment.

Once there are more applicants than remaining slots, excess applicants will be held on a waitlist. In the event a treatment group participant declines GI, two waitlist applicants will be randomly selected to treatment and control.

For one of the pregnant population sites, Urban does not conduct the randomization, and the site runs the randomization process directly. Every two weeks, the site randomly selects 40% of new applicants from each county to be in the treatment group, and those that are no longer eligible are assigned to the control group. The individuals included in each randomization process include all newly eligible applicants as well as previous applicants that did not get selected for GI but are still eligible (<27 weeks gestation). This maintains a treatment: control group ratio of 1:1.5. Randomization is conducted using numbers randomly generated from a uniform distribution.

Individual

No

We anticipate enrolling a total of 3,021 participants in the study (774 FFY and 2,247 pregnant people).

We anticipate enrolling a total of 3,021 participants in the study (774 FFY and 2,247 pregnant people).

We anticipate enrolling 774 FFY (395 treatment, 379 control) and 2,247 pregnant people (1,156 treatment and 1,091 control).