Something that worked somewhere else

If you are wondering whether any of these programs actually work, here is the short version. A few have been tested properly, most of those found nothing, and the ones that found something moved it a little. Not one of them got anybody's income above about $5,000 a year.

That is the whole honest list. Below it is, with the numbers as the researchers wrote them, and links so you can check us.

One note on the numbers. When you see "p=0.005" it means the odds of seeing a gap that big by pure luck are about 5 in 1,000. Small p, probably real. Big p, probably noise.

Tennessee: a caseworker who keeps showing up

Between October 2010 and October 2012, 1,322 people ages 18 to 24 in Tennessee, who had left foster care or juvenile justice custody as teenagers or were aging out at 18, were split at random into two groups. One group was offered intensive one-on-one case management, support and counseling. The other was not. Random assignment is what makes this worth your time: the two groups started out the same, so whatever gap shows up at the end came from the program.

Housing got steadier in year one. Housing instability was scored 0 to 4, counting how many of these you hit in the past year: homeless, couch surfing, unable to pay rent, lost housing because you could not pay. Program group 0.98. Control group 1.20. A gap of 0.22, p=0.005.

Money is more complicated. Asked directly, the program group reported $611 more in year one, $4,099 against $3,488 (p=0.043). State wage records did not back that up on averages: over two years, $5,240 against $5,016, p=0.555. But the same wage records did show more of the program group clearing low bars. Earning $2,500 or more: 5.1 points higher in year one (p=0.035), 6.3 points higher in year two (p=0.016). The average did not move. The bottom of the distribution did.

Nothing moved in education, social support, or arrests. Just over half of both groups had a diploma at one year. About a quarter of both enrolled in college at some point in two years. Just under half of both were arrested. All of it is in the MDRC summary report (Tables 1 and 2). The program is now called YVLifeSet.

Massachusetts: the one that moved college

Four federally funded independent living programs have been tested with random assignment: employment services in Kern County, California; case management and mentoring in Massachusetts; tutoring and mentoring in Los Angeles County; and classroom life skills training, also in Los Angeles County. Three of the four found nothing.

The one that found something was the Massachusetts Adolescent Outreach Program. 194 young people in intensive foster care, assigned at random, followed for two years. Using National Student Clearinghouse enrollment records, 48.9 percent of the Outreach group enrolled in college across more than one academic year, against 30.8 percent of the control group. The final report is on the ACF site (direct PDF).

Staying in care until 21

Chapin Hall's Midwest Study followed more than 700 young people in Iowa, Wisconsin and Illinois. Illinois was the only one of the three where you could stay in foster care to 21.

At 19, 12.2 percent of the Iowa and Wisconsin group had been homeless, against 4.5 percent in Illinois. At 21, 20.8 against 16.2, and the difference was no longer statistically significant. By 23 or 24, 29.9 against 28.9. The researchers' own conclusion: extending care delays homelessness rather than preventing it. Their summary is here and the 2010 brief is here.

Hold this one loosely. It compares two state policies, not two randomly assigned groups, and it is from 2010.

New York: money, no strings

A hundred young people leaving foster care in New York City got $1,000 a month for twelve months starting June 2023. CIDI, a research office inside the Mayor's office, compared 88 of the recipients against 168 similar young people picked by statistical matching, using administrative records.

More recipients earned taxable income: 69 percent against 48 percent, p=0.016. They also worked fewer quarters (median 2 against 3, p=0.007) and earned less (median $5,254 against $6,304, p=0.002). No difference in cash assistance use, shelter use, involvement with child welfare as a parent, or justice system involvement. The report is on nyc.gov.

The part most pages leave out

Washington State runs Passport, a scholarship plus wraparound campus support for former foster youth. The legislature told its research institute to measure the impact. In 2012 that institute wrote: "A true comparison group evaluation was not possible for this study." It listed why. The program covered the whole state, so there was no untreated area to compare against. Applicants who did not get an award were different from those who did. College-ready young people who never applied could not be identified in the records. And the program itself changed partway through. So they measured Passport students against general college performance benchmarks instead. Their report is here.

Twelve years later, in June 2024, the evaluation prepared for the Washington Student Achievement Council still asks for research comparing participants "with those of a credible comparison group," and still suggests the legislature authorize "a more rigorous impact study that estimates the causal effects of Passport."

That is not a scandal. It is normal. Most programs aimed at you have never been measured against anything, which is why most of what gets said about them is guesswork with a logo on it.

What we take from this

Two things.

The programs that worked, worked small. A fifth of a point on a four-point housing scale. Five or six percentage points more people clearing $2,500 in a year. That is real, and it is worth having, and it is not a way out.

And the money is the money. Around $5,000 a year in Tennessee. Around $5,300 to $6,300 in New York. Every one of these studies is describing people living far under the poverty line before the program and far under it after.

Your Heart Our Hands does not run a tested program and we are not going to tell you we do. We are volunteers who fix the thing in front of us this week, and we hand off to Thrive and other partners for the long work. We are showing you the evidence because it is yours, and because it tells you exactly how much anyone offering you a program is in a position to promise.

Before you act on any of this

We checked this page on 21 August 2026. Programmes get changed, expanded and cancelled, and they get re-evaluated. A finding from one year is not a promise about this one.

Every link here goes to the source we actually read, so you can open it and see what it
says today instead of what it said when we wrote this. If the source and this page
disagree, the source is right and we are wrong. Tell us and we will fix it.

If you would rather not chase it down on your own, call or text us at 602-888-0341
and we will look it up with you. That is not a referral to somewhere else. It is us,
on the phone, finding the current answer with you.