Panel Paper:
Adapting Conditional Cash Transfer Programs for the US: Lessons from Two Different Experiments
*Names in bold indicate Presenter
In 2007, New York City’s Center for Economic Opportunity (CEO) launched a CCT program called Opportunity NYC–Family Rewards (“Family Rewards 1.0”) in six high-poverty communities. It was the first comprehensive CCT program in a developed country. Based on early lessons emerging from the evaluation, CEO led an effort to design and test a modified CCT program (dubbed “Family Rewards 2.0”) in the Bronx and Memphis, which was launched in 2010. MDRC helped design both models and conducted the evaluations.
Family Rewards 1.0 enrolled low-income families with children in elementary, middle, or high school, and for three years it offered cash rewards tied to a pre-specified set of activities/outcomes concerning children’s education, family preventive health care, and parents’ employment. It was evaluated through a randomized trial involving about 4,800 families (and 11,000 children). The average family earned over $3,000 in incentives per year from the program (almost $9,000 in total), substantially reducing material hardship and poverty. Interim findings showed that program produced some positive effects on some human capital outcomes for some subgroups of participants (including impacts on graduation for initially reading proficient 9thgraders), but had little positive effect on most outcomes.
This paper presents final impact results (covering up to six years for certain outcomes), showing that the pattern of earlier effects persisted. New exploratory analyses point to possible positive effects on school performance for siblings who at program entry were too young for the education incentives, and some positive effects on full-time college enrollment rates for 9th graders who were reading proficient at program entry.
Family Rewards 2.0 differed from the original in several ways. First, it was targeted to families receiving TANF or SNAP benefits and had a child entering 9th or 10th grade. Second, it was much simpler, offering 8 rather than 22 incentives. Third, it included a proactive advisement role for the staff. Active case management was proscribed for staff in Family Rewards 1.0 (to allow for a test of “incentives only”), but implementation research found that many families who wanted to earn certain rewards did not know how to go about meeting the conditions.
The evaluation of Family Rewards 2.0 involves a total of 2,400 families who randomly assigned to the program or control group in the Bronx and Memphis. Impact results will be available in 2016 (prior to the APPAM conference). This paper will discuss those findings, which cover up to 3.5 years of follow-up, and compare the results to those from Family Rewards 1.0. It will draw general lessons from both studies about the potential for CCT programs to reduce current and longer-term poverty for adults and children in the US.
