Family Rewards was an innovative approach to poverty reduction in the United States that was modelled on the conditional cash transfer (CCT) programs common in lower- and middle-income countries. The program offered cash assistance to poor families to reduce immediate hardship, provided they met certain criteria related to family health care, children’s education, and parents’ work, in the hope of reducing poverty over the long term. The first version of Family Rewards was evaluated in New York City in 2007. The lessons learned from that evaluation led to the next iteration of the model (“Family Rewards 2.0”).

MDRC evaluated Family Rewards 2.0 through a randomized controlled trial involving about 1,200 families in each city, half of whom could receive the cash rewards and half of whom could not. This report presents the program’s costs and the economic value of the estimated effects over four years.

Key Findings

Family Rewards 2.0 spent a little over a dollar ($1.07) to transfer one dollar ($1.00) to families in the form of a reward payment. These rewards produced positive effects on some outcomes, but left others unchanged. In the end, the program was beneficial for participating families but the economic value of these effects was typically less than the cost of the program to taxpayers.

• Over about three years, the program spent $13,459 on the typical participating family. Nearly half of this amount (48.3 percent) was for rewards paid to participants. The remainder paid staff to actively advise families on how to earn rewards as well as process rewards.

• The program was estimated to produce positive benefits for participating families, taxpayers, and society as a whole. These results were driven primarily by the conditional cash transfer value and a positive impact on the average self-rated health status.

• The program had a positive net present value for participating families and a negative net present value for taxpayers and society (meaning, the present value of benefits per family was less than the present value of program costs per family).

• A Monte Carlo analysis showed that if the results were repeated many times, the program produced positive benefits for society 73.9 percent of the time.

• The Monte Carlo analysis also indicated that the program’s net present value for society (that is, the benefits less the program costs per family) was positive 10.9 percent of the time.

• As operated, Family Rewards 2.0 did not produce positive net present value for taxpayers. Key impacts would need to change dramatically in order to do so.

The findings show that the level of effort required to support participants and process rewards, as well as the value of potential impacts on targeted outcomes, are primary drivers of success for CCT programs. Conditional cash payments are more likely to produce benefits in excess of program costs for taxpayers and society when the level of effort required to administer reward payments is low and the potential value of impacts on targeted outcomes is high.