Program

Dec 30, 2016

OVERVIEW

Family Rewards, a part of the Opportunity NYC initiative, is a community-based conditional cash transfer (CCT) program that aims to help improve children’s education, families’ health care, and parents’ work outcomes for low-income families by offering cash rewards as incentives.  A random-assignment study compared families who received cash assistance in return for meeting outcome-related goals to those who received no cash assistance. At the 18-month follow-up, significant positive impacts were found for reducing current poverty and economic hardship, increasing academic achievement for high school students who were better prepared for high school, increasing parent and high school student use of health services, increasing parents having bank accounts, increasing savings, increasing parent full-time employment and average earnings, and increasing the proportion of parents who married, as well as the proportion who divorced. There were no impacts on the educational achievement of elementary or middle school students.

At the 30-month follow-up, significant impacts were found for reducing families’ current poverty and economic hardship, increasing parents’ savings, reducing parents’ reliance on families and friends for cash loans, increasing self-reported full-time employment, increasing families’ use of dental care services, increasing high school students’ graduation rates, improving school outcomes for students who were already stronger readers, increasing teens’ time spent on academic activities, increasing parents’ spending on their teenagers and saving for their future education, and reducing problem behaviors among some teens.  No impacts were found for school outcomes for elementary or middle school students, families’ use of preventive medical care, most health outcomes, employment in or earnings from jobs covered by unemployment insurance, parental monitoring, teen disclosure to parents, teens’ levels of depression, anxiety, or delinquent behaviors, college completion beliefs, school behavioral engagement, academic efficacy, or teens’ intrinsic motivation to learn.

At the 42-month follow-up, after the end of the three year program period, significant impacts were found for increasing family income, material well-being, savings, the number of times seeing a dentist, current employment, and full-time employment. For students in the ninth grade at baseline, significant impacts on increasing enrollment, attendance, credits earned, and Regents exams passed were found for students who had been proficient on the eighth-grade English Language Arts and math tests.

A study at the end of year six concluded that, overall, positive impacts were modest. The largest impacts were for reducing poverty and material hardship during the three years that the cash rewards were offered, particularly for families who entered the program in more severe poverty. Impacts on behaviors were either small or limited to particular subgroups. Some impacts were encouraging, such as the increase in high school graduation rates for ninth-graders who demonstrated proficiency in reading before entering high school, or the increase in dental care for all types of families. But there were no impacts on education among younger students, and only a few small impacts on other outcomes related to health care and parents’ employment.   Participants in the treatment group who had lower educational achievement, employment, and income at baseline experienced negative employment and earnings outcomes in years one to five.

DESCRIPTION OF PROGRAM

Target population: Legally resident urban families with incomes at or below 130 percent of the federal poverty level, and at least one child in the fourth, seventh, or ninth grade.

Family Rewards is a community-based conditional cash transfer (CCT) program that aims to help low-income families break the generational cycle of poverty by providing income supplements, resources, and incentives to reduce current poverty and help families to invest in their human capital. Family Rewards offers cash assistance to families in return for meeting goals related to their children’s education, family health care, and parents’ work and training. Seedco, a national intermediary organization, helps train and support community-based organizations, called “neighborhood partner organizations” (NPOs), to implement Family Rewards.  Families complete activity requirements in each of three key areas (children’s education, families’ preventive health care, and parents’ employment); submit forms documenting their accomplishments to Seedco; and then can submit claim forms and coupons to receive cash rewards. Cash rewards vary in amount, depending on the activity or accomplishment completed, and range from $20 to $600. Most reward payments are made to parents, although some education-related payments are made, in part or in full, to high school students.  Throughout the three-year program, the NPOs offer workshops teaching families about the benefits of participating in Family Rewards, provide one-on-one help with submitting coupons, and remind families about available rewards. NPOs also host social events, such as summer barbeques and holiday parties, to create a sense of community among participating families.

EVALUATIONS OF PROGRAM

Study 1:

Riccio, J., Dechausay, N., Greenberg, D., Miller, C., Rucks, Z., & Verma, N. (2010). Toward reduced poverty across generations: Early findings from New York City’s conditional cash transfer program. New York, NY: MDRC.

Evaluated population: Approximately 4,800 families (11,000 children) legally residing in the United Sates, with at least one child in the fourth, seventh, or ninth grade, and an income at or below 130 percent of the federal poverty level, participated in the study.  Of the participating families, 81 percent were one-parent families, 87 percent rented an apartment or home, and 87 percent received some kind of public benefits or housing assistance. For 77 percent of the families, English was the primary language spoken at home, and families on average had 2.5 children.  Fifty-one percent of parents were African American, 47 percent were Latino, one percent were white, and one percent were another race or ethnicity. For 30 percent of the participating families, the parent’s highest level of education was a high school diploma or GED; nine percent had an Associate’s degree or degree from a two-year college, nine percent had a Bachelor’s degree, and 50 percent had less than a high school diploma.  Fifteen percent of the children were in special education and 13 percent were in English Language Learner (ELL) programs.

Approach: Approximately 4,800 families applied to participate in the program, mostly as volunteers, but with some families specially recruited to make the sample more representative of the families living in the targeted study areas. Families were randomly assigned through a lottery to either a treatment group that was offered incentives or a control group that was not offered incentives.  There were no significant differences between families in the treatment and control groups in terms of demographics.  For the analyses, the sample size was 3,082 families, with 1,574 families in the treatment group and 1,508 in the control group.

Data were collected from administrative records, a survey of parents administered at about 18 months after random assignment, and in-depth qualitative interviews with families and Family Rewards staff.  Teens’ academic outcomes were assessed at the end of year one.  Data were collected on many outcomes in the categories of poverty and hardship, banking and savings, parents’ use of health services, parents’ health status, high school students’ use of health services, parents’ employment, parents’ education and training, students’ school attendance, and students’ school achievement.

Results:  The study found significant impacts on measures of current poverty, small or modest impacts on some measures of human capital, and no impacts on a number of other important outcomes.

In the area of poverty and hardship, at post-test, the percentage of families assigned to the treatment group that had a household income at or below the federal poverty level was significantly less than the percentage of families assigned to the control group; 58.9 percent of families in the treatment group had a household income at or below the federal poverty level, compared with 70.0 percent of the families in the control group. The study also found that significantly fewer of the families in the treatment group had total household incomes that were less than 50 percent of the federal poverty level than those in the control group; 16.7 percent of the households in the treatment group had household incomes that were less than 50 percent of the federal poverty level, compared with 30.0 percent of the families in the control group.  Those in the treatment group scored significantly lower on having insufficient food than those in the control group; 14.8 percent of those in the treatment group sometimes or often did not have enough food to eat, compared with 22.1 percent of those in the control group. Participants in the treatment group scored significantly lower on not having enough to make ends meet at the end of the month than those in the control group; 34.1 percent of those in the treatment group did not have enough to make ends meet at the end of the month, compared with 41.8 percent of those in the control group.

In the area of banking and savings, the study found that those in the treatment group scored significantly higher in agreeing strongly or somewhat agree with their financial situation being better than last year; 62.7 percent of those in the treatment group strongly or somewhat agreed with their financial situation being better than last year, compared with 44.5 percent of those in the control group.  Significantly more of the families in the treatment group currently had a bank account than those in the control group; 73.3 percent of families in the treatment group currently had a bank account, compared with 51.8 percent of those in the control group. participants assigned to the treatment group were found to cash a check at a check cashing business at least once a month significantly less often than those in the control group; 29.2 percent of those in the treatment group had cashed a check at a check casher at least once a month, compared with 36.5 percent of those in the control group. Participants in the treatment group were found to have significantly higher average savings than those in the control group; those in the treatment group had $575 in average savings, compared with $354 for those in the control group.

In the area of parent’s health, those in the treatment group scored significantly lower on having a period with no health insurance coverage since random assignment than those in the control group; 16.1 percent of those in the control group had a period without health care, compared with 19.4 percent for those in the control group.  The study found a significantly positive impact for parents in the treatment group having a usual source of healthcare; 94.9 percent of those in the treatment group had a usual source of health care, compared with 91.4 percent of those in the control group. The study found a significant decrease in using the emergency room for routine health care for those in the treatment group, compared with parents in the control group; 3.3 percent of those in the treatment group used the emergency room for routine health care, compared with 5.3 percent of those in the control group. Participants in the treatment group scored significantly higher on seeing a personal doctor in the past 12 months; 86.1 percent of those in the treatment group, saw a personal doctor, compared with 82.3 percent of those in the control group.

Participants in the treatment group were found to score marginally significantly higher on currently being treated for any medical condition; 47.2 percent of those in the treatment group, compared with 44.4 percent of those in the control group. Participants in the treatment group were also found to marginally significantly more often self-rate their health as excellent than those in the control group; 15.8 percent of those in the treatment group self-rated their health as excellent, compared with 13.5 percent of those in the control group.  The study also found a significant impact on parents in the treatment group having had at least one dental visit, compared with parents in the control group; 86.0 percent of those in the treatment group had had at least one dental visit, compared with 83.0 percent of those in the control group.  Similarly, parents in the treatment group were found to have had at least two dental visits significantly more than those in the control group; 67.4 percent of those in the treatment group had had at least two dental visits, compared with 57.9 percent of those in the control group.

In the area of high school students’ use of health services, the study found that significantly fewer of those in the treatment group used the hospital emergency room as their usual source of care when sick than those in the control group; 9.8 percent of those in the treatment group used the emergency room, compared with 15.9 percent of those in the control group.  Also, 86.3 percent of those in the treatment group used another place as their usual source, compared with 80.1 percent for those in the control group. High school students in the treatment group were found to have had at least two dental visits significantly more than those in the control group; 70.3 percent of those in the treatment group had had at least two dental visits, compared with 57.2 percent of those in the control group. No significant impacts were found for the number of teens who had had a health checkup or received shots in the past 12 months, or who had a usual source of care when sick.

In the area of parental employment, a significant negative impact was found for those in the treatment group ever having been employed in year one, according to Unemployment Insurance (UI) records, than those in the control group; 56.2 percent of those in the treatment group had been employed, compared with 58.6 percent of those in the control group.  Again according to UI records, the study found a marginally significant impact on average quarterly employment; those in the treatment group had an average quarterly employment of 48.9 percent, compared with 50.3 percent for those in the control group.  Using data from surveys, the study found a significant positive impact on currently working for those in the treatment group, compared with those in the control group; 59.9 percent of those in the treatment group reported currently working, compared with 54.3 percent of those in the control group.  Furthermore, the study found a significant impact on working full time for those in the treatment group; 48.6 percent of those in the treatment group reported working full time, compared with 43.0 percent in the control group. In the area of parents’ education and training, the study found a marginally significant impact on having any trade license or training certification for those in the treatment group, compared with those in the control group; 54.2 percent of those in the treatment group had a trade license or training certification, compared with 51.2 percent of those in the control group. The study found a significant impact on participants in the treatment group having an associate’s degree, compared with those in the control group; 10.2 percent of those in the treatment group had an Associate’s degree, compared with 7.7 percent of those in the control group. The study found no significant program impacts on parents’ average earnings in Year 1; the number of parents who had ever participated in education, training, or any employment activity; and whose highest level of education was a bachelor’s degree.

In the area of family composition, participants in the treatment group scored significantly lower on being single (never married) than those in the control group; 45.5 percent of those in the treatment group were single, compared with 50.1 percent of those in the control group.  Participants in the treatment group scored significantly higher on being married and living with spouse than those in the control group; 19.0 percent of those in the treatment group were married living with spouse, compared with 15.6 percent of those in the control group. Participants in the treatment group scored significantly higher on being divorced than those in the control group;15.2 percent of those in the treatment group were divorced, compared with 12.3 percent of those in the control group.

In the area of educational outcomes, the study found no significant impacts for the fourth- or seventh-grade cohorts.  For participants in the treatment group who were part of the ninth-grade cohort, the study found a significant impact on having an attendance rate of 95 percent or higher in year two, compared with those in the control group; 28.8 percent of those in the treatment group had 95 percent or higher attendance rate, compared with 23.7 percent of those in the control group.

Subgroup analysis of the ninth grade cohort by the students’ proficiency on the eighth grade standardized math test found significant positive impacts for those in the treatment group from more proficient subgroup on having an attendance rate of 95 percent or higher in Year 2, compared with those in the control group; 51.1 percent of those in the treatment group had 95 percent or higher attendance rate, compared with 36.2 percent of those in the control group.  The analysis also found a significant impact for those in the treatment group from the more proficient group to have remained in ninth grade in year two, compared with those in the control group; 3.0 percent of those in the treatment group remained in ninth grade in year two, compared with 8.8 percent of those in the control group.  The subgroup analysis also found a significant impact for those in the treatment group from the more proficient subgroup on having earned 22 credits in years one and two, compared with those in the control group; 72.7 percent of those in the treatment group had earned 22 credits, compared with 64.5 percent of those in the control group.  For the more proficient group, the study found a significant impact on having passed at least two Regents exams in years one and two for those in the treatment group, compared with those in the control group; 77.6 percent of those in the treatment group had passed at least two exams, compared with 71.7 percent of those in the control group.

Study 2:

Morris, P., Aber, J. L., Wolf, S., & Berg, J.. (2012). Using incentives to change how teenagers spend their time: The effects of New York City’s conditional cash transfer program. New York, NY: MDRC

Evaluated population: A subsample, consisting of 511 teens who were in ninth grade at baseline, and their parents, was taken from the same population as in Study 1.

Approach: A randomly selected subsample of the families with children in grades four, seven, and nine were interviewed at 18-month follow-ups, providing baseline information.  From among these participants, 716 parent-teen pairs were selected to participate in phone interviews at 30-month follow-ups; 511 teens with parental consent, and their parents, participated in this step. The teens and their parents were interviewed separately by telephone, approximately two years after enrollment in Family Rewards.  Data were collected on how teenagers spent their time, how the teenagers and their parents spent their money, interactions between the parents and teenagers, the teenagers’ mental health and problem behaviors, and the teenagers’ academic engagement, motivation, and sense of competence in school. These measures may act as mechanisms of action for the impacts on other Family Rewards outcomes.

Results: In the area of parent spending, the study found a marginally significant positive impact on parent monthly spending for those in the treatment group, compared with those in the control group; parents in the treatment group spent $695, on average, monthly, compared with $624 for those in the control group.   A significant impact was found for parents’ monthly leisure spending; parents in the treatment group spent $109, on average, monthly, compared with $82 for those in the control group. Similarly, a significant impact was found for parent productive spending; parents in the treatment group spent $156, on average, monthly, compared with $125 for those in the control group.  The study found that parents in the treatment group saved significantly more for their children’s future education than those in the control group; 54.1 percent of parents in the treatment group saved, compared with 41.5 percent of those in the control group.

In the area of teenager spending, the study found that total spending by teenagers in the treatment group was significantly less than for those in the control group; teenagers in the treatment group spent $230, on average, monthly, compared with $301 for those in the control group.  Teenager maintenance spending for those in the treatment group was also found to be significantly less than for those in the control group; teenagers in the treatment group spent $162 monthly, on average, compared with $213 for those in the control group.  Teenager leisure spending for those in the treatment group was also found to be significantly less than for those in the control group; teenagers in the treatment group spent $35 monthly, on average, compared with $51 for those in the control group

In the area of teenager problem behaviors, the study found that significantly fewer of the teenagers in the treatment group had exhibited aggressive behaviors than those in the control group; 14.1 percent of teenagers in the treatment group were aggressive, compared with 24.1 percent of those in the control group. The study found that significantly fewer of the teenagers in the treatment group had any substance use in the past month than those in the control group; 17.5 percent of teenagers in the treatment group had any substance use, compared with 32.6 percent of those in the control group. The study also found that the teenagers in the treatment group reported significantly fewer friends using substances in the past month than those in the control group.

The study found no significant impacts on the level of parents’ maintenance spending; teenagers’ productive spending; parent-teenager interactions; or on teenagers’ approaches to schooling, their mental health, or delinquent behaviors.

Study 3:

Riccio, J., Dechausay, N., Miller, C., Nunez, S., Verma, N., Yang, E. (2013). Conditional cash transfers in New York City – The continuing story of the Opportunity NYC-Family Rewards Demonstration. New York, NY: MDRC.

Evaluated population: The evaluated population was the same as in Study 1.

Approach: The full sample was evaluated at the end of Family Rewards, although sample sizes for analyses varied because of missing data.  For some measures, results cover three years after random assignment, while for other measures the follow-up period was somewhat longer.  The evaluation findings were based on analyses of a variety of administrative data; results from a survey of parents administered about 42 months after they entered the study; and qualitative in-depth interviews with families and Family Rewards staff.

Results: In the area of income and poverty, the study found that household income both during year three and the early post-program period was significantly higher for those in the treatment group, compared with those in the control group; those in the treatment group had average monthly household income of $1,973 during year three and $1,700 in the post-program period, compared with $1,620 for those in the control group in year three and in the post-program period.  The percentage of families assigned to the treatment group that had a household income at or below the federal poverty level during year three was significantly less than the percentage of families assigned to the control group; 56.0 percent of families in the treatment group had a household income at or below the federal poverty level, compared with 68.2 percent of the families in the control group. The study also found that significantly fewer of the families in the treatment group had total household incomes that were less than 50 percent of the federal poverty level than those in the control group; 16.3 percent of the households in the treatment group had household incomes that were less than 50 percent of the federal poverty level, compared with 27.4 percent of the families in the control group.

In the area of material hardship, the study found that those in the treatment group scored significantly lower on having insufficient food than those in the control group; 15.3 percent of those in the treatment group sometimes or often did not have enough food to eat, compared with 20.7 percent of those in the control group. Participants in the treatment group scored significantly lower on not having enough to make ends meet at the end of the month than those in the control group; 35.4 percent of those in the treatment group did not have enough to make ends meet at the end of the month, compared with 41.0 percent of those in the control group. Participants in the treatment group scored significantly lower on not having paid full rent or mortgage in the past year than those in the control group; 40.0 percent of those in the treatment group had not paid full rent or mortgage, compared with 44.1 percent of those in the control group. The study also found that those in the treatment group scored significantly higher in agreeing strongly or somewhat with their financial situation being better than last year; 51.4 percent of those in the treatment group strongly or somewhat agreed with their financial situation being better than last year, compared with 46.6 percent of those in the control group.

In the area of banking and savings, the study found that significantly more of the families in the treatment group currently had a bank account than those in the control group; 64.0 percent of families in the treatment group currently had a bank account, compared with 46.6 percent of those in the control group. The study also found that families participating in the treatment group were significantly more likely to have any savings than those in the control group; 24.6 percent of families in the treatment group had any savings, compared with 16.8 percent of those in the control group. Similarly, families participating in the treatment group were significantly more likely to have savings in excess of $500 than those in the control group; 12.5 percent of families in the treatment group had more than $500 in savings, compared with 9.2 percent of those in the control group. Parents assigned to the treatment group were found to borrow cash from family or friends less often than those in the control group; 47.3 percent of those in the treatment group had borrowed from family or friends, compared with 52.5 of those in the control group.

In the area of parents use of health services and health status, the study found that those in the treatment group scored marginally significantly lower on having a period with no health insurance coverage in the past 12 months than those in the control group; 15.3 percent of those in the control group had a period without health care, compared with 17.6 percent for those in the control group.  The study also found a significant impact on parents in the treatment group having seen a dentist for any reason, compared with parents in the control group; 85.4 percent of those in the treatment group seen a dentist, compared with 75.3 percent of those in the control group. Similarly, parents in the treatment group were found to have had at two or more dental checkups in the last 12 months than those in the control group; 45.2 percent of those in the treatment group had had at least two dental visits, compared with 33.5 percent of those in the control group.

In the area of high school students’ use of health services and health status, the study found a significant impact on those in the treatment group having seen a dentist for any reason, compared with students in the control group; 93.8 percent of those in the treatment group seen a dentist, compared with 89.1 percent of those in the control group.  Similarly, students in the treatment group were found to have had at two or more dental checkups in the last 12 months than those in the control group; 62.9 percent of those in the treatment group had had at least two dental visits, compared with 44.1 percent of those in the control group.

In the area of parents’ employment, using data from surveys, the study found a significant positive impact on currently working for those in the treatment group, compared with those in the control group; 56.0 percent of those in the treatment group reported currently working, compared with 49.6 percent of those in the control group. The study also found a significant impact on working full time for those in the treatment group; 44.4 percent of those in the treatment group reported working full time, compared with 39.5 percent in the control group.

In the area of children’s education, the study found that for students in grade nine at baseline, those who were in the treatment group scored significantly higher on having an attendance rate of 95 percent or higher in Year 3 than those in the control group; 25.1 percent of those in the treatment group had an attendance rate of at least 95 percent, compared with 21.9 percent for those in the control group.

Subgroup analysis of the ninth grade cohort by the students’ proficiency on a standardized English-Language Arts (ELA) test in the eighth grade found that there were significant positive impacts for those in the treatment group from the more proficient subgroup on graduating within four years, compared with those in the control group; 74.8 percent of those in the treatment group graduated within four years, compared with 66.9 percent of those in the control group.  The analysis also found a significant impact on being enrolled in grade 12 in year four for those in the treatment group, compared with those in the control group; 78.4 percent of those in the treatment group were enrolled in grade 12, compared with 68.2 percent of those in the control group. The analysis also found that participating in the treatment group had a significant impact on attendance rates in years three and four, compared with those in the control group; 83.9 and 76.7 percent in years three and four for those in the treatment grouped, compared with 77.1 and 71.6 percent for those in the control group. The subgroup analysis also found a significant impact for those in the treatment group from the more proficient subgroup on the average number of credits earned in years one to four, compared with those in the control group; 44.3 credits for those in the treatment group, compared with 40 for those in the control group.  The analysis found a significant impact for those in the treatment group on having earned at least 33 credits in years one to three, compared with those in the control group; 68.4 percent of those in the treatment group had earned 33 or more credits, compared with 55.4 percent of those in the control group. Similarly, the analysis found a significant impact for those in the treatment group on having earned at least 44 credits in years one to four, compared with those in the control group; 66.1 percent of those in the treatment group had earned 44 or more credits, compared with 56.6 percent of those in the control group. For the more proficient group, the study found a significant impact on having passed at least five Regents exams in years one to four for those in the treatment group, compared with those in the control group; 72.5 percent of those in the treatment group had passed at least five exams, compared with 63.1 percent of those in the control group.

Subgroup analysis also found three significant impacts for ninth grade students in the treatment group who were proficient on the eighth grade math test, compared with those in the control group.  The analysis found that participating in the treatment group had a significant impact on attendance rates in year three, compared with those in the control group; 82.8 percent in year three for those in the treatment grouped, compared with 78.2 percent for those in the control group.  The subgroup analysis also found a significant impact for those in the treatment group from the more proficient subgroup on the average number of credits earned in years one to four, compared with those in the control group; 43.5 credits for those in the treatment group, compared with 41.3 for those in the control group. Finally, the analysis found a significant impact for those in the treatment group on having earned at least 33 credits in years one to three, compared with those in the control group; 67.4 percent of those in the treatment group had earned 33 or more credits, compared with 59.8 percent of those in the control group.

No significant impacts were found during the early post-program period on household income, not paying full rent or mortgage, parents cashing check at check cashers at least once a month, parents’ other use of health services, high school student’s other use of health services, and parents’ employment status as per unemployment insurance records.  For students overall, students who were not proficient on the eighth grade ELA test, and students who were not proficient on the eighth grade math test, the study found no significant impacts for students who were in ninth grade at baseline.  For students in ninth grade at baseline who were proficient on the eighth grade ELA test, no significant outcomes were found for being enrolled in any grade in year four, or for average attendance rate in year four. For students in ninth grade at baseline who were proficient on the eighth grade math test, no significant outcomes were found for graduating within four years, being enrolled in any grade in year four, being enrolled in grade 12 in year four, average attendance rate in year four, average number of credits earned in years one to four, having earned at least 33 credits in years one to three, earning at least 44 credits in years one to four, or passing at least five Regents exams in years one to four.

Study 4:

Riccio, J., Miller, C. (2016). New York City’s first conditional cash transfer program – What worked, what didn’t. New York, NY: MDRC.

Evaluated population: The evaluated population is the same as in Study 1.

Approach: This report recaps the history and design of Family Rewards and summarizes the main research findings, synthesizing previously published findings, presenting subgroup analyses, and presenting new results from administrative data up to six years after the date when each family entered the study. Data came from administrative records and two waves of surveys of parents at 18 months and at 42 months after random assignment, and from a special survey of teenagers at two years after random assignment. Other data, including findings from in-depth interviews with families and staff, as well as program fiscal data, provided evidence on program operations and costs, families’ reward receipt patterns, and families’ understanding and views of the program.

Results:  In the area of income and poverty, subgroup analysis found that for those families whose income was at or above 50 percent of the federal poverty line, participating in the treatment group had a significant impact on household income during year three, compared with those in the control group; those in the treatment group had an average monthly income of $2,093, compared with $1,771 for those in the control group, and 51.1 percent of those in the treatment group had an annual income at or below the poverty line, compared with 62.1 percent for the control group. For those whose income was below 50 percent of the poverty line at baseline, the study found a significant impact on average monthly income during year three and in the early post-program period for those in the treatment group, compared with those in the control group; those in the treatment group had an average monthly income of $1,781, compared with $1,409 for those in the control group, and 63.5 percent of those in the treatment group had an annual income at or below the poverty line, compared with 77.2 percent for the control group.  The study found a significant impact on the number of families with average savings above $500 for those in the treatment group, compared with those in the control group; 10.3 percent of those in the treatment group had savings above $500, compared with 6.0 percent of those in the control group. For those families participating in the treatment group, the study found a significant impact on reducing the number of times that families did not have enough to eat, compared with those in the control group; 14.0 percent of those in the treatment group sometimes did not have enough to eat, compared with 23.2 percent of those in the control group. The study found a significant impact on reducing the number of times that families did not usually have enough money to make ends meet, compared with those in the control group; 36.4 percent of those in the treatment group usually did not have enough money to make ends meet, compared with 45.2 percent of those in the control group. For those participating in the treatment group, the study found a significant impact on reducing the number of times that families did not pay full rent or mortgage, compared with those in the control group; 35.7 percent of those in the treatment group sometimes did not pay full rent or mortgage, compared with 46.4 percent of those in the control group.

In the area of education, the study found that participating in the treatment group had a significant impact on having an attendance rate of 95 percent or higher in year two for students in grade nine at the time of random assignment, compared with those in the control group; 28.8 percent of those in the treatment group had high attendance, compared with 23.7 percent of those in the control group.  Similarly, a marginally significant impact was found for year three; 25.1 percent of those in the treatment group had high attendance, compared with 21.9 percent of those in the control group.

Subgroup analysis by students’ performance on the English Language Arts test in eighth grade, found a significant impact for those who were proficient on the test, compared with those who were not proficient, for those in the treatment group on having graduated within six years, compared with those in the control group; 81.7 percent of those who were proficient and participated in the treatment group graduated within six years, compared with 72.2 percent of those who were proficient and in the control group. The study also found a significant impact on average number of credits earned in years one to six for those in the treatment group, compared with those in the control group; students in the treatment group earned 45.2 credits in years one to six, on average, compared with 41.2 credits for those in the control group.  Similarly, the study found a significant impact on having passed at least five Regents exams in years one to six for those in the treatment group, compared with those in the control group; 74.0 percent of those in the treatment group had passed at least five exams, compared with 65.4 percent of those in the control group.

For the sample as a whole, no significant impacts were found for ever having enrolled at any post-secondary institution within six years after baseline, but subgroup analysis found significant impacts on post-secondary enrollment for those participants who were proficient on the eighth  grade ELA test, compared with those who were not proficient. The study found a marginally significant impact on ever being enrolled at a four-year institution within six years after baseline for those in the treatment group, compared with those in the control group; 46.0 percent of those in the treatment group had enrolled at a four-year institution, compared with 38.8 percent of those in the control group. The study found a significant impact on the highest level ever enrolled being full-time for those in the treatment group, compared with those in the control group; 49.4 percent of those in the treatment group had enrolled full-time, compared with 38.8 percent of those in the control group. The study also found a significant impact on being enrolled in a four-year institution full-time for those in the treatment group, compared with those in the control group; 40.9 percent of those in the treatment group had enrolled full-time, compared with 31.2 percent of those in the control group.

At the six-year follow-up, a significant positive impact was found for children who were ages two to seven at the time of random assignment and who participated in the treatment group on scoring at the proficient level or higher on ELA test, compared with those in the control group; 18.1 percent if those in the treatment group scored proficient, compared with 12.8 percent of those in the control group. Similarly, a marginally significant impact was found for those children participating in the treatment group on scoring at the proficient level on the math test, compared with those in the control group; 21.4 percent of those in the treatment group scored proficient or above, compared with 15.9 percent of those in the control group.

In the area of health care services, subgroup analysis found that for the subgroup of parents who self-rated their health as excellent, very good, or good at baseline, participating in the treatment group resulted in a significant impact on their having seen a dentist for any reason in the past twelve months, compared with those in the control group; 86.3 percent of those in the treatment group had seen a dentist, for any reason in the past twelve months compared with 76.8 percent of those in the control group. The study also found that participating in the treatment group resulted in a significant impact on having had two dental check-ups or more, compared with those in the control group; 45.1 percent of those in the treatment group had at least two dental check-ups, compared with 34.9 percent of those in the control group.  The study found a marginally significant impact for those in the treatment group currently having and being treated for any medical condition, compared with those in the control group; 44.1 percent of those in the treatment group were being treated, compared with 39.8 percent of those in the control group. The study also found a marginally significant impact for those in the treatment group currently having and being treated for any asthma, compared with those in the control group; 13.9 percent of those in the treatment group were being treated for asthma, compared with 10.8 percent of those in the control group.

For those who self-rated their health as fair or poor at baseline, the study found a significant impact on their having seen a dentist for any reason in the past twelve months; 81.5 percent of those in the treatment group had seen a dentist, compared with 68.8 percent of those in the control group. The study also found that participating in the treatment group resulted in a significant impact on having had two dental check-ups or more, compared with those in the control group; 45.7 percent of those in the treatment group had at least two check-ups, compared with 27.8 percent of those in the control group.

The study found a significant impact for those in the treatment group on self-rating their health as excellent or very good at follow-up, compared with those in the control group; 12.6 percent of those in the treatment group rated their health highly, compared with 6.4 percent of those in the control group.  For participants in the treatment group, the study found a significant negative impact on currently having and being treated for asthma, compared with those in the control group; 19.4 percent of those in the treatment group had asthma, compared with 30.8 percent of those in the control group.

In the area of employment and earnings covered by unemployment insurance, a follow-up at five years found a significant negative impact for those in the treatment group on having ever been employed in years one to five; 66.8 percent of those in the treatment group had ever been employed, compared with 69.1 percent of those in the control group.  The study also found a marginally significant negative impact on average quarterly employment in year five on jobs covered by the New York State UI program, compared with those in the control group; those in the treatment group had an average quarterly employment rate of 44.9 percent, compared with 46.7 percent for those in the control group. Subgroup analysis by whether the respondent had a high school diploma or GED at baseline found that for those who had no diploma or GED, participating in the treatment group had a marginally significant negative impact on ever being employed in years one to five; 57.5 percent of those in the treatment group were ever employed, compared to 60.7 percent of those in the control group.  The study found a significant negative impact on average quarterly employment in years one to five; 35.5 percent for those in the treatment group, compared with 38.5 percent for those in the control group.  The study also found a significant negative impact on total earnings in years one to five; $34,579 for those in the treatment group, compared with $37,512 for those in the control group.  Subgroup analysis by whether the respondent was employed at baseline found that for those who were unemployed at baseline, participating in the treatment group had significant negative impacts on ever being employed in years one to five; 43.5 percent of those in the treatment group were ever employed, compared 47.4 percent of those in the control group.  The study found a significant negative impact on average quarterly employment in years one to five; 18.7 percent for those in the treatment group, compared with 21.1 percent for those in the control group.

Subgroup analysis by income at baseline being above or below 50 percent of the federal poverty line found that for those with income less than 50 percent of the federal poverty line, participating in the treatment group had a significant negative impact on ever being employed in years one to five; 39.0 percent of those in the treatment group were ever employed, compared with 45.3 percent of those in the control group.  The study found a significant negative impact on average quarterly employment in years one to five for those with income less than 50 percent of the federal poverty line who participated in the treatment group, compared with those in the control group; average quarterly employment was 18.3 percent for those in the treatment group, compared with 21.8 for those in the control group.  The study also found a significant negative impact on total earnings in years one to five for those with income less than 50 percent of the federal poverty line who participated in the treatment group, compared with those in the control group; those in the treatment group earned $14,624, on average, compared with $17,512 for those in the control group.

SOURCES FOR MORE INFORMATION

References

Riccio, J., Dechausay, N., Greenberg, D., Miller, C., Rucks, Z., & Verma, N. (2010). Toward reduced poverty across generations: Early findings from New York City’s conditional cash transfer program. New York, NY: MDRC.

Morris, P., Aber, J. L., Wolf, S., & Berg, J. (2012). Using incentives to change how teenagers spend their time: The effects of New York City’s conditional cash transfer program. New York, NY: MDRC.

Riccio, J., Dechausay, N., Miller, C., Nunez, S., Verma, N., & Yang, E. (2013). Conditional cash transfers in New York City – The continuing story of the Opportunity NYC-Family Rewards Demonstration. New York, NY: MDRC.

Riccio, J., & Miller, C. (2016). New York City’s first conditional cash transfer program – What worked, what didn’t. New York, NY: MDRC.

Miller, C., & Deitch, V. (2016). New York City’s First Conditional Cash Transfer Program: What Worked, What Didn’t — Supplemental Data on Impacts and Costs. New York, NY: MDRC.

Websites:        http://www.mdrc.org/publication/learning-together

http://www.mdrc.org/project/opportunity-nyc-demonstrations

KEYWORDS: Children, Adolescents, Elementary, Middle School, High School, Co-ed, High-Risk, Urban, Community-Based, Welfare/Public Assistance, Reading/Literacy, Mathematics, Academic Achievement/Grades, Employment/Earning, Job Training/Readiness, Health Status/Conditions , Other Physical Health

Program information last update on 12/30/16.

Subscribe to Child Trends

Short weekly updates of recent research on children and youth.