This is the fifth and final report from a multi-year evaluation of Ohio’s Learning, Earning, and Parenting (LEAP) Program.(1)  Developed and operated by the Ohio Department of Human Services (ODHS), LEAP is a statewide initiative that employs financial incentives in an attempt to increase school enrollment and attendance among pregnant teenagers and custodial teen parents on welfare (almost all of them are women). LEAP, which began operating in 1989, requires these teens to stay in school and attend regularly or, if they have dropped out, to return to school or enter a program to prepare for the General Educational Development (GED), or high school equivalency, test. The program thereby strives to increase the proportion of teens who graduate from high school or receive a GED, find jobs, and ultimately achieve self-sufficiency. These longer-term goals are important because, even though teen parents make up fewer than 10 percent of all Aid to Families with Dependent Children (AFDC) case heads, families started by women who first gave birth as teenagers account for approximately 50 percent of all long-term AFDC recipients.

During the period of this study (the rules have recently been modified), teens who met LEAP’s requirements had their welfare checks increased — $62 for school enrollment and an additional $62 each month they attended school regularly (2) — and teens who did not (without an acceptable reason) had $62 deducted from their welfare grant every month until they complied with program rules. Those who exceeded the allowed number of total absences in a month but not the allowed number of unexcused absences qualified for neither a bonus nor a "sanction" (as such grant reductions are called). Teens could be temporarily exempted from LEAP’s requirements for medical reasons, to care for an infant, or if child care or transportation was unavailable. Teens were no longer subject to LEAP’s requirements when they reached the age of 20, left AFDC, or received a high school diploma or a GED. During 1992 — approximately the midpoint in the period covered by this report — a teen living on her own with one child (the most common situation) was eligible for a monthly AFDC grant of $274. Thus, a bonus raised her grant to $336 and a sanction reduced it to $212. If she went from being sanctioned to receiving a bonus, she would experience a 58 percent increase in her welfare grant.

Teens’ enrollment and attendance are monitored by case managers, who explain the program’s rules, offer guidance, and authorize assistance with child care and transportation teens may need to attend school. LEAP itself provides no other services, although many Ohio high schools have special programs, called GRADS (Graduation, Reality, and Dual-Role Skills), which are designed to assist teen parents in managing their dual roles as parents and students. Also, as part of the evaluation, an enhanced version of LEAP was put in place on a pilot basis in some Cleveland high schools; it offered school-based services such as child care and intensive case management as well as special GED classes and other services for teens who were not complying with LEAP.

The LEAP evaluation began with the start of the program itself and encompassed 12 Ohio counties, in which the large majority of LEAP teens resided. Previous reports described the program’s operations and analyzed LEAP’s effectiveness in achieving its education goals and in moving teens toward employment and self-sufficiency. The primary purpose of this final report is to present longer-term (four-year) effects of LEAP on the teens’ employment, earnings, and AFDC receipt and to assess the program’s cost-effectiveness.(3)

The report is based on a study of 4,151 teens who were identified as eligible for LEAP, in all 12 research counties, during the program’s second year of operation (August 1990 through September 1991). As part of the study, these teens were randomly assigned, by chance, to one of two groups: a program group, which was subject to the LEAP program, or a control group, which was not. Data were collected on both groups during a four-year follow-up period. For employment and welfare outcomes, this follow-up period started in the first month of the calendar quarter in which the teen was randomly assigned. For outcomes measured with the two follow-up surveys, the follow-up period started in the month of random assignment itself. This means, for example, that four-year employment outcomes for a teen randomly assigned in December 1990 would cover the period from October 1990 through September 1994. On the other hand, three-year data on school completion would cover December 1990 through November 1993.

Because teens were assigned to the two groups through a random process, there were no systematic differences between the groups’ members when they entered the study; any differences that emerged between the two groups during the follow-up period can reliably be attributed to the LEAP program. These differences are referred to in this report as the program’s effects, or "impacts."

In September 1996, based on its operational experience and informed by the flow of findings from the evaluation, ODHS implemented changes in LEAP to build upon and improve the program’s achievements. For example, with the goal of increasing graduation rates, an additional $62 bonus for completion of each grade (except 12th grade) and a graduation bonus of $200 have been added; with the goal of changing the behavior of hard-to-reach teens, those who are sanctioned for six consecutive months now have their needs and their children’s needs deducted from the calculation of the welfare cash benefit each month until they comply with LEAP’s rules. (If they are on another person’s case, that person’s portion of the cash benefit is unaffected.) Although this report does not include a study of these recent changes, it does discuss the results in light of them.

The larger national context has changed as well, with enactment of the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 — the landmark welfare law that replaced the entitlement to AFDC with block grants to states. Under the law, unmarried custodial teen parents under the age of 18 who have not graduated from high school or received a GED or comparable credential may not receive federally funded welfare assistance unless they attend high school or a program that prepares them to earn an alternative education or training credential. Prior to the law’s enactment, some 25 states had already instituted some form of school requirement for teen parents on welfare. As one of the first programs implementing a school attendance mandate for the state’s entire teen parent welfare caseload, and as the first to be evaluated with a rigorous, random assignment design, LEAP served as a model for many of these subsequent efforts. The 1996 welfare law is likely to intensify interest in the program and in the evaluation results as those who design and develop new state programs and policies seek lessons applicable to their own goals and circumstances.

The LEAP evaluation was conducted by the Manpower Demonstration Research Corporation (MDRC) under contract with the Ohio Department of Human Services. Additional funding for the study was provided by the Ford Foundation, the Cleveland Foundation, BP America, the Treu-Mart Fund, the George Gund Foundation, the Procter & Gamble Fund, and the U.S. Department of Health and Human Services.

The Findings in Brief

Previous reports found that LEAP was successfully implemented, itself a genuine achievement. The sheer scope of LEAP — a statewide initiative intended to reach the entire target population — was virtually unprecedented, and the existing management information system (called CRIS) was not designed to permit the ready identification and tracking of eligible teens or the administration of the bonus and sanction system itself. LEAP also required entirely new linkages between the county welfare departments and the education system, which provided the enrollment and attendance data needed to implement the bonuses and sanctions. Over time, the program achieved smooth operations, with the phasing in from 1991 to 1992 of a highly sophisticated statewide public assistance computer system (CRIS–Enhanced). But because the research sample entered the program before LEAP was fully implemented, the reported results may be a conservative estimate of the program’s potential.

LEAP’s immediate goals were to induce dropouts to return to high school or to enroll in GED programs, and to induce those still enrolled to remain in high school. By linking bonuses and sanctions to school attendance, the program also sought to promote better attendance among those enrolled. To a considerable degree, LEAP met these goals: It significantly increased school enrollment and attendance, the outcomes directly linked to the incentives provided, for both in-school teens and dropouts. During a three-year follow-up period, the program also achieved positive impacts on school progress (completion of the 9th, 10th, and 11th grades).

LEAP’s longer-term goals included higher rates of high school graduation and GED receipt, which it was hoped would increase teens’ employment and reduce their receipt of welfare. Here, the results were more mixed. In general, the impacts on school completion and employment were more positive for teens who were in school when the program identified them as eligible (the "intially enrolled" teens) than for those who had dropped out before LEAP reached them (the "not initially enrolled" teens). Specifically, the program did not increase high school graduation rates except for the initially enrolled group in Cleveland. LEAP did have positive effects on GED receipt, but only for the initially enrolled group.

In spite of its limited school completion impacts, LEAP was successful in increasing the employment rates of initially enrolled teens throughout the four-year follow-up period, and had positive impacts on their earnings during the first two years, after which the control group caught up. No employment impacts were found for the not initially enrolled teens. Both groups experienced significant reductions in welfare receipt, which were comparable to welfare savings in welfare-to-work programs for adults. LEAP’s impacts on earnings and welfare receipt also were calculated separately for each county. In general, there was no statistically significant variation in the impacts across the various counties.

All in all, the LEAP program benefited initially enrolled teens by increasing their school attendance, GED receipt, and work experience, whereas impacts for the not initially enrolled were limited to school enrollment and attendance. LEAP achieved these results while at the same time being cost-neutral for the welfare department.

Table 1 highlights some of the key impacts, which are discussed further in the remainder of this summary.(4)

The Data Sources and Sample for This Report

Part of this final report reviews and summarizes analyses presented in previous reports. These analyses used several data sources: (1) a survey of 1,188 teens, administered approximately one year after each of them had entered the study (i.e., had been randomly assigned to the program group or control group), and focused primarily on the teens’ early experiences in LEAP and their school enrollment status at the one-year point; (2) a review of 263 LEAP case files from Cuyahoga, Franklin, and Hamilton counties (Cleveland, Columbus, and Cincinnati) to study LEAP’s implementation in general and the bonus/sanction process in particular; and (3) a three-year survey of 913 teens in seven counties three years after random assignment to measure education outcomes (including school attendance, graduation rates, and GED receipt) and employment and welfare outcomes. New analyses presented in this report focus on longer-term (four-year) employment and earnings outcomes and AFDC receipt, and are based on administrative data from the statewide Unemployment Insurance and welfare systems, respectively. For the first time, data for the impact analysis were available for all 12 research counties in which teens were randomly assigned to the LEAP program. The 12 counties are Cuyahoga, Franklin, Hamilton, Jefferson, Lawrence, Lorain, Lucas, Montgomery, Muskingum, Stark, Summit, and Trumbull. They were selected through a weighted random assignment procedure that assured the inclusion of the three counties with the largest estimated LEAP caseloads — Cuyahoga, Franklin, and Hamilton. About two-thirds of the statewide teen population targeted by LEAP resided in these 12 counties.

The sample for these new analyses included 4,151 teens (3,479 program group members and 672 control group members), randomly assigned between mid-August 1990 and September 1991, i.e., those who entered the study in the second (and last) year of random assignment. This sample was chosen for two reasons. First, teens who entered the study later experienced a more mature (though still evolving) LEAP program than those who entered in the program’s first year of operation. Second, for administrative reasons, AFDC and employment and earnings data were not available for the early part of the follow-up period: The employment data cover the last three-and-a-half years of the four-year period; the AFDC data cover years 3 and 4 only. By limiting the analysis to those randomly assigned after mid-August 1990, it was possible to present a consistent impact analysis, covering the same length of follow-up for all sample members. The three-year survey sample — the primary data source for the previous report — was also drawn from teens who were randomly assigned in the second year of random assignment, and, therefore, is a subsample of the research sample for the present report.

Teens’ Initial School Enrollment Status as a Context for the Findings

Many of this report’s analyses are broken down by school enrollment status at random assignment. Findings are presented separately for teens who were enrolled in high school or in a GED program when they entered the study (usually labeled "initially enrolled" in this report, and constituting about 55 percent of the sample) and for teens who were not enrolled in school or in a GED program (labeled "not initially enrolled" or also referred to as "dropouts"). Even though, after random assignment, many teens moved from one status to the other (i.e., either re-entered school or a GED program when they were dropouts, or dropped out after having been initially enrolled), the previous report found that initial school status was an important predictor of program success.

In previous reports, it was hypothesized that the incentives provided through the LEAP program may have been sufficient to keep more of the enrolled teens in school, but may not have been strong enough to induce more of the teens who had left school to return and remain enrolled. Doing so could have represented a major (and, possibly, costly) change in the lives of the teens, perhaps especially for those who had been out of school a long time. Such teens may not have been willing to make this change in return for the monetary incentives provided. Also, the fact that some teens were not initially enrolled, while others were, may reflect underlying obstacles to school enrollment, which the teens would have had to overcome in order to return to school. For example, the not initially enrolled teens tended to be older, were less likely to live with a parent, and had more children to care for.

Findings on Program Implementation

• The state and county welfare departments successfully implemented LEAP statewide, forging a link between welfare departments and schools and putting the incentive system in place.

As discussed in detail in the 1993 report and summarized in this one, LEAP’s incentive structure was successfully implemented in each of the 12 research counties. Program operations improved over time, which meant that most teens in the research sample were exposed to a more efficient and predictable LEAP program in the latter part of the follow-up period. The key was full implementation of a statewide computer system that made identifying and tracking teens easier and carrying out bonuses and sanctions largely automatic. Since the end of the follow-up period, counties have also continued to improve teens’ access to child care and other support services.

• Almost all eligible teens (93 percent) were touched by LEAP’s incentives, with 75 percent earning at least one bonus and 56 percent qualifying for at least one sanction.

As discussed in previous reports, a review of a sample of LEAP case files in three counties found that fully 93 percent of these teens earned at least one bonus or sanction, with the average teen qualifying for about six grant adjustments (3.5 bonus payments and 2.8 sanctions) during her first 18 months in LEAP. During this 18-month period, there were more bonuses than sanctions: 37 percent of teens earned only bonuses; 18 percent qualified for only sanctions; and 38 percent earned at least one bonus and one sanction. In other words, 75 percent of teens earned at least one bonus and 56 percent qualified for at least one sanction. As time passed and the teens got older, those who were still eligible for LEAP received more sanctions than bonuses, probably because teens who had graduated or received a GED by month 18 (generally cooperative teens, who earned frequent bonuses) were no longer subject to LEAP, leaving a higher proportion of frequently sanctioned teens still subject to the program.

• The majority of teens with multiple sanctions reported diminished spending on essentials for their families, especially clothing and food. Most teens with multiple bonus payments reported spending a large share of the additional money on their children.

Teens who were sanctioned at least four times reported in the three-year LEAP survey that the resulting welfare grant reductions had a material effect on their families: 58 percent said that their families had fewer essentials (most often clothing, food, and medicine) because of the grant reductions. Moreover, the sanctions reportedly affected the children at least as much as their teenage parents. Teens replaced part of the income they lost to sanctions by borrowing money (usually from their parents), applying for other forms of public assistance (most frequently Food Stamps), and seeking additional child support. Two-thirds of the teens postponed paying bills, most often utilities bills or rent.

Among teens who received at least four bonus payments, close to 90 percent reported using the additional money on essentials, especially for their children. Almost a quarter also reported being able to pay for some "luxuries" such as new clothing and outings (e.g., to the movies or to the zoo) for their children. These teens also were better able to pay their bills and to save some money, which they said was later used to obtain special items for their children, buy household essentials, and cover unexpected emergencies.

Findings on School Enrollment, Graduation, and GED Receipt

• LEAP’s immediate goals — and the behaviors directly targeted by its financial incentives — were to improve the teens’ enrollment and attendance in school or in a GED program. After one year of follow-up, LEAP achieved a substantial increase in these outcomes, for both initially enrolled and not initially enrolled teens. The program also produced modest, but statistically significant, increases in completion of the 9th, 10th, and 11th grades.

As shown in Table 1, LEAP increased the number of months the average teen attended high school during the first year of follow-up, from 4.2 to 4.8 months — an increase of 14.3 percent relative to the control group. There were increases for both the initially enrolled teens (from 6.6 months for the control group to 7.3 months for the program group) and the not initially enrolled (from 1.0 months to 1.5 months). The latter also increased their enrollment and attendance in GED programs, from 0.9 months for the control group to 1.7 months for the program group.

These impacts on school enrollment and attendance were followed by modest increases in completion of the 9th, 10th, and 11th grades.

• LEAP’s effect on school completion — its longer-term education goal — was found to be limited when measured three years after teens entered the program. The program increased GED receipt among teens who were enrolled in school when they entered LEAP. There were no impacts on high school graduation or GED receipt among teens who were not enrolled in school when they entered LEAP.

LEAP’s subsequent education outcomes were less promising. For the initially enrolled teens, the program increased GED completion rates (from 4.4 percent for the control group to 10.0 percent for the program group), but did not increase high school graduation rates, except in Cleveland (based on an analysis of school records). There is some evidence that these gains in Cleveland may have been partly attributable to the enhanced LEAP services provided on a pilot basis in some of its high schools (as described in detail in the 1994 report).

For the not initially enrolled, the impacts on enrollment and attendance in GED programs did not translate into a higher rate of GED receipt, nor did the 11th-grade completion impact lead to an increased rate of high school graduation, at least not within the period covered by the three-year follow-up survey.

• Two-thirds of the teens in the sample did not receive a high school diploma or a GED certificate within a three-year follow-up period. However, approximately one-sixth of the teens were still in school at the end of that period and could have graduated or received a GED subsequently.

A number of factors probably explain the low rates of high school completion shown in Table 1. Some teens may not have completed all the courses required for graduation, or may have opted to take a GED test instead of further pursuing a high school diploma. The teens’ feelings about school and their own future doubtless played a part. In an earlier survey, a large proportion of LEAP teens reported that their schools were unsafe, inflexible, and unsupportive. Some also viewed their economic prospects as dim, with or without a high school diploma or a GED. Other studies have pointed to the situational and emotional problems that can make school attendance difficult for teenage single mothers.

However, in light of the teens’ youth, when school completion and enrollment are considered together, significantly more LEAP teens than control group members (51.6 percent compared with 46.5 percent) had graduated from high school, received a GED, or were in high school or a GED program at the time of the three-year survey. By the end of the three-year follow-up period for school outcomes, 30.4 percent of teens in the sample were under age 20, and 17.5 percent of program group members, compared with 14.5 percent of control group members, were in high school or a GED program. Thus, LEAP’s impacts on high school graduation and GED receipt may have increased somewhat after the three-year follow-up period.

It is also important to consider LEAP’s high school graduation impacts in the context of the overall graduation rates in the same locales. According to official data, the high school graduation rates for all students in five large school districts (Cincinnati, Cleveland, East Cleveland, Columbus, and Toledo) ranged between 27 and 45 percent in 1994. Lifting graduation rates among LEAP teens to the prevailing levels in their schools would be a noteworthy achievement.

Findings on Employment and Earnings

• Teens in both the program group and the control group experienced substantial growth in their employment rates and earnings during the four-year follow-up period. By the end of follow-up, four out of five teens had worked in a job covered by Unemployment Insurance. Overall quarterly employment rates for such employment increased from 17 percent in the third quarter of year 1 to 40 percent in the last quarter of year 4. If employment not covered by Unemployment Insurance could have been taken into account, the employment rates and earnings presented in this report would have been higher.

In spite of their responsibilities as parents, their youth, and their educational and economic disadvantages, a large proportion of teens included in the LEAP evaluation were employed (often full time) at various points throughout the follow-up period. Quarterly employment rates increased steadily for both LEAP teens and those in the control group. Almost 40 percent of all the teens were employed at some point during the last three months of follow-up (as shown in Table 2) and average quarterly earnings had grown to $568 (or $1,457 for those employed). Therefore, even though few sample members could be considered self-sufficient by the end of the four-year follow-up, many had begun to develop ties to the labor market, and relatively few had not worked at all. These findings attest to the motivation of many young parents in the sample to find work and become part of the formal labor market.

• The LEAP program increased employment for initially enrolled teens, but not for teens who were not initially enrolled. These impacts were strongest early in the follow-up period.

Table 2 summarizes the program’s impacts on employment. The table shows that program group members who were initially enrolled in school were more likely to be employed than their control group counterparts during years 1, 2, and 4 of the follow-up period. For the full sample, these impacts on employment were small (and statistically significant only in year 2), because they averaged positive impacts for the initially enrolled group with small negative impacts for teens who were not initially enrolled.

Employment impacts for teens who were enrolled in school at the time of random assignment were quite substantial, especially in years 1 and 2 of follow-up. In fact, these employment impacts are comparable to those found for adult welfare recipients in welfare-to-work programs evaluated by MDRC. These impacts are remarkable, especially because they occurred in spite of these teens’ greater involvement in school during this time, suggesting that, contrary to what one might expect, increased school attendance may not reduce opportunities to work — at least part time — for these teens.

• LEAP increased initially enrolled teens’ earnings during the first two years of follow-up, but the impacts became smaller in later years.

Table 3 presents the program’s impacts on earnings (averaged in are zero earnings for those teens who were not employed at all). Teens who were enrolled in school at random assignment experienced quite substantial earnings impacts through year 2 of follow-up. Earnings in the second half of year 1 were $101 (41.9 percent) higher for initially enrolled teens in the program group than for their counterparts in the control group. In year 2, earnings impacts for this group were $228 (a gain of 28.0 percent). These earnings gains not only reflect an increase in the number of initially enrolled teens who were ever employed during these periods, but also indicate higher earnings for those who were. Earnings impacts were not sustained in later years because teens in the control group caught up with teens in the program group and increased their earnings as well. However, the overall four-year earnings impact ($544) remained positive for the initially enrolled teens, even though it was not quite statistically significant.

Findings on AFDC Receipt

• Rates of AFDC receipt remained high throughout the four-year follow-up period, but were declining over time. More than 60 percent of all teens were on welfare for at least one of the last three months of follow-up, and close to 30 percent received AFDC continuously during the last two years.

In any given month within the four-year follow-up period, a majority of teens in the program and control groups were receiving welfare. Quarterly rates of AFDC receipt declined from 100 percent at random assignment to 79.1 percent after two years, reaching 60.9 percent in the last quarter of year 4, with program group members being somewhat less likely to receive AFDC than control group members in at least one month in most quarters. However, at the end of the follow-up period, almost 22 percent of teens (or one in three of those receiving welfare) combined welfare with (some) work.

• During the final two years of follow-up, LEAP reduced welfare receipt by increasing the number of teens who were not receiving any AFDC and reducing the number who were receiving AFDC in every single month.

Figure 1 shows a distribution of the number of months of AFDC receipt during years 3 and 4 of the follow-up period. The black bars represent program group members, and the white bars represent control group members. The figure shows that LEAP increased the number of teens who left AFDC altogether during this time (shown in the first pair of bars) and decreased the number of teens who received AFDC in every one of the 24 months represented (shown in the last pair of bars).

• LEAP reduced both the number of teens receiving AFDC and the amount of AFDC they received. Reductions were somewhat smaller for those not initially enrolled than they were for teens who were enrolled in school when they entered LEAP.

Table 4 summarizes LEAP’s impacts on AFDC during years 3 and 4 of the follow-up period (data were unavailable for earlier years). During this time, LEAP reduced the average number of months of AFDC receipt by three-quarters of a month (0.76 months), which amounts to a 4.7 percent reduction relative to the control group average of 16.03 months. During the first six months of the last year of follow-up, the reduction in the rate of AFDC receipt peaked at 5.1 percentage points — a reduction comparable to those achieved by welfare-to-work programs for adult welfare recipients. This reduction in the rate of AFDC receipt was also reflected in the amount of AFDC received by teens in the LEAP program. The average amount of AFDC received was reduced by $275, or 5.0 percent, during years 3 and 4 of the follow-up period.

In contrast to the employment findings, AFDC impacts did not vary a great deal by school enrollment status at random assignment, although impacts for teens who were not initially enrolled were somewhat smaller and failed to reach statistical significance in some cases. They also declined faster toward the end of the four-year follow-up period.

• During the last two years of follow-up, LEAP caused reductions in the amount of AFDC received that exceeded the program’s positive effects on earnings.

In years 3 and 4 of follow-up, program group members’ cash income from AFDC and earnings was $345 less than that available to teens in the control group, whose cash income was $8,693 over those two years. This 4.0 percent reduction (not quite statistically significant) occurred because earnings gains among program group members were not large enough to offset reductions in AFDC payments. The loss of cash income from AFDC and earnings was greatest for teens who were not initially enrolled ($586 on average) because they experienced welfare reductions without any offsetting earnings gains. The corresponding loss for initially enrolled teens was only $101, which was not statistically significant. However, some of the reductions in combined cash income from AFDC and earnings may have been offset by the Earned Income Tax Credit (EITC), increases in the receipt of Food Stamps, or other unmeasured sources of income.

The Costs and Benefits of LEAP

• The LEAP program was relatively inexpensive, with an upfront investment by the Ohio Department of Human Services of $1,388 per teen, or $747 for 12 months. Costs were higher for initially enrolled teens than for those not initially enrolled.

Over the four-year follow-up period for this report, program group members were in LEAP an average of 22.3 months, at a net cost to ODHS of $1,388 per program group member, or $747 for 12 months. (Net cost refers to the average cost per program group member minus the average cost per control group member. Although control group members were not eligible for LEAP, they were eligible for its child care services.) The net costs were higher for the initially enrolled group ($1,659) than for those who entered the program as school dropouts ($1,067). This is because the former were younger and thus in LEAP longer (24.0 months, on average, compared with 19.9 for those not initially enrolled). These costs translate into 12-month net costs of $830 (for the initially enrolled) and $643 (for those not initially enrolled).

Table 5 shows the net cost to ODHS per LEAP teen by component. The case management/county administrative cost ($1,140 per program group member), which accounted for most of the cost, includes expenditures for all program-related activities on behalf of LEAP teens: orientation, initial and yearly assessments, referrals for child care, providing transportation stipends, monitoring school enrollment and attendance, and arranging for bonuses and sanctions. In addition, case managers frequently offered guidance in their ongoing contact with teens. The case management/county administrative cost also includes expenditures for administrative supervision and general overhead.

LEAP’s incentives themselves did not add to the program’s total net cost because, on average, teens received slightly more sanctions than bonuses, leaving only the administration of the incentives as a cost to ODHS. The cost to ODHS of transportation for LEAP teens to attend school was $134 per program group member, and child care costs to ODHS averaged $258 per program group member, slightly more than the control group average of $185; 7.6 percent of program group members and 5.5 percent of control group members used child care paid for by LEAP, which was limited to licensed providers and was available only to teens attending school.

• Because of LEAP-induced savings in AFDC, Food Stamps, and related Medicaid expenditures, ODHS recovered its investment in the program over the four-year follow-up period.

Benefits of the LEAP program were estimated using the available administrative earnings and welfare data, supplemented with approximations of taxes paid, tax credits received, Food Stamps received, and Medicaid expenditures incurred. As shown in Table 6, over four years ODHS recovered its investment in LEAP, with a return of 99¢ per $1 invested. (In this context, net gains and losses refer to what was spent and saved on program group members compared with control group members.) The program was somewhat more cost-effective for initially enrolled teens.

A related analysis takes into account costs and financial benefits to taxpayers as a result of the LEAP program. Again, these are "net" costs and benefits — the average for the program group minus the average for the control group — over the four-year follow-up period. From this "taxpayer perspective," the costs were the ODHS costs plus $332 incurred by other agencies (mainly $518 for the additional LEAP-generated attendance in high school and in GED programs, which was partly offset by savings in other employment-related services). Most of the financial benefits were the savings in AFDC, Food Stamps, and Medicaid. As shown in Table 6, taxpayers recovered 75¢ of every $1 invested in LEAP, even after factoring in the additional education costs attributable to LEAP’s getting more teens to remain in or return to school.

• LEAP teens experienced a net loss of $1,110 over the four years.

As suggested by the earlier discussion of the program’s impacts on cash income from AFDC and earnings, the story was different for LEAP teens, who experienced a net loss of more than $1,100 during the four-year follow-up period. As shown in Table 6, both initially enrolled and not initially enrolled teens experienced losses (e.g., reductions in AFDC receipt, Food Stamps, and Medicaid eligibility) that were not offset by gains (e.g., from earnings and tax credits). Even though the initially enrolled group experienced some positive earnings impacts, reductions in AFDC and other public benefits were also greater for this group. On the other hand, LEAP teens may experience long-term benefits from the additional education and education credentials they received. These long-term benefits may outweigh the losses apparent in the four years covered by the evaluation.

Implications for Policy and Program Design

• The LEAP results suggest that it is sensible to include a LEAP-like approach as one element of a state’s welfare strategy.

LEAP has proved itself to be a promising approach to increasing school involvement among teen parents on welfare. As other states continue to develop their teen parent strategies, they may want to pursue similar approaches. The main challenge — one that ODHS has addressed in its recent modifications of LEAP — is not only to increase school attendance but, beyond that, to improve high school graduation rates so that these young mothers have a better chance of success in the labor market and a route to self-sufficiency.

• The LEAP experience highlights the importance of meeting the implementation challenges inherent in such programs, particularly the need for ongoing case management and well-designed management information systems.

The LEAP program relies on case managers who are dedicated to working with teen parents, and who are supported by a computerized case information system. Without ongoing case management, a program like LEAP would be very difficult to implement: The attendance monitoring role that case managers play is central to the program, as is the role of providing assistance and information to teens.

In order to implement programs like LEAP, it is also necessary to develop a well-designed management information system. The need to integrate welfare data, school data, and case management information to provide correct and timely incentives places a substantial burden on the information system available to program administrators. With the 1996 federal welfare law requiring school attendance by unmarried, custodial, minor teen parents receiving federal Temporary Assistance for Needy Families (TANF) funds, other states will be pushed to develop programs similar to LEAP, and many will confront system development tasks similar to those that faced ODHS.

• The importance of initial school enrollment status as a predictor of program success underscores the need to prevent teen parents from dropping out of school. Recent program changes address this problem by limiting LEAP’s pregnancy and age-of-child exemptions, which could have accounted for some teens’ dropping out after becoming pregnant.

This report shows that teens who were in school when they were reached by LEAP had better outcomes and experienced stronger program effects than teens who had dropped out of school before LEAP reached them. Apparently, it is more difficult to work with teens after they have dropped out of school already. Therefore, it is important to develop interventions and policies that prevent teens from dropping out in the first place.

One way that ODHS has approached this issue is by changing the LEAP program’s exemption policy. Previously, LEAP teens were exempt from the program after the first trimester of pregnancy and up to three months following the birth of their child. This means that a teen could have been out of school for nine months, which would have set her back academically and would have increased the risk that she would drop out altogether. As part of the program changes implemented in 1996, the pregnancy exemption was eliminated (except in cases of medical need), and the post-partum exemption was reduced to six weeks from three months. It is hoped that these changes will reduce dropout among LEAP teens.

• LEAP’s impact on school attendance was encouraging, and would be strengthened if more teens responded to the program’s incentive structure. This outcome might be achieved by changing the incentive structure and its implementation, and by directly addressing teens’ reservations about going to school.

School attendance might be further boosted by enhancing the teens’ general understanding of the LEAP incentives, reinforcing their awareness of the short-term financial benefits of meeting LEAP’s requirements (partly by modifying procedures so that the bonuses can be paid more rapidly), and continuing to stress the longer-term benefits of completing one’s education. The size of the incentive could also be increased. With sanctions and bonuses combined, LEAP’s incentive was already substantial, but ODHS’s decision to dramatically increase sanctions for teens who do not comply with LEAP’s regulations for six consecutive months is likely to increase compliance, albeit simultaneously increasing the risk to teens (and their families) who are unable or unwilling to meet LEAP’s requirements.

Some teens may not stay in school (or re-enroll in school) despite LEAP’s incentives because they feel that the cost of staying in school or going back is simply too high to be offset by the incentives offered to them. Some are working, and may experience a loss of income if they return to school. Others do not want to go back to school because they want to spend time with their children or do not want to leave them in the care of strangers. Yet others say that they consider school a dangerous place, feel unwelcome there, or find it to be a poor environment in which to learn. All of these practical considerations and negative feelings represent a "cost" of returning to school. Reducing these costs could further increase school attendance. For example, teens could be allowed to receive reimbursement for child care options they are comfortable with (instead of relying entirely on licensed day care), and the school environment could be improved.

Finally, it is important to acknowledge that many teen parents have academic problems or have dropped out before they became parents, which suggests that their academic problems may not be related to their status as teen parents. Other types of assistance, or more general school reform measures, may be needed to help teens who drop out for academic reasons prior to becoming teen parents.

• In addition to stimulating school enrollment and attendance, programs like LEAP should provide incentives that specifically reward academic progress and school completion. Special school-based services may be needed to improve school outcomes for teen parents who respond to incentives and return to school.

Rewarded for "seat time" with attendance bonuses, some of the teens who responded to the LEAP program may not have had an incentive to actually make academic progress, earn credits, and work toward graduation. In fact, teens who regularly attended school, and made steady academic progress, were in a sense penalized for graduating or earning a GED because at that point they could no longer receive bonus payments for continuing to attend. This suggests that programs like LEAP should develop incentives that emphasize school performance and graduation in addition to seeking increases in school enrollment and attendance. For example, ODHS has recently implemented grade completion bonuses, which are paid when teens successfully complete one grade and are promoted to another (except grade 12), and a graduation bonus when teens successfully complete high school or obtain a GED. Alternatively, bonuses could be based on grades, which reflect both attendance and academic performance.

LEAP’s high school completion impacts may also be improved by providing tutoring and remedial classes that help teens clear the hurdles associated with earning a high school diploma, such as required courses and school-wide achievement tests. Finally, incentives and counseling should emphasize the importance of completing high school, as opposed to dropping out to pursue a GED or to seek employment instead.

• Special attention is needed for teen parents who had dropped out but who returned to school because of LEAP, and for older teens who were far behind in their schooling when they dropped out.

The results of the LEAP evaluation indicate that teens who were not enrolled in school or in a GED program when they became eligible for LEAP were at particularly high risk of making some academic progress but then failing to graduate from high school. Special attention is needed for these teens, including stronger efforts by the schools themselves, to foster the retention and academic progress of teen parents who are induced to return to school with financial incentives and start out at a disadvantage.

Experience with LEAP also made it clear that it often is neither feasible nor desirable to place long-term dropouts who are not of compulsory school age into low grade levels that fit their academic skills (e.g., placing an 18-year-old in the 7th grade). To address this issue, ODHS’s recent changes in the program included the provision that if no appropriate education activity can be found for these teens, they will be transferred to the JOBS welfare-to-work program and will be subject to JOBS participation requirements.

• The 1996 federal welfare law poses important challenges for policymakers who develop and administer teen parent programs like LEAP. In addition to getting teen parents to attend school, programs must now address work requirements and time limits on receipt of federally provided welfare assistance that will affect young parents after they graduate or age out of programs like LEAP. Therefore, these programs may have to combine their education message and incentives with more employment-oriented activities.

As states continue to operate or launch learnfare programs for teen parents on welfare under the new welfare rules, they will need to consider the employment and welfare prospects facing these teens once they leave school. As teens graduate or reach the age of 20, they will likely become subject to the same work requirements and time limits that may apply to adult recipients of welfare. Therefore, it may be necessary for LEAP and other such programs to implement services and incentives that encourage teens to work, or make connections with the world of work, outside regular school hours. Fortunately, the employment data presented in this report show that many teens engaged in temporary and part-time employment, even without an intervention that specifically tried to make them do so. LEAP and similar programs could build on these teens’ apparent eagerness to work by offering opportunities for paid internships and part-time work during the summer and after school. By integrating these opportunities into LEAP and other programs, cooperation with those programs could become attractive to more teen parents on welfare, thereby enhancing the overall effectiveness of these programs.

• The problems facing teen parents on welfare are substantial and complex. LEAP successfully addressed one problem area by increasing these teens’ school attendance. However, more needs to be done to improve the self-sufficiency of these teens and to reduce the persistent poverty among them. The LEAP evaluation shows that there are no easy answers. More experimentation is necessary, both within and outside the context of the LEAP program.

By providing both positive and negative financial incentives, the LEAP program uses an innovative approach to induce teen parents to attend school. This report showed that this approach holds promise. However, by design it addresses only part of the problem facing teen parents on welfare, and only some of the causes of their potential long-term dependency. It will be important for governments and program administrators to continue to experiment with different approaches to changing the behavior and improving the outcomes for teen parents and their children. While some of these may be targeted specifically at teens and the problems of teenage parenthood, others may have to address the broader social and economic circumstances that affect the lives and opportunities of low-income young people and the choices they make.

By creating, implementing, rigorously testing, and continuously seeking to improve the LEAP program, the State of Ohio and the Ohio Department of Human Services took a pioneering step. The program’s designers and administrators well knew how complex and daunting was the issue of teen parenting and the associated long-term welfare receipt, but they also knew how little had been proven to work in this area, and how critical was the need for promising new approaches. ODHS has already put to work lessons from this evaluation — about both the program’s strengths and limitations — by making far-reaching changes to LEAP. For Ohio and the nation, tracking how this modified program works will be important to the development of responsible and effective social welfare policy.

Notes

1.  The four previous reports are: Dan Bloom, Hilary Kopp, David Long, and Denise Polit, LEAP: Im-plementing a Welfare Initiative to Improve School Attendance Among Teenage Parents, 1991; Dan Bloom, Ve-ronica Fellerath, David Long, and Robert G. Wood, LEAP: Interim Findings on a Welfare Initiative to Improve School Attendance Among Teenage Parents, 1993; David Long, Robert G. Wood, and Hilary Kopp, LEAP: The Educational Effects of LEAP and Enhanced Services in Cleveland, 1994; and David Long, Judith M. Gueron, Robert C. Wood, Rebecca Fisher, and Veronica Fellerath, LEAP: Three-Year Impacts of Ohio’s Welfare Initiative to Improve School Attendance Among Teenage Parents, 1996.

2.   For high school students, “regular attendance” was defined as having no more than two unexcused absences and no more than four total absences in a month.

3.  Sample members are referred to as “teens” in this report, but by the end of the four-year follow-up period, most were in their early twenties.

4.  In this and the other impact tables, statistical significance is indicated by asterisks.  A “statistically significant” result is one that has less than a 10 percent probability of having occurred simply by chance and not as a result of the program.

The Manpower Demonstration Research Corporation's evaluation of Ohio's Learning, Earning, and Parenting (LEAP) Program was funded in part by the Ford Foundation, the Cleveland Foundation, BP America, the True-Mart Fund, the George Gund Foundation, the Proctor & Gamble Fund, and the U.S. Department of Health and Human Services. This is the fifth and final report in the evaluation.Dissemination of MDRC's Work is also supported by MDRC's Public Policy Outreach Funders: the Ford Foundation, the Ambrose Monell Foundation, the Alcoa Foundation, and the James Irvine Foundation.