| I. Introduction
Connecticut’s Jobs First program, which began
operating in January 1996, was one of the first welfare reform
initiatives to impose a statewide time limit on welfare receipt.
Today, almost all states have established time limits on cash
assistance benefits, either for adults or for entire families,
and the 1996 federal welfare law has imposed a nationwide
60-month time limit on federally funded benefits (with limited
exceptions). Jobs First has attracted national attention because
it includes all the features that are central to most states’
current welfare programs, it has one of the shortest time
limits in the nation, and it is one of the few programs of
its kind that has been subject to a rigorous evaluation, including
an assessment of effects on participants’ children.
This is the final report in an independent evaluation
of Jobs First conducted by the Manpower Demonstration Research
Corporation (MDRC), under contract with the Connecticut Department
of Social Services (DSS). The evaluation was also funded by
the U.S. Department of Health and Human Services, the Ford
Foundation, the Smith Richardson Foundation, and other organizations
listed at the front of the report.
Jobs First limits families to 21 cumulative months of
cash assistance unless they receive an exemption or extension.
The program also includes unusually generous financial work
incentives and requires recipients to participate in employment-related
services targeted toward quick job placement. Jobs First operates
statewide, but this study focused on two welfare offices,
Manchester and New Haven, which together include about one-fourth
of the state’s welfare caseload.
To assess what difference Jobs First made, the study
compared the experiences of two groups of people: the Jobs First group, whose members were subject to the welfare reform
policies, and the Aid to Families with Dependent Children (AFDC) group, whose
members were subject to the prior welfare rules. To ensure
that the groups would be comparable, about 4,800 welfare applicants
and recipients were assigned at
random to one or the other group between January 1996
and February 1997. Because the two groups had similar kinds
of people, any differences that emerged between the groups
during the study’s four-year follow-up period can reliably
be attributed to Jobs First rather than to differences in
personal characteristics or changes in the external environment.
These differences are referred to as impacts or effects.
The Jobs First evaluation differs from many earlier random
assignment studies in which individuals subject to a mandatory
welfare-to-work program were compared with people in a control
group who were not required to participate in employment services
(but could do so voluntarily). In this case, members of the
AFDC group were subject to the program that existed before
Jobs First began, which included some emphasis on employment
and self-sufficiency and provided some employment-related
services to recipients but was less mandatory than Jobs First.
Thus, the study is assessing what difference Jobs First made
above and beyond the effects of the state’s previous
welfare-to-work program. (In October 2001, after the follow-up
period for this study, Connecticut implemented substantial
changes in Jobs First.[1]
This report does not evaluate the new policies.)
II. Findings in Brief
The Jobs First evaluation was conducted during a period
characterized by unusually low unemployment rates, a decline
of almost 60 percent in Connecticut’s welfare caseload,
and publicized changes in state and national welfare policies.
These factors shaped the outcomes for the AFDC group, many
of whom found jobs and left welfare without the program, creating
a high benchmark for Jobs First to surpass. In addition, while
the key components of Jobs First were put in place in Manchester
and New Haven, start-up problems and specific features of
the program design prevented it from being implemented very
intensively. Thus, the evaluation results represent a conservative
estimate of the program’s potential. Nevertheless, Jobs
First produced several important effects:
- Just
over half the Jobs First group reached the time limit
during the study period. About two-thirds of those recipients
received an extension of their benefits, generally because
they had very low income and were deemed to have made
a good-faith effort to find work.
- Over the four-year period, roughly one-third of Jobs First group members’
cases were closed because of the time limit. Most parents
whose grant was closed because of the time limit were working.
Although some people received multiple extensions of benefits,
the vast majority of cases that received an extension on reaching
the time limit were no longer receiving benefits three years
later.
- On
average, over the four-year study period, Jobs First increased
employment, earnings, and income and did not affect cash
assistance receipt.
Over four years, Jobs First group members earned an average
of about $1,800 (7 percent) more than their AFDC group counterparts.
The two groups received about the same amount in average cash
assistance benefits, but the Jobs First group received a little
more in Food Stamp payments. Over the study period, the Jobs
First group had about $2,400 (6 percent) more total income
from earnings, cash assistance, and Food Stamps, compared
with the AFDC group.
- Jobs
First made progress toward its key goal of replacing welfare
with work. By the end of the four-year period, Jobs First
group members were more likely to be working and less
likely to be receiving welfare than their AFDC group counterparts.
The pattern of Jobs First’s effects changed over
time. In Jobs First, all earned income is disregarded (not
counted) in calculating recipients’ cash grants (and
Food Stamp benefits) as long as that income is below the federal
poverty level. This allows working parents to retain their
full cash grant in months in which their income would have
made them ineligible for assistance under the prior (AFDC)
rules. As a result, before anyone reached the time limit,
Jobs First increased the fraction of people receiving cash
assistance. It increased average annual cash assistance payments
during the first two years of the follow-up period by 16 percent
($558). The program also increased employment and earnings
in the pre-time-limit period. Because Jobs First participants
had both higher earnings and higher public assistance payments,
their average total income from these sources was 12 percent
higher than the AFDC group average during the two years following
study entry.
When members of the Jobs First group began to reach the time limit, the
program began to decrease cash assistance receipt and payments.
By the end of the four years, only 19 percent of the Jobs
First group were receiving welfare, compared with 28 percent
of the AFDC group. Employment and earnings gains continued
throughout the period, but because
of the cash assistance reductions, the income gains diminished:
In the last three months of the study period, the two groups’
average income from earnings, cash assistance, and Food Stamps
was almost identical, although a larger fraction of the Jobs
First group than of the AFDC group were working and not on
welfare (51 percent, compared with 42 percent).
- The program’s impacts on employment and
earnings were concentrated among individuals facing greater barriers
to employment.
Among individuals who were long-term welfare recipients,
had no recent work history, and did not have a high school diploma ―
making up 12 percent of study participants ― the Jobs First group
had about $3,600 (37 percent) more earnings than the AFDC group over four
years. After people began reaching the time limit, the program substantially
decreased welfare payments for this subgroup.
In contrast, Jobs First had little effect on employment
and no effect on earnings among individuals with the fewest barriers to
employment (high school graduates with recent work history who were not
long-term welfare recipients). The program’s primary impact for this
subgroup was to allow those who would have worked anyway to continue receiving
public assistance in the pre-time-limit period, thereby raising their
income. After they began reaching the time limit, Jobs First began to
reduce welfare receipt.
- Like most programs studied, Jobs First had no
consistent effect on a wide range of indicators of material well-being.
Levels of hardship remained high for families in both groups.
According to responses to a survey administered three
years after sample members’ entry into the study, Jobs First produced
no impacts on a wide range of measures of material hardship, although
it had a mixed effect on living conditions. Relative to the AFDC group,
the Jobs First group reported fewer problems in their neighborhood during
the year before they were interviewed, suggesting that some may have moved
to better neighborhoods. At the same time, they were also more likely
to have been homeless during the year before their interview. Although
the level of homelessness in each of the research groups was low (2 percent
of the AFDC group and 3 percent of the Jobs First group), the increase
is of concern. Analysis found that some of the Jobs First group members
who became homeless had rather steep drops in income during the year before
the interview, possibly as a result of Jobs First policies such as sanctioning
(benefit reductions because of noncompliance with program rules) and the
time limit.
- Jobs First had a few positive effects on the
behavior of elementary school children, concentrated among 5- to 8-year-olds,
and had mixed effects for adolescents.
Among children who were 5 to 12 years old at the end
of the third year of follow-up, Jobs First children were more likely than
their AFDC peers to be in child care. Parents (but not teachers) reported
that Jobs First children exhibited fewer behavioral problems and more
positive behaviors; these effects were concentrated among children who
were 5 to 8 years old. According to parents’ and teachers’ ratings,
Jobs First did not affect performance or engagement in school for 5- to
12-year-olds.
Among children 13 to 17 years old, Jobs First had both
positive and negative effects. Unlike most similar programs studied, Jobs
First increased the use of child care for adolescents, primarily after
school, from grandparents. Parents reported that adolescents in the Jobs
First group were less likely than those in the AFDC group to have been
convicted of a crime. They also reported, however, that adolescents in
Jobs First had lower school achievement than those in the AFDC group.
- Over five years, the government’s investment
in Jobs First was not offset by decreased welfare payments. The investment
generated substantial gains in income and services for Jobs First participants.
The program’s net cost for employment services
and related support services — the cost of these services over and
above what was spent on the AFDC group ― was only about $2,250 per
person over five years. This is relatively low compared with the cost
of other welfare-to-work programs, because most Jobs First participants
took part in short-term job search activities. The government also spent
more on Jobs First group members, compared with the AFDC group, for Food
Stamps and Medicaid benefits. These investments were not offset by welfare
savings, because the Jobs First and AFDC groups received about the same
amount in welfare payments. In sum, relative to the AFDC program, Jobs
First cost the government about $4,150 per person over five years. Program
participants gained income from increased earnings and Food Stamp payments
and lower tax payments (because of the Earned Income Credit). They also
received more child care assistance, Medicaid benefits, and employment-related
fringe benefits.
III. Implications of the
Evaluation’s Findings
The Jobs First evaluation provides some of the first
information on the implementation and impacts of a welfare reform strategy
that included a time limit on benefit receipt. On average, Jobs First’s
combination of employment and support services, generous work incentives,
and time limits increased employment and earnings and, after people began
reaching the time limit, decreased cash assistance receipt. As a result,
the program increased the proportion of people who worked and did not
receive welfare. Jobs First also modestly increased participants’
income, although this was not an explicit program goal. Importantly, the
Jobs First experience shows that, at least under certain circumstances,
time limits can be implemented without causing the widespread severe consequences
predicted by some critics of the policy. Then again, the program did not
substantially improve families’ well-being, as some proponents of
time limits had hoped.
When drawing conclusions based on the Jobs First evaluation,
however, it is important to remember that the program is an unusual hybrid
and was implemented in a specific manner. First, Jobs First has one of
the shortest time limits in the nation, but, during the period studied,
those who had very low income when they reached the limit typically received
benefit extensions. Second, the program includes an unusually generous
earned income disregard, which allowed many working parents in the study
to retain their entire welfare grant at least temporarily. Third, Jobs
First provides employment-services to help people find jobs, but the program
was not implemented very intensively. The effects of Jobs First reflect
the complex interactions of these components. Finally, the evaluation
period was characterized by an unusually strong economy, which likely
fostered job-finding and helped reduce the chances that Jobs First would
harm vulnerable families.
[1]Changes include limiting the circumstances under which recipients
can be granted more than three 6-month benefit extensions and imposing
a new 60-month limit.
|
