| Vermont’s Welfare Restructuring Project (WRP) was one of the
earliest statewide welfare reform programs initiated under
waivers of federal welfare rules granted before the passage
of the 1996 federal welfare reform law. The program, which
operated from 1994 to 2001, was designed to increase work
and reduce reliance on welfare. WRP required that welfare
recipients work in a wage-paying job after they had received
cash assistance for a specified number of months (30 months
for single-parent families and 15 months for two-parent families).
Recipients received help finding jobs and were offered minimum-wage
community service jobs if they could not find unsubsidized
employment. If a recipient did not comply with the work requirement,
the state took control of her grant, used the money to pay
her bills, and required her to attend frequent meetings at
the welfare office.The program also included a set of financial
incentives that were intended to encourage and reward work.
WRP served as a model for Vermont’s current welfare program,
which took effect in mid-2001.
This is the final report in a long-term evaluation
of WRP conducted by the Manpower Demonstration Research Corporation
(MDRC) under contract to the State of Vermont. The evaluation
was also funded by the U.S. Department of Health and Human
Services and the Ford Foundation. The evaluation used data
from the entire state but focused in depth on 6 of Vermont’s
12 welfare districts. The results from the WRP evaluation
provide important evidence about one of the many diverse strategies
that states adopted to reform welfare in the 1990s.
In order to assess what difference WRP made,
parents who were applying for or receiving cash assistance
in Vermont between July 1994 and December 1996 were assigned,
at random, to one of three groups: (1) the WRP group, whose
members received the financial work incentives and were subject
to the work requirement; (2) the WRP Incentives Only group,
whose members received the incentives but were not subject
to the work requirement; or (3) the Aid to Needy Families
with Children (ANFC) group, whose members remained subject
to the pre-WRP welfare rules, which included neither the incentives
nor the work requirement. MDRC followed all three groups for
six years, using computerized records and a survey. Any differences
that emerged over time in the groups’ outcomes (for example,
in their employment or welfare receipt) can reliably be attributed
to WRP’s policies; such differences are known as impacts,
or effects.
The evaluation also included a study of the
implementation of WRP and an assessment of its financial costs
and benefits for the government and for participating families.
The study mainly focused on single-parent families, who make
up most of Vermont’s welfare caseload.
Key Findings
Key findings from the evaluation include:
- The full WRP program increased employment and reduced
reliance on cash assistance for single-parent families,
particularly in the period after some parents became subject
to the work requirement.
WRP was implemented in an exceptionally healthy
economic climate; Vermont’s unemployment rate was even lower
than the national rate throughout the study period. As a result,
a very large proportion of the ANFC group (87 percent) worked
at some point during the six-year study period, even without
any work requirements or special financial incentives.
Nevertheless, WRP increased employment. The
employment gains were small early in the study period, before
anyone had reached the work requirement, but they grew larger
after the 30-month point. At the peak — in the beginning of
the fourth year of the follow-up period — the employment rate
for the WRP group was 10 percentage points higher than for
the ANFC group (58 percent, compared with 48 percent). Employment
gains persisted throughout the rest of the six-year period,
although the size of the effects diminished over time. Over
the six-year period, WRP increased average annual earnings
by 9 percent ($508). Most of the people who went to work because
of WRP worked full time or nearly full time, in jobs paying
at least $7.50 an hour.
WRP had little effect on cash assistance receipt
until the 30-month point, when it began to reduce the amount
of assistance that families received. Later, the program began
to reduce the number of families receiving any cash assistance.
By the end of the follow-up period, only 18 percent of the
WRP group were receiving assistance, compared with 24 percent
of the ANFC group. WRP reduced cash assistance payments by
28 percent ($449) per year during the last two years of the
study period.
- WRP had little effect on family income, material hardship,
children’s school performance, or other family and child
outcomes.
The WRP group’s higher earnings were largely
offset by their lower cash assistance payments; as a result,
except for a brief period during the third year of the follow-up
period, average income for the WRP group was no higher than
average income for the ANFC group. However, consistent with
the program’s goals, WRP group members derived a greater share
of their income from earnings and a smaller share from public
assistance.
A survey that was administered 42 months into
the follow-up period examined WRP’s impacts on a range of
outcomes, including families’ financial assets, neighborhood
quality, food security, and children’s school performance
and behavior. Because such impacts are typically driven by
changes in income, it is not surprising that WRP generated
few effects on these outcomes.
- The program’s work requirement was needed in order to
generate impacts. WRP’s financial incentives alone did not
lead to increases in employment or income, probably because
the incentives were not substantially different from incentives
under the prior rules.
WRP included two types of financial incentives
that were designed to encourage and reward work. First, WRP
changed the welfare rules to allow recipients to earn somewhat
more without losing eligibility for cash assistance (this
is known as an enhanced earnings disregard). Recipients
could also own a more valuable (and hence more reliable) car
and could accumulate more savings from earnings without losing
eligibility for assistance. Second, the program extended transitional
supports for recipients who were leaving welfare for work
— for example, by providing three years of transitional Medicaid
coverage instead of the single year of coverage mandated under
prior rules.
Other studies have found that financial incentives
alone can increase work and income, but this was not the case
in WRP. The WRP Incentives Only group was no more likely to
work than the ANFC group and did not have higher income. However,
in assessing this result, it is important to note that WRP’s
incentives — while probably important to many families — were
not substantially different from the incentives and rules
that applied to the ANFC group. For example, at most levels
of earnings, WRP’s enhanced earnings disregard during the
first four months of work was actually somewhat less generous
than the disregard available under the prior rules. Similarly,
because Vermont provides unusually generous child care and
health insurance subsidies for all low-income working families,
the ANFC group was eligible for supports that were not dramatically
different from WRP’s transitional benefits.
- WRP increased employment among most subgroups, but the
increases were largest for the most disadvantaged sample
members. WRP increased income for the least disadvantaged
sample members.
Among individuals who were long-term welfare
recipients, had no recent work history, and did not have a
high school diploma — some 9 percent of the study’s participants
— the WRP group earned an average of 31 percent ($870) more
per year over the six-year follow-up period than the ANFC
group. Because WRP increased earnings but did not reduce welfare
receipt among sample members with the fewest barriers to employment
(high school graduates with recent work history who were not
long-term welfare recipients), the program raised their income
(by an average of 7 percent, or $696, per year).
- WRP’s work requirement was implemented as planned, but,
contrary to initial expectations, very few community service
employment positions were needed.
When WRP was designed, planners believed that
its success would hinge on the state’s ability to create a
large-scale community service employment (CSE) program for
recipients who could not find unsubsidized jobs after receiving
benefits for 30 months. In fact, the work requirement was
implemented largely as intended, but the maximum number of
people working in CSE slots statewide never exceeded 70 in
any one month. Only 3 percent of the single-parent WRP group
members (and 4 percent of the two-parent WRP group members)
ever worked in a CSE position during the six-year study period.
Few CSE slots were needed for two main reasons.
First, most recipients were never subject to the work requirement:
Only 46 percent of the single-parent WRP group received cash
assistance for 30 months or more. This figure was nearly the
same for the ANFC group (45 percent), suggesting that the
strong economy and broad changes in Vermont’s welfare system
that affected all three research groups were the key factors
that spurred people to leave welfare sooner than anticipated.
Second, of those who reached the 30-month point, most who
were required to work found unsubsidized jobs; most single
parents were required to work only part time, and jobs were
readily available in most areas of the state. Some others
were exempted from the work requirement or were sanctioned
(penalized) for failing to comply with the requirement.
- The net cost of WRP was quite low, and the government’s
spending on the program was more than offset by reduced
public assistance payments; in other words, WRP saved money
for taxpayers.
The WRP group received few services that were
not also available to the ANFC group. Both groups were eligible
to participate in the state’s welfare-to-work program (the
WRP group was required to participate in Months 29 and 30
of benefit receipt), and both groups received child care assistance
and other supports if recipients worked or participated in
activities while on welfare. As noted earlier, supports for
those who exited welfare were also similar for the two groups.
Thus, the main net costs associated with WRP
— that is, costs over and above those incurred for the ANFC
group — were for relatively inexpensive job search services
provided to recipients who reached the work requirement and
for support services for parents who were participating in
activities or working while on welfare. (More WRP group members
than ANFC group members worked and participated in activities.)
Thus, the net cost of WRP was only about $1,300 per person
over six years. The program saved about $1,700 per person
in cash assistance and Food Stamp benefits over six years,
more than offsetting its cost.
- WRP generated few effects for two-parent families with
an unemployed parent.
WRP’s work requirements for two-parent families
with an unemployed parent were not substantially different
from requirements under the prior rules. Even before Vermont
implemented WRP, principal wage-earners in two-parent families
were required to work or participate in employment activities
throughout their time on welfare — although WRP required full-time
work after 15 months of assistance. WRP eliminated most of
the nonfinancial criteria that restricted eligibility for
two-parent families under ANFC.
WRP did not affect employment or earnings
for two-parent families with an unemployed parent. The financial
incentives increased cash assistance receipt somewhat during
the first four years of the follow-up period, but the effect
did not last. WRP did not substantially affect income, material
hardship, or outcomes for children among these families.
Conclusions and Implications for Policy
The results of the WRP evaluation illustrate
that there are diverse paths to the broadly accepted goals
of increased employment and reduced reliance on public assistance.
Unlike other states, Vermont did not require single parents
on welfare to work until they had received benefits for 30
months, did not use grant reductions or closures to enforce
these requirements, did not require full-time work for most
single parents, and did not set time limits on cash assistance
receipt. Nevertheless, WRP increased employment and, eventually,
reduced welfare payments. Because the program’s net cost was
low, WRP actually saved money for taxpayers — an unusual achievement
for any social program. And, at least within a strong economy,
Vermont was able to impose a work requirement for welfare
recipients without creating a large subsidized employment
program.
Although WRP increased work, it did not make
families better off financially and did not substantially
improve their material well-being. Like previously studied
programs that have increased parents’ employment levels but
not their income, WRP also did not substantially affect participants’
children. However, it is worth noting that low-income families
in Vermont may be better off than those in some other states:
Vermont’s welfare grant levels are among the highest in the
nation, and the state offers an unusually generous set of
supports for low-income working families.
Vermont’s new welfare program — implemented
in mid-2001 — builds on WRP and remains distinctive from programs
in many other states. In response to WRP’s small effects before
any recipients reached the work requirement, the new program
requires recipients to participate in work or work-related
activities as soon as they are deemed to be “work-ready” or
after 12 months of welfare receipt, whichever happens first.
The program also uses financial penalties to enforce its requirements,
although the penalties are less severe than in most other
states. Vermont remains one of only two states that have not
established a time limit on welfare receipt.
|
