Sanctions and Time Limits:
State Policies, Their Implementation, and Outcomes for Families
Dan Bloom, Manpower Demonstration Research Corporation, and Ladonna Pavetti,
Mathematica Policy Research, Inc.
The welfare reforms of the 1990s dramatically expanded the range of circumstances in
which families can have their welfare benefits reduced or canceled. In particular, sanctions
(financial penalties for noncompliance with program requirements) and time limits on benefit
receipt have become central features of most state TANF programs. While sanctions have long
been used to enforce program requirements, sanction policies under TANF have taken on much
greater significance than in the past, in part owing to the emergence of “full-family” sanctions
that remove all, rather than part, of a family’s cash grant. Time limits, on the other hand, are
more recent and have set the 1990s reforms apart from previous efforts. Since both policies
potentially involve the loss of all cash assistance, they warrant careful monitoring to understand
both their implementation and their effects on the behavior and well-being of families.
Although sanctions and time limits are both being used to restructure the nature of the
social contract for poor families, they operate in different ways. Sanctions generally are tied to
specific expectations and are imposed only when clients fail to meet those expectations. A key
goal of sanctions then is to instill in clients a realization that there are consequences associated
with the choices they make. Rather than being tied directly to a client’s behavior, time limits
rely on the passage of time. Consequently, effective implementation of time limits requires
instilling in clients a realization that their current choices may have profound consequences on
their well-being sometime in the future.
Although sanctions and time limits operate differently from the perspective of recipients
and administrators, understanding their implementation and outcomes presents similar analytic
challenges. In the current environment, states and/or localities are the critical decision-makers
when it comes to sanction and time limit policies, making it difficult to systematically document
and assess variation in the design of these policies and even harder to characterize their
implementation. In addition, while a few rigorous evaluations have focused on welfare reforms
that include sanctions and/or time limits, and other studies have examined the circumstances of
families who were sanctioned or reached time limits, none of these studies were designed to
isolate the impact of sanctions or time limits per se on individuals’ behavior or well-being. As a
result, relatively little is know with certainty about the effects of these two critical policies.
This paper describes states’ policy choices regarding sanctions and time limits, examines
some of the issues involved in implementing the policies, and reviews the available research on
how sanctions and time limits are affecting recipients’ behavior and well-being. The final section
draws some policy conclusions and identifies areas for consideration during reauthorization.
Sanction and Time Limit Policies: Federal Framework and State Choices
State sanction and time limit policies are influenced by federal TANF rules, but the block
grant structure leaves states with great flexibility to shape these policies. Given this flexibility, it
is important to consider not only specific policy choices a state has made but also how these
policies interact and influence one another. In particular, the design of a state’s sanction policy is
likely to affect the number and characteristics of families affected by time limits.
Federal law requires states to impose at least a pro-rata (partial) benefit reduction on
families who do not satisfy work and child support enforcement compliance requirements. The
sanction must remain in place until the family complies with the requirement. States have the
option to impose more stringent penalties and can expand the penalties to program requirements
other than work and child support enforcement. In addition, TANF sanctions can, under certain
circumstances, affect Food Stamps benefits for the entire family and Medicaid coverage for non-
Most states have chosen to implement sanction policies that are much stricter than those
that existed under the Job Opportunities and Basic Skills (JOBS) program and stricter than those
required by Federal law. In particular, the majority of states have adopted full-family sanctions
where the penalty for non-compliance is loss of all cash assistance. A total of 37 states impose
full-family TANF sanctions for noncompliance with work, and/or other personal responsibility,
requirements. In 15 of these 37 states, a full family sanction is imposed immediately, while in
the remaining 22, the grant is initially reduced to send a warning signal to the family that
cooperation is expected. In seven states (Delaware, Georgia, Idaho, Mississippi, Nevada,
Pennsylvania and Wisconsin), continued or repeated noncompliance may result in a lifetime bar
from receiving cash assistance (Goldberg and Schott 2000).
Only six states still use the sanctions that were in place under JOBS, eliminating the non-
compliant adult from the grant. The remaining eight states have increased the financial penalty
for non-compliance but do not completely eliminate a family’s cash assistance grant. Some of
these states only provide assistance to families in the form of vendor payments.
Time Limit Policies
Federal law restricts states from using federal TANF funds to provide assistance to most
families for more than 60 cumulative months, starting from the date of implementation of the
state’s TANF program (“assistance” is generally defined as payments designed to meet a
family’s ongoing basic needs). States are free to set time limits of less than 60 months, but also
may use TANF funds to assist families beyond the 60-month point, as long as those families do
not constitute more than 20 percent of the average monthly state caseload (calculated after
families begin reaching the 60-month limit – 2001 at the earliest). States also may use state
maintenance of effort (MOE) funds to assist families who have reached the 60-month point. For
the most part, federal law does not specify particular categories of families who are exempt from
the 60-month limit, other than child-only cases that do not include any adult recipients.
As with sanction policies, there is great variation in state time-limit policies. The vast
majority of states have imposed time limits, but relatively few have developed limits that are
more stringent than the federal limit. A total of 43 states (including DC) have imposed a
“termination” time limit that can result in the cancellation of a family’s entire welfare grant.
Twenty-six of these states have imposed a 60-month termination time limit. The other 17 states
– accounting for about one-fourth of the national caseload – have imposed termination time
limits of less than 60 months. Six of these states have imposed a lifetime limit of less than 60
months; the others have imposed “fixed period” time limits – for example, a limit of 24 months
on benefit receipt in any 60-month period.
At the other end of spectrum, eight states – including California – have not imposed
termination time limits. Six of the 8 have “reduction” time limits that eliminate benefits for
adults but maintain benefits for children, and the other two (Michigan and Vermont) currently
have no time limits on benefits for either adults or children. States are free to pursue these
policies, although all states will need to use state maintenance of effort funds to assist families
who pass the federal time limit and exceed the cap on exemptions.
The Relationship Between State Sanction and Time Limit Policies
Table 1 categorizes and compares states’ approaches to sanctions and time limits. For
both policies, state approaches are categorized as stringent, moderate or lenient. The largest
number of states (12) have both stringent time limits and sanctions in place. An additional seven
states use a more moderate approach for both policies. The largest concentration of TANF cases
(40.4 percent), however, can be found in the five states (California, Indiana, Maine, New York
and Rhode Island) that have the most lenient sanctions and time limits in place. Most of the
remaining states rely on a moderate approach to at least one of the policies.
In order to fully interpret the role sanctions and time limits have played in reforming the
welfare system, it is critical to understand the ways in which these policies interact. When full
family sanctions are in place, families who do not comply with program requirements are
completely removed from the welfare rolls. The families who remain – and who are likely to
reach a time limit – are those who are employed but not earning enough to work their way off
welfare or those who have looked for but not found employment – in other words, families who
have ”played by the rules.” In states that do not have full family sanctions (or rarely impose
them), time limits will affect both families who play by the rules and those who do not.
Implementing Sanctions and Time Limits
Sanctions and time limits provide program administrators and front line staff with
concrete tools for changing expectations about the receipt of government assistance. Sanctions
are intended to send a message that receipt of assistance is contingent on meeting a set of
specified set of expectations, while time limits send a message that the government will provide
assistance during times of crises, but not ongoing income maintenance. Presented together,
sanctions and time limits can be used to create a sense of purpose and urgency that might be
more difficult with each policy alone. A common message that case managers attempt to convey
is that recipients need to take advantage of the opportunities available to them now because a
time will come when they will no longer be able to receive cash assistance.
Putting sanction and time limit policies into practice is a complex endeavor. In order to
impose sanctions or time limits in a meaningful way, states need to create the foundation for a
mandatory employment program – they must design a mechanism for determining which
activities recipients will be assigned to, ensure that slots are available, and develop systems for
monitoring recipients’ compliance with program requirements. Within this mandatory
employment framework, key implementation challenges include: (1) deciding who should be
subject to work requirements and the sanctions and time limits that accompany those
requirements; (2) communicating sanction and time limit policies to recipients; (3) encouraging
and supporting compliance with program mandates; and (4) deciding whether and how to
support families who lose cash benefits due to sanctions and time limits.
Deciding Who Is Subject to Sanctions and Time Limits
PRWORA does not explicitly define which families should be subject to work
requirements (and therefore, sanctions) or time limits. States are, however, required to meet
work participation rates that are based on the work activities of all families headed by an adult,
except those with a child under the age of one if a state chooses to exclude them. States are also
subject to the 60-month TANF time limit, but may use TANF funds to assist families beyond the
60-month point, subject to the 20 percent cap described earlier.
Most states have chosen to exempt some families from work requirements and/or time
limits. For example, 44 states exempt from work requirements adult recipients who are caring
for a young child, although these exemptions are typically much narrower than those that existed
under the JOBS program. Thirty-four states exempt disabled or incapacitated recipients from
work requirements, many of them following or modifying exemption policies that were in place
under JOBS (State Policy Documentation Project, 2000; Thompson et al. 1998).
In many states, the families who are exempt from work requirements are also exempt
from time limits, but 18 states exempt no families from time limits (including those who may be
exempt from work requirements). These states – most of which have 60-month time limits –
presumably plan to wait until families begin reaching the time limit to decide which families, if
any, should receive benefit extensions. Some states may avoid defining time-limit exemptions in
advance because they want to maintain a sense of urgency for all families.
Defining and putting exemption policies into practice, particularly those that target
disabled parents or household members, is a complicated endeavor. Although significant
numbers of TANF recipients face substantial personal and family challenges, many of them are
not so severe that they make work totally unfeasible (Olson and Pavetti 1996). These recipients
may, however, need more assistance to find and maintain work than most TANF work programs
are prepared to provide. When this is the case, exemptions protect vulnerable families from the
negative consequences associated with sanctions and time limits, but also may keep them from
receiving the services they need to become self-sufficient.
The small number of states that exempt few or no TANF recipients from work
requirements define allowable work activities broadly, including such activities as substance
abuse and mental health treatment (Thompson et al. 1998). However, most states are only
beginning to integrate these and other strategies for the hard-to-employ into their TANF
employment programs. Other issues that make the implementation of exemption policies
complex are caseworkers’ skill levels and workloads. Most caseworkers have limited training
and/or skills to identify families who may be unable to comply with work requirements. In
urban areas where caseloads tend to be especially high, unless a family explicitly requests that
they be exempt from program requirements, caseworkers often do not have the time to conduct a
thorough assessment to determine whether or not a family may have difficult meeting the
Communicating Sanctions and Time Limits to Clients
Although sanctions and time limits both have the potential to reduce the number of families
receiving cash assistance, that is not their primary goal. Instead, their goal is to encourage
TANF recipients to go to work sooner than they would in the absence of these policies. In order
for this to happen, recipients must know what is expected of them, understand the consequences
associated with those expectations and believe that they will be adversely affected by any
negative actions resulting from their behavior.
In the case of sanctions, the main message that workers aim to communicate to clients is that
participation in work activities is required in order to receive the cash assistance for which they
are eligible based on their current income and assets. For time limits, the message is far more
complicated as states hope to use time limits to motivate clients to take steps toward self-
sufficiency long before they exhaust their months. Depending on the state, or sometimes local
office, clients may be encouraged to leave welfare to “bank” their months for use in
emergencies, or they may be encouraged to stay on assistance, combining work and welfare, in
order to benefit from earned income disregards, which have been expanded in about 40 states.
Some welfare offices have problems with the most basic tasks – informing recipients about
sanctions and time limits and explaining how they work. Often these messages are given along
with other information regarding receipt of benefits, such as when changes in circumstances
must be reported and when and how benefits will be issued, making it easy for key messages to
get lost amidst the masses of information that clients receive when they apply for benefits.
In the case of time limits, some offices have developed ways to inform and remind recipients
about limits – for example, sending frequent reminder letters showing the number of months
remaining, asking recipients to cross months off on a calendar during redetermination interviews
– but have found that it is much more difficult to translate a vague sense of urgency about a
future event into concrete short-term actions. Staff often report that many recipients are not
strongly motivated by the time limit when it is still far away (Brown, Bloom, and Butler, 1997).
Because sanctions are tied so directly to specific program requirements, one might expect
fewer difficulties communicating sanctions to clients. However, a recent Inspector General’s
report found that even though local offices explained sanctions to clients repeatedly and in a
logical format, TANF clients did not fully understand them (U.S. Department of Health and
Human Services 1999).. Clients often knew that they might lose their benefits if they didn’t do
what was expected of them, but they rarely understood what benefits they would lose and for
how long. A study of sanctioned families in Tennessee found that 34 percent of families did not
understand what they were required to do (Overby 1998). Other studies have found similarly
high rates of confusion among clients.
The importance of communicating expectations to clients cannot be overstated. In studies
of the implementation and performance of the Work Incentive (WIN) and JOBS programs, Mead
(1985 and 1997) found that work programs that sanctioned many cases tended to perform poorly
in terms of job placements and other performance measures. Mead speculates that this is
because these offices did a poor job of making program expectations clear in more effective and
informal ways. The offices that performed well were offices that made work expectations clear;
they threatened sanctions, but rarely needed to impose them.
Encouraging and Supporting Compliance with Program Requirements
The way in which sanctions and time limits are enforced is likely to play a significant
role in determining the role sanctions and time limits play in creating a work-oriented assistance
system and who is adversely affected by them. In the case of both sanctions and time limits,
local welfare offices face critical choices about how to create an effective balance between
enforcing sanction and time limit policies and engaging in efforts to promote compliance with
program requirements. If sanction and time limit policies are not enforced, no one will take them
seriously, nullifying their effect. Efforts to promote compliance are likely to reduce the number
of families who lose benefits due to sanctions or time limits and increase the number of families
who take appropriate steps to achieve self-sufficiency.
Creating an effective balance between enforcement and engagement efforts is especially
important in implementing sanction policies. Many families who are sanctioned never appear in
the welfare office, making it difficult to know whether they have failed to comply because they
are unwilling to do so or because they do not understand what is expected of them or are
experiencing personal or family challenges that make compliance difficult. In the latter cases,
extra efforts may be required to make sure that these families understand what it is they are
required to do and that they are capable of meeting those requirements.
In an effort to encourage workers to impose sanctions when clients do not comply with
program requirements, but also ensure that sanctions are not imposed inappropriately, some
states have adopted strict review procedures that must be followed before a sanction can be
levied. For example, Utah requires a “case staffing” where the case manager, client, and key
service providers meet to develop a plan to avoid the sanction. In Tennessee, a customer review
required prior to termination of benefits reduced the number of cases sanctioned incorrectly by
30 percent (Overby 1998). Virtually all states have maintained some form of a grievance
procedure where clients can appeal sanctions they believe to be imposed in error.
The majority of states have not had any clients reach time limits yet, so there is limited
information available on how states are likely to balance enforcing time limits and promoting
compliance. As states near time limits, some are developing more intensive job search programs
for families who are nearing the time limit and have not yet found employment; the goal of these
programs is to reduce the number of families who reach the time limit without employment.
The more critical decision that states will face when families begin to reach the time limit
is when to grant extensions and when to terminate benefits. In states where TANF recipients
have reached time limits, employed recipients who have reached time limits have usually had
their cash benefits terminated. Canceling these families’ welfare grants is sometimes seen as
uncontroversial because they have a means of support and would have already lost their benefits
if earnings disregards had not been enhanced (although such families may face difficulties if they
lose their jobs and are unable to return to assistance).
There is greater variation and generally greater concern over canceling benefits of
families who reach time limits without jobs. Presumably many of these families have looked for
but been unsuccessful in finding employment; otherwise, in many states they would have already
lost benefits due to full family sanctions. In Connecticut, nearly all recipients who reached the
time limit with income below the welfare payment standard were deemed to have played by the
rules, and were granted at least one six-month benefit extension. In contrast, in Massachusetts
and Virginia, relatively few extensions have been granted (although many of the recipients who
reached the time limits in those states were employed). Some of the variation relates to the way
different states define “playing by the rules” and how they operationalize these definitions in
their processes for reviewing cases that reach time limits. A key challenge is to develop systems
that can review large numbers of cases efficiently and preserve some discretion to consider the
unique circumstances of individual families – all the while ensuring that families are treated
Providing Assistance After Sanctions or Time Limits Are Imposed
States have considerable flexibility to provide assistance to families after they have been
sanctioned or have reached a time limit. Thus, in an effort to protect vulnerable families while
maintaining stringent work requirements and an emphasis on providing cash assistance for a
limited period, some states have developed or plan to develop alternative assistance programs for
families who are sanctioned or reach time limits. For sanctioned families, these programs aim to
help families come into compliance with program requirements and return to the welfare rolls.
For families affected by time limits, these programs aim to provide families with alternatives for
meeting their basic needs.
Cuyahoga County's (Ohio) Safety Net program provides an example of the kinds of
programs that are being developed to help sanctioned families. Families who have been
sanctioned are contacted by community agencies through phone calls and home visits in an effort
to re-engage them in work activities. During the first 10 months of implementation, 46 percent
of sanctioned families referred to the program were assessed and provided with information and
services. Almost all of these families were able to participate in work activities and have their
cases re-opened as a result of the program (Goldberg and Schott 2000).
Connecticut, the state where the largest number of families have reached a time limit, has
the most highly developed system for providing assistance to families who have lost benefits due
to a time limit. Families who are terminated from assistance with income below the payment
standard (i.e., because they were deemed noncompliant) are referred to the Safety Net, a program
run by nonprofit organizations that assists these families in finding jobs and meeting their basic
needs (using vouchers in some cases). Those who are terminated with income above the
payment standard can receive temporary rental assistance and, in some cases, are allowed to
return to cash assistance if they experience an involuntary drop in income. Although families in
New York will not reach a time limit for some time, the state has already indicated that it intends
to provide vouchers to families who reach the 60-month time limit to cover their basic needs.
Another approach to providing assistance to families who reach the time limit is to
provide them with a publicly funded job. Cuyahoga County, Ohio, where several thousand
families were scheduled to reach a time limit in October 2000, has announced that subsidized
transitional jobs will be provided to employable recipients who reach the time limit and are
unable to find unsubsidized jobs. Vermont has, since 1994, implemented a “work trigger” time
limit in which recipients are required to work after 30 months on welfare, and the state provides
subsidized community service jobs to those who cannot find regular jobs. With the strong
economy (and a part-time work requirement for many recipients), very few community service
positions have been needed.
How are Sanctions and Time Limits Affecting TANF Recipients?
Sanctions and time limits may affect recipients in a variety of ways. When individuals
are sanctioned or reach a time limit, the removal of their benefits may lead to changes in their
behavior or well-being. The policies may also affect recipients through deterrence: individuals
may change their behavior to avoid being sanctioned or in anticipation of reaching a time limit.
Such effects are difficult to measure. Several states have surveyed individuals who left
welfare owing to sanctions or time limits, but it impossible to tell how these families would have
fared if they had been allowed to remain on welfare. A few rigorous random assignment studies
have examined welfare reforms that included sanctions or time limits, but all of the reforms that
were tested also included other elements (for example, expanded earnings disregards), and the
studies were not designed to isolate the impact of sanctions or time limits per se.
With these limitations in mind, this section looks first at what is known about how
sanctions and time limits affect recipients’ behavior. It then looks at the numbers of families who
have lost benefits due to sanctions and time limits and then draws some common themes from
the surveys of individuals who left welfare because of sanctions or time limits.
Effects on Recipients’ Behavior
Sanctions and time limits can clearly reduce welfare caseloads and spending when they
are imposed, but both policies are designed to generate changes in behavior – to motivate
recipients to find jobs, participate in employment-related services, and so forth.
Effects of sanctions. There is limited information on the effects of sanctions that have
been implemented under TANF. However, information from earlier studies can provide some
insight into what the effects of sanctions might be. In the absence of direct experimental data,
one way to examine the impact of sanctions is through cross-study comparisons. Findings from
the 11 programs in the National Evaluation of Welfare-to-Work Strategies (NEWWS) suggest
that programs need to enforce work-related mandates in order to achieve high rates of
participation in employment activities. In programs that did not closely monitor attendance and
rarely imposed sanctions, participation rates were only slightly higher for those subject to the
program than for those in a control group that faced no mandates. However, among the programs
that enforced mandates, higher sanctioning rates were not associated with higher participation
rates (Hamilton and Scrivener, 1999).
Another way to obtain some evidence on the impact of sanctions it to look at how
individuals respond when sanctioned. (Of course, this approach ignores those who comply with
program mandates due to the threat of a sanction – the deterrent effect mentioned earlier.)
According to caseworkers, recipients often do not take a sanction seriously until the sanction is
imposed. However, the available data suggests that most people do not comply with program
requirements even after a sanction is imposed. In Iowa, half of all sanctions were canceled
because families complied with the requirements, but most studies that have looked at rates of
compliance find rates closer to 30 percent (GAO 2000).
Another important question is whether more severe sanctions – and, in particular, full
family sanctions – generate larger impacts on behavior. Under the JOBS program, staff often
complained that partial sanctions were not severe enough to affect most recipients (especially
because, at the time, Food Stamp benefits generally rose to partly offset the cash sanction).
Three studies that attempt to address this issue conclude that more stringent sanctions do lead to
lower caseloads or increased employment exits. However, a closer look at individual state
performance suggests that full family sanctions may not be necessary to produce these outcomes.
A study that examined the impact of waiver policies on exits from welfare found that
more stringent sanction policies are associated with increased employment exits from welfare
(Hofferth, Stanhope and Harris 2000). A second study estimates that states with an initial full-
family sanction will have a caseload reduction rate that is 25 percentage points higher than states
with weak sanctions (Rector and Youssef 1999). A third study that examined the relationship
between welfare reform policies, governmental quality and caseload changes found qualitatively
similar results (Mead 2000).
A closer look at Indiana's experience suggests that full family sanctions may not be
necessary to promote high levels of employment among welfare recipients. Indiana is one of
only a few states to continue to use the JOBS sanction, eliminating only the non-compliant adult
from the grant. In 1998, Indiana outperformed all other states in its rate of employment
placements as measured for consideration for a high performance bonus. In 1999 and 2000,
Indiana ranked 13th among the 50 states in its placement rate.
Taken as a whole, these data suggest that sanctions are influencing the behavior of many
TANF recipients, but that many others have not responded even to the loss of all of their cash
assistance. Families may not come into compliance with program requirements because they are
unclear about what they must do in order to come into compliance or because they are unable to
comply due to personal or family circumstances. They may also find employment on their own
rather than return to the welfare system for support. Some, such as those who are attending
college, may choose to pursue their current activities rather than looking for work.
Effects of time limits. One could hypothesize that time limits might change behavior in
several ways. Some people might stay off welfare to avoid starting their clock. Those on
assistance might try to get off faster to “bank” some of their months. Finally, those are cut off at
a time limit might start working or otherwise change their behavior to replace the lost income.
Some believe that the “anticipatory” impacts of time limits are quite large – that many
people have resisted going onto welfare or left more quickly to avoid using up their scarce
months of assistance. Most of the evidence is anecdotal, but one study using Current Population
Survey (CPS) data estimated that time limits may account for 16 to 18 percent of the recent
caseload decline even though few people have actually reached a time limit (Grogger, 2000).
The evidence from random assignment studies tells a somewhat different story. In
evaluations of welfare reform waiver programs in Connecticut, Delaware, Florida, and Virginia,
welfare applicants and recipients were assigned at random to a program group subject to a
termination time limit (along with other reforms) or to a control group that was not. The results
were fairly consistent: all four programs generated employment impacts in the period before
program group members began reaching the time limits, but it is impossible to say how much of
this impact was attributable to time limits, as opposed to other program features. More to the
point, however, in almost all study sites, program group members were no more likely than
control group members to leave welfare before people actually began reaching the time limits
(Bloom et al., 2000a; Fein and Karweit, 1997; Bloom et al., 2000b; Gordon and Agodini, 1999).
At first glance, this pattern suggests that clients were not "banking" their months for
future emergencies. In fact, however, all four of the programs combined time limits with
expanded earned income disregards, and it is possible that two program components worked in
opposite directions: the expanded disregards held some people on welfare longer, while other
program features (including, perhaps, the time limits) caused other people to leave welfare faster.
Further analysis of data from the Florida study suggests that this may have been the case
(Grogger and Michalopoulos, 1999). In addition, a study in Vermont that was designed to
isolate the impact of a “work trigger” time limit found that the limit generated small but
statistically significant increases in employment and reductions in welfare receipt even before
anyone reached it (Bloom et al., 1998).
In any case, the random assignment studies suggest that the anticipatory impacts of time
limits are modest – and if impacts are modest for those most directly affected, it seems likely that
any impacts on welfare applications would be even smaller. Implementation studies suggest
several explanations for this pattern. First, in many of the programs, line staff did not
consistently encourage welfare recipients to bank their available months. This is, in part, a
symptom of the conflict between disregards and time limits. In essence, staff were expected to
simultaneously encourage recipients to find jobs and stay on welfare (to benefit from the
disregards) and to find jobs and leave welfare (to stop their clock). Interestingly, the one
Virginia county where staff encouraged clients to bank their months was the only site to generate
reductions in welfare receipt in the period before people began reaching the limit. (Seven states,
including Illinois, have eased the conflict between time limits and disregards by stopping the
clock during months in which a recipient is employed at least a minimum number of hours.)
Second, all of the studies examined early time-limit programs, focusing on some of the
first recipients to hear a time-limit message. Staff reported that many recipients did not believe
that time limits would be implemented. Conversely, the time-limit message affected some
control group members in each study, reducing the chances that impacts could be measured.
Third, focus groups held in several states found that many recipients were aware of time
limits – and believed they would be implemented – but had difficulty translating this awareness
into concrete short-term steps; many were understandably focused on day-to-day concerns
(paying bills, keeping teenage children out of trouble). To the extent that they discussed welfare
reform, they focused on policies such as work requirements that affected them more immediately
(Brown, Bloom, and Butler, 1997). Ironically, there is some evidence that time limits can
prolong welfare stays: recipients may adopt the time limit as their own schedule for leaving
welfare, rather than trying to leave sooner.
Finally, although it is least rigorous way to examine the question, it is clear that time
limits are not necessary to achieve welfare caseload reductions. For example, in Michigan (with
no time limit) and Oregon (with a time limit that is not emphasized), caseloads have fallen by
more than 60 percent since 1994.
There is even less evidence available to test the hypothesis that parents respond to time
limits after they reach them – by finding jobs to offset the lost income. Post-time limit follow-up
studies have found that some parents begin working or increase their hours after reaching time
limits, but it is impossible to say whether this is attributable to the cancellation of their benefits.
Only two of the random assignment studies have followed sample members beyond the
point when program group members began reaching termination time limits. In Connecticut, the
program’s impact on employment did not change when people began reaching the limit, although
most of the people who were cut off were already working while on welfare. In Florida’s pilot
Family Transition Program (FTP), employment impacts grew slightly around the time people
began reaching the time limit, but then declined. As in Connecticut, many of those terminated at
the time limit were already working. In addition, a detailed study of recipients whose grants
were canceled found that a number of those who did not work after the time limit were being
supported by family – both before and after leaving welfare – and may not have needed to work.
The Circumstances of Families Who Have Been Sanctioned or Reached Time Limits
Before discussing the results of post-welfare surveys, this section briefly discusses how
many families have been directly affected by sanctions or time limits and examines what is
known about the characteristics of such families.
The number of families affected. Although precise figures are not available, it is clear
that the number of families nationwide who have been sanctioned is much larger than the
number who been terminated from welfare because of time limits.
A recent General Accounting Office (GAO) study found that, in 1998, an average of
112,700 families each month (4.5 percent of the national TANF caseload) received reduced
benefits owing to a sanction (GAO 2000). The study also estimated that 16,000 families per
month lost benefits due to full-family sanctions in 1998. However, in any given month, many
families remain off welfare because of full-family sanctions that were imposed in earlier months.
One study estimated that 540,000 families lost benefits due to full-family sanctions from 1997 to
1999, and that 370,000 remained off assistance at the end of 1999. The same study noted that
seven states reported that sanctions accounted for one-fifth or more of their case closures in 1999
(Goldberg and Schott, 2000). Studies that examine the experiences of a cohort of recipients over
time also suggest that sanction rates are often quite high: one-quarter to one-half of families
subject to work requirements are sanctioned over a 12 to 24-month period (Hamilton and
Scrivener 1999, Fein and Lee 1999, and Pavetti et al. 1999).
In contrast, the number of families whose benefits have been canceled owing to a time
limit is still relatively small – perhaps 60,000 nationwide as of mid-2000 (unpublished data from
the Center on Budget and Policy Priorities). The figure is low for three main reasons. First, as
noted earlier, 26 states have a 60-month time limit, and no families will reach those limits until
late 2001 at the earliest; another 8 states do not have termination time limits. Thus, there are
only 17 states (most of them relatively small) in which families could have reached a time limit.
In fact, the vast majority of families who have reached a time limit are in three states:
Connecticut, Massachusetts, and Louisiana.
Second, in the states with shorter time limits, relatively few recipients are quickly using
up their months of eligibility. For example, in Virginia, of 3,051 cases that were subject to the
24-month time limit by June 1996, only 454 (15 percent) had reached the limit by June 1998.
The speed with which recipients use up their months of eligibility in a particular state is affected
by several factors including welfare grant levels and earnings disregards, time-limit exemption
rules, the nature of work requirements, and the prevalence of full-family sanctions.
Third, as noted earlier, a number of states – including Connecticut, the state where the
largest number of families have reached a time limit – have granted extensions to a substantial
fraction of the recipients who have reached time limits.
The characteristics of recipients affected. Families who are sanctioned are
heterogeneous, but most studies find that hard-to-employ families are over-represented. In
Tennessee, 60 percent of sanctioned families did not have a GED or a diploma compared to 40
percent of families who left TANF for work. In South Carolina, 36 percent of recipients who
were high school dropouts were sanctioned compared to 22 percent of high school graduates.
Long-term recipients were over-represented among sanctioned families in South Carolina.; 38
percent of long-term recipients were sanctioned compared to 21 percent of shorter-term
recipients. Several studies also indicate that many families who are sanctioned experience
personal and family challenges that may make it difficult for them to comply with program
requirements at a higher rate than other welfare recipients (Goldberg and Schott 2000; GAO
2000). These personal and family challenges include chemical dependency, physical and mental
health problems and domestic violence. Many families also experience logistical barriers such as
not being able to find child care or not having access to transportation.
The characteristics of families leaving welfare due to time limits are likely to vary from
state to state, for the reasons mentioned earlier. Four studies have compared the characteristics
of families who were cut off at time limits with those of families who left welfare before
reaching limits (or who reached time limits and received extensions). It appears, not surprisingly,
that individuals who reached time limits were more likely to have been long-term recipients
before becoming subject to the limits. Three of the studies found that families reaching time
limits had more children, on average – recipients with more children need to earn more in order
to lose eligibility for welfare. The three studies with data on housing status found that individuals
who reached time limits were more likely to live in public or subsidized housing. It is not clear
whether this findings reflects characteristics of the individuals who live in subsidized housing or
a difference in the financial work incentives they face.
At the same time, it is also clear that the families who are cut off at time limits are not
necessarily the most disadvantaged. For example, in Connecticut, where relatively large
numbers of extensions were granted, those who were cut off at the time limit were less
disadvantaged, on average, than those reached the limit but received extensions. Similarly, in
Florida’s pilot FTP, some of the participants who faced the most serious barriers to employment
received medical exemptions and did not reach the time limit.
In-depth interviews with about 50 people who were terminated at FTP’s time limit reveal
some diversity that is masked by aggregate data. While many of the interviewees had tried
without success to find jobs while on welfare, others had been attending college (sometimes
without the program’s knowledge) or were supported by a parent or partner and did not see an
urgent need to work (FTP did not use full-family sanctions during most of the study period).
Also, administrative data showed that nearly 30 percent of those who reached the time limit had
worked in all four quarters of their last year on welfare; many of these individuals would have
become ineligible for welfare earlier had it not been for the expanded earned income disregard.
Post-welfare circumstances. There are two sets of data available to assess the
circumstances of families who left welfare because of sanctions or time limits. The two random
assignment studies that followed their samples past termination time limits found similar results:
both programs increased average income before program group members began reaching the
tlimits (owing to the presence of expanded disregards), but the income gains disappeared in the
post-time limit period and, in fact, both programs seem to have reduced income for small groups
of families during that period. However, the Florida FTP study, which is complete, found no
evidence that the program increased material hardship.
The other source of data are a group of targeted “welfare leavers” studies focusing on
families who left welfare owing to sanctions or time limits (see Blank and Schmidt in this
volume for a broader review of leaver studies). Respondents were typically surveyed 6 to 18
months after exit (the best surveys achieved response rates of at least 75 percent) and/or tracked
with administrative records. Most of the studies either compared respondents’ circumstances
before and after leaving welfare (in most cases, by asking respondents to remember their lives
while on welfare) or compared the post-welfare circumstances of sanction or time-limit families
to the circumstances of “voluntary” welfare leavers. Although clearly better than having no
comparison, neither of these approaches reliably assesses the impact of being cut off welfare.1
The survey findings are mixed, but several themes emerge: First, a substantial proportion
of the families who left welfare due to sanctions or time limits are struggling. Although each
survey measured hardship differently, anywhere from one-third to one-half of respondents
reported that they were having trouble making ends meet. Between 15 percent and 25 percent
reported that they sometimes or often did not have enough to eat or were rated as “severely food
insecure” according to a widely-used scale. On the other hand, it is not clear that the sanction
1 Studies summarized in this section include: Bloom et al., 2000b; Gordon et al., 1999; Hunter-Manns
and Bloom, 1999; Massachusetts Department of Transitional Assistance, 2000; Melton and Bloom, 2000;
Richardson et al., 1999; Richardson et al., 2000, GAO 2000, Nixon et al. 1999.
and time-limit families are experiencing much greater levels of hardship than other, broader
groups of welfare leavers; the difference, of course, is that time-limit and sanctioned leavers may
not be eligible for further assistance. Finally, it is reassuring that extreme housing-related
hardship has been rare: no study that examined the post-welfare circumstances of families
affected by sanctions or time limits found rates of homelessness above three percent.
Second, while employment rates are relatively high among most categories of leavers,
sanctioned leavers are substantially less likely to work than individuals who left welfare for other
reasons; this is consistent with the fact that hard-to-employ families appear to be over-
represented, as discussed earlier. Summarizing data from nine studies of sanctioned families,
GAO estimates that an average of 41 percent of sanctioned families were working after they left
TANF (GAO 2000). Studies that compare employment rates for sanctioned families and other
welfare leavers find as much as a 27 percentage point difference in the rate of employment
among sanctioned and non-sanctioned leavers. For example, 40 percent of sanctioned families
in Arizona were working one quarter after leaving welfare compared to 55 percent of non-
sanctioned leavers; the comparable figures in South Carolina were 31 percent for sanctioned
families and 58 percent for other leavers (Goldberg and Schott 2000).
The time-limit story is more complicated, and depends on the design and implementation
of each state’s limit. For example, in Connecticut, where most people who reached the time limit
without jobs were granted extensions, the employment rate among time-limit leavers (about 80
percent) is much higher than among other leavers. A similar result was found in Virginia, where
recipients were required to work long before reaching the time limit (and were given workfare
slots if necessary), and were subject to full-family sanctions if they failed to cooperate; thus,
those who remained on welfare for 24 months were quite likely to be working. Most of the
studies found that the employment rate among time-limit leavers increased somewhat in the post-
time limit period although, once again, it is hard to know whether this is related to the
termination of welfare benefits.
Third, despite the focus on employment as the key indicator of post-welfare well-being,
the studies provide mixed evidence about whether families in which the former recipient is
employed are in fact better off. Employed families typically report higher average income, but
not necessarily lower levels of hardship. In North Carolina and Virginia, for example, time-limit
leavers who were employed were more likely to report that they were having trouble paying for
food. Similarly, in Florida, employed respondents were less likely to have health coverage and
more likely to experience food insecurity. As noted earlier, many of the parents who were not
employed received substantial assistance from family. It is not clear whether they received this
help because they were unable to work, or whether the assistance made it possible for them not
to work. There is also no way to know how long the assistance will remain available. Finally, it
is important to note that many former recipients are not necessarily supporting their households
alone: the post-time limit studies found that 35 percent to 44 percent of respondents were living
with at least one other adult when interviewed.
Fourth, it is clear that, in addition to family supports, public assistance programs –
notably Food Stamps, housing assistance, and Medicaid – are playing a critical role in supporting
at least some of the families who were sanctioned or reached time limits. Without these non-time
limited supports, levels of serious hardship might be much higher.
In Connecticut, Florida, and Massachusetts, time-limit leavers were much more likely
than other leavers to continue receiving Food Stamps after leaving welfare. In addition, between
one-sixth and one-third of respondents in the post-time limit surveys reported that they had
received food from a food pantry, soup kitchen, or similar source since leaving welfare.
The situation for sanctioned families is more complicated. States have the option to
reduce a non-compliant family’s Food stamp benefit and eliminate Medicaid for the non-
compliant adult, however, it is unclear how often these benefits are affected. Seven studies that
examined continued participation in Food Stamps and Medicaid found relatively high rates of
participation in both programs, from 57 to 71 percent of sanctioned recipients received Food
Stamps and 59 to 88 percent received Medicaid (GAO, 2000). However, there is speculation that
welfare reform in general, and sanction policies in particular may be contributing to steep
declines in participation in the Food Stamp and Medicaid programs (Dion and Pavetti 2000).
In short, the rather limited data available to date suggest that most of the families who
have left welfare owing to full-family sanctions and time limits are struggling and far from self-
sufficient, but that the policies do not appear to have caused widespread, severe hardship. The
critical question is whether families will be able to continue to make ends meet without cash
assistance over the long-term, particularly if economic conditions worsen.
Shaping a Post-TANF Agenda for Sanction and Time Limit Policies
Prior to the passage of PRWORA and the creation of TANF, there was considerable
debate and concern over time-limiting cash assistance for families with children. Relying on the
past behavior of welfare recipients, experts estimated that large numbers of families could
potentially be affected by time limits. In stark contrast, sanctions received little attention. Now
that both policies have been implemented, there is clear evidence that many more families have
lost cash assistance due to sanctions than time limits. While this could change as time limits
draw near in more states, it appears that some of the families that many feared would be affected
by time limits may have already left the welfare rolls due to full family sanctions.
And yet, there still appears to be much greater attention focused on the circumstances of
the relatively small number of families who have lost benefits due to time limits than on the
much larger number who have been affected by sanctions. Because the design and
implementation of a state’s sanction policy will significantly influence who is affected by time
limits, examining the effects of time limits without taking into account who has already lost
benefits due to sanctions presents a skewed picture. In short, any assessment of the impact of
one of the policies must consider the influence of the other.
Given that we have much left to learn about how sanctions and time limits are
influencing families’ choices and affecting their lives, it is difficult to identify specific policy
changes that should be considered during the reauthorization of TANF – particularly because the
current structure already provides states with considerable flexibility in both areas.
However, the information available on the implementation of and outcomes associated
with sanctions and time limits does suggest a need to focus on two broad policy issues: how to
address the needs of hard-to-employ families and how to support working families within a time-
limited environment. The available studies show that families who lose benefits due to sanctions
are harder-to-employ than welfare recipients as a whole, while many families who reach time
limits are employed, but not earning enough to make ends meet.
Over the past decade, most states have implemented work first programs that require all
or most able-bodied welfare recipients to find work as quickly as possible. Most research
suggests that the combination of these programs, the policies that supported them, and a strong
economy encouraged large numbers of welfare recipients, including some who had been on
welfare for long periods of time, to enter the labor market. But most states are only now
beginning to explore alternative employment strategies for families who need more assistance to
make the transition to employment than most work first programs provide. While states are free
to implement these strategies and to use TANF funds to do so, many of these activities, such as
job readiness programs that last longer than six weeks or substance abuse and mental health
treatment cannot count toward a state’s work participation rate, possibly making some states
hesitant to implement them. Adoption of a broader range of employment-related activities could
potentially help to address some of the needs of families who have lost benefits due to sanctions,
as well as addressing the needs of longer-term recipients who remain on the welfare rolls and
may be affected by time limits.
With a broader range of allowable activities in place – and with full-family sanctions
already in effect in most states to enforce participation requirements – Congress might even
reconsider the need for a federal time limit (states would, of course, be free to continue their own
time limits), or at least consider modifications to the 20 percent exemption to make it easier for
states to continue benefits beyond the 60-month point for families engaged in activities designed
to promote self-sufficiency. As discussed earlier, it seems clear that time limits are not a
necessary ingredient for successful state welfare reforms and, in a system in which recipients are
seriously required to work or engage in other productive activities, time limits seem most likely
to affect the recipients who least need a “push” – those facing the most serious barriers to
employment, and those who are working but unable to earn their way off assistance.
The latter group is most directly affected by the conflict between time limits and
disregards, discussed earlier. A few states have already addressed this conflict by stopping the
clock for working families, but most will simply stop providing assistance to working families
when they reach the time limit, undercutting the income-enhancing effects of their disregards and
leaving families vulnerable if they lose their jobs.
While stopping the time limit clock for working families is a creative solution to the
conflict between time limits and disregards – and perhaps should be more strongly encouraged
under federal TANF policies – it is only one of the possible ways to assist the working poor.
Although the welfare system today is quite different than the system in place prior to PRWORA,
in most states, it remains focused on addressing the immediate basic needs of non-working
families. As discussed elsewhere in this volume (see, for example, Michalopoulos and Berlin), it
remains an open question whether the current system can be further transformed so as to be more
responsive to the needs of working families or whether we should be seeking alternative
strategies for assisting these families.
Bloom, Dan, Laura Melton, Charles Michalopoulos, Susan Scrivener, and Johanna Walter. 2000a. Jobs First:
Implementation and Early Impacts of Connecticut’s Welfare Reform Initiative. MDRC.
Bloom, Dan, Charles Michalopoulos, Johanna Walter, and Patricia Auspos. 1998. WRP: Implementation and Early
Impacts of Vermont’s Welfare Restructuring Project. MDRC.
Bloom, Dan, James Kemple, Pamela Morris, Susan Scrivener, Nandita Verma, and Richard Hendra. 2000b. FTP:
Final Report on Florida’s Initial Time-Limited Welfare Program. MDRC.
Brown, Amy, Dan Bloom, and David Butler. 1997. The View from the Field: As Time Limits Approach, Welfare
Recipients and Staff Talk About Their Attitudes and Expectations. MDRC.
Dion, Robin and LaDonna Pavetti. 2000. Access to and Participation in the Food Stamp Program. Washington,
D.C.: Mathematica Policy Research.
Edelhoch, Marilyn, Qiduan Liu and Linda Martin. 1999. The Post-Welfare Progress of Sanctioned Clients: A
Study Using Administrative and Survey Data to Answer Three of Four Important Questions. Columbia, South
Carolina: South Carolina Department of Human Services.
Fein, David and Jennifer Karweit. 1997. The ABC Evaluation: The Early Economic Impacts of Delaware’s A Better
Chance Welfare Reform Program. Abt Associates.
Fein, David and Wang Lee, 1999. The ABC Evaluation: Carrying the Using the Stick: Financial Sanctions in
Delaware’s A Better Chance Program. Abt Associates.
Fraker, Thomas, Lucia Nixon and Carol S. Prindle. 1997. Iowa's Limited Benefit Plan. Washington, D.C.:
Mathematica Policy Research, Inc. and Iowa City, Iowa: The Institute for Social and Economic Development.
Goldberg, Heidi, and Liz Schott. 2000. A Compliance-Oriented Approach to Sanctions in State and County TANF
Programs. Center on Budget and Policy Priorities.
Gordon, Anne, and Carole Kuhns, Renee Loeffler, and Roberto Agodini. 1999. Experiences of Virginia Time Limit
Families in the Six Months After Case Closure: Results for an Early Cohort. Mathematica Policy Research.
Gordon, Anne and Roberto Agodini. 1999. Early Impacts of the Virginia Independence Program: Final Report.
Mathematica Policy Research.
Grogger, Jeff. 2000. Time Limits and Welfare Use.” Unpublished paper.
Grogger, Jeffrey, and Charles Michalopoulos. 1999. “Welfare Dynamics Under Time Limits.” NBER Working
Paper No. W7353. National Bureau of Economic Research.
Hamilton, Gayle and Susan Scrivener. 1999. Promoting Participation: How to Increase Involvement in Welfare-to-
Work Activities. MDRC.
Hofferth, Sandra L., Stephen Stanhope and Kathleen Mullan Harris. 2000. "Exiting Welfare in the 1990s: Did
Public Policy Influence Recipients' Behavior?" Unpublished Paper.
Hunter-Manns, JoAnna and Dan Bloom. 1999. Connecticut Post-Time Limit Tracking Study: Six Month Survey
Massachusetts Department of Transitional Assistance. 2000. After Time Limits: A Study of Households Leaving
Welfare Between December 1998 and April 1999.
Mead, Lawrence M. 1985. "Expectations and Welfare Work: WIN in New York State," Polity 18, no. 2 Winter):
Mead, Lawrence M. 1997. "Optimizing JOBS: Evaluation versus Administration," Public Administration Review
57, no. 2 (March/April). 113-23.
Mead, Lawrence M. 2000. “Governmental Quality and Welfare Reform.” Paper presented at the 2000 Annual
Meeting of the American Political Science Association, Marriott Wardman Park, August 31-September 3.
Melton, Laura and Dan Bloom. 2000. An Analysis of Welfare Leavers in Connecticut. MDRC.
Nixon, Lucia A., Jacqueline F. Kauff and Jan L. Losby. 1999. Second Assignments to Iowa’s Limited Benefit Plan.
Washington, D.C.: Mathematica Policy Research, Inc.
Olson, Krista and LaDonna Pavetti. 1996. Personal and family Challenges to the Successful Transition from Work.
Washington, D.C.: The Urban Institute.
Overby, Russell. 1998. Summary of Surveys of Welfare Recipients Employed or Sanctioned for Noncompliance.
Memphis TN: University of Memphis.
Pavetti, LaDonna, Nancy Wemmerus and Amy Johnson. 1999. Implementation of Welfare Reform in Virginia: A
Work in Progress. Washington, DC: Mathematica Policy Research, Inc.
Rector, Robert E. and Sarah E. Youssef. 1999. The Determinants of Welfare Caseload Decline. Washington, D.C.:
The Heritage Foundation.
Richardson, Philip, Richard White, Mark Tecco, Susan LaFever, and Mary Ann Kerttula. 1999. Study of Families
Leaving Work First After Reaching the 24-Month Time Limit. MAXIMUS.
Richardson, Philip, Kim Reniero, Mark Tecco, Susan LaFever, and Greg Schoenfeld. 2000. Study of Families
Leaving Work First Due to Time Limits: Results of the Second Follow-Up Surveys. MAXIMUS.
Schott, Liz. 2000. Ways States Can Serve Families That Reach Welfare Time Limits. Washington, D.C.: The
Center on Budget and Policy Priorities.
State Policy Documentation Project. 2000. www.spdp.org. Website administered by the Center on Budget and
Policy Priorities and the Center on Law and Social Policy.
Thompson, Terri S., Pamela A. Holcomb, Pamela Loprest and Kathleen Brennan. 1998. State Welfare-to-Work
Policies for People with Disabilities: Changes Since Welfare Reform. Washington, D.C.: The Urban Institute.
U.S. General Accounting Office. 2000. Welfare Reform: State Sanctions Policies and Number of Families Affected.
U.S. Department of Health and Human Services, Office of Inspector General. 1999. Temporary Assistance for
Needy Families: Educating Clients about Sanctions. .