While much attention has been paid to efforts to help welfare recipients
find jobs, policymakers and program administrators are increasingly
focused on designing interventions that will help welfare
recipients stay employed over the long-run.
Past research has shown many welfare recipients have
relatively unstable work patterns. While many do find work, some eventually lose
their jobs and return to the welfare rolls. In addition, past experience indicates that most welfare recipients
work in relatively low quality jobs with low wages, few benefits,
and little opportunity for advancement. Because of time limits
on the receipt of federal cash assistance through the Temporary
Assistance for Needy Families (TANF) program, there is a new
urgency to develop programs and policies that will help individuals
find and keep jobs and stay off the rolls.
To
help policymakers design programs that will help individuals
sustain employment and potentially move up into better jobs,
it is necessary to understand the general patterns of employment
for welfare recipients who find jobs and what factors may
be related to achieving success.
Towards this end, this paper provides a description
of the employment patterns of welfare recipients over a four-year
follow-up period and the employment patterns and characteristics
of those who are most and least successful in sustaining employment.
This paper adds new insights into the employment patterns
of welfare recipients because of unusually rich and detailed
data sources on both the characteristics of individuals who
find jobs, the type of jobs they obtain, and their overall
employment patterns. It
also has a relatively long follow-up period of up to four
years and tracks a diverse range of welfare recipients.
The National Evaluation of Welfare-to-Work Strategies
(NEWWS) is a large scale random assignment evaluation examining
the effectiveness of 11 mandatory welfare-to-work programs
operated in seven locales. Three of the sites in the evaluation –
Atlanta, Grand Rapids, and Riverside – operated two different
programs simultaneously.
One was an employment-focused program known as “Labor
Force Attachment” (LFA) which emphasized moving individuals
quickly into jobs and the other was an education-focused program
known as “Human Capital Development” (HCD) which
emphasized investing in education and training before entering
the labor market. In
each site, to measure the effectiveness of the programs, individuals
were randomly assigned to a program group which received services
and to a control group that did not. The effects of the program are measured by
comparing outcomes of the program group and the control group.
Individuals were randomly assigned to research
groups over approximately a two-year period in each site. Random assignment for the sample studied in
this paper began in June, 1991, in Riverside, and ended in
December, 1993, in Columbus and Portland.
The results presented in this paper cover the calendar
period of June, 1991 (the first sample member’s entry
into the study), through December, 1997. Unlike under TANF, individuals in the studied
programs did not face a time limit on eligibility for welfare
assistance for most of the follow-up period for this study. The programs, however, shared TANF’s primary
goal of moving welfare recipients into paid work and off assistance.
This paper provides a descriptive analysis of
employment outcomes based on data for program group
members only from six of the NEWWS evaluation sites. Overall, the analysis reports on the experiences
of 17,055 program group members.
Each site is weighted equally and results are presented
separately for employment-focused programs (Riverside LFA,
Grand Rapids LFA, Atlanta LFA, and Portland) and education-focused
programs (Riverside HCD, Grand Rapids HCD, Atlanta HCD, Detroit,
and Columbus). Because
the results for employment-focused and education-focused programs
are similar, in most cases the employment patterns for individuals
enrolled in an employment-focused program are presented and
discussed in the text and the results for individuals enrolled
in education-focused programs are presented in the Appendix.
This paper uses two primary data sources.
The first is Unemployment Insurance (UI) records which
provide quarterly employment and earnings data over a four-year
follow-up period. These
records include employment obtained both within and outside
of the counties in which sample members were randomly assigned.
UI earnings records, however, are not available for
out-of-state earnings or for jobs that are not usually covered
by the UI system, such as self-employment, domestic service,
or informal child care — work which may have been “off
the books” — or for employers who do not report
earnings. Further, UI records report earnings by calendar
quarter — and do not provide start and end dates of employment
— and thus often overstate how long people actually worked. In addition, when using UI records to track trends in earnings over
time, one cannot distinguish among the several changes in
job characteristics (number of hours or weeks of work or hourly
wages) that may have affected quarterly earnings.
The second data source — used in selected
analyses in this paper — is a survey administered two-years
after random assignment for a subsample of 4,709 program group
members. The survey provides more detailed information
on employment and other issues but for a shorter two-year
follow-up period.
This
section examines employment patterns for the sample of program
group members who found jobs during the four-year follow-up
period. Overall, many
sample members worked at some point during the four-year follow-up
period: 81 percent
of employment-focused program group members and 78 percent
of education-focused program group members worked for pay
at some point during this time (results not shown).
(While this paper primarily focuses on the experiences
of sample members who found jobs, the characteristics of the
group of welfare recipients who did not find jobs are discussed
briefly later in the paper).
A. Duration of Employment over the Follow-up Period
An
important measure of success in sustaining employment is the
number of quarters program group members were employed over
the four-year follow-up period. As discussed above, because
this measure is calculated with quarterly UI data, the duration
of employment may be overestimated. People who began working
or changed jobs during a calendar quarter probably experienced
weeks of joblessness during the quarter, which UI records
do not capture.
As
shown on Table 1, employment-focused
program group members who worked during the follow-up period
were employed for an average of 8.5 quarters – or just
over half of the four-year (or 16 quarter) follow-up period.
Education-focused program group members were employed for
slightly less time – 8.1 quarters. A vast majority (60
to 70 percent) started working during the first year after
they enrolled in the welfare-to-work program.
The
average of quarters employed includes an initial period of
joblessness when many program group members were participating
in program activities designed to help them find jobs. In
theory, education programs may require a longer initial investment
of time – with the hope that they will result in more
stable and/or higher paying employment in the long run. As
shown on Table 1, individuals in employment-focused
activities typically started working after 2.9 quarters while
those in education-focused activities started working after
a slightly longer period – 3.6 quarters. To “level
the playing field” and judge employment and education-focused
on a comparable basis, it is useful to examine the proportion
of the follow-up period employed once individuals found jobs.
Table 1 shows the results for the two
types of programs are similar – individuals in both types
of programs were employed for almost two-thirds of the remaining
follow-up period on average.
Overall, the NEWWS sample members experienced a diverse range of work
patterns. A significant
portion of the program group members worked very little during
the follow-up period while a roughly similar proportion worked
for a large majority of the four-year period (see Table
1). Individuals
fell equally into four major groups.
-
Approximately
one-quarter were employed for 25 percent or less of the
four-year follow-up period (1-4 quarters).
Roughly half of these individuals (15 percent)
were employed in only two quarters or less of the follow-up
period (not on table).
-
Approximately
one-quarter were employed between 26 and 50 percent of
the follow-up period (5-8 quarters).
-
Approximately
one-quarter were employed between 51 and 75 percent of
the follow-up period (9-12 quarters).
-
The
remaining quarter were employed for more than 75 percent
of the four-year follow-up period (13-16 quarters).
Very few – roughly 8 percent — were employed
in all quarters of the follow-up period (not on table).
Later
sections of this paper examine factors that are related to
whether individuals are more or less successful in sustaining
employment.
B.
The Number and Duration of
Employment Spells
Another
important issue in understanding the post-employment experiences
of welfare recipients is whether individuals worked continuously
or in multiple spells with significant breaks in employment.
As discussed above, because this measure is calculated using
quarterly UI data, the number of spells may be underestimated.
Breaks in employment that do not span at least an entire quarter
are not reflected by UI data.
This
analysis shows that the vast majority of program group members
had one or two spells of employment within the four-year period.
As shown on Table 1, a significant portion
(43 percent) of the sample members enrolled in employment-focused
programs had only one spell of employment. However, over
one-third had two spells of employment, 17 percent had three
spells, and six percent had four or more spells.
Figure
1 provides more detail on employment spells by tracking the
employment experiences of 100 typical employment-focused program
group members through different employment spells over the
four-year follow-up period (results for education-focused
program group members are presented in the Appendix). This
analysis focuses on sample members who found jobs during the
follow-up period and thus 100 of them have an initial spell
of employment. As shown, roughly one-fifth (22/100) of these
individuals stayed employed until the end of the follow-up
period. The remainder eventually became jobless, with approximately
three-quarters (57/78) going into a second (see
Figure 1) employment spell and one-quarter
(21/78) not working again during the follow-up period. Similar
patterns are observed over the remaining spells of employment.
Of those working in a second spell of employment, approximately
one-third (20/57) remained employed through the end of the
follow-up period. Of those who lost this job, roughly two-thirds
found another job (24/36) and one-third never worked again
(13/36) during the follow-up period.
The
analysis also examined how long individual spells of employment
lasted. Table 2 shows that the
first employment spell for employment-focused program group
members lasted 5.3 quarters on average.
However, almost one-quarter of those who had one spell
of employment were still working at the end of the follow-up
period. Thus, the
first employment spell is truncated and underestimates the
actual length of the spell.
(Results for other employment spells are similar.)
Table 2 also shows that most of those who ended
an employment spell did so relatively quickly.
The first spell of employment lasted 3.5 quarters on
average for those who stopped working — and roughly one-third
were employed for a quarter or less. Those who did find a job after they lost their
initial job did so after a relatively long period of unemployment
– 2.8 quarters. (Results
for other employment spells are similar.)
Overall, while some individuals do experience multiple spells of employment
over a four-year period, most (over three-quarters) have only
one or two spells of employment during this period.
Individuals who ended their employment spells did so
after a relatively short period of working.
While some did go on to subsequent jobs, it was usually
after a relatively long joblessness spell.
C. Earnings and
“Successful” Employment Spells
Equally
important as the duration and continuity of employment is
the quality of the jobs in which welfare recipients work.
The quality of jobs over the four-year follow-up period
is examined using a number of measures.
Because UI records do not provide wages and hours worked,
we first examine the extent to which program group members
earned $2,500 or more per quarter – roughly equivalent
to working full-time at the current minimum wage. It should be noted that quarters with low earnings
may also reflect situations when an individual worked only
a portion of the quarter because they started or stopped a
job.
On
average, individuals earned less than $2,500 per quarter,
indicating that many individuals were working less than full-time
or at low wages. As shown on Table 3,
employment-focused program group members who worked earned
$2,315 on average per quarter of employment.
Average earnings were also very low for a substantial
portion: roughly 30 percent averaged less than $1,000 per
quarter. Only 20 percent
averaged earnings of $3,000
or more per quarter.
While
earnings levels were relatively low, program group members
did experience some growth in earnings over the four-year
follow-up. For example, average earnings in the first measured quarter of employment
for employment-focused program group members
were
$2,178 and climbed to $2,847 during the last measured quarter
– an increase of 30 percent (this analysis excludes quarters
when individuals start and stop employment).
Because UI records do not provide information on hours
worked or hourly wage levels, it cannot be determined whether
the earnings gains occur through an increase in wages or an
increase in hours worked, or both.
We
also examined whether individuals experienced at least one
employment spell that was “successful” based on
its duration and earnings. Overall, roughly 30 percent were employed in at least one spell
that lasted over 4 quarters and where individuals earned $2,500
or more per quarter. The
extent to which individuals had experienced “success”
by the end of the follow-up period was also examined.
A significant portion – 40 percent – were
employed in all four quarters of year 4. However, only twenty-seven percent earned $10,000
(equivalent of full-time, minimum wage employment for an entire
year) in the fourth year of follow-up.
Overall, while some program group members experienced
at least one successful employment spell (based on a range
of measures), many did not.
IV. Patterns
of Employment, Earnings, and Welfare Receipt for the Most and
Least Successful in Sustaining Employment
This
section examines the patterns of employment, earnings, and
welfare receipt (see Tables 4 and 5)
for employment-focused program group members who were more
and less successful in sustaining employment.
Results for education-focused program group members
are presented in the Appendix and
are generally similar.
Most Successful (worked for over 75 percent of the four-year
follow-up period): Members of this group found a job relatively
quickly – on average they were unemployed less than half
a quarter initially. Over
one-half worked in one spell of employment and one-third worked
in two. Very few were able to work continuously through multiple employment
spells. Over 90 percent were still employed at the end of
year 4. (Table 4)
On average, the group who was most successful in sustaining employment
had relatively high earnings compared to other program group
members. Their earnings averaged $3,000 per quarter
over the follow-up period and close to 20 percent averaged
$4,000 or more per quarter (see Table 5).
Members earned $2,500 or more per quarter in over half
of the quarters they worked. They also experienced a substantial growth
in earnings over the four-year follow-up period of close to
50 percent, although it cannot be determined whether this
earnings growth occurred through an increase in hourly wages
or hours worked, or both.
Those
who were most successful in sustaining employment continued
to rely on public assistance when they obtained jobs, although
to a much lesser degree than the other groups (Table 5).
They received cash assistance and Food Stamps in 33
percent and 47 percent, respectively, of the quarters in which
they worked. These data may not always reflect instances where
individuals are combining work and welfare.
Rather, they could reflect situations where an individual
leaves a job and goes back on welfare.
A majority – over 60 percent
– received cash assistance in the months they
were not working (a result that did not vary much across the
four groups).
Moderately Successful (employed between 51 and 75 percent of the
follow-up period): The majority of this group started
working in the first year of the follow-up period, although
some became employed in the second year. They
were much more likely to experience multiple spells of employment.
One-third had two spells of employment and over one-third
had three or more. This group experienced a moderate level of earnings (averaging $2,129
per quarter) and some earnings growth (approximately 25 percent). This group had higher rates of cash assistance
and Food Stamps receipt in the quarters in which they worked
than those who were most successful.
Limited Success (employed between 26 and 50 percent of the follow-up
period): These program group members exhibited similar
employment patterns to those who were moderately successful: nearly three-fourths of this group had two
or more spells of employment.
However, this group had lower levels of earnings ($1,553
per quarter), minimal earnings growth (8 percent), and used
public assistance programs at higher rates when they were
working.
Least Successful (employed 25 percent or less of the follow-up
period): Members
of this group experienced the longest spells of initial joblessness. They started working well into the follow-up
period (i.e. during the second year of follow-up period). Almost two-thirds of this group worked in only
one employment spell over the follow-up period. One-quarter of these individuals started working so late in the
follow-up period that they were still employed at the end
of the follow-up period, indicating the actual length of time
employed for this group is somewhat truncated.
The least successful group had very low earnings on
average ($1,093 per quarter), very little earnings growth,
and high rates of cash assistance and Food Stamps receipt
when they were working (these benefits were used in over two-thirds
of the quarters they were working).
These results are likely to reflect that the least
successful group often was not employed throughout the entire
quarter.
In sum, this analysis clearly shows that those
who were most successful in sustaining employment exhibited
very different employment and earnings patterns than those
who were moderately and least successful.
Interestingly, they achieved their success by staying
in their initial employment spell or experiencing only one
break in employment. This
group also experienced substantial earnings growth over the
follow-up period. Those
who were least successful also only had one spell of employment
but they were unable to continue working over the long-run.
This group had very low earnings and, accordingly,
had relatively high rates of participation in cash assistance
and Food Stamps. The
middle groups were more likely to experience multiple spells
of employment and joblessness and had moderate levels of earnings
and earnings growth.
V.
Individual
and Job Characteristics for the Most and
Least Successful in Sustaining Employment
It
is important for policymakers and program administrators to
understand which factors are associated with sustained and
with sporadic employment. To shed light on this issue, this section examines
several key questions for groups who were successful, moderately
successful, had limited success, and were least successful
in sustaining employment.
First, do these groups differ in background characteristics
likely to affect success in the labor market, such as educational
attainment or prior work experience? Second, did members of the most successful
group obtain better jobs initially, i.e. with higher hourly
wages and fringe benefits?
Finally, was the most successful group more likely
to use transitional Medicaid and child care?
The
findings described here should be seen as suggestive of the
factors that are important to sustaining employment and do
not establish a causal relationship.
This analysis also does not assess which job characteristics
or individual characteristics are most important in
determining whether an individual can sustain employment. For example, more job-ready individuals may
find better first jobs, which enables them to sustain employment.
A.
Individual Characteristics
Table
6 summarizes variation among the four groups across key
demographic characteristics. It also shows demographic characteristics for
those who did not find jobs during the follow-up period –
these are discussed later in the paper.
Demographic data were collected at the time of random
assignment.
Education
and basic skills. The
most successful group had the highest levels of education
and skill levels. Almost three-quarters had received their high
school diploma or GED before random assignment (compared to
55 percent for those who were least successful).
Most of the differences across groups is explained
by the receipt of a high school diploma, not a GED.
The group who was most successful in sustaining employment
was also twice as likely to have a technical or two-year degree
(10 percent vs. 4 percent). Differences across skill levels were significant
but not as large: 62
percent of those who were most successful in sustaining employment
passed both the reading and math literacy test compared to
52 percent who were least successful.
Employment
and welfare history.
Work history is also an important factor in an individual’s
ability to sustain employment.
Compared to the least successful group, those who were
most successful in sustaining employment were more than twice
as likely to have a recent work history at the time of random
assignment. Forty
percent (compared to 16 percent) of this group were employed
in the quarter before random assignment.
Sixty-two percent had earnings in the year before they
were randomly assigned, compared to 34 percent who were least
successful. Differences
across the subgroups in prior welfare history were not large,
although those who were least successful were somewhat more
likely to be long-term recipients (received welfare five years
or more). Interestingly,
those who were most successful in sustaining employment were
somewhat more likely to live in subsidized housing.
Personal
barriers. Variation across a range of personal barriers
to employment — depression, family and personal problems,
family attachment (measuring a preference to stay home with
children rather than returning to work), and locus of control
(measuring the degree to which a person feels in control of
her life) – was examined by responses to a Private Opinion
Survey, completed at random assignment. As expected, those who were most successful
in sustaining employment had fewer personal barriers to employment
compared to those who were least successful.
There were major differences between the two groups
in terms of the proportion who reported family or personal
problems (10 percent vs. 20 percent), family attachment (20
percent vs. 33 percent), and locus of control (17 percent
vs. 26 percent). Differences
were not as great on the depressive symptoms scale, with between
33 and 40 percent of program group members
having a high or moderate barrier.
Other
demographic characteristics.
Among those welfare recipients who found work, there
were not notable differences across the groups in terms of
age, ethnicity, marital status, and number and age of children.
B.
Initial Job Characteristics
and Other Employment-Related Issues
Table
7 summarizes variation in initial job characteristics
as well as in the use of transitional Medicaid and child care
benefits for those who were most and least successful in sustaining
employment.
Initial
job characteristics.
Wages, benefits, and hours of the initial job
individuals obtained were examined through a survey conducted
two years after random assignment. The initial job is examined in detail because
program operators may be able to influence the initial job
individuals take when they are on cash assistance.
For example, the NEWWS evaluation in Portland produced
large impacts on earnings and employment in part by encouraging
individuals to take “good” jobs. In addition, as shown above, the individuals
who were most successful in sustaining employment were often
still employed in their initial spell of employment at the
end of the follow-up period, indicating that the characteristics
of this job may have some effect on whether individuals are
able to keep working.
-
Hours
worked. There was not a large variation in terms of the average number
of hours worked in the initial job after random assignment,
although there were some differences in the distribution
of hours worked. The group who was most successful in sustaining
employment was more likely to work in full-time jobs:
-
60
percent were working 40 or more hours per week (compared
to roughly 48 percent for other groups). The group who was least successful, compared to the most successful,
also had a higher proportion working 20 hours or less
per week (18 percent vs. 6 percent).
-
Wages.
The average wage in the initial job was not substantially
different across the groups:
the average wage for the least successful group
was $6.34 per hour and the wage for the most successful
group was $6.76 per hour.
In addition, the wages for the least successful
group were actually higher on average than those who had
moderate or limited success in sustaining employment.
This could indicate that the least successful group
was more likely to consider the pay when making a decision
to take a job. A
combination of moderately higher hours and wages resulted
in the most successful group having higher average earnings
per week.
-
Use
of transitional child care and other child care issues.
The group who was most successful in sustaining employment
was also more likely to use transitional child care, although
their usage rates were relatively low given that members
averaged two-thirds of their quarters of employment off
of welfare (see Table 5).
Approximately 22 percent used this benefit within
two years, compared to 12 percent in the least successful
group. The group
that was most successful in sustaining employment was
more likely to pay out of pocket for child care expenses
(in part this could be because they worked and earned
more). There were
no major differences across the groups in the extent to
which they experienced child care problems that caused
them to be absent or late for work.
-
Employer-sponsored
health insurance and other benefits.
Of the group who was most successful in sustaining
employment, over twice as many had employer-provided medical
insurance in their initial job, compared to those who were
least successful (44 percent vs.17 percent). The group that was most successful was also
over twice as likely to have other employer benefits such
as paid sick and vacation leave.
-
Use
of transitional Medicaid.
Usage rates for transitional Medicaid were twice as
high for the group who was most successful in sustaining employment.
Forty–seven percent used transitional Medicaid at some
point during the two-year follow-up period, compared to 26
percent for the group who was least successful.
The findings here on the role of transitional Medicaid
in sustaining employment are particularly noteworthy.
While it is likely that both individual characteristics
(such as work experience or educational attainment) and finding
“good jobs” with health insurance are linked to sustaining
employment (making it difficult to determine which is the
most important factor), this is less likely to be true for
transitional Medicaid. Unlike
finding a job with health insurance, participation in a publicly-funded
program like transitional Medicaid should depend less on an
individual’s work experience, skills and or other characteristics.
Thus, although this analysis does not establish a causal
link, transitional Medicaid does appear to be an important
factor in sustaining employment.
Overall,
the analysis indicates that health insurance – whether it
is provided from public or private sources – is important
in sustaining employment. Over 75 percent of the most successful group were covered by private
or public health insurance at the end of the two-year follow-up
period for the survey (compared to 30 percent for the least
successful group). This difference is primarily due to differences
in health insurance coverage for those who were employed and
left welfare.
VI. Characteristics
of the Hardest-to-Employ: Individuals
Who Never Worked in the Follow-up Period
As discussed above,
approximately one-fifth of NEWWS program group members did
not work during the follow-up period, according to UI records.
Because this group is most likely to quickly reach
time limits on cash assistance, there is an interest in understanding
the characteristics and potential employment barriers of this
group. Toward this
end, as shown on Table 6, the demographic characteristics
of those who did not work during the follow-up period are
examined.
Age
and family status. Compared to the individuals who found jobs
during the follow-up period, this was an older group with
older children. Close to 40 percent of this group was over
35 (compared to 25 percent for the groups that worked for
at least 1 quarter) and only 13 percent were under the age
of 25 (compared to over 20 percent for the groups that worked
for at least 1 quarter). This group also had older children, as might be expected given their
age. Only 16 percent
had a child under two (compared to roughly 25 percent for
the groups that worked), and close to one-half had a youngest
child who was over six (compared to 40 percent).
This group was also more likely to have been married
prior to random assignment.
Education
and basic skills. Those who never worked had relatively low education
and basic skills levels.
Roughly one-half had a high school diploma or GED certificate,
compared almost three-quarters of those who worked more than
75 percent of the follow-up period.
They were also very likely to have low basic skills.
Over 56 percent scored “low” on a basic skills
literacy or math test (compared to 38 percent of the most
successful group), with most doing poorly on the math section
on the test. Over
one-third were at high educational risk (no high school diploma
and failed both the literacy and math test).
This group was half as likely to have enrolled in education
and training compared with at the time of random assignment,
compared to those in most other groups.
Welfare
and employment history. Program group members who never worked after
random assignment were most likely to have a long history
of welfare receipt: 42
percent had received welfare for five years or more (compared
to 27 percent of the most successful group). Only 19 percent had earnings in the year before
random assignment (compared to over 60 percent of the most
successful group), although 64 percent had worked full-time
for 6 months or more for one employer at some point in their
lives. Interestingly,
this group was equally
likely to have grown up in a household that received AFDC
and less likely to have had a child as a teenager than the
groups that were able to obtain employment.
Personal
barriers. There were differences across the groups in
terms of the family barrier, locus of control, and family
attachment scales. Notably,
approximately 30 percent of those who never worked had a barrier
according to the family or personal problem scale and according
to the locus of control scale (compared to 10 and 17 percent,
respectively, of the most successful group), and almost 40
percent had a barrier on the family attachment scale (compared
to 20 percent of those who worked steadily).
It is important
to note, however, that even within this relatively disadvantaged
group of non-workers, a diverse range of welfare recipients
is represented, including some with higher skills and education
and no reported personal barriers.
VII. Summary
This analysis
has identified a number of interesting findings about the
work patterns of welfare recipients who find jobs. Work is common: approximately 80 percent of the NEWWS sample members worked at some
point during the four-year follow-up period. On average, the group who found jobs was employed just over half
of the follow-up period.
A majority experienced one or two spells of employment.
Those who ended an employment spell did so relatively
quickly: the average spell length for this group was
three quarters with roughly one-third becoming unemployed
within one quarter. Those who did find a job after they lost
their initial job did so after a relatively long period of
unemployment – 2.8 quarters. Individuals who lose their initial job may
be an important target group for post-employment services.
NEWWS
sample members also experienced a diverse range of work patterns. One-quarter of the employed sample worked for a substantial portion
of the four-year follow-up period (more than 75 percent). However, one-quarter of the employed sample
worked 25 percent or less of the follow-up period. Individuals who were most and least successful in sustaining employment
had different types of employment and earnings patterns.
-
Those
individuals who worked relatively continuously throughout
the follow-up period started working relatively early
in the follow-up period and were most likely to be in
their initial or second job after random assignment. This group also experienced substantial earnings
growth over the four-year follow-up period: earnings grew by close to 50 percent during this time.
-
The
group of individuals who were least successful in sustaining
employment were more likely to have only one relatively
short spell of employment and to start working later in
the follow-up period. This group had very low levels
of earnings and almost no growth in earnings during the
follow-up period.
-
Those
who worked for a “moderate” portion of the follow-up
period were more likely to experience multiple employment
spells. This group had low to moderate levels of quarterly
earnings and earnings growth.
In
order to help policymakers understand which factors are important
to who succeeds in sustaining employment and who does not,
the analysis examined the variation in the composition of
each of these groups in terms of their demographic characteristics
as well as the characteristics of their initial jobs and their
use of transitional Medicaid and child care.
While the analysis does not establish a causal relationship
between the characteristic and the ability to sustain employment,
it does indicate factors that are linked to this outcome.
Several findings are notable about this analysis.
-
The
group that was most successful in sustaining employment
had higher education and basic skill levels, more recent
work history, and fewer personal barriers (particularly
family or personal problems, family attachment, and locus
of control), compared to the group that was least successful. There were small or negligible differences
between the groups based on age, ethnicity, marital status
and children, and depression levels; and a somewhat larger
difference in prior welfare receipt.
-
Health
insurance – whether provided through public or private
sources – appears to be critical if individuals are
to sustain employment. The availability of employer-provided health insurance was twice
as prevalent in the initial job among those who were most
successful in sustaining employment, compared to those
who were least successful. The most successful group was also twice as
likely to have used transitional Medicaid. This indicates
that health insurance is important in sustaining employment.
-
The
group who was most successful in sustaining employment
was also more likely to use transitional child care, although
their usage rates were relatively low given their employment
status: Approximately
20 percent used this benefit within two years, compared
to 12 percent in the least successful group.
-
There
were not large differences in the hourly wages of the
initial job between those who were most and least successful
in sustaining employment and only moderate differences
in the hours worked.
-
Those
who did not work after random assignment were most likely
to have low education and basic skills, limited work experience,
and to score high on scales measuring family attachment,
family problems, and locus of control. In addition, this
was an older group of individuals who had somewhat older
children and tended to be long-term welfare recipients
with little recent work experience at the time of random
assignment. It is important to note, however, that even
within this defined group of non-workers, a diverse range
of welfare recipients is represented – including
some with higher skills and education and no reported
personal barriers.
Berlin,
Gordon L. 2000.
Encouraging Work Reducing Poverty: Impacts
on Work Incentive Programs.
New York: Manpower
Demonstration Research Corporation.
Dodoo,
Martey. 2000.
Evaluating Alternative Welfare-to-Work Approaches: Four-Year Impacts for Eleven Programs. Washington, D.C.: U.S. Department of Health
and Human Services, Administration for Children and Families
and Office of the Assistant Secretary for Planning and
Evaluation; and U.S. Department of Education, Office of
the Under Secretary and Office of Vocational and Adult
Education.
Freedman,
Stephen, Daniel Friedlander, Gayle Hamilton, JoAnn Rock,
Marisa Mitchell, Jodi Nudelman, Amanda Schweder, and Laura
Storto. 2000. Evaluating Alternative Welfare-to-Work
Approaches: Two-Year
Impacts for Eleven Programs.
Washington, D.C.:
U.S. Department of Health and Human Services, Administration
for Children and Families and Office of the Assistant
Secretary for Planning and Evaluation; and U.S. Department
of Education, Office of the Under Secretary and Office
of Vocational and Adult Education.
Freedman,
Stephen. 2000.
Four-Year Impacts of Ten Programs on Employment Stability and Earnings
Growth. Washington,
D.C.: U.S. Department
of Health and Human Services, Administration for Children
and Families and Office of the Assistant Secretary for
Planning and Evaluation, and U.S. Department of Education,
Office of the Under Secretary and Office of Vocational
and Adult Education.
Friedlander,
Daniel, and Gary Burtless. 1995. Five
Years After: The Long-Term Effects of Welfare-to-Work Programs.
New York: Russell Sage Foundation.
Knox,
Virginia, Cynthia Miller, and Lisa A. Gennetian.
2000. Reforming
Welfare and Rewarding Work:
A Summary of the Final Report on the Minnesota
Family Investment Program. New York:
Manpower Demonstration Research Corporation.
Martinson,
Karin. 2000.
The Experiences of Welfare Recipients Who Find Jobs. Washington, D.C.: U.S. Department of Health and Human Services, Administration for
Children and Families and Office of the Assistant Secretary
for Planning and Evaluation; and U.S. Department of Education,
Office of the Under Secretary and Office of Vocational
and Adult Education.
Michalopoulos,
Charles, and Christine Schwartz, with Diana Adams-Ciardullo.
2000. What
Works Best for Whom:
Impacts of 20 Welfare-to-Work Programs by Subgroup.
Executive Summary. U.S. Department of Health and Human Services,
Administration for Children and Families and Office of
the Assistant Secretary for Planning and Evaluation; and
U.S. Department of Education, Office of the Under Secretary
and Office of Vocational and Adult Education.
Rangarajan,
Anu, Peter Schochet, and Dexter Chu. 1998. Employment
Experiences of Welfare Recipients Who Find Jobs:
Is Targeting Possible?
Princeton, NJ: Mathematica Policy Research, Inc.
Rangarajan,
Anu, and Tim Novak. 1999. The Struggle to Sustain Employment: The Effectiveness of the Post-Employment Services
Demonstration. Final
Report. Princeton, NJ: Mathematica Policy Research, Inc.
Scrivener,
Susan, Gayle Hamilton, Mary Farrell, Stephen Freedman,
Daniel Friedlander, Marisa Mitchell, Jodi Nudelman, and
Christine Schwartz. 1998. Implementation,
Participation, Costs, and Two-Year Impacts of the Portland
(Oregon) Welfare-to-Work Program.
Washington, D.C.: U.S. Department of Health and Human Services,
Administration for Children and Families and Office of
the Assistant Secretary for Planning and Evaluation; and
U.S. Department of Education, Office of the Under Secretary
and Office of Vocational and Adult Education.
Strawn,
Julie, and Karin Martinson. 2000. Steady
Work and Better Jobs: How to Help Low-Income Parents Sustain Employment
and Advance in the Work Force.
New York: Manpower
Demonstration Research Corporation.
|
