Number of Visitors Who have viewed this page since 20-Feb-2006
Linking Welfare Recipients to Jobs:
Connections between Client Abilities, Previous Work
and Education History, Social Isolation and Placement in a Rapid Attachment
Program
Rapid attachment
to the workforce programs play a prominent role in welfare to work strategies
under the
Job search and
job placement programs have long been a staple of states’ strategies to reduce
welfare dependence (Handler and Hasenfeld 1991, Rose 1995). While studies show that these programs
produce modest increases in employment and earnings for their participants,
none suggest that they provide a road out of poverty (Gueron 1987, Riccio,
Friedlander and Freedman 1995, Friedlander and Hamilton 1993). As most welfare research shows, the majority
of the population on public assistance moves between welfare and low wage
employment on a regular basis (Bane and Ellwood 1994, Edin and Lein 1997). The
primary research studies on placement programs do not show if these programs
simply hasten movement back into low wage employment. In fact, none of studies looks carefully at
the nature of the jobs found through these programs. Using data from an administrative database
from a rapid attachment program in
Research on job
placement programs also provides little information on the role of previous occupational and educational
background in placement. For example,
the SWIM and GAIN studies report that having a high school diploma increases
earnings, but no other information is available ( Riccio, Friedlander and
Freedman 1994, Friedlander and Hamilton 1993).
By tracing the career and training histories of program participants,
this study examines the connections between previous employment and placement
through a rapid attachment program.
Since the database also contains information on case manager assessments
of participants’ attitude, dress and presentation, as well as measures of math
and reading ability, this research shows how participant characteristics affect
the ability of these programs to place participants.
Scholars of
persistent poverty point to the importance of segregation and social isolation
in keeping welfare recipients out of stable employment (Massey and Denton 1993,
Wilson 1996). However, few studies
actually trace the connection between where welfare recipients live and where
they work. By examining zip code data on
neighborhoods and employer locations, this paper provides a preliminary look at
the relationship between social isolation and employment through a rapid
attachment program.
My earlier
research on agencies which serve welfare populations suggests that the agency’s
characteristics and connections to employers also impact on the ability of that
organization to place its clients (Schneider 1997). By examining the relationship of placements
to the agency’s physical location and their previous experience to the kinds of
placements found through this program, this paper suggests that understanding
rapid attachment program results must include evaluating these factors.
Through an
analysis of this database, the paper generally addresses the following
questions:
1.
Who
are the people served by a rapid attachment program? What are their employment and educational
histories? What are their actual
educational skills and how prepared are they to enter the workforce? Are they isolated from main-stream labor
markets?
2.
How
do these characteristics affect who is placed into a job by a rapid attachment
program and the kinds of jobs that they find?
3.
What
other factors (transportation, the local labor market, agency experience and
connections) influence the nature of the jobs found through these programs?
Rapid Attachment in
Participation in
a rapid attachment program is the first step in the
Like most rapid
attachment programs, the one studied here included a combination of job
readiness workshops, structured and independent job search, and job placement
activities. Job readiness workshops
generally focus on such issues as interviewing, resume preparation, appearance,
time management and related issues. The
premise of this rapid attachment model is that welfare recipients can find jobs
if they are supported in job search activities and given some short term
instrumental training in the soft skills of job seeking such as communications,
dress and presentation in an interview.
This research examines the way that participants with different work and
educational experience fare in this kind of program.
Sample and Method
The data analyzed
here comes from the administrative database for this program during its first
year. It includes all 718 people in the
program from February 1996 to February 1997.
The database included the following information: 1) demographic
information on participants, 2) government program utilitization data, 3)
information on substance abuse and criminal histories, 4) education and
training experience for participants, 5) work experience and location of employers,
6) age, sex and childcare needs of children, 7) interviewer assessments of
presentation, attitude, dress, interviewing technics, TABE Math and Reading
test scores and 8) information on placement status, type of employment, salary,
hours, benefits and location of employers.
While this
database presented a rich source of information on program participants, it
also demonstrated the perils of working with administrative databases. In many cases, key information was lacking or
incorrectly coded. For example, 44
percent of the cases were either missing data or coded as “other” for the key
variable of race. Information on start
and end dates for employment and training programs were blank, making it impossible to perform a time
line analysis of work history as in previous studies (Schneider 1997b). Many of the job and training program
categories were coded incorrectly.
Fortunately, the actual job titles and training program names were
available, allowing the opportunity to recode these data. As a result, while this database provided
important information on placement through a rapid attachment program, the
research can not answer many questions regarding the role of race or several
other key factors in employment.
In addition to
the database, this paper draws on my experience working with this agency and
with others in this locality to understand the employment field which the
program worked within. I use this
information - along with conversations with agency program administrators, in
the placement section. However, it is
important to note that I did not observe this program in process and this paper
does not constitute and evaluation of the program. Instead, I use these data to understand the
pattern of placements in terms of locality and type of jobs found.
Analysis
primarily consists of frequencies and cross tabulations on nominative
data. Findings reported are mostly
significant at the p.000 or p.001 level.
As in other studies, I performed cluster analysis to categorize program
participants by demographic characteristics and work experience. I also used cluster procedures to analyze the
types of jobs found by participants.
Finally, I used logistical regression analysis to look at the role of
various factors in determining who was and was not placed by this program. Regression analysis was also run to look at
the relationship between placement wages and hours and various participant
characteristics.
Participant Characteristics, Work and
Training Histories
Basic Demographics
The sample
population consisted mostly of women who had never married. Eighty-six percent were female, 68 percent
were single, 8 percent were currently married and the rest separated or
divorced. Men were more likely to be
married: 34 percent of the men in the sample were currently married compared to
four percent of the women. The majority had small families: 66 percent had two
children or less. Forty-nine percent
were the only adult in their household, 26 percent lived with another adult,
eight percent were teen parents with no adult in the household, and the
remainder lived in households with more than two adults. Sixty-four percent said that their housing
was stable.
These basic
demographics proved remarkably similar to my earlier research on people in a
wide array of training programs in
While much of the
race data was missing, the population seemed to consist primarily of people of
color. For those who responded to
questions regarding race, 47 percent were African American, 6 percent white, 4
percent Latino, .6 percent Asian and 42 percent other. People who did not live in stable housing
were more likely to answer “other” to race.
By looking at the districts where program participants lived and
comparing that information to census maps of racial segregation in
The majority of
the program participants fell into two age groups. Forty percent were between the ages of 22 and
25 and another forty percent between 36 and 45.
As in earlier studies, age also had little impact on who was placed and
the kinds of jobs found by program participants.
Education is
often named as the key factor influencing the work prospects of low income
people (Harlan and Steinberg 1989, Kingfisher 1996, Riccio, Friedlander and
Freedman 1994, Friedlander and Hamilton 1993).
A high school diploma is considered particularly crucial. Fifty-four percent of the program
participants had a high school diploma or GED.
Forty-nine percent had attended a training or educational program after
leaving high school. Nearly nine percent
of the program participants had attended college. As discussed in detail below, completing a
diploma or GED did influence work history and placement. Additional training proved far more nebulous.
The population in
this study proved to have less education than those in the Social Network
Study. Sixty-eight percent of that
population had a high school diploma or GED and eighty-three percent had
attended a training program after high school (Schneider 1997b: 7-8). Of those who had taken additional training,
sixty-five percent of the rapid attachment program participants had only
attended one training program and 38 percent two programs. In comparison, 38 percent of the Social Network
Study participants had only attended one program, 33 percent two and 12 percent
three or more (Schneider 1997b: 8).
However, the rapid attachment program participants high school
completion rate was identical to that of AFDC recipients participating in JOBS
program activities (personal communication, Pennsylvania Department of Public
Welfare).
These differences suggest two implications for
scholars focusing on the links between education and work. First, results from this study can probably
be generalized to the population participating in most similar programs. Second, as the fact that over 30 percent more
Social Network Study participants had attended training programs than
participants in the Rapid Attachment program demonstrates, the population that
seeks training as an avenue out of welfare is more likely to have focused on
education earlier in their lives and to continue to turn to education programs
as a solution to poverty throughout their work histories.
The rapid
attachment data also included information on other significant barriers to
employment such as addictions, disabilities and criminal history. Seven percent of the participants reported
having a history of drug abuse, but only 1 percent had been in a drug
rehabilitation program. Information on
alcohol abuse was largely missing. Nine
percent reported that they were disabled.
Most of the disabled people were older: 22 percent of those aged 45 and
older had disabilities, many due to work injuries. However, these two barriers had no impact on
whether or not an individual was placed by the program. In addition, 4 percent reported that they had
criminal records and 2 percent were on parole.
Access to
transportation also influences where an individual can work. Fourteen percent of the program participants
reported that they had a drivers license.
While those who could legally drive seemed to have more extensive work
histories, having a drivers license had little impact on placement through this
program.
Work and Education Histories
As in previous
studies of welfare recipients in Philadelphia (Schneider 1997), most of the
people in the rapid attachment program had significant experience with the
world of work. Seventeen percent
reported no previous employment, 20 percent one job, 25 percent two jobs, 17
percent three jobs, 12 percent four jobs, and eight percent more than four
jobs.
Table 1: Most Frequent
Types of Employment
As outlined on
table one, the types of jobs held by program participants proved both similar
and different to those in my earlier studies.
As in other studies of employment for low wage workers (Edin and Lein
1997), the majority of the rapid attachment program participants worked in low
paid, service sector jobs. These
included cashier, nursing assistant, housekeeping, sales and restaurant work.
In comparison to the Social Network Study, fewer rapid attachment participants
had worked as cashiers, but many had worked as housekeepers or in other
domestic occupations. This difference
suggests that the Social Network Study participants were more attached to
formal sector employment which required communications and math skills while 15
percent of the rapid attachment program participants worked in more informal
settings which required little formal education and communication skills.
The other
significant difference between the two studies involved the number of people in
blue collar and factory work. Twice as
many Social Network Study participants had worked as security guards, in
maintenance or construction than in the Rapid Attachment program. Twenty-three percent of the Social Network
Study participants had worked in factories compared to 15 percent in the Rapid
Attachment database. These differences
are primarily due to the fact that the Social Network Study included a group of
white, male displaced workers which are largely absent from the rapid
attachment program.
In both studies,
not everyone had worked in low-wage, service sector employment. Twenty-three percent had worked in data-entry
or clerical jobs in the Rapid Attachment program and 27 percent of the Social
Network Study participants had held similar employment. Since clerical jobs can either be similar to
low-wage service sector employment or provide stable employment for women,
additional information was used to evaluate the nature of work history for this
population. In addition, nine percent of
the Social Network Study participants and eleven percent of the Rapid
Attachment participants had worked in professional or professional entry-level
jobs such as social worker, caseworker or nurse. These data highlight the fact that people
with significant work histories end up on welfare.
The wages and
benefits for all of the people in the Rapid Attachment program were very
low. Only 17 percent had earned between
seven and nine dollars an hour at any point in their career. Six percent had earned between nine and
eleven dollars and only four percent had ever held a job that paid over eleven
dollars an hour. Employment provided
little means to support a family and seldom enough to save for hard times.
Employment
vulnerability was also evident in the percentage of the population that had
applied for or received unemployment insurance benefits. Overall, 22 percent had applied for
unemployment and 12 percent had actually received benefits. Ability to get unemployment varied
significantly by gender and race. Thirty
percent of the men had applied for unemployment benefits and 23 percent had
received them. In comparison, only 20
percent of the women had even applied for benefits and only 10 percent had
qualified for unemployment. Forty-three
percent of whites had applied for unemployment, compared to 29 percent of
African Americans, 25 percent of Latinos and 8 percent of those categorized as
other. The percentages actually
receiving benefits differed even more dramatically. Thirty percent of the whites had received
unemployment, compared to 16 percent of African Americans, 4 percent other, and
no Latinos. These findings highlight the
fact that people of color and women work in the unstable, small employer firms
of the secondary sector (Gordon, Edwards and Reich 1982).
People left jobs
for a variety of reasons. Thirteen
percent had quit jobs due to pregnancy, 12 percent due to personal conflicts on
the job, 12 percent because the job did not pay enough, 9 percent because the
job had no future, and 15 percent quit for other reasons. Very few program participants had been fired:
2 percent for attendance problems and 6 percent for personal problems. Many had been laid off: 6 percent because the
company had relocated, 23 percent for various reasons, and 16 percent for other
reasons. Reasons for leaving a job had
no connection to placement through this program.
My previous
research identified four types of people who were on public assistance: 1)
limited work experience, 2) low wage workers, 3) displaced workers and 4)
immigrants and refugees (Schneider 1997).
These different populations had varying experience with work, education
and welfare. Those with limited work
experience had worked only one job for less than a year or never held a
job. Low wage workers fit the profile of
most welfare recipients. They alternate
between low-wage, service sector jobs with few benefits and welfare. Displaced workers had stable employment
histories, primarily in blue collar occupations, clerical or professional
jobs. Immigrants and refugees had a
range of experience, but most were on welfare because they had limited access
to mainstream labor markets.
In the Rapid Attachment study I sought to
replicate this finding through cluster analysis of types of employment, wages,
and reasons for leaving jobs. Cluster
analysis groups cases with similar characteristics, allowing for categories to
arise inductively from the data.
Unfortunately, data on dates of employment were missing, which did not
allow me to measure the duration of employment for program participants in this
study. Nevertheless, four clusters
emerged from these data which fit the patterns from the earlier research quite
well. Since the Rapid Attachment program
included few immigrants and refugees (three percent said that they had a
language barrier and only two were Asian refugees), this group was not found in
the Rapid Attachment study population. A small percentage of the cases were
reassigned from the computer created clusters in order to place the majority of
those who had lost jobs due to industry relocations and those with professional
training and experience together.
Percentages for the major cluster variables are listed on table 2.
As in earlier
research (Schneider 1997), the minority of the study population had limited
work histories. Only 17 percent of the
population fell into this category.
Most had never worked. Only 29
percent had a high school diploma. These
people also were more likely to live in households with three or more adults,
indicating family structures which may require care giving to sick adults or
children or extreme poverty which required households to double up. However, contrary to assumptions that this
most vulnerable group lacks work experience simply because they lack education,
it is important to note that 6 percent of this group had attended college and
thirteen percent had attended a training program.
The majority of
the study population were low wage workers. Overall, 73 percent of the study
participants fell into this category.
Since I could not distinguish between people who had only worked one job
for a short time and those with more significant histories, I kept the initial
clusters which grouped this population into two subcategories: those who had
worked one to three jobs (60 percent)
and those who had worked three to five jobs (13 percent). In both cases, the types of employment were
similar: cashier, clerical, housekeeper, sales, restaurant work. The majority had earned wages below five
dollars an hour and had little wage progression throughout their careers. Fifty-three percent of those who had worked
one to three jobs had a high school diploma or GED and 77 percent of those who
had held three to five jobs had this credential. Thirteen percent of those who had worked
three to five jobs had attended college and seven percent of those working one
to three jobs had been to college. The
majority had not lost their jobs due to company relocations.
The remaining ten
percent of the program participants were categorized as displaced workers. As in the Social Network Study, this group
had a different work trajectory. The
Social Network Study found that first jobs for displaced workers were often
early work experience unrelated to future careers (Schneider 1997b: 21-22). The Rapid Attachment population supported
this observation. While 26 percent of
the displaced workers had worked as housekeepers and 22 percent in sales, their
employment histories were not confined to this kind of work. A full 44 percent also held clerical jobs, 56
percent were in professional, entry level positions and 57 percent worked as
professionals.
This is the only
group in this study population which showed a wage progression. However, most displaced workers had gone on to low
wage employment near the end of their careers.
The average wage difference between the first and last job for this
group was less than one dollar. This
pattern is also similar to that in the Social Network Study (Schneider
1997b).
The jobs help by
displaced workers were stable and well paid enough to provide unemployment
compensation. Forty-five percent of
displaced workers had applied for unemployment benefits and 25 percent had
received unemployment as opposed to 17 percent or less who had received
unemployment benefits for the other groups.
While education
played a role in their success, it was clearly not the only factor in the
different work histories. Roughly the
same percentage had completed high school as those who had worked numerous low
wage jobs. Seventeen percent had
attended college as compared to 13 percent of the low wage workers who had held
three to five jobs.
The Social
Network Study suggested that displaced workers tended to come from the stable
working and middle class with good networks to jobs that last. As discussed elsewhere (Schneider 1997, 1997d),
the working and middle class often possess social capital, in the form of
connections to good jobs and educational programs, family resources which can
sustain them in their careers and social skills appropriate for the
“mainstream” labor market learned through growing up in a working or middle
class environment. To a certain extent,
the class and racial characteristics of the individual’s home neighborhood can
indicate access to social capital which can help people find and retain
jobs. The largest percentage of
displaced workers were found in a demographic cluster which included African
American women with stable housing who were separated or divorced. Sixty-three
percent lived in the stable working to middle class neighborhoods of Germantown
(16 percent), Southwest Philadelphia (14 percent), Fishtown, Bridesburg, Port
Richmond and Frankford (13 percent), Olney (14%) and East Falls, Manyunk,
Roxborough (6 percent). These findings
suggest that displaced workers came from environments with social capital which
could aid them in their careers.
The displaced
worker population in this study is roughly half that in the Social Network
Study. These differences are due to the
fact that the Rapid Attachment program only included one subgroup of the
displaced worker category. In the Social
Network Study, displaced workers included white men displaced from primary
sector blue collar employment and people of color who had worked their way into
stable employment through education and affirmative action (Schneider 1997:
14-17). The Rapid Attachment program
only included people in the second category.
Seventy-two percent of the displaced workers in this study were African
American, seven percent Latino, four percent white and seventeen percent
other.
The role of
education also reflected similar patterns for the displaced persons of color
and limited work experience groups in the Social Network Study. In that study, two groups of people went
through three training programs or more: those with limited work experience who
seemed to be on a training track to nowhere and displaced workers of color who
had used training early in their careers to enter stable employment and now
returned to education after their jobs ended (Schneider 1997: 24-37). While the lack of dates for employment and
training makes it impossible to determine the role of training in previous work
history, the information on number of training programs holds for this
population. Thirteen percent of those
with limited work histories had attended three training programs or more and
fifteen percent of the displaced workers had gone to multiple programs compared
to between five and seven percent for the other work type categories.
Social Isolation
Poverty Map
African American Map
Latino Map
Scholars often
assume that persistent poverty is linked to social isolation (Massey and Denton
1994, Wilson 1996, Galster 1996). Returning to the concept of social capital,
Wilson in particular, infers that people that are isolated from mainstream
labor markets and live among people who do not hold stable employment will not
have the connections to find good jobs or the appropriate social skills to be
hired or retain employment. As Stack
(1974) and Edin and Lien (1997) demonstrate, low income populations in
segregated neighborhoods have many social resources in the form of friends and
family which help them survive on welfare and low wage work. However, these significant social networks
may not provide the connections and social skills to successfully build careers
in the center city and suburban labor markets. In this thesis, geographic
social isolation represents the lack of social capital which can bridge into
the labor markets where employers are hiring. Using zip code data from this database,
this paper contains a preliminary analysis of the role of social isolation in
work history.
While study
participants lived throughout the city, significant concentrations lived in the
high poverty neighborhoods of North Philadelphia/Kensington (19%) and
Fairmount/Spring Garden (16%). These
neighborhoods are also mostly African American and Latino.
Social isolation
was determined by identifying people who had never worked outside of their
neighborhood. This included those with
limited work histories and those whose entire careers were in their home neighborhood. Twenty-two percent of the study population
fit into this category. Ninety-eight
percent had limited work histories and the rest were people who continuously
found work near their homes. The later
group tended to be white working class who lived in the traditional factory
neighborhoods along the rivers in Philadelphia.
As discussed below, social isolation did play a role in placement.
Even for those
who did not live and work exclusively in their home neighborhoods, the
relationship between work and home seemed very important for many people in
this study. Table 3 shows the
relationship between home and work in the work history of program
participants. Work history shows two
patterns. For the stable working and
middle class neighborhoods of the Southwest, Northeast, Germantown/Mt. Airy,
and Olney/Logan, people increasingly leave their home neighborhoods for work in
center city and throughout the rest of the area. If finding work outside of
one’s home community indicates connections to mainstream job networks, this
work pattern shows an ability for people
in stable working and middle class neighborhoods to cross these
boundaries. It also indicates
neighborhoods which serve as bedroom communities versus those that serve as
both work and home.
Table 3: Neighborhood
vs. Workplace Location
Significant
percentages of the remaining population work in the same neighborhood as they
live. This is particularly true of the
traditional factory working class neighborhoods along the rivers. Between 80 and 67 percent of the people who
live in East Falls, Manyunk and Roxborough live and work in the same
neighborhoods. Between forty-five and
fifty-six percent of those living in Fishtown, Port Richmond and Frankford also
work there. Similar relationships hold
true for the more impoverished neighborhoods.
For may program participants, the patterns of behavior, dress and speech
of work and home coincide. Limited
ability to move beyond this environment could play a role in future job
placement.
Case Manager Evaluations
The Rapid
Attachment database also contained codes for case manager assessments of
ability to present oneself to an employer, interview techniques, eagerness to
find work, and general attitude. The
database also contained information from a TABE test on reading and math. However, the reading and math tests were not
scored “because all of these people were assumed to be job ready (by the
commonwealth)” (Personal communication).
Instead, those scores list the number of questions that the individual
answered correctly on each part of the test.
The study
revealed that displaced workers were able to present themselves better than
other groups. Fifty-one percent of the
displaced workers were listed as excellent for presentation compared to 22
percent for those with limited work experience, 19 percent for low skill
workers with one to three jobs, and 31 percent for low skill workers with three
to five jobs. On the other end of the
scale, only 2 percent of the displaced workers were rated deficient, compared
to 19 percent for those with limited work histories.
Displaced workers
also interviewed for jobs better.
Forty-three percent of the displaced workers were rated as “excellent”,
compared to 20 percent for low skill workers with three to five jobs, 14
percent for low skill workers with one to three jobs, and 10 percent of those
with limited work experience. Only 6
percent of the displaced workers were listed as “deficient” compared to 26
percent of those with limited work histories.
All of these
groups were equally eager to find work and all had similar attitudes toward the
program. While these presentation
abilities distinguished between the different groups when they entered the
program, none of these factors played a significant difference in who was
placed in a job. Other factors combined
to make a difference. This finding
suggests that, while the “soft skills” of presentation and communication may
make a difference in career histories, they were not the significant factor in
finding a job in a rapid attachment program.
While the TABE
scores have limited meaning, they did show some predictable differences. Displaced workers had higher scores than
those with limited work histories.
People who worked as housekeepers had much lower scores than most other
jobs. People in clerical and
professional employment scored higher.
In an earlier
paper, I compared job training and placement programs for low income people to
business assumptions about creating a successful product. Program participants become the raw materials
for a job placement program. But, since
we are not just creating widgets, these raw materials have different attributes
which influence the ability of any program to place them into jobs (Schneider
1997c). Their work histories,
educational credentials, the way that they present themselves, and their
ability to use math and English all play a role in this process. The next section looks at the role of these
various background characteristics in job placement from a rapid attachment
program.
Placement
Before discussing
the placement results for participants in the rapid attachment program, it is important to understand the contracting
context for this program. As described
elsewhere (Schneider 1997c), any employment or training program works
closely within the guidelines of its contract.
The contracts for this rapid attachment program were extremely
strict. Agencies only received full
payment if they placed their participants into jobs. While any job counted in
this program, agency standards were to find jobs which paid $6.00 an hour with
benefits and which lasted at least 90 days.
The contract stipulated a placement goal of at least 50 percent. The agency’s responsibility was to meet these
contract goals in any way possible.
These included direct placement of participants and encouraging them to
find work on their own.
While this
project does not evaluate the ability of this particular program to meet these
goals, it does show the role of the participants’ background characteristics in
this process. The final section examines
some attributes of the agency’s networks and the local labor market which also
influence this process.
Placement Status
This first
section looks at who was placed through this program. Data on placement status were missing for 28
percent of the participants, or 202 people.
Presumably, many of these individuals were current participants in the
program. Of the remaining participants,
47 percent were placed in jobs, 46 percent were not placed, .2 percent quit for
health reasons and 7 percent quit for other reasons.
Percentage placed
varied by whether or not a person had a high school diploma, social isolation
and worker type. Fifty-five percent of
those with a high school diploma found jobs compared to 36 percent of those who
did not have this credential. Eleven percent
of those without a high school diploma quit, versus 5 percent of those with a
diploma.
Social isolation
profoundly affected who completed the program.
Forty-three percent of those who quit were categorized as socially
isolated. Only 34 percent of the people
who were socially isolated were placed in a job.
Displaced workers
were most likely to be placed. A full 69
percent of this group found jobs. Only
30 percent of those with limited work experience were placed. Those with low wage work experience fell in
the middle: 47 percent of those with one to three jobs and 52 percent of those
with three to five jobs found work.
Those with limited work experience were also most likely to quit. Fourteen percent of this population quit compared
to six percent for other groups.
College education
contributed both to placement and quitting the program. Sixty-four percent of those who had attended
college were placed and 18 percent of this group quit.
People with work
experience in several key occupations were more likely to be placed as
well. Fifty-nine percent of those with
clerical work experience, 55 percent of those with sales experience, 68 percent
of the professionals and 61 percent of those with professional entry level
experience were placed.
However, when
evaluated in a logistical regression formula, none of these variables proved
particularly significant in determining who was placed by this program. Only two variables were statistically
significant at all: the number of questions answered on the TABE math test and
being a displaced worker. Together,
these two variables accounted for less than 25 percent of the variance. Clearly,
background experience and education play a limited role in determining
who finds work out of a rapid attachment program.
Job Placement Characteristics
These various
variables play an even more complicated role in the kinds of work found by
participants. Fifty-six percent of those
placed found full time jobs (35-40 hours per week) and 44 percent found
part-time work. Twenty-seven percent of
the jobs were 20 hours a week or less.
Thirty-three percent of the jobs offered health insurance and another 7
percent provided health insurance after several months. This meant that majority of those placed
still relied on medicaid. The average
wage was $7.09, with a minimum wage of $4.25 and a maximum of $17.63.
Table four lists
the primary types of jobs found by program participants. As with their previous work experience, many
found work in clerical jobs and as nursing assistants. The same occupations seen in the list of
previous employment appear again, but many jobs are less important. Far fewer were placed in jobs as cashiers, in
housekeeping, in factory work and in restaurants. Only one person found a professional job
while in this program.
These
similarities and differences reflect a combination of past work experience, the
nature of the local labor market, and the placement strengths of the
agency. Previous work experience played
the biggest role. Sixty-one percent of
those who found clerical work as a placement job had worked in computer or
clerical jobs in the past. Those with
health care experience primarily found work as nursing assistants or in
childcare. Forty-four percent of the
nursing assistants had worked in health care in the past. Thirty-six percent of those who found work in
childcare also had a health care background.
Forty-seven percent of those who found sales jobs had sales
experience. People with factory work
experience found jobs as security guards (100 percent), factory workers (44
percent) and factory drivers (50 percent).
Twenty-one percent of those who found cashier jobs and 31 percent of
those placed in restaurant jobs had worked in a restaurant before. Only cashier and housekeeper had no
relationship to previous employment.
The nature of the
labor market is also reflected in the kinds of placement jobs. Philadelphia has steadily been losing factory
work for many decades (Goode and Schneider 1994: 29-40). While the city has experienced a general loss
of jobs, the few growth occupations include clerical, nursing assistant and
other service sector industries. Job
placements in these fields partly reflect these trends.
However,
significant placements into clerical and nursing assistant positions also
reflect the training and placement history of the agency. Milofsky and Hunter (1995) describe
non-profits of this type as linked through network associations to similar
agencies and businesses. In other
papers, I discuss the way that this kind of social capital impacts on the
ability of any agency to meet placement goals (Schneider 1997c, 1997d). As with
individuals, agencies have access to labor markets through their
connections. Part of the role of any
agency running a rapid attachment program is to link program participants into
the labor market through their established connections. This agency was involved in training, with a
long history of sponsoring training programs in clerical and health care,
particularly nursing assistant. The
agency necessarily has contacts in this area.
In addition, many of the clerical placements were within other divisions
of the agency, thus drawing on its own need for workers to meet placement
goals. Analysis
of the employers which hired people from this program supports this
theory. In a separate run, the agency
provided a list of the name and number of placements for all employers who have
hired people from this program from 1996 until March 1998. This included 384 jobs with 200 separate
employers. Two patterns were
evident. On the one hand, blocks of
placements, from three people to twenty were placed with several
employers. Most of these employers were
health care providers, a few were clerical, and some were blue collar or
clerical jobs like UPS. Fifteen
employers accounted for 121 jobs, or 31 percent of the placements. Nine percent of the placements overall were
in-house and 31 percent of the large block placements were within the agency.
Table 5: Location of
Placement Job
These major
employers share several characteristics.
None were major firms in the city like the large hospitals, hotels or
universities. Nor was this agency able
to make many successful contacts with the suburban firms which have most
employment opportunities in the Philadelphia area. The agency talked about limited success in
making connections outside of the city.
The majority of the block employers were
midsized firms which evidently had some ties to the agency. This was also true of the entire list of
placements. The list of employers
included several agencies which work with the agency running this program as
well as government offices with ties to this agency. Geography played a role both in the large
block employers and overall placements.
Fifty-four percent of the placement jobs were in center city, many near
the agency. The rest of the job placements cover a range of
employers, most hiring just one person.
Many of these jobs are service sector retail, clerical and some health
care. The list also includes government,
factories and a variety of other firms.
Placements from these
programs came about in two ways. The
first involved direct placement by the agency and the other involved
participants finding jobs on their own.
Many of the single placements probably reflect this second trend. Here we see a combination of people relying
on their own networks, agency guidance, and work close at hand. Given that the agency was located in center
city, it is no surprise that the majority of jobs were found within short
distance from the agency sponsoring the program.
As table five
shows, people were no longer finding jobs in their local communities. Locality based employment networks did not
provide employment for most of these participants. This is no surprise given that these networks
had failed for a sufficient amount of time for most program participants to end
up on welfare. The only contradiction to
this trend were the socially isolated people in East Falls, Manyunk and
Roxborough who managed to return to those networks once again to find
employment.
Placements
instead reflect primary locations for different kinds of work and
transportation routes. Center city is
the hub for most public transportation and the logical place to seek work for
many people who lack cars. Center city
is also one major place within the city for clerical and health care
employment.
Overbrook/Westbrook
is on a bus line and is the location of several restaurants and hotels. In fact, 32 percent of the jobs found in this neighborhood were
restaurant work. Thirty-five percent of
the jobs found in Manyunk and the Northeast were also restaurant work,
reflecting this kind of service sector employment in these neighborhoods.
Manyunk has become a trendy neighborhood hosting upscale restaurants and
galleries. The Northeast is a suburb
within the city which contains a number of chain restaurants and strip mall
employment opportunities. Nineteen
percent of the jobs in the Northeast were sales jobs, and 25 percent each in
Olney and East Falls, neighborhoods with local services as bedroom communities.
Table 6: Placement
Job Characteristics
The nature of the
work found by program participants reveals a disturbing trend in relation to
the kinds of the jobs which participants with different background
characteristics may find through a rapid attachment program. Table 6 summarizes the types of jobs found
through this program and the average wage for each. The majority found work in the service sector
(cashier, restaurant, sales), helping professions (nursing assistant, child
care, teacher’s aid) or clerical jobs.
Many of these jobs were part-time, with no insurance, little difference
from work held before entering this mandated program. Rapid attachment was able to help them find
work, but largely put them back into low skill jobs. The fact that most of those with professional
and professional entry level experience found work in clerical, sales or
helping professions highlights the fact that under time pressure, people with
good experience will end up in low-paid, no benefit secondary sector jobs.
Participants with
various characteristics reacted to the rapid attachment program
differently. These findings are perhaps
the most disturbing trend in this study.
Regression on wages and hours showed absolutely no relationship with
previous experience or worker type. In
fact, having a diploma alone had no effect on either wages or working
conditions. College training showed no significant difference. Other
training also made no difference.
Overall,
displaced workers earned about one dollar more than the other groups, with a
median wage of $8.00 per hour as opposed to $7.00 for other groups. However, looking closely at the job type
characteristics reveals that most displaced workers were taking work which was
either part time or has no insurance.
Thirteen percent took clerical full time jobs with no insurance, 22
percent took blue collar part time jobs with insurance, 13 percent took other,
part-time jobs with no insurance. The
strategy seemed to be to take the easiest job to find, which often meant not
getting the most stable, family sustaining employment.
Those with
limited work experience or only one to three
low skill jobs found work in part time helping professions with no
insurance and blue collar part-time jobs with no insurance. Twenty percent of the low skill workers with
one to three jobs found helping profession part time jobs and 19 percent found
blue collar part time jobs. Forty-two
percent of those with limited work experience found work in these two
categories (21 percent for each category).
None of those with limited work experience found jobs with
insurance. The most vulnerable program
participants often found work in the worst jobs.
However, many of
the most secure, good paying jobs went to those with only one to three low
skill jobs. Seventy-nine percent of
those in clerical, full time jobs with insurance and 83 percent of those in
helping profession jobs with insurance fell into this category. None of the
displaced workers found full time clerical work with insurance, the kind of
employment characteristic of numerous displaced workers before losing their
stable jobs. Since one would ordinarily
expect the displaced workers to get these better jobs, another principal may be
operating here. While there is no way to
verify this suspicion in this study, my ethnographic work with program
participants suggests that these individuals with limited work experience may
have avoided taking the less stable employment because of the potential lack of
wages and benefits. Since they have
relied more on the agency for placement than the displaced workers or those
with more work experience in low-skill jobs, they also may have benefited more
from the program priority to find full time work with insurance than those with
better personal resources. In this case,
a few of the more vulnerable participants with some work experience actually
found better jobs than their previous work experience through the rapid
attachment program. It is important to
note, however, that only nine percent of the people in this work type category
found these stable jobs. In most cases,
rapid attachment meant quickly finding unstable, secondary sector employment.
Conclusion
Analysis of this
rapid attachment program reveals several findings related to the role of this
welfare to work strategy as a method to move welfare recipients into stable
employment. On a positive note, many
people found jobs through this program, including 34 percent of those with
limited work experience. Placements in
this program also moved many program participants into firms located in center
city, which may give them access to “mainstream” labor markets as opposed to
neighborhood based employers who had been a major source of employment in the
past. However, it is important to note
that these individuals may simply be finding low-paid, part-time secondary
sector jobs in center city as opposed to their local neighborhoods. The location of the work may change, but not
the nature of the work. The kinds of
work connections may or may not lead to more stable, lucrative employment.
People who had worked in personal services like housekeeping before found work
in jobs which were more likely to provide social security. If these placements
truly represent a first step into stable employment, these shifts in job
location and type of employment could eventually lead to incorporation into the
wider labor markets of the city.
However, results
from this analysis show that rapid attachment is not changing the work
experience potential for many of the participants. Those most likely to find work had
significant, stable work histories before entering the program and the majority
found employment in the same fields as their previous experience. Rapid attachment, therefore, means
re-attachment to similar employment, often in secondary sector jobs offering
part-time hours and no health insurance benefits. It is impossible to tell if these
“entry-level” jobs will turn into full time employment. Earlier research suggested that displaced
workers were able to parlay part-time clerical work into full time jobs earlier
in their careers and may be able to follow the same trend now (Schneider
1997,1997b). Those with low-skill work
experience, particularly those without a high school diploma, were not able to
use “entry-level” jobs as a stepping stone (Schneider 1997). The previous work experience trends in this
study and others suggest that re-attachment simply continues the cycle of
poverty.
These data also
suggest that rapid attachment can not replace the need for background skills,
communications skills and experience.
Those who were socially isolated were more likely to drop out of the
program. Those with limited work
experience were least likely to be placed and often found the worst jobs. Ability to present oneself well had no affect
by itself in determining who found work.
This may indicate that the program’s job readiness courses are working
to correct these initial deficits.
However, given that displaced workers presented themselves better than
other groups, it is more likely that the combination of good work experience
and appropriate presentation skills sell a particular participant to employers.
Taken together,
these data suggest that rapid attachment may be a first step back into paid
employment for welfare recipients which fulfills the work experience
requirements of the new Federal law.
However, if people are expected to rely on government aid for only five
years in a lifetime, it can not be the only strategy to reform welfare. Other strategies are needed for those with
limited work experience, particularly when this combines with social
isolation. Many of these program
participants who found jobs are likely to find that they still can not make
ends meet, or that jobs end due to the seasonal or unstable nature of the employment. The most important government interventions
may need to occur after the rapid attachment program, through a strategy
of continued case management to retain employment and placement support to find
more stable, better paid jobs. This
strategy must also support appropriate education combined with work. Now that states are moving past initial
implementation of TANF, it becomes increasingly important that they pay
attention to changing the nature of the work found by their program
participants if there is any hope of ending welfare as we know it.
References
Cited
Bane, Mary Jo and David Elwood.
(1994). Welfare Realities: From Rhetoric to Reform. Cambridge, MA:
Harvard University Press.
Edin, Kathryn and Laura Lein.
(1997). Making Ends Meet: How Single
Mothers Survive Welfare and Low Wage Work.
New York: Russell Sage Foundation.
Freedman, Stephen, and Daniel
Friedlander. (1995). The JOBS Evaluation: Early Findings on Program Impacts
in Three Sites. Prepared with U.S.
Department of Health and Human Services and the U.S. Department of Education.
New York: Manpower Demonstration Research Corporation.
Friedlander, Daniel and Gayle
Hamilton. (1993). The Saturation Work Initiative Model in San Diego: A Five
Year Follow Up Study. New York: Manpower Demonstration Research
Corporation.
Galster, George. (1996). Polarization, Place and Race. In Race,
Poverty and American Cities edited by John Charles Boger and Judith Welch
Wegner. Chapel Hill: The University of
North Carolina Press, 186-227.
Goode, Judith and Jo Anne
Schneider.(1994). Reshaping Ethnic and Facial Relations in Philadelphia:
Immigrants in a Divided City. Philadelphia, Pennsylvania: Temple University
Press.
Gordon, David M., Richard Edwards and
Michael Reich. (1982). Segmented Work, Divided Workers: The historical
transformation of labor in the United States. Cambridge, Massachusetts:
Cambridge University Press.
Gueron, Judith. (1987). Reforming
Welfare with Work. New York: Ford Foundation.
Handler, Joel F. and Yeheskel
Hasenfeld. (1991). The Moral Construction of Poverty: Welfare Reform in
America. Newbury Park: SAGE Publications.
Harlan, Sharon L. and Ronnie J.
Steinberg, eds. (1989). Job Training for Women: The Promise and Limits of
Public Policies. Philadelphia: Temple University Press.
Kingfisher, Catherine.(1996).Women
in the American Welfare Trap.
Philadelphia: University of Pennsylvania Press.
Massey, Douglas S. and Nancy A.
Denton. (1993). American Apartheid: Segregation and the Making of the
Underclass. Cambridge, Massachusetts: Harvard University Press.
Mead, Lawrence M. (1992). The
New Politics of Poverty: The Nonworking Poor in America. New York: Harper
Collins Publishers.
Milofsky, Carl and Al Hunter
(1995). Where Nonprofits Come From: A Theory of Organizational Emergence. Unpublished paper presented at the Southern
Sociological Society Meetings, Atlanta, GA.
Riccio, James, Daniel Friedlander
and Stephen Freedman. (1994). Gain: Benefits, Costs and Three Year Impacts
of a Welfare to Work Program. New York: Manpower Demonstration Research
Corporation.
Rose, Nancy E. (1995). Welfare
or Fair Work: Women, Welfare, and Government Work Programs. New Brunswick,
New Jersey: Rutgers University Press.
Schneider, Jo Anne.(1997).Pathways
to Opportunity: The Role of Race, Social Networks, Institutions and
Neighborhood in Career and Educational Paths for People on Welfare.
Unpublished paper presented at the 27th Annual Urban Affairs Association
Meetings, Toronto, Canada.
(1997b). Social
Networks, Career and Training Paths for Participants in Education and Training
Programs: Technical Report.
Philadelphia: Institute for the Study of Civic Values.
(1997c). We’re Not Just Making Widgets: Non-profit Training
Providers and Welfare Reform in Philadelphia. Unpublished manuscript.
(1997d).
Welfare to Network. Demos Quarterly, 12, 30-32.
Stack, Carol. (1974). All Our
Kin: Strategies for Survival in a Black Community. New York: Harper.
Wilson, William Julius.(1996). When Work Disappears. New York:
Alfred A. Knopf.
Draft 4/98: For discussion only do not quote without
permission of author
Linking Welfare Recipients to Jobs:
Connections between Client Abilities, Previous Work
and Education History, Social Isolation and Placement in a Rapid Attachment
Program
Presented at the 28th Annual Urban Affairs Association
Meetings
Fort Worth Texas, April 1998
Jo Anne Schneider
Department of Sociology
and Anthropology
University of Wisconsin-Parkside
900 Wood Road, P.O. Box 2000
Kenosha, Wisconsin 53141-2000
Joanne.schneider@uwp.edu
