November 16, 1998 Welfare Recipients Who Find Jobs:
What Do We Know About Their Employment and Earnings?
by Sharon Parrott
Table of Contents
I. Overview
II. A Brief Description of Research Methodology
III. Key Findings Regarding Employment and Earnings
IV. Policy Implications
V. References
VI. AppendixClick here for PDF version of this report
State cash assistance programs increasingly are oriented toward work. A large proportion of parents receiving assistance are now required to prepare for employment, seek work, or hold jobs. Many states are now considering how best to help parents find jobs and support their families.
The success of state efforts in these areas is likely to hinge at least in part on how clearly states understand the employment situation of parents who do find jobs. Recent studies of families that have left welfare and evaluations of state welfare-to-work programs provide information about the earnings of welfare recipients who find jobs and characteristics of the jobs they hold. These include studies of recipients who find jobs in California, Delaware, Florida, Georgia, Indiana, Maryland, Minnesota, Michigan, Ohio, Oregon, South Carolina, and Wisconsin.(1) While each report studies a somewhat different group of parents and covers different time periods, all of the studies demonstrate that recipients who find jobs typically work a substantial number of hours per week but are paid low wages. The jobs held by parents who have left welfare or parents combining work and assistance often fail to provide basic benefits such as paid sick days, vacation leave, and health benefits.
- Employed former recipients and recipients combining work and welfare typically are paid less than $8 per hour and a substantial portion earn less than $6 per hour.
- Those who find jobs tend to work a substantial number of hours typically more than 30 hours per week during weeks in which they are employed.
- Despite this relatively high number of weekly work hours, recipients who find jobs typically earn between $2,000 and $2,700 per quarter (or between $8,000 and $10,800 annually), a total well below the poverty line for a family of three.(2) The low quarterly earnings figures suggest that many recipients who find jobs have periods of unemployment or very low earnings during a three-month period.
- Not surprisingly, recipients with better skills have higher earnings when they find jobs than do recipients with very low skills. A three-city study found that recipients who have a high school diploma or GED (General Equivalency Degree) have quarterly earnings that are between 19 and 29 percent higher than those who lack diplomas.
- In addition to low wages, most recipients who find jobs do not receive paid vacation or sick leave from their employer or employer-sponsored health insurance. Many such parents also will not be covered by the federal Family and Medical Leave Act (FMLA) which requires employers to provide up to 12 weeks of unpaid leave for employees who need time off for certain reasons such as to care for a child with a serious health condition or to attend to their own serious health condition. Only employees who have worked for their current employer for at least one year and who worked at least 1,250 hours approximately 25 hours per week in the last year are covered under the FMLA. Thus, many recipients who find jobs may not have access to paid or unpaid leave even when a family member is seriously ill.
These findings suggest that recipients who find jobs are likely to have incomes that are inadequate to meet their families' basic needs. Earnings alone are likely to be particularly inadequate for families in which the parent has very low skills. For many families, a combination of earnings, cash and in-kind government income support, and in the case of single-parents child support from the non-custodial parent will be necessary to make ends meet. Many states have freed-up resources due to declining welfare caseloads. The findings described in this report highlight the importance of states using these and other resources to provide supports to working families whose income from their earnings alone leave them unable to meet their basic needs.
A Brief Description of Research Methodology
Recent studies of employment and earnings patterns for welfare recipients who find jobs have relied upon one of two basic approaches. One set of studies including those conducted in Indiana, Maryland, South Carolina, Washington State, Wisconsin as well as Cuyahoga County (which encompasses Cleveland, Ohio) and the city of Milwaukee have collected information from or about former welfare recipients who left their state's cash assistance program.(3) A second set of studies initiated before enactment of the 1996 federal welfare law including recent evaluations in California, Delaware, Florida, Georgia, Michigan, Minnesota, and Oregon follow a group of welfare recipients who were subject to new rules under welfare reform "demonstration" programs. The earnings and aid receipt of this group (called the "program group") were compared to the earning and aid receipt of a group of recipients in a "control group" who continued to receive assistance under the standard rules of the AFDC program. These evaluations include information on the employment rates and earnings of recipients who found jobs.(4)
In contrast to the studies of former recipients, these evaluations provide earnings information for all recipients who find jobs, including those who continue to receive cash assistance while they work.
Both types of studies employed varying methods to gather information about the employment rates and earnings of recipients who found jobs.
- Unemployment insurance data: Four of the studies that followed former welfare recipients studies in Maryland, Wisconsin, the city Milwaukee, and Cuyahoga County, Ohio and all of the evaluations of demonstration programs reviewed here gathered information about the earnings of recipients by using data collected through the unemployment insurance (UI) system. (The unemployment insurance system collects quarterly earnings information that is used to determine eligibility for unemployment insurance benefits for individuals who become unemployed.)
Because most workers' earnings are reported by employers to state UI programs on a quarterly basis, this methodology has the advantage of providing basic information on parents' quarterly earnings and on the types of employers for whom they work without having to locate and contact the parents themselves. The disadvantage of this approach is that the UI data shed light on only a limited array of research questions. More detailed information regarding hourly wage rates, hours worked per week, and benefits provided by employers cannot be obtained from UI administrative records.
- Surveys: Several studies survey parents who once received assistance or participated in the demonstration program. These include studies of former recipients in Indiana, South Carolina, and Washington State and some evaluations of welfare demonstration programs such as those done in Florida, Portland, Oregon, and Minnesota. These surveys typically ask a range of questions including whether a parent is employed, what kind of job she has, how much she earns, and what types of fringe benefits are offered by her employer. The advantage of survey methodology is that information can be collected from parents on a broad range of topics related to their employment and earnings. The reliability of this information, however, depends greatly upon the extent to which researchers are able to locate and contact former recipients. The survey response rates in Indiana (71 percent), South Carolina (76 percent), and Washington state (62 percent) as well as those done as part of rigorous evaluations of demonstration programs were high enough to yield useful results.(5)
It is important to note that none of the studies reviewed here provide information on the employment rates and earnings of parents who, despite meeting the financial eligibility for a state's cash assistance program, do not apply for assistance or whose applications for assistance are denied. The extent to which parents in families "diverted" from cash assistance programs find jobs or have other sources of income from which they meet basic needs are important research questions.
Data from the Department of Labor Indicates
Increased Employment Rates Among Single MothersNational data now indicate that employment rates among single mothers (the group most likely to receive welfare) have risen over the past several years. Between 1994 and 1997, unpublished data from the Bureau of Labor Statistics show that the number of separated, divorced, and never-married mothers who held jobs at some point during the year increased by 14 percent 867,000 more single mothers held jobs in 1997 than did in 1994.(6) While this increase is not entirely due to increased employment among former recipients, many of the additional single mothers who are now working would likely be receiving assistance if they were not employed. There are many likely reasons why employment rates among single mothers has increased a strong economy increased the availability of jobs for less skilled workers, increased supports for low-income working families such as an expanded earned income tax credit and expanded Medicaid eligibility for children in low-income working families increased the returns to work, and an increased emphasis on work within welfare programs required parents to search and prepare for employment. It is difficult to disentangle the relative importance of these factors.a
____________________
a See "Can the Labor Market Absorb Three Million Welfare Recipients?" by Gary Burtless, Brookings Institution Working Paper, June 1998.
Key Findings Regarding Employment and Earnings
1. Most of the studies that track former welfare recipients have found that between half and three-quarters of parents are employed shortly after they leave the rolls.
Studies conducted in Indiana, Maryland, South Carolina, Wisconsin, Cuyahoga County, Ohio, and the city of Milwaukee found that between half and three-quarters of recipients who left welfare were working.(7)
- Among Indiana families that had received assistance at some point between May 1995 and May 1996 but were no longer receiving assistance when surveyed in the early part of 1997, 64.3 percent included employed parents.
- About half of the former Maryland recipients who left the state's welfare program between October 1996 and September 1997 were employed in the calendar quarter after they left the state's cash assistance program.
- Nearly 70 percent of parents who left South Carolina's welfare program between April and June 1997 were working when interviewed between February and June of 1998. It is important to note that the only families included in this study were those that did not return to the cash assistance program between the time they left the welfare program and the time of the interview.
- Some 72 percent of former recipients in Wisconsin who left the program between August 1995 and July 1996 worked in the quarter following their exit from welfare. About two-thirds of all former recipients had earnings of more than $500. Former recipients who did not return to the welfare program for at least one year following their exit had higher employment rates. Some 74 percent of these former recipients worked in the first quarter following their exit and some 80 percent were working in the fourth quarter following their exit from welfare.(8)
- The Cuyahoga County, Ohio (the county that encompasses Cleveland) study measured the quarterly earnings of recipients who left welfare in 1996. Earnings were measured in each of the four quarters following the quarter in which the family left welfare. The study found that slightly more than half of the former recipients were working in the quarters after they left welfare. For example, among adults who left the program in the first quarter of 1996, some 56 percent were working in the second quarter following the exit.
- Among Milwaukee, Wisconsin families that received assistance in December 1995 but were no longer receiving assistance in September 1996, 65.7 included parents who were working in the fourth quarter (October-December) of 1996.(9)
A study of former recipients in the state of New York, however, found lower rates of employment among former recipients. Among former welfare recipients with children (New York also provides some cash assistance benefits to poor single individuals and childless couples), only 30 percent were working in the quarter after they left welfare.(10) It is unclear why the employment rate among former New York recipients is so substantially lower than that found in all of the other studies of former recipients.
2. Recipients who find employment typically work a significant number of hours.
Surveys of recipients who find jobs usually ask employed (or recently employed) parents how many hours per week they work. These studies typically find that recipients who find jobs work more than 30 hours per week. While recipients work a large number of hours when they are employed, quarterly earnings data from these and other studies suggest that, over a three-month period, many recipients experience weeks in which they either do not work at all or work less than the number of hours they report as typical.
Table 1 provides data from six studies on the number of hours recipients who find jobs work. For each study, the table describes the group of recipients for whom hours worked data were gathered. The column "Hours Worked" provides information on the hours worked by recipients who found jobs. Some studies provide information on the average number of hours worked by recipients who found jobs while others provide information on the distribution of hours worked. The table presents the data as they are presented in the studies.
Studies that measure the hours worked by employed former recipients and those that measure the hours worked by all recipients who found jobs including those who combine work and aid receipt find that parents typically work more than 30 hours per week. The Indiana study measured the hours worked by employed former recipients and those combining work and welfare separately. Those recipients combining work and aid receipt work substantially fewer hours than those no longer receiving assistance. This is not surprising many families in which a parent works full-time at minimum wage will have incomes above the eligibility limits for Indiana's welfare program.(11)
Table 1: Number of Hours Worked Per Week By Recipients Who Find Jobs Area Studied
Group Studied
Hours Worked
Escambia County, Floridaa
recipients who found jobs, including some who combined work and aid receipt and some who left welfare
93% worked 20+ hours/week
74% worked 30+ hours/week
47% worked 40+ hours/weekIndiana working former recipients 62% worked 35+ hours/week
working current recipients 35% worked 35+ hours/week
5 Urban Counties in Minnesota long-term urban recipients who found jobs, including some who combined work and aid receipt and some who left welfare average hours worked per week:
31Portland,Oregon recipients who found jobs, including some who combined work and aid receipt and some who left welfare
average hours worked per week:
35South Carolina working former recipients
average hours worked per week:
36Washington State former recipients who were working at the time of the survey or had worked in the prior 12 months
average hours worked per week:
34Notes: a Escambia County encompasses the city of Pensacola.
3. Recipients who find employment typically are paid low wages.The studies that surveyed recipients who found jobs provide data on the hourly wage rates of recipients who found jobs. While there is some variation among the studies, all find that recipients who find jobs earn low wages, typically below $8 per hour and often below $6 per hour. As would be expected, states such as Oregon and Washington State that have higher-than-average wage levels generally also have higher wage rates among recipients who find jobs.(12)
Table 2: Hourly Wages of Recipients Who Find Jobs Area Studied
Group Studied Hourly Wages Escambia County,
Floridarecipients who found jobs,including some who combined work and aid receipt and some who left welfare
64% earned less than $6/hour
77% earned less than $7/hour
93% earned less than $10/hourIndiana working former recipients 39% earned less than $6/hour
80% earned less than $8/hour5 Urban Counties in Minnesota long-term urban recipients who found jobs, including some who combined work and aid receipt and some who left welfare average hourly wage:
$6.55Portland, Oregon recipients who found jobs, including some who combined work and aid receipt and some who left welfare
average hourly wage:
$7.34South Carolina working former recipients
average hourly wage:
$6.44Washington State former recipients who were working at the time of the survey or had worked in the prior 12 months
average hourly wage: $8.42
median wage: $7.40aNotes: a The median wage in the Washington State study was significantly lower than the average wage. The median wage figure means that among former recipients in Washington State, half earned less than $7.40 while half earned more than this amount. This indicates that a group of recipients earn wages substantially higher than the median figure, which increases the average.
4. Studies that measure parents' earnings over three-month periods find earnings levels well below the poverty line.As discussed above, some evaluations of welfare demonstration programs and some studies of former recipients use data from the unemployment insurance system to measure parents' quarterly earnings that is, their earnings over a three month period. Evaluations of demonstration programs in California, Delaware, Florida, Georgia, Minnesota, Michigan, and Oregon consistently find that recipients who find jobs earned an average of between $2,000 and $2,700 per quarter, or between $8,000 and $10,800 annually. Studies of former welfare recipients in Maryland, Milwaukee, and Cuyahoga County, Ohio find very similar quarterly earnings levels.
These quarterly earnings data are important because they show working recipients' earnings over a three-month period. Many recipients who find jobs do not work every week during a three-month period, a fact which is captured by quarterly earnings measures but not by surveys that ask parents about their hourly wages and the hours they usually work when employed.
As noted, average quarterly earnings figures in studies measuring earnings in this manner typically fall between $2,000 and $2,700, or between $154 and $210 per week. By contrast, the South Carolina evaluation found that, on average, employed former recipients earned $230 per week and in Washington State, the typical (or median) employed former recipient earned $250 per week. (The average weekly earnings of Florida recipients who found jobs was lower at $207 per week.)(13)
While the findings from the studies that measure quarterly earnings and those that measure weekly earnings may seem to be in conflict, they are not. When individuals are surveyed and asked about their weekly earnings (or their hourly wage and number of hours worked per week), only those individuals who worked in the week the survey asks about provide data on their earnings. Thus, the figures represent the average weekly earnings of recipients or former recipients when they are working.(14) The quarterly earnings figures, by contrast, represent parents' earnings over a three-month period. All parents who worked at some point over the three-month period are included when determining average quarterly earnings of those who worked. Therefore, the quarterly earnings of a parent who was employed during part of the quarter and unemployed during part of the quarter would be included when determining average quarterly earnings of parents employed at some point during the quarter.
As is discussed later, some studies indicate that a large number of recipients who find jobs work in temporary employment agencies. Employees for such agencies often have must wait significant periods of time between assignments. The quarterly earnings of a temporary agency employee could be well below what might be expected if only her weekly earnings during the weeks she was employed were examined.
Thus, the data presented in this report on average quarterly earnings probably provide a better sense of families' earnings over the course of a year, while the data on weekly and hourly earnings provide a better sense of the types of jobs recipients find and the hours they work during periods when they are employed.
Table 3: Quarterly Earnings of Former Welfare Recipients Area Studied Time Period over which earnings were measured Quarterly Earnings Maryland Number of quarters following exit: 1a Calendar period over which recipients studied had exited welfare program: October 1996 - September 1997
Average quarterly earnings:
$2,384(avg. monthly: $794)
Wisconsin Number of quarters following exit: 3b Calendar period over which recipients studied had exited welfare program: August 1995 - July 1996
Average quarterly earnings:
$2,563(avg. monthly: $854)
Cuyahoga County, Ohio Number of quarters following exit: 3c Calendar period over which recipients studied had exited welfare program: January 1996 - December 1996
21% earned less than $1,000 per quarter
40% earned less than $2,000 per quarter
Milwaukee, Wisconsin This study followed parents who received assistance in December 1995 but were not receiving aid in September 1996. Earnings were measured in the fourth quarter of 1996. 42% earned less than $2,500 per quarter
76% earned less than $4,000 per quarter
Notes:a The Maryland study followed some former recipients for two quarters following their welfare exit. In the second quarter following the exit, 53 percent of former recipients were working. Employed former recipients earned an average of $2,439 in the second quarter following their welfare exit.
b This study measured earnings for each of the five quarters following a recipient's welfare exit. The third quarter earnings figures are shown in the table. Additional information is available in the Appendix.
c This study measured earnings for each of the four quarters following a recipient's welfare exit. The third quarter earnings figures are shown in the table. Additional information is available in the Appendix.Table 3 summarizes the quarterly earnings findings from three studies of former recipients done in Maryland, the state of Wisconsin, Cuyahoga County, Ohio (the county that encompasses Cleveland), and Milwaukee. The table includes a description of the point at which earnings were measured relative to when the family left the state's welfare program. Table 4 summarizes the findings from six evaluations of state welfare demonstration programs in which the quarterly earnings of all recipients who found jobs were measured, including those who continued to receive cash assistance. Table 4 also describes the point at which earnings were measured relative to when the recipient was assigned to the waiver program. For example, the Delaware evaluation provides information on participants' earnings four quarters or one year after the recipient was first assigned to Delaware's waiver program.(15) In both tables, the column labeled "quarterly earnings" provides the quarterly earnings among those families in which an adult was employed at some point during the quarter.
Table 4: Quarterly Earnings of Welfare Recipients (Including Those No Longer Receiving Assistance and Those Combining Welfare and Work) Who Find Employment Area Studied Time Period over which earnings were measured Quarterly Earnings Quarterly Earnings as a Percent of Poverty Linea Los Angeles, California
Types of families in demonstration: sample of current recipients and new applicants Number of quarters after families assigned to demonstration program: 3
Calendar Period earnings data reflect: fourth quarter 1996
$2,085 (avg. monthly:
$695)67%
Delaware
Types of families in demonstration: sample of current recipients and new applicants Number of quarters after families assigned to demonstration program: 4
Calendar period earnings data reflect: late 1996 and early 1997
$2,115
(avg. monthly: $705)
66%
Escambia County, Florida
Types of families in demonstration: sample of current recipients and new applicants Number of quarters after families assigned to demonstration program: 8
Calendar period earnings data reflect: mid-1996 and early 1997
$2,046
(avg. monthly: $682)
65%
Michigan
Types of families in demonstration: sample of current recipients only Number of quarters after families assigned to demonstration program: 16
Calendar period earnings data reflect: late 1994 and mid 1996
$2,723
(avg. monthly: $907)
90%
5 Urban Counties in
MinnesotaTypes of families in demonstration: sample of single-parent recipients in urban counties who had received assistance 24 of the 36 months prior to being assigned to the Minnesota Family Investment Program Number of quarters after families assigned to demonstration program: 6
Calendar period earnings data reflect: late 1995 to mid-1996
$2,098
(avg. monthly: $699)
67%
Portland,
OregonTypes of families in demonstration: sample of applicants who did not find jobs upon initial job search and recipients Number of quarters after families assigned to demonstration program: data reflect the average quarterly earnings per quarter worked during the first two years following random assignment
Calendar period earnings data reflect: random assignment occurred between February 1993 and December 1994
$2,136
(avg. monthly: $712)
70% Notes: a The poverty line is adjusted each year based on inflation. The poverty line figures used in these calculations reflect the measure for the year in which earnings were measured. Finally, both tables show the average monthly earnings of recipients who are employed at some point during the quarter examined. While included in the table because monthly earnings are often easier to understand than quarterly earnings figures, it is important to note that many families will have varying earnings in each of the three months of the period.
5. Earnings of welfare recipients who find jobs typically are higher for those who had a high school diploma at the time of program enrollment than for those who did not have a diploma.Data from a three-site study of welfare-to-work programs conducted by the Manpower Demonstration Research Corporation show that among recipients who find jobs, those with a high school diploma have significantly higher earnings.(16) The three site evaluation was designed to determine the relative merits of welfare-to-work programs that focused on immediate employment versus those that focus on upgrading skills. In each of the three sites studied, recipients participated in either a welfare-to-work program that emphasized immediate employment (called a "labor force attachment" program) or one that emphasized upgrading skills through education and training (called a "human capital development" model). Recipients were assigned to the welfare-to-work programs during either 1991 or 1992. As part of this evaluation, MDRC measured average quarterly earnings data for two categories of recipients who found jobs those who had a high school diploma or GED at the time they entered the program and those who did not.
Table 5: Average Quarterly Earnings Among Working Participants With and Without a High School Diploma or GED Site Studied Average Quarterly Earnings Among Working Participants With a High School Diploma or GED Average Quarterly Earnings Among Working Participants Without a High School Diploma or GED Atlanta, Georgia $2,126 $1,644 Grand Rapids, Michigan $1,838 $1,541 Riverside, California $2,663 $2,075 The data presented in Table 5 represent average quarterly earnings of employed adults during the first year after they entered the program. The data reflect the earnings of participants in the labor force attachment programs.(17) While neither group had high earnings, those with a diploma or GED had significantly higher earnings on average between 19 percent and 29 percent higher than those who lacked a diploma. This suggests that when recipients with very low skills find jobs, their earnings are likely to be far below what is needed to meet their families' basic needs.
6. Recipients who find jobs typically find jobs in sales, food preparation, clerical support, or other service jobs.Five studies those of former recipients in Maryland, South Carolina, Washington State, Wisconsin, and a Milwaukee study of jobs held by recipients and former recipients provide information on the industries and occupations of recipients who find jobs. The Maryland, Wisconsin, and Milwaukee studies categorized jobs by the industry or type of firm a parent worked in (i.e., department store, restaurant, temporary employment agency) while the other two studies categorized jobs by the type of work the former recipient performed. While characterizing the jobs somewhat differently, all four studies indicate that a substantial portion of former recipients work in sales, food preparation, or clerical support jobs and a large portion work for temporary employment agencies. For example:
- More than one-third of the working former recipients in Maryland found employment in the wholesale and retail trade industry. The most common types of employers in this industry were eating and drinking establishments, department stores, and grocery stores. In addition, 24 percent of working former recipients found employment in the "organizational services" industry that is, for-profit, non-profit, and governmental entities that provide services such as health care or social services.
- In South Carolina, nearly half 46 percent of employed former recipients worked in "service occupations." These occupations included food and beverage preparation, lodging and related services, miscellaneous personal services, etc. More than one-quarter worked in clerical or sales occupations defined as including computing and account recording, typing and filing, and stock clerking.
- The Washington State study found that 17 percent of employed former recipients worked in clerical jobs, 14 percent worked in general labor or construction jobs, and the majority of remaining former recipients working in sales or services jobs.
- The statewide study of former Wisconsin recipients also finds a substantial proportion of parents working in retail and service jobs as well as working in temporary agencies. Nearly 10 percent of working former recipients were employed by temporary agencies in the first quarter following their exit from welfare. Some 15 percent worked in hotels or restaurants, 23 percent worked in retail trade or personal or business services and nearly 15 percent worked in health services. The report documents a substantial movement between industries. For example, among those former recipients who worked for temporary agencies in the first quarter following their exit from welfare, only about one-quarter were working for such agencies in the fifth quarter following their welfare exit.
- A Milwaukee study reviewed the jobs held between January 1996 and March 1997 by single parents who had received AFDC in December 1995.(18)Some of those who worked during the January 1996 to March 1997 period received cash assistance during this period while others had left the welfare rolls. Some 30 percent of all jobs held by Milwaukee recipients who found jobs were in temporary employment agencies. This figure understates the extent to which these parents were employed at some point in such an agency. More than four in ten 42 percent of all parents who worked during the 15 month period studied worked at some point in a temporary agency.
7. Many welfare recipients who find jobs do not receive employment-related benefits such as paid vacation and sick leave.Two studies the Florida evaluation and a study of former Washington state recipients provide some information on the extent to which recipients who find jobs receive paid sick and vacation leave. Among Florida recipients who found jobs, only 36 percent worked in jobs that offered paid sick days, 46 percent received paid vacation days, and 43 percent worked for employers that offered health benefits.(19) A Washington State study of former recipients found similar rates of receipt of such fringe benefits. This lack of basic benefits such as paid sick leave and vacation days is particularly problematic for working parents. Without paid sick or vacation leave, a parent could lose a significant portion of her monthly income if a child was sick for two days with the flu or the parent needed to take time off to attend a teacher's conference.
It is important to note that many recipients who find jobs will not be covered by the federal Family and Medical Leave Act (FMLA) which requires employers to provide up to 12 weeks of unpaid leave for employees who need time off for certain reasons. For example, the FMLA requires employers to provide unpaid leave to employees who need to care for a child with a serious health condition or to attend to their own serious health condition. However, only employees who have worked for their current employer for at least one year and who worked at least 1,250 hours approximately 25 hours per week in the last year are covered under the FMLA. Thus, many recipients who lack paid leave also may not have access to unpaid leave, even when a family member is seriously ill.(20)
8. Many welfare recipients who find jobs do not receive employer-provided health insurance.The Florida evaluation and the studies of former recipients in Indiana, Washington State, and South Carolina each provide some information on health care coverage of recipients who find jobs or leave welfare. While each study asks the health insurance questions differently, they all indicate that many welfare recipients find jobs that do not offer employer-sponsored health care coverage. The studies of former recipients show that many employed former recipients are without any form of health insurance.
The Florida evaluation and the Indiana study of families that left welfare provide information on the extent to which those recipients who found jobs worked in jobs in which health benefits were offered to employees. In both cases, the surveys asked working respondents whether their employers offered health insurance. The surveys did not ask whether respondents actually were covered by an employer-sponsored health care plan. Because someone who chooses not to participate in an employer-sponsored health plan will be classified as someone who works in a job that "offers" health insurance, these figures over-estimate the proportion of individuals who actually receive employer-provided coverage.
- Some 61 percent of employed former recipients in Indiana worked for employers that offered health coverage to their employees. The Indiana study also surveyed families that were still receiving assistance. Among working recipients, only 39 percent worked in jobs that offered health insurance.
- Among those recipients in Florida who found employment (some of whom still received assistance while others did not), 43 percent worked for employers who offered health insurance coverage.
Some employees offered employer-provided health insurance choose not to participate in the health plans. Employees decline coverage for a variety of reasons some choose not to participate because the premiums employees must pay are too costly or the benefit package is too limited while others do not participate because they have access to health care coverage through their spouses' employer-sponsored plan or through Medicaid. Data compiled by KPMG Peat Warwick indicate that families had to contribute an average of $1,615 a year for employer-based family coverage in 1996, an amount that makes coverage inaccessible for many low-income working families.(21) Cost-sharing requirements even for employee-only coverage (as distinct from full family coverage) can be prohibitive. The surveys do not ask respondents whether the employer provides health benefits to the employee only or also to his or her dependents.
Among former recipients in Washington State who were working at the time of the survey or had worked at some point during the prior 12 months, 37 percent worked for employers that "provided" health benefits. The wording of the question makes it unclear whether this represents the proportion who received health benefits through their employer or the proportion offered such coverage.
The South Carolina study of former welfare recipients also gathered some information on health insurance coverage. The health insurance data were not reported separately for working and non-working former recipients. The South Carolina study interviewed 391 households (about two-thirds of former recipients were working). These households included a total of 820 children and 550 adults. (Some of these adults and children may not have been part of the welfare "case" that is, they may not have been part of the assistance unit receiving assistance.)
- Some 84 percent of the children were covered by health insurance. Among those covered, 88 percent were covered by Medicaid and 12 percent were covered by private insurance.(22)
- Only 58 percent of the adults in these households were covered by health insurance. Among those covered, 73 percent were covered by Medicaid, 25 percent were covered by private insurance, with the remainder being covered by Medicare.(23)
Policy Implications
The findings outlined above suggest that recipients who find jobs continue to struggle to meet their basic needs despite working a substantial number of hours per week when employed. As the number of non-working families receiving basic cash assistance falls, states should consider investing in efforts that support low-income working families. Policies that provide supports to low-income working families such as income supplements, child care assistance, help in meeting their transportation needs, health care coverage to children and parents, and skill upgrading opportunities can help parents meet their families' basic needs, retain employment, and find better-paying and more stable jobs.
- Increased income support for working poor families: While some 42 states have changed the way earnings are treated when determining a family's cash assistance benefit generally allowing working poor families to receive more assistance than under former AFDC rules in more than 40 states, families lose eligibility for cash assistance before their earnings reach the poverty line.(24) States with freed up resources could increase the amount of assistance available to working poor families by further reducing the amount by which benefits fall when earnings rise or by increasing basic benefit levels. States can also provide income support for low-income working families in a program separate from TANF. For example, a state can provide worker stipends to supplement earnings or pay for work-related expenses, assure child support to children whose non-custodial parent fails to pay support owed, or establish a state earned income tax credit.
- Education and training opportunities for low-wage workers: Most states have adopted a "Work First" approach in their welfare-to-work programs. This approach focuses on helping parents find jobs immediately rather than on education and training that might increase their chances of finding higher-paying, more stable jobs. Parents who have found employment but are working in low-paying, unstable jobs could benefit from education and training opportunities that allow them to upgrade their skills while they work so they can ultimately attain stable, higher-paying jobs.(25)
- Support services: Low-income working parents are likely to have more success maintaining employment if they have stable child care and transportation arrangements. While many states have expanded the availability of child care assistance in the past several years, freed-up resources stemming from caseload decline provide new opportunities to expand the availability of child care subsidies and to invest resources in expanding the supply and improving the quality of child care. Similarly, transportation remains a serious obstacle to maintaining steady employment for many low-income working parents. Resources can be used to provide direct transportation subsidies such as bus passes as well as to help working poor parents purchase cars and afford both insurance and maintenance costs associated with owning a car.
- Health insurance: While low-income children often remain eligible for publicly-funded health insurance when their families leave welfare either through Medicaid or new child health programs parents often become uninsured. The welfare law allows states to expand Medicaid coverage to low-income working parents. Providing ongoing health care coverage to low-income working parents can make leaving welfare for work a more viable option for many parents. Moreover, the opportunity to receive regular health care could promote job retention among low-income working parents by helping them avoid or shorten illness that could cause them to miss work. For some parents in need of ongoing medical care, coverage will eliminate the need to choose between forgoing essential health care in order to keep a job and leaving a job to qualify for Medicaid.
The following is a list of the studies used in this report. They are listed alphabetically by the state(s) studied.
1. Los Angeles, California
The Los Angeles Jobs-First GAIN Evaluation: Preliminary Findings on Participation Patterns and First-Year Impacts, by Stephen Freedman, Marisa Mitchell, and David Navarro, Manpower Demonstration Research Corporation Working Paper, August 1998. This evaluation is available on the web going to www.mdrc.org and clicking on "publications" and scrolling down to "Los Angeles's Jobs-First GAIN Program."
Geographic Region Covered: Los Angeles County, California
2. Riverside, California
Atlanta, Georgia
Grand Rapids, MichiganEvaluating Two Welfare-to-Work Program Approaches: Two-Year Findings on the Labor Force Attachment and Human Capital Development Programs in Three Sites, by Gayle Hamilton, et al., Manpower Demonstration Research Corporation, December 1997.
Geographic Regions Covered: Riverside, California; Atlanta, Georgia; and Grand Rapids, Michigan.
3. Delaware
The ABC Evaluation: The Early Economic Impacts of Delaware's A Better Chance Welfare Reform Program, by David Fein and Jennifer Karweit, Abt Associates, December 1997.
Geographic Region Covered: Recipients in five of the state's 13 welfare offices participated in the state's welfare demonstration project. The five offices in Carroll's Plaza, Georgetown, Hudson, Thatcher, and Williams were located throughout the state.
4. Florida
The Family Transition Program: Implementation and Interim Impacts of Florida's Initial Time-Limited Welfare Program, by Dan Bloom, et al., Manpower Demonstration Research Corporation, March 1998.
Geographic Region Covered: Escambia County, Florida this county includes Pensacola, Florida.
5. Atlanta, Georgia
See #2 above.
6. Indiana
The Indiana Welfare Reform Evaluation: Who is On and Who Is Off? Comparing Characteristics and Outcomes for Current and Former TANF Recipients, Abt Associates, September 1997.
Geographic Region Covered: Survey was representative of the entire state.
7. Maryland
Life After Welfare: Second Interim Report, School of Social Work, University of Maryland, March 1998.
Geographic Region Covered: Statewide
8. Michigan
Final Impact Report: The Evaluation of To Strengthen Michigan Families, by Alan Werner and Robert Kornfeld, Abt Associates, Inc., September 1997.
Geographic Region Covered: Kalamazoo County, Madison Heights district welfare office in Oakland County (a suburb of Detroit), and two district offices (McNichols/Goddard and Schaefer/Six Mile) in Wayne County (the county that encompasses Detroit).
9. Grand Rapids, Michigan
See #2 above.
10. Minnesota
Making Welfare Work and Work Pay: Implementation and 18-Month Impacts of the Minnesota Family Investment Program, by Cynthia Miller, et al., Manpower Demonstration Research Corporation, October 1997.
Geographic Region Covered: The evaluation followed recipients in three urban counties Hennepin (which includes Minneapolis), Anoka, and Dakota and four rural counties. The findings used in this report are those for long-term recipients from the three urban counties.
11. Cuyahoga County, Ohio
Work After Welfare: Employment in the 1996 Exit Cohort, Cuyahoga County, by Claudia Coulton, Marilyn Su, Neil Bania, and Edward Wang, Center on Urban Poverty and Social Change Briefing Report No. 9803. This report is available on the web at: http://povertycenter.cwru.edu/br9803work_.PDF.
Geographic Region Covered: Cuyahoga County, Ohio which includes the city of Cleveland.
12. Portland, Oregon
Implementation, Participation Patterns, Costs, and Two-Year Impacts of the Portland Welfare-to-Work Program, by Susan Scrivener, et al., Manpower Demonstration Research Corporation, May 1998.
13. South Carolina
Survey of Former Family Independence Program Clients: Cases Closed During April Through June, 1997, South Carolina Department of Social Services, July 1998.
Geographic Region Covered: Survey was representative of the entire state.
14. Washington State
Washington's TANF Single Parent Families Shortly After Welfare Survey of Families Which Exited TANF Between December 1997 and March 1998, Department of Social and Health Services, July 1998.
Geographic Region Covered: Survey was representative of the entire state.
15. Wisconsin (statewide study)
Post-Exit Earnings and Benefit Receipt Among Those Who Left AFDC in Wisconsin, by Maria Cancian, Robert Haveman, Thomas Kaplan, and Barbara Wolfe, Institute for Research on Poverty, University of Wisconsin Madison, October 1998. http://www.ssc.wisc.edu/irp/research/home.htm
Geographic Region Covered: The state of Wisconsin
16. Milwaukee, Wisconsin
Employment and Earnings of Milwaukee County Single Parent AFDC Families: Establishing Benchmarks for Measuring Employment Outcomes Under "W-2," by John Pawasarat, University of Wisconsin-Milwaukee Employment and Training Institute www.uwm.edu/Dept/ETI/afdcearn.htm
The Employer Perspective: Jobs Held by the Milwaukee County AFDC Single Parent Population (January 1996 - March 1997), by John Pawasarat, University of Wisconsin - Milwaukee Employment and Training Institute. This report is available on the web: www.uwm.edu/Dept/ETI/employer.htm.
Geographic Region Covered: Milwaukee, Wisconsin
Florida
The Family Transition Program: Implementation and Interim Impacts of Florida's Initial Time-Limited Welfare Program, by Dan Bloom, et al., Manpower Demonstration Research Corporation, March 1998.
The evaluation of Florida's Family Transition Program (FTP) included a survey of families two years after they had first been assigned either to FTP or to the control group. The families surveyed had been assigned to FTP between December 1994 and February 1995, thus the survey was conducted between December 1996 and February 1997. Some of the families surveyed were still receiving assistance while others were not. The survey included questions about adults' current or most recent job. (The results described below represent the responses of the "program" group, not the "control" group. While a larger proportion of program group members found jobs, the characteristics of those jobs, including their wage levels, did not differ substantially from those found by the control group.)
Earnings of Recipients Who Found JobsSurvey respondents who had worked at some point since they initially participated in FTP were asked about the hourly wage they earned in either their current or most recent job. The column labeled "Percent of Survey Respondents" in Table 6 below shows the proportion of respondents whose earnings in their current or most recent job fall within each of the various ranges. The column labeled "Cumulative Percent" shows the proportion of respondents who earn no more than the wage in the corresponding row. For example, the table shows that 32.2 percent of respondents who were (or had been) employed earned no more than $4.99 per hour.
Table 6: Florida FTP Participants' Wages in Current or Most Recent Job Hourly Wages Earned in Current or Most Recent Job
Percent of Survey Respondents
Cumulative Percent
Less than $4.25 17.7% - $4.25 - $4.99 14.5% 32.2% $5.00 - $5.99 31.9% 64.1% $6.00 - $6.99 12.9% 77.0% $7.00 - $9.99 15.7% 92.7% $10.00 or More 7.3% 100.0% How to read this table: 14.5 percent of survey respondents earned between $4.25 and $4.99 per hour in their current or most recent job. 32.2 percent of respondents earned below $5.00 per hour while 64.1 percent earned less than $6.00 per hour.
Hours Worked By Recipients Who Found JobsWhile respondents reported low wages, they also reported working a substantial number of hours.
- Some 47.1 percent reported working 40 or more hours per week, almost three-quarters worked at least 30 hours per week, and 93 percent worked at least 20 hours per week.(26)
Weekly EarningsRespondents also provided information on their weekly earnings in their current or most recent job.
- On average, respondents earned $207 per week. If a parent in a family of three worked 52 weeks earning $207 per week, the family's earnings would equal just $10,764 or 14 percent below the poverty line for a family of three in 1996. As is discussed below, most FTP participants who found jobs did not receive paid vacation or sick leave from their employers making it unlikely that a parent could work earn this amount every week.
Table 7: Florida FTP Participants' Weekly Earnings in Current or Most Recent Job Weekly Earnings Percentage Distribution Cumulative Distribution Less than $100 11.6% $100 - $149 17.3% 28.9% $150 - $199 24.9% 53.8% $200 - $299 29.7% 83.5% $300 or more 16.5% 100.0% How to read this table: 17.3 percent of survey respondents earned between $100 and $149 per week in their current or most recent job. 28.9 percent of respondents earned below $149 per week while 53.8 percent earned less than $199 per week.
Employer-Provided Fringe Benefits
Table 8 below shows the proportion of Florida recipients who found jobs who were offered various types of employer-provided benefits.
- Only slightly more than one-third of those recipients who found jobs received paid sick leave and only 46 percent received paid vacation days.
- Among recipients who found employment, 43 percent worked for employers who offered health insurance coverage.
Table 8: Proportion of Florida FTP Participants Who Found Jobs That Provided or Offered Different Types of Benefits Employer-Sponsored Benefit
Percent of Recipients Who Found Jobs That Offered Benefits (some recipients who found jobs continued to receive aid while others left welfare)
Paid Sick Days 36.3% Paid Vacation Days 46.0% Health Benefitsa 43.3% Dental Benefitsa 34.9% Tuition Assistance or Paid Training Classes 24.4% a The survey from which these data were obtained asked respondents whether their jobs offered health and dental benefits, not whether the respondent was actually covered by those policies. An employee might choose not to participate in an employer health or dental plan if, for example, the employee decided he or she could not afford the premiums.
IndianaThe Indiana Welfare Reform Evaluation: Who is On and Who Is Off? Comparing Characteristics and Outcomes for Current and Former TANF Recipients, Abt Associates, September 1997.
Abt Associates Inc. conducted a survey in early 1997 of 1,600 families that had participated in Indiana's welfare program between May 1995 and May 1996. The report on the survey findings provides data separately for families still receiving assistance at the time of the survey and families no longer receiving aid.
Percent of Former Recipients Who Were Working
- Nearly two-thirds 64.3 percent of those no longer receiving aid were working at the time of the survey.
Earnings of Former Recipients
- Among those working, 80.1 percent earned less than $8 per hour. Nearly four in ten working former recipients earned less than $6 per hour.
- Some 16 percent of those still receiving assistance were working at the time of the survey. As would be expected, these parents were more likely to work part-time and earned even lower wages than working former recipients. Two-thirds of working recipients earned less than $6 per hour.
Hours Worked By Employed Former Recipients
- 61.7 percent of employed former recipients in Indiana worked at least 35 hours per week. The number of hours worked among working parents who continued to receive assistance was substantially lower as would be expected. Some 35 percent of employed recipients worked 35 hours or more per week.
Employer-Sponsored Health Insurance Coverage
- Some 61 percent of employed former recipients in Indiana worked for employers that offered health coverage to their employees. The Indiana study also surveyed families that were still receiving assistance. Among working recipients, only 39 percent worked in jobs that offered health insurance.
MarylandLife After Welfare: Second Interim Report, School of Social Work, University of Maryland, March 1998.
This study followed a random sample of Maryland families whose welfare cases closed between October 1996 and September 1997. The study uses data available through the unemployment insurance system to determine the employment rates and earnings of families no longer receiving assistance. The study provides information about adults' earnings in each of the two quarters immediately following their families' exit from welfare as opposed to reporting families earnings for a particular calendar quarter.
Percent of Former Recipients Who Were Working
- About half of the adults in families that left Maryland's assistance program between October 1996 and June 1997 worked in the quarter following the exit from welfare.
Earnings of Former Recipients
- Those recipients who worked in the quarter following their welfare exit, earned an average of almost $2,400 during the quarter (or an average of $800 per month). This level of earnings is equal to almost 75 percent of the 1997 poverty line for a family of three.
- The study was able to track two quarters of post-welfare earnings for some former recipients. When earnings in the second quarter following the welfare exit were examined, the researchers again found that about half of the adults were working and they were earning roughly $2,400 over the quarter. Among those working in the second quarter following their welfare exit, some 85 percent were also working in the first quarter following their welfare exit. In other words, nearly all of those working in the second quarter also worked in the first quarter following their exit from welfare. This indicates that a large portion of those who left welfare and worked in the first quarter following their exit were still working in the second quarter and that few recipients who were not working in the first quarter following their welfare exit found jobs in the second quarter.
Subsequent Receipt of Cash Assistance
- The study was able to determine the proportion of families that had left welfare that subsequently received assistance. Among the 535 families for which a full year of data on subsequent aid receipt was available, less than one-quarter had subsequently received cash assistance.
Industry in Which Former Recipients Work
- More than one-third of the working former recipients in Maryland found employment in the wholesale and retail trade industry. The most common types of employers in this industry were eating and drinking establishments, department stores, and grocery stores.
- Some 24 percent of working adults found employment in the "organizational services" industry that is, for-profit, non-profit, and governmental entities that provide services such as health care or social services.
- Almost 20 percent found jobs in "personal/business" services such as temporary agencies, hotels, and companies that provide security services.(27)
Minnesota
Making Welfare Work and Work Pay: Implementation and 18-Month Impacts of the Minnesota Family Investment Program, by Cynthia Miller, et al., Manpower Demonstration Research Corporation, October 1997.
The evaluation of the Minnesota Family Investment Program (MFIP) included a survey of families that had been assigned to the MFIP program between September and December 1994. The families were surveyed one year after they had been assigned to MFIP. The families surveyed were long-term recipients recipients who had received assistance in 24 of the 36 months prior to being assigned to MFIP who lived in urban areas.
Almost two-thirds of the survey respondents were employed. (Some of those who were employed continued to receive welfare while others left the program entirely.) Those employed earned an average of $6.55 per hour and worked an average of 31 hours per week or about 13 percent below the poverty line for a family of three in 1995. More than four in ten working recipients had held at least two jobs since they had been assigned to MFIP suggesting significant job turnover.
Cuyahoga County, OhioWork After Welfare: Employment in the 1996 Exit Cohort, Cuyahoga County, by Claudia Coulton, Marilyn Su, Neil Bania, and Edward Wang, Center on Urban Poverty and Social Change Briefing Report No. 9803.(28)
This study used administrative welfare data and unemployment insurance data to track the employment and earnings of adults who left the Cuyahoga County cash assistance rolls in 1996. In total, there were 18,570 exits of adults from the welfare program in Cuyahoga County (a small number of these exits reflected adults who left the program, subsequently received aid during the year, and then left a second time). The researchers tracked these adults' employment rates and earnings for four quarters following their exit from the welfare program.
Percent of Former Recipients Who Were Working
- In each of the four quarters following welfare exits, slightly more than half of the former recipients were working. For example, among adults who left the program in the first quarter of 1996, some 56 percent were working in the second quarter following the exit.
Earnings of Former Recipients
- In each of the four quarters following welfare exits, between 21 and 23 percent of those working earned less than $1,000.
- In each of the four quarters following welfare exits, between 40 and 43 percent of those working earned less than $2,000 per quarter. The proportion earning below $2,000 did decline somewhat over time. That is, while 43.1 percent of those who were working in the first quarter following their exit from welfare earned less than $2,000, among those working four quarters after their exit from welfare 39.7% earned less than this level.
- By the fourth quarter following the welfare exit, some 42 percent of former recipients were earning more than $3,000 per quarter.
Subsequent Receipt of Cash Assistance
- Among those former recipients who worked in the first quarter following their welfare exit, 22 percent received cash assistance during the following year. This was a lower rate of subsequent welfare receipt than among those who left the program but were not working in the first quarter following their exit. Among working former recipients who returned to the welfare program, one-quarter combined work and welfare upon returning to the cash assistance program.
Portland, OregonImplementation, Participation Patterns, Costs, and Two-Year Impacts of the Portland Welfare-to-Work Program, by Susan Scrivener, et al., Manpower Demonstration Research Corporation, May 1998.
The evaluation of the JOBS program in Portland, Oregon provides information on hourly and weekly earnings of recipients who found employment. Portland's JOBS program is part of the National Evaluation of Welfare-to-Work Strategies, being conducted by MDRC, which is evaluating welfare-to-work programs in seven cities across the country. A sample of applicants (who had not found jobs through an initial job search) and recipients were assigned to either the program group or the control group. Program group members were required to participate in welfare-to-work activities while control group members did not face a participation requirement. The welfare-to-work program in Portland focused on placing recipients in jobs, rather than on education and training, but did make heavier use of short-term training and education than many "Work First" programs.
Applicants and recipients were assigned to the welfare-to-work program between February 1993 and December 1994. The study followed recipients for two years after they initially were assigned to the program. The Portland program produced some of the largest impacts on employment rates and earnings measured in a welfare-to-work program. That is, the program group members had significantly higher employment rates and earnings than members of the control group and the difference was larger than has been measured in most welfare-to-work programs. At the end of two years, clients were surveyed.
Percent of Recipients Who Were Working
- At the time of the survey, nearly half of those in the program group were employed (35 percent of those in the control group were employed).
Hourly Earnings of Employed Recipients
- Employed program group members worked for an average hourly wage of $7.34. (This was significantly higher than the wages earned by employed control group members who earned an average of $6.48 per hour.)
- The focus on attaining higher-paying jobs appeared to have some effect on job quality. Employed program group members earned an average of 86 cents more per hour than did employed control group members.
Hours Worked By Employed Recipients
- Employed program group members worked an average of 35 hours per week 80 percent worked full-time.
Weekly Earnings
- On average, employed program group members earned $260 per week. If these parents were able to maintain this level of earnings for 52 weeks, their earnings would roughly equal the poverty line for a family of three.
South CarolinaSurvey of Former Family Independence Program Clients: Cases Closed During April Through June, 1997, South Carolina Department of Social Services, July 1998.
The South Carolina Department of Social Services conducted interviews with almost 400 families that had left the state's welfare program between April and June of 1997. The interviews were conducted between February and April of 1998. Some types of families were excluded from the study. Specifically, only those families in which the adult was required to look for work (or the adult was voluntarily seeking work) while they were receiving assistance were included in the study. (In South Carolina, families headed by an incapacitated or pregnant adult and families in which the youngest child is under one year of age are not required to participate in work activities.) While these were the only families included in the study, this does not mean that all of the families left the program because they found employment. Finally, only those families that had not received cash assistance since their case had been closed were included in the study. Thus, this study provides information on the employment characteristics of adults who left the state's welfare program and did not receive assistance in the next 8-12 months.
Percent of Former Recipients Who Were Working
- At the time of the survey, more than two-thirds 69.6 percent of the adults interviewed were working.(29), (30)
- In addition to asking questions about individuals' current employment status, the survey also asked individuals who were not currently working whether they had worked at all since leaving welfare. Only 12.5 percent of all of those who left the rolls (and had not received cash assistance between the time they left the program and the survey) had never worked since leaving welfare.
Hours Worked By Employed Former Recipients
- These former recipients worked an average of 36 hours per week at an average wage of $6.44 per hour.
Occupations of Employed Former Recipients
- Nearly half 46 percent of employed former recipients in South Carolina worked in "service occupations." These occupations included food and beverage preparation, lodging and related services, miscellaneous personal services, etc.
- More than one-quarter worked in clerical or sales occupations defined as including computing and account recording, typing and filing, and stock clerking.(31)
Health Insurance
The South Carolina study of former welfare recipients also gathered some information on health insurance coverage. The health insurance data was not provided separately for working and non-working former recipients. The South Carolina study interviewed 391 households (about two-thirds of former recipients were working). These households included a total of 820 children and 550 adults. (Some of these adults and children may not have been part of the welfare "case" that is, they may not have been part of the assistance unit receiving assistance.)
- Some 84 percent of the children were covered by health insurance. Among those covered, 88 percent were covered by Medicaid and 12 percent were covered by private insurance.
- Only 58 percent of the adults in these households were covered by health insurance. Among those covered, 73 percent were covered by Medicaid, 25 percent were covered by private insurance, with the remainder being covered by Medicare.
Washington StateWashington's TANF Single Parent Families Shortly After Welfare Survey of Families Which Exited TANF Between December 1997 and March 1998, Department of Social and Health Services, July 1998.
Washington State surveyed 560 single parent families that had left the TANF program between December 1997 and March 1998. Among these families, 58 percent reported leaving the program due to increased earnings while another 10 percent left due to increased income from other sources. An additional 12 percent left due to a conflict with program requirements.(32)
Percent of Former Recipients Who Were Working
- More than two-thirds of those who left the program were working at the time of the survey which was conducted in April and May of 1998. Some 78 percent reported working at some point during the prior 12 months indicating that some former recipients were employed prior to leaving the welfare rolls. (In Washington State, recipients who find jobs can remain eligible for assistance until their monthly earnings exceed $1,000.)
Earnings of Former Recipients Who Worked At Some Point
- The report provides earnings information for the 78 percent of former recipients who were either employed at the time of the survey or had been employed at some point in the prior 12 months. The average wages of these former recipients were somewhat higher than in other studies. On average, former recipients who were working or had worked during the prior year earned $8.42 per hour, though the median wage was significantly lower. Among those former recipients who were employed or had worked during the last year, the median wage was $7.40 per hour. This means that half of such former recipients earned less than this amount while half earned more than this wage level.
Hours Worked By Employed Former Recipients
- As found in other studies of former recipients, recipients who were working or had worked during the previous 12 months worked a substantial number of hours. On average, such former recipients worked 34 hours per week at their current or most recent job.
Occupations in Which Employed Former Recipients Work
- The Washington State survey asked former recipients who were working at the time of the survey or had worked at some point in the past 12 months about the occupations in which they worked. Table 9 shows the occupations in which these former recipients were concentrated.
Table 9: Occupations of Former Recipients in Washington State Occupation
Proportion of former recipients who were employed at the time of the survey or who had worked during the prior 12 months
Administrative support/clerical/general office 17% General labor/construction/equipment operation 14% Retail and other sales 14% Health care 11% Food and beverage services 9% Child care/personal services 9% Janitors/maids 5% Teacher aids/educational services 5% Other 14% Note: Figures do not add to 100 percent due to rounding. Paid Sick and Vacation Leave
- The survey also asked about the employer-provided benefits of those former recipients who were employed at the time of the survey or who had worked at some point during the 12 month period prior to the survey. Among these former recipients, only 35 percent reported that their employer provided paid sick leave and only 44 percent received paid vacation days when they were working.
Health Insurance
- Among former recipients in Washington State who were working at the time of the survey or had worked over the prior 12 months, 37 percent worked for employers that "provided" health benefits. The wording of the question makes it unclear whether this represents the proportion who received health benefits through their employer or the proportion offered such coverage.
Wisconsin (Statewide)
Post-Exit Earnings and Benefit Receipt Among Those Who Left AFDC in Wisconsin, Maria Cancian, et al., Institute for Research on Poverty, University of Wisconsin Madison, October 1998.
This study followed Wisconsin AFDC recipients who left welfare between August 1995 and July 1996. Using data from the Unemployment Insurance system as well as administrative records from public assistance programs, the study followed former recipients for five quarters after they left the cash assistance program.
Percent of Former Recipients Who Were Working
Table 10 shows the proportion of recipients who left welfare between August 1995 and July 1996 who worked in each of the five quarters following their exit from the program. The Table presents data on two different groups of recipients who left welfare. The column labeled "All Leavers" shows the proportion of all recipients who left welfare who worked in each of the five quarters following their exit regardless of whether the recipient subsequently received cash assistance. The column labeled "Continuous Leavers," by contrast, provides information on the employment rates of those former recipients who did not subsequently receive aid during the five quarters following their welfare exit. Some 70 percent of recipients who left welfare between August 1995 and July 1996 did not subsequently receive assistance in the following five quarters.(33)
Table 10: Employment Rates of Former Wisconsin Recipients Quarter Following Exit All Leavers Continuous Leavers First Quarter 72.4% 74.2% Second Quarter 72.5% 77.2% Third Quarter 73.3% 79.0% Fourth Quarter 74.3% 79.6% Fifth Quarter 75.8% 81.1% Earnings of Former Recipients
Table 11 shows median quarterly earnings of employed former Wisconsin recipients. As in the previous table, this table shows median quarterly earnings figures for "All Leavers" and "Continuous Leavers." In both cases, however, only those former recipients with earnings were considered when median earnings were calculated.
Table 11: Median Earnings of Employed Former Wisconsin Recipients Quarter Following Exit All Leavers Continuous Leavers First Quarter $2,383 $2,583 Second Quarter $2,437 $2,682 Third Quarter $2,460 $2,715 Fourth Quarter $2,602 $2,845 Fifth Quarter $2,632 $2,861 How to read this table: The median earnings for all recipients who left welfare between August 1995 and July 1996 and who were employed in the first quarter following their exit from welfare was $2,383. That is, half of the former recipients earned more than this level and half earned less than this level. The median earnings figure for those former recipients who did not subsequently receive assistance during the five quarters following their exit from welfare were higher. In the first quarter following their exit from welfare, the median earnings of these "continuous leavers" was $2,583. The authors of this study conducted statistical tests to determine whether certain characteristics of former recipients were associated with higher earnings. The authors found that, "the following factors seemed to be most closely associated with higher earnings:
- Human capital (having more education and having prior work experience) was positively and significantly associated with higher earnings;
- Having more children was associated with higher earnings, but having children who were very young children discouraged earnings;
- Having been sanctioned or having a family member on SSI appeared to reduce earnings;
- Legal immigrants had significantly higher earnings than did those native-born;
- Living in a county with a low unemployment rate was associated with substantially higher earnings."(34)
It is important to note that the statistical tests performed by the authors enabled them to test the independent effect of various factors on earnings. Thus, when the authors conclude that having been sanctioned is associated with lower earnings, this means that recipients who had been sanctioned had lower earnings than recipients who had not been sanctioned independent of other differences such as differences in education levels that might have existed between recipients who had been sanctioned and those who had not.
Subsequent Receipt of Cash Assistance
- Roughly one-third of those who left assistance between August 1995 and July 1996 received cash assistance at some point during the fifteen months (or five quarters) following their exit from welfare.
Industry in Which Former Recipients
Table 12 shows the industries in which former recipients worked in the first quarter following their exit from welfare. The industries are listed in order according to the median earnings of former recipients in those industries. The first industry listed temporary agencies is the industry with the lowest median earnings among former recipients in the first quarter following their welfare exit while the last industry listed the financial, insurance and real estate industry is the industry with the highest median earnings among former recipients.
The study finds that many former recipients change industries during the first five quarters following their exit from welfare. The study states, "No more than 40 percent of the leavers who started in an industrial classification with relatively low median earnings stayed in the same classification across all quarters of observation. For those who started in classifications with the very lowest median earnings (Temporary Agencies, Agriculture/Forestry/Mining, and Hotels/Lodging) many more moved up than down the classification hierarchy."(35)
Table 12: Industries in Which Former Wisconsin Recipients Worked in the First Quarter Following Their Exit From Welfare Industry
Proportion of Employed Former Recipients Working in Industry Median Earnings of Former Recipients in First Quarter Following Welfare Exit Agriculture, Forestry, Mining .6% $1,536 Restaurants 11.8% $1,630 Hotels, Lodging 3.1% $1,666 Temporary Agencies 9.6% $1,782 Retail Trade 13.0% $1,960 Other Services 1.6% $1,980 Personal Services 2.5% $2,198 Business Services 6.7% $2,220 Wholesale Trade 2.7% $2,550 Social Services, Public Administration, Education 14.1% $2,665 Non-durable Manufacturing 5.5% $2,809 Construction .5% $2,867 Transportation, Communications, Public Utilities 3.0% $2,877 Health Services 14.6% $2,947 Durable Manufacturing 6.9% $3,093 Financial, Insurance, Real Estate 3.5% $3,284 Note: Figures may not add to 100 percent due to rounding.
Milwaukee, Wisconsin
Employment and Earnings of Milwaukee County Single Parent AFDC Families: Establishing Benchmarks for Measuring Employment Outcomes Under "W-2," by John Pawasarat, University of Wisconsin-Milwaukee Employment and Training Institute
The Milwaukee study tracked families that received AFDC in December 1995 and were not receiving assistance in September 1996. Using data from the unemployment insurance system, these families' earnings were measured in the last quarter of 1996 (October-December) and the first quarter of 1997 (January-March). Because of the manner in which the sample of families was selected, the study measured the earnings of some families immediately following their exit from the welfare program while other families had left welfare eight months prior to the quarter in which their earnings were measured. Thus, the data in the report on the earnings of Milwaukee families that left welfare is best understood as information that reflects the earnings of adults who recently left welfare.
Percent of Former Recipients Who Were Working
- Among the families studied, about one-third 34.3 percent were not working in the fourth quarter of 1996 (October-December).
Earnings of Former Recipients
- Among those families in which the parent was working, 41.6 percent had quarterly earnings of less than $2,500.
- Stated another way, four in ten recipients who left welfare between December 1995 and September 1996 and had some earnings, earned less than $833 per month during the last three months of 1996. (In 1996, the current poverty line for a family of three was $12,516 per year, or $1,043 per month. For a family of four, the figures are $16,036 and $1,336, respectively.)
- One-third of the working families in the study earned between $2,500 (the equivalent of full-time, minimum wage earnings at the time) and $4,000 (roughly the poverty line for a family of four at the time) in the October-December 1996 quarter.
- Only 16 percent of the working families in the study earned more than $4,000 in the fourth quarter of 1996.
Industries in Which Former Recipients Work
A second Milwaukee study Employer Perspective: Jobs Held by the Milwaukee County AFDC Single Parent Population (January 1996 - March 1997) by John Pawasarat of the University of Wisconsin-Milwaukee Employment and Training Institute reviewed the jobs held between January 1996 and March 1997 by single parents who received AFDC in December 1995. Some of those who worked during the January 1996 to March 1997 period continued receiving cash assistance while others had left the welfare rolls. The 18,126 recipients who worked at some point during the period studied held a total of 42,120 jobs.
- Some 30 percent of all jobs held by these parents were in temporary employment agencies. This figure understates the extent to which these parents were employed at some point in such an agency. More than four in ten 42 percent of all recipients who worked during the 15 month period studied worked at some point in a temporary agency.
- Nearly one-quarter of all jobs held by parents who worked during the period studied 23 percent were in retail trade firms (restaurants, department stores, etc.) while 13 percent were in service firms such as hotels, auto repair stores, or business services (such as printing companies or building maintenance firms).
- The study found substantial job turnover. Of those jobs held in 1996 that were not in temporary agencies, 41 percent paid less than $1,000 in total wages. That is, 41 percent of the jobs were held by the parent for such a short period of time that the parents earned less than $1,000 during their tenure in that job.
End Notes:
1. Many of these studies follow recipients in a particular region of the state. The references section of this report provides the citations for all of the studies used in this report as well as a description of the geographic regions studied.
2. Throughout this report, earnings are compared to the poverty line for families of 3 and 4. It is important to note that while a family's earnings might be below the poverty line, the family might also receive the earned income tax credit and food stamps. The combination of earnings, the EITC, and the food stamps can lift a family with earnings below the poverty line out of poverty.
3. In these studies, families were tracked regardless of the reason they no longer were receiving assistance. Families may leave assistance because their earnings make them ineligible for aid, but they also may leave for reasons such as being subject to a full-family sanction, failing to follow procedural requirements that lead to case closure, having unearned income (such as social security income) that makes the family ineligible for aid, or no longer having a child young enough to qualify the family for assistance.
4. Because current TANF-related programs states are running are more similar to the waiver programs states implemented than to the former standard AFDC program, the data presented in this report represent the earnings of program-group members, not control group members. In general, waiver programs which often included more rigorous work requirements and greater financial incentives to work increased the proportion of parents working. However, working program-group members typically had earnings that were similar to the earnings of working control group members.
5. In contrast, some recent studies of former recipients conducted in Kentucky and New Mexico obtained responses from less than 20 percent of the former recipients they tried to contact, making it difficult to gain reliable information from these studies.
6. See "Can the Labor Market Absorb Three Million Welfare Recipients?" by Gary Burtless, Brookings Institution Working Paper, June 1998.
7. Throughout this report, data from several studies are presented to illustrate consistent findings. While the data presented point to a consistent finding, the findings from each study are not necessarily comparable. That is, because of different methodologies used to determine which group of families would be studied and the time period over which earnings would be measured, readers should not use figures here to draw conclusions about the differences in the employment rates and earnings of recipients or former recipients between states. For example, the Maryland study found that about half of former recipients worked in the quarter following their exit while the South Carolina study found that nearly 70 percent of former recipients worked. However, the South Carolina study followed a group of families that had left welfare and had not returned to the program for about nine months while the Maryland study followed all families that exited, regardless of whether they subsequently returned to the program. Such differences which are not always evident from the text of this report - make comparisons between findings presented here unwise. It is important to note that differences in findings between the study of former recipients in the entire state of Wisconsin and the city of Milwaukee also should not be interpreted to mean that former recipients in Milwaukee systematically differ from those in the rest of the state. While both studies used administrative data and data from the Unemployment Insurance system, methodological differences make such comparisons ill-advised.
8. The estimates of the proportion of former recipients who work are highest in the statewide study of former Wisconsin recipients. This study, however, uses a somewhat different methodology than many of the other studies when calculating employment rates of former recipients which is likely to make the Wisconsin estimates appear higher than those in other states. The difference stems from the way in which researchers choose to treat former recipients for whom no employment or public assistance-related information is available.
In some quarters, a former Wisconsin recipient may not appear in any of the databases used in the study that is, the recipient may not appear in the Unemployment Insurance database (suggesting that the individual is unlikely to have worked in Wisconsin in that quarter) and also may not appear in any public assistance database. In some cases, the individuals not found in any records may be working in another state, working in a job not covered by the Unemployment Insurance system, or may not be working at all. In this study, former recipients for whom no information was available in a particular quarter were excluded from the sample when the employment rates of former recipients were calculated. If these former recipients are less likely to be working than former recipients for whom data are available, this methodology is likely to overstate employment rates. Other studies including the Milwaukee, South Carolina, and Maryland studies do not adopt this approach. These studies include all former recipients in the sample when determining the employment rates of former recipients. This is likely to understate employment rates among former recipients. Both approaches are reasonable, but the approach used can affect the measured employment rates of former recipients.
9. The findings presented here from the Milwaukee study of former recipients and the statewide Wisconsin study of former recipients should not be compared. While the two studies use similar methodologies, the way in which families were selected for the research sample, the time period over which earnings were measured, and the way in which the researchers treated certain families for whom no information could be found differ. These differences make the findings difficult to compare.
10. See the sixth edition of the quarterly performance measures for the New York State Social Services System.
11. Indiana's treatment of earnings mirrors that of the former AFDC program. In Indiana, a family of three in which a parent has worked fewer than 5 months In Indiana, is eligible for assistance until earnings exceed $550. A family of three in which the parent has worked more than four but fewer than 12 months are eligible for assistance until the family's earnings equal $410 while a family in which the parent has worked for more than 12 months becomes ineligible for assistance when the family's earnings exceed $380. See One Year after Federal Welfare Reform: A Description of State Temporary Assistance for Needy Families (TANF) Decisions as of October 1997, Gallagher et al., Urban Institute, June 1998.
12. A recent Center study found that after Oregon's state minimum wage rose in two stages to $6 an hour in January 1998, the average earnings of newly employed welfare recipients climbed 76 cents an hour, a nine percent increase after adjusting for inflation. See, New Findings from Oregon Suggest Minimum Wage Increases Can Boost Wages for Welfare Recipients Moving to Work, by Ed Lazere, Center on Budget and Policy Priorities, May 1998.
13. The weekly earnings figures for South Carolina and Washington State were calculated by multiplying the average hourly wages of employed former recipients by the average number of hours worked by these recipients. While not a precise way to measure average weekly earnings, it is unlikely that the actual figures differ substantially from these estimates.
14. Moreover, the surveys often ask - implicitly or explicitly - for the number of hours the parent typically works. While many parents in any particular week work fewer-than-typical hours, most survey respondents are likely to answer according to the typical number of hours they work.
15. Because families that were already receiving assistance participated in the waiver programs and were included in the evaluation, measuring earnings four quarters after assignment to the program is not the equivalent to measuring earnings four quarters after a family first received assistance. Some families participating in the program may have been receiving aid for a long period of time while others may have been new applicants.
16. Evaluating Two Welfare-to-Work Program Approaches: Two-Year Findings on the Labor Force Attachment and Human Capital Development Programs in Three Sites, by Gayle Hamilton, et al., Manpower Demonstration Research Corporation, December 1997.
17. The pattern of earnings among recipients who participated in human capital development programs was similar. The average earnings of those recipients with a high school diploma who found jobs exceeded the average earnings of those without a diploma by 18 percent in Atlanta and 40 percent in Grand Rapids. In Riverside, California, the human capital development approach was only available to those without a high school diploma.
18. The Employer Perspective: Jobs Held by the Milwaukee County AFDC Single Parent Population (January 1996 - March 1997), by John Pawasarat, University of Wisconsin - Milwaukee Employment and Training Institute. This report is available on the web: www.uwm.edu/Dept/ETI/employer.htm. This paper was written by the same researcher who studied the earnings of former Milwaukee recipients.
19. It is likely that receipt of fringe benefits such as paid vacation and sick leave is lowest when a parent first begins a job. Some jobs that offer such benefits to some employees do not offer them to new employees. Some of the parents surveyed in Florida and in Washington State are likely to be relatively new employees in their current jobs. A recent study by Harvard researchers measured the receipt of paid sick and vacation leave by women who worked more than half-time, were between the ages 28 to 35, and had received welfare at some point between 1978 and 1992. Among those women who had received welfare for two years or less during that time period, 47 percent did not receive paid sick leave and 28 percent received no paid vacation. While still showing a high rate of non-receipt of these fringe benefits, the figures do show higher rates of receipt than those reported in the Florida and Washington State studies. One of the reasons for this difference is likely to be that, because many of the women in the Harvard study had not received assistance for a significant period of time, that group had a higher average number of years of work experience and had remained with their current employer for a longer average period of time. See, "The Work-Family Balance: What Hurdles Are Parents Leaving Welfare Likely to Confront," by S. Jody Heymann and Alison Earle in the Journal of Policy Analysis and Management, Vol. 17, No. 2, 1998.
20. See Employee Rights Under the Family and Medical Leave Act (FMLA) by Catherine K. Ruckelshaus, National Employment Law Project.
21. See Paying More and Losing Ground: the Employer Cost-Shifting Is Eroding Health Coverage of Working Families, commissioned by the AFL-CIO (Lewin Group, Inc. 1998).
22. Federal law requires states to provide Medicaid to children under the age of six with family income below 133 percent of the poverty line, as well as to older children born after September 30, 1983, with family income below 100 percent of poverty. The requirement to phase in coverage of older children assures that by the year 2002 all children under the age of 19 will be eligible for Medicaid if they have family income below the poverty line. At present, the requirement means that states must cover children between the ages of six and about 14 with family income below the poverty level. These are federal minimum requirements; a majority of states have expanded coverage above these eligibility standards, and more can be expected to do so as a result of the child health block grant included in the Balanced Budget Act of 1997.
23. Parents not receiving cash assistance can be eligible for Medicaid. First, parents in families that meet the income and resource standards of the state's AFDC program in place in 1996 are eligible for Medicaid regardless of whether the family receives any form of cash assistance. Moreover, parents may also be eligible for Medicaid for a transitional period for up to 12 months if their incomes rise above this level because their earnings increase or because the child support paid by a non-custodial parent increases. The Personal Responsibility and Work Opportunity Reconciliation Act also gave states the option of extending Medicaid coverage to a broader group of low-income parents. See Taking the Next Step: States Can Now Expand Health Coverage to Low-income Working Parents Through Medicaid, by Jocelyn Guyer and Cindy Mann, Center on Budget and Policy Priorities, August 19, 1998.
24. This figure assumes the family has been working more than four months. See pages 8-9 of One Year after Federal Welfare Reform: A Description of State Temporary Assistance for Needy Families (TANF) Decisions as of October 1997, by L. Jerome Gallagher et al., Urban Institute, June 1998.
25. Washington State has recently implemented a program to help low-income working parents improve their skills and find better jobs. The program offers low-income parents working at least 20 hours per week the opportunity to participate in training programs including those based at community colleges. The program also links parents to child care assistance for hours parents are working and hours parents are participating in training.
26. The average hours per week worked was one area in which the difference between the control and program groups was substantial. While 47.1 percent of program group members worked at least 40 hours per week in their current or most recent job, only 37.9 percent of control group members worked this many hours. However, the difference narrows when comparing the proportion of respondents who worked at least 30 hours per week 74.2 percent of program group members and 70.2 percent of control group members worked at least 30 hours per week in their current or most recent job.
27. The remaining 22 percent of working former recipients were employed in a range of "other" industries. For example, less than seven percent were employed in manufacturing firms and less than five percent worked in the transportation industry. See Life After Welfare: Second Interim Report.
28. This report is available on the web: http://povertycenter.cwru.edu/br9803work_.PDF.
29. South Carolina has conducted three such studies of former recipients. The other two studies of families that left the rolls between October and December 1996 and those that left between January and March 1997 also surveyed families eight to 12 months after they had ceased receiving assistance. Only those families that did not receive aid during that time period were included in the survey. The earlier two studies reported somewhat lower employment rates among former recipients at the time of the survey. Among those who left the rolls between October and December 1996, some 59 percent were employed at the time of the follow-up survey. Among those who left the program between January and March 1997, some 64.6 percent were employed at the time of the follow-up survey.
30. The South Carolina study finds higher rates of employment among former recipients than is found in other studies. This is likely due in part to the study design in which families that left the program but received assistance again before the time of the follow-up survey were excluded from the sample.
31. See Survey of Former Family Independence Program Clients: Cases Closed During April Through June, 1997.
32. These represent former recipients characterizations of why they no longer receive assistance. The state welfare agency may code the reasons for case closures differently.
33. For a discussion of one of the reasons the measured employment rates in the statewide Wisconsin study may by higher than that in other studies of former recipients, see footnote 7.
34. Post-Exit Earnings and Benefit Receipt Among Those Who Left AFDC in Wisconsin, pg. 22.
35. Post-Exit Earnings and Benefit Receipt Among Those Who Left AFDC in Wisconsin, pgs. 24-25.
Related analyses: