On Election Day, November 8, 2016, while Americans were focused on who would be moving into the White House, the U.S. Department of Labor (DOL) slyly released a major experimental impact evaluation that found the federal government’s primary job-training programs to be ineffective. The peculiar timing and months-long delay of releasing a report that was finalized in May 2016 occurred despite the DOL’s official policy of releasing reports within two months of a report’s completion.
The “Workforce Investment Act (WIA) Gold Standard Evaluation” study assesses only two of the WIA’s three major funding streams. The evaluation assessed the WIA Adult and Dislocated Worker funding streams, while ignoring that of the WIA Youth. Perhaps the DOL could not risk another scientifically rigorous evaluation finding youth job-training programs to be ineffective.
The WIA Adult program is available for individuals 18 and older, while the Dislocated Worker program provides services primarily to workers displaced, terminated, or laid off. In fiscal year (FY) 2015, Congress appropriated $775 million and $1.24 billion for the Adult and Dislocated Worker programs, respectively, and over $900 million was allocated to the Youth program.
The study found that federal job training was largely ineffective at raising the earnings of participants. While the DOL is supposed to offer training in high-demand occupations, the vocational services provided by the department, according to WIA participants, are unlikely to lead to the intended occupations. This mismatch between training and occupations actually obtained is a clear indication of how the federal job-training system is out of touch with the needs of employers.
While advocates of federal job-training programs are quick to highlight any beneficial impacts, policymakers need to acknowledge that these programs also cause harm.
- WIA participants that received the most extensive services were less likely to obtain health insurance and pension or retirement benefits.
- WIA participants in vocational training were slightly more likely to leave their training prior to completion compared to similar individuals that received the least amount of WIA services.
- The households of dislocated workers that received the full job-training services earned several thousand dollars less than their counterparts that were offered skills assessments, job-search assistance, and career counseling.
- A majority of WIA participants that were offered job training believed their training was unrelated to eventually finding employment.
Such dismal results could explain why the publication was released on the very evening when Americans were eagerly anticipating the presidential election results. Simply put, the evaluation results continue a decades-long trend of ineffective employment and training programs funded by the DOL.
The WIA Gold Standard Evaluation almost did not take place. When Congress passed the WIA in 1998, it required the DOL to complete an evaluation of the program’s effectiveness by 2005. After the WIA was signed into law by President Clinton, the Administration left the responsibility for starting the evaluation to his successor. The Bush Administration ignored the 2005 deadline and did not award the contract for conducting the randomized controlled trial until President Bush was preparing to leave office.
The new report presents the short-term effectiveness (15 months after random assignment) of WIA Adult and Dislocated Worker programs. A follow-up report assessing the 30-month findings is expected to be released in 2017.
History of Failure
The bar is already set low for what federal taxpayers can expect from the DOL’s employment and training programs. The last large-scale experimental evaluation of the Department’s major job-training programs is decades old. Conducted in 16 sites across the nation during the late 1980s and early 1990s, the Job Training Partnership Act (JTPA) evaluation tracked program effects for more than 20,000 adult men, adult women, and out-of-school youths over the course of 30 months. The JTPA was Congress’s attempt to reform the ineffective programs administered under the Comprehensive Employment and Training Act. According to testimony at congressional hearings, the JTPA was intended to help alleviate the shortage of skilled workers that were needed by industry. This “skills gap” has been used for decades to justify the need for federal employment and training programs.
Overall, the performance of JTPA programs is widely considered to be a failure. While adult females had several positive outcomes, the results were generally not large enough to be considered very meaningful in any policy sense. Further, JTPA programs were largely ineffective in raising the incomes of adult males or male and female youths.
Congress responded to the ineffectiveness of the JTPA by “reforming” the nation’s job-training system through the passing of the Workforce Investment Act of 1998. During the signing ceremony, President Clinton asserted, “This is the crowning jewel of a lifetime learning agenda, the Workforce Investment Act to give all our workers opportunities for growth and advancement.”
In contrast to the JTPA, the WIA emphasized a “one-stop” approach in which most information on the availability of federally funded employment and training programs in an area are located in a single location. The one-stop centers serve as central locations where those seeking job assistance can find referrals for various services, including those not administered by the DOL. Approximately 1,850 one-stop centers operate across the nation. While WIA programs are administered differently, these supposedly new programs offered virtually the same services provided by the JTPA. Despite the asserted need to provide training for “high-skilled” jobs, employers primarily use one-stop centers to fill low-skill positions.As such, employers view the labor available from these centers as being predominantly low-skilled.
In 2014, Congress passed the Workforce Investment and Opportunity Act (WIOA) that, again, “reformed” the nation’s ineffective job-training system. When Congress passed the WIOA, President Barack Obama proclaimed,
Today’s vote helps ensure that our workers can earn the skills employers are looking for right now and that American businesses have the talent pool it takes to compete and win in our global economy.
Commenting on the one-year anniversary of the passage of the WIOA, then-chairman of the House of Representatives Committee on Education and the Workforce John Kline (R–MN) asserted, “This law will help our friends and neighbors find good-paying jobs by improving existing federal workforce development programs and fostering the modern workforce.”
For decades, Members of Congress and Presidents alike have claimed to have reformed the previously flawed federal job-training system with new legislation. Each successive reform of the federal job-training system has been preceded by acknowledgements that the previous reforms failed. These numerous reforms have never lived up to the political promises offered by federal job-training advocates.
Why the WIA Gold Standard Evaluation Findings Are Relevant
While WIA programs were reauthorized with the passage of the WOIA in 2014, the results of the WIA Gold Standard Evaluation are very relevant to policymakers today for three important reasons.
- The evaluation assessed the performance of WIA Adult and Dislocated Worker programs that are very similar to the programs administered today.
- The evaluation used random assignment to allocate individuals to three types of services.
- The evaluation has a high degree of external validity because the job-training sites were randomly selected.
First, the evaluation assessed the performance of WIA Adult and Dislocated Worker programs that are very similar to the programs administered today. “While WIOA made important changes to the workforce system,” according to the authors, “it did not significantly change the basic set of services that the local areas offered, nor who was eligible to receive them.” Further, “many of the important changes explicitly introduced by WIOA reflected changes in local areas were already making under WIA.”
Second, the WIA impact evaluation used random assignment to allocate individuals to three types of services:
- Core services offered mostly information and online tools for participants to plot their careers and find employment;
- Core-and-intensive services involved greater staff involvement in career assistance than core services, such as skills assessments, workshops, job-search assistance, career counseling, placement in work experience positions and short-term prevocational training; and
- Full-WIA services, provided after the core and intensive services, mainly provided Individual Training Accounts (ITAs) to purchase training in high-demand fields from approved providers.
The experimental design of the impact evaluation is the only way to determine with a high degree of certainty the effectiveness of social programs. The participants effectively assigned to the core services group is the control group, while the intervention groups are the core-and-intensive and full-WIA services groups.
To assess the effect of the WIA, the evaluation compared the impact of providing services based on three comparisons.
- The first comparison contrasted the full-WIA with core services to estimate the effect of offering full-WIA services, other than job-search information. For this evaluation, the core services group serves as close as possible to a true control group because the services provided, such as job opening lists, are the types of activities that any job searcher can perform without the assistance of the federal government.
- The second comparison contrasted the full-WIA services with core-and-intensive services. This comparison estimates the marginal effect of offering training in addition to the core-and-intensive services.
- The third comparison of core-and-intensive to core services estimated the marginal impact of offering intensive services.
Over 35,000 individuals were involved in the study, with 31,304 participants in the full-WIA group, 2,181 participants in the core-and-intensive group, and 2,180 in the core group. For the impact analysis, the final sample consisted of 1,176 full-WIA, 1,684 core-and-intensive, and 1,669 core group members.
Starting in November 2011 with a national unemployment rate of 8.6 percent and ending in the spring of 2013 with a rate of 7.6 percent, random assignment took place during a time of declining unemployment rates. Thus, WIA participants were seeking employment services during a period of improving employment conditions.
The study participants consisted of a diverse population:
- Sixty-one percent were women;
- Sixty-two percent were racial or ethnic minorities; and
- Seventy percent had a high school degree or General Educational Development (GED) credential.
Third, the WIA Gold Standard Evaluation has a high degree of external validity because the 28 job-training sites were randomly selected. As a large-scale experimental evaluation based on multiple sites, this evaluation avoids problems of simplistic generalizations. Federal social programs should be evaluated in multiple sites to test them in the various conditions in which they operate and in the numerous types of populations that they serve. Thus, the WIA impact evaluation results are nationally representative of the effectiveness of the WIA Adult and Dislocated Worker programs.
While the 15-month outcomes assessed in the report are short-term, the findings are highly suggestive of the long-term performance of these programs. Ultimate judgments on the long-term effectiveness of the Adult and Dislocated Worker programs should be reserved until the long-term results are released in 2017. However, if the current findings and history are any guide, policymakers should not place much hope in the forthcoming long-term study finding success.
Because of the large number of outcomes being assessed, the authors of the evaluation guarded against finding impacts to be statistically significant due to chance by using two tactics.
- The most important outcome of interest is earnings in the fifth quarter because it is “the latest measure of earnings available in the 15-month follow-up period and thus the best indication of future earnings.” The fifth quarter outcomes are also important because the potential effect of job training is not likely to manifest during the earlier quarters.
- The authors adjusted the threshold for defining statistical significance to account for multiple significance testing. This adjustment corrects for the increased probability of falsely finding statistically significant results when assessing many outcomes. Thus, the adjustment of the statistical significance threshold guards against the increased likelihood of incorrectly finding statistically significant effects.
Impact on Training Completion and Credential Receipt
Unsurprisingly, members of the full-WIA group were more likely to complete training programs and receive occupational credentials than were members of both the core-and-intensive and core groups. For example, 30 percent of the full-WIA group completed training participation, compared to 15 percent for the core group—a statistically significant difference of 15 percent. In addition, the full-WIA group was more likely to receive training funding from state employment agencies, compared to the core group.
WIA-full and core group members had credential receipt rates of 23 percent and 11 percent, respectively—a statistically significant difference of 12 percent. Additionally, the full-WIA group was more likely to complete training programs and receive credentials than the core-and-intensive group. However, the core-and-intensive group was no more or less likely than the core services group to pass these milestones.
The full-WIA and core-and-intensive groups were significantly more likely to complete specific vocational training and receive vocational certificates than the core group. Further, the full-WIA group had higher rates of completing vocational training and obtaining vocational certificates than the core-and-intensive services group. However, the full-WIA group was no more or less likely to finish educational programs, compared to the core-and-intensive and core groups. The same held true when this outcome was the comparison of the core-and-intensive services group to the core services group. Members of the full-WIA group were more likely to leave vocational training prior to completion (4.6 percent) compared to the core group (3.1 percent)—a statistically significant difference of 1.5 percent.
Impact on Earnings and Employment
Did the statistically significant increases in completing training programs and receiving occupational credentials lead to increased earnings and employment for the full-WIA and core-and-intensive services groups? Overall, the short-term results strongly suggest that WIA employment and training services are ineffective. Tables 1 and 2 summarize the major findings for the fifth quarter impacts.
Full-WIA Compared to Core Services. The most important test of the WIA’s effectiveness is the comparison of the full-WIA to core services on earnings and employment. During the five quarters of the follow-up period, members of the full-WIA group failed to have statistically different earnings than the earnings of the core group members. In the fifth quarter, the full-WIA group earned, on average, about $250 more than the core services group—a statistically insignificant difference. Despite being more likely to enroll in training, receive one-on-one assistance, and other employment services, members who participated in the full-WIA saw no effect on earnings.
When the results of fifth quarter earnings are separated by those participating in the WIA Adult and Dislocated Worker programs, WIA participation still had no effect on earnings. While full-WIA Adult and Dislocated Worker participants earned, on average, $423 and $39 more than similar members of the core services group, respectively, these differences are not statistically significant.
For employment, members of the full-WIA group were more likely to acquire employment during the fifth quarter. During this quarter, the full-WIA group had an employment rate of 68 percent, compared to the 62 percent rate of the core services group—a statistically significant difference of 6 percent. Despite being more likely to obtain employment in the fifth quarter, the full-WIA services were not associated with increased earnings.
The fifth quarter employment rates of those participating in WIA Adult and Dislocated Worker programs were not statistically different than the rates for similar members of the core services group. Full-WIA Adult participants had an employment rate of 68 percent, while the rate for Dislocated Worker participants was 71 percent. In comparison, core services participants in the Adult and Dislocated Worker programs both had an employment rate of 63 percent.
For quarterly hours worked, participation in full-WIA yielded no differences compared to core services during all five quarters. In the fifth quarter, the full-WIA group worked, on average, 14 additional hours compared to the core group—a statistically insignificant difference. While the full-WIA group was more likely to be engaged in productive activities than the core group in the first quarter, participation had no effect on being involved in productive activities, such as being employed or in training, during the second through fifth quarters. This outcome is an odd finding because full-WIA participants had the advantage of having access to training services through ITAs.
As for characteristics of their most recent job, differences between the full-WIA and core services groups were largely nonexistent. Members of the full-WIA group who found employment were no more or less likely to obtain health insurance, pension or retirement benefits, paid vacation and holidays, paid sick days, and tuition assistance, compared to similar members of the core services group. Interestingly, the average hourly wage rates of the full-WIA and core group were $12.63 and $12.92, respectively—a statistically insignificant difference suggesting that employers placed little value on the human capital gained from full-WIA participation. If WIA services improved the productivity of participants, then one would expect these individuals to garner higher hourly wages.
How do WIA participants believe that the services provided to them resulted in finding jobs? Only 43 percent of full-WIA participants believed that their training resulted in them obtaining employment, despite ITAs being intended to be used for training in high-demand occupations. This means that a solid majority of 57 percent believe that the provision of full-WIA services was unrelated to finding employment. In addition, the core-and-intensive and core services groups were no more or less likely to believe that the assistance provided to them led to employment. Despite being required to provide training in occupations highly demanded by local businesses, the DOL has failed miserably at this task.
Participants in the WIA were largely unable to find employment in occupations related to their training. Only 32 percent of full-WIA participants found occupations in the area of their training. Thus, 68 percent were unable to find employment in their intended occupations. Full-WIA participants were no more or less likely to find employment in their planned occupation than the other groups.
Full-WIA Compared to Core-and-Intensive Services. According to the evaluation authors, “The availability of WIA-funded training did not lead to an increase in earnings in any of the first five quarters after random assignment.” In particular, the full-WIA group earned $3,767 on average compared to $4,089 for the core-and-intensive group—a statistically insignificant difference of –$322. Thus, the marginal impact of adding training to intensive services is negligible.
For employment, the findings were similar to the earning outcomes. Members of the full-WIA group were no more or less likely to find employment over the 15-month follow-up period. During the fifth quarter, the full-WIA group had an employment rate of 68 percent—a not substantially different rate from that of the core-and-intensive services group.
For quarterly hours worked, participation in the full-WIA yielded no differences compared to the core-and-intensive services participants. In the fifth quarter, the full-WIA group worked, on average, 14 fewer hours than the core-and-intensive group. However, this difference was statistically indistinguishable from zero. Similarly, the full-WIA group was just as likely to be engaged in productive activities as the core-and-intensive group.
Differences in characteristics of the most recent job between the full-WIA and core-and-intensive groups were largely nonexistent, except for a few notable exceptions. Members of the full-WIA group who found employment were less likely to obtain health insurance and pension or retirement benefits, compared to similar members of the core-and-intensive group. This finding strongly suggests that the marginal effect of adding training to intensive services is harmful. Also of interest are the average hourly wage rates of the full-WIA and core-and-intensive groups were $12.63 and $13.12, respectively—a statistically insignificant difference suggesting that the human capital benefits derived from full-WIA participation was no different than the lesser service.
Core-and-Intensive Compared to Core Services. During the first four quarters of the follow-up period, members of the core-and-intensive services group failed to have statistically different earnings than the earnings of the core group members. However, the core-and-intensive services group earned $600 (17 percent) more than the core services group during the fifth quarter—a statistically significant impact.
For employment, the findings were similar to the earning outcomes. Members of the core-and-intensive group were no more or less likely to find employment during the first four quarters. During the fifth quarter, the core-and-intensive group had an employment rate of 70 percent, which was substantially different from the 62 percent rate for the core services group—a statistically significant difference of 8 percent.
For quarterly hours worked, participation in core-and-intensive services yielded no differences compared to core services for the first four quarters. In the fifth quarter, the core-and-intensive group worked, on average, 29 additional hours than the core group. This difference was statistically significant. In contrast, the core-and-intensive group was no more or less likely to be engaged in productive activities than the core group.
As for characteristics of their most recent job, members of the core-and-intensive group had more consistent benefits than those of core services. For their most recent job, members of the core-and-intensive group were more likely to obtain jobs with health insurance, paid holidays, paid sick days, pension or retirement benefits, and tuition assistance or reimbursement, compared to similar members of the core group. Also of interest is the average hourly wage rates of the core-and-intensive and core groups were $13.12 and $12.92—a statistically insignificant difference of $0.20 suggesting, again, that the human capital benefits derived from the WIA program has no measurable benefit.
Impact on Self-Sufficiency
The availability of WIA-funded services has the potential to affect household income and participation in public assistance programs, such as Temporary Assistance for Needy Families (TANF) and the Supplemental Nutrition Assistance Program (SNAP). For example, if a household head earns new job skills through WIA participation and these skills increase earnings, then the individual may be less likely to receive welfare benefits. Alternatively, if a household head pursues job training, rather than finding employment, this delay in obtaining employment, may cause the individual to spend more time collecting welfare benefits. Participation in the Adult and Dislocated Worker programs would hopefully lead to households being less dependent on welfare programs.
Overall, the WIA services, regardless of the level of services provided, had no effect on self-sufficiency. According to the evaluation, the full-WIA group did not have different participation rates during the 15-month follow-up in cash assistance programs, including TANF and Social Security, SNAP, and the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC), when compared to the other types of WIA services. For example, 47 percent of households in the full-WIA group received SNAP benefits, compared to 41 percent of households in the core services group—a statistically insignificant difference of 6 percent. The average amount of cash assistance and SNAP for each household did not statistically differ for each group. Thus, the offering of full-WIA services did not increase self-sufficiency, compared to the lesser service levels.
For total annual household income—defined as the participant’s earnings, earnings from other individuals in their household (such as a spouse), welfare transfers from government assistance programs, and Unemployment Insurance (UI) benefits—the offering of full-WIA services failed to have an effect when compared to the lesser service levels.
One crucial omission from the WIA Gold Standard Evaluation is the fact that the 15-month findings failed to assess the individual impact of services on the receipt of UI benefits. WIA services may delay or expedite UI participation, which has important ramifications for a cost-benefit analysis of federal job training. For example, an unemployed person steered into training through an ITA may pass up on immediate employment opportunities, thus prolonging the individual’s collection of UI benefits.
The average household income for the full-WIA group was $21,795, compared to $23,664 for the core-and-intensive services group. However, this negative impact of $1,869 is statistically indistinguishable from zero. Further, households in the full-WIA group had incomes that averaged $3,894 less than the average incomes of core services households. This negative difference was also statistically insignificant. The household income findings were similar when the comparison was between core-and-intensive and core services.
When total annual household income was analyzed by participation in the Adult or Dislocated Workers programs, there were two statistically harmful significant impacts. On average, households in the full-WIA Dislocated Worker group earned $6,154 and $4,793 less than their counterparts in the core and core-and-intensive Dislocated Worker program households, respectively. The Adult program households saw no difference in annual total household income.
The harmful effect of the Dislocated Worker program on earnings is similar to the damaging effects of the Trade Adjustment Assistance (TAA) program. TAA provides overly generous government benefits to American workers who lose their jobs because foreign companies prove more competitive than their American employers. TAA encourages recipients to participate in job training. Overall, the notion that TAA improves the employment and earning outcomes of displaced workers has little empirical support. In fact, TAA participants are more likely to earn less after participating in the program.
The Failed Promise of Reform
The WIA Gold Standard Evaluation confirms the Government Accountability Office’s previous assessment that federal job-training programs are ineffective. Taken as a whole, the performance of federal employment and training programs is certainly underwhelming. The occasional positive outcomes notwithstanding, these benefits tend to be concentrated among the receipt of training, GEDs, and trade-related certifications. The receipt of these credentials too frequently fails to translate into increased earnings and wages. However, when the occasional beneficial impacts on employment and earnings do arise, these impacts are generally not large enough to have any meaningful policy significance. Based on the lack of impact on hourly wages, full-WIA and core-and-intensive services do not boost the human capital and productivity of participants.
While the DOL is supposed to offer training in high-demand occupations, the vocational services provided by the department, according to WIA participants, are unlikely to lead to the intended occupations. This mismatch between training and occupations actually obtained is a clear indication of how the federal job-training system is out of touch with the needs of employers. Yet, during every reauthorization cycle, federal policymakers promise to fix this problem.
While advocates of federal job-training programs are quick to highlight any beneficial impacts, policymakers need to acknowledge that these programs also cause harm.
- Participants in the full-WIA group who found employment were less likely to obtain health insurance and pension or retirement benefits, compared to similar members of the core-and-intensive group.
- Participants offered training were less likely to obtain employment that offered crucial benefits.
- Participants in WIA vocational training were slightly more likely to leave their training prior to completion compared to similar individuals who received the least amount of WIA services.
- Households in the full-WIA Dislocated Worker group earned several thousand dollars less than their counterparts in the core and core-and-intensive member households.
- A majority of full-WIA participants believed their training was unrelated to eventually finding employment.
With so many null effects—meaning no impact on employment and earning outcomes—participation in the DOL’s job-training services is largely a waste of time for the majority of job seekers and an enormous squandering of taxpayer dollars. Despite decades of “reform,” the DOL is still failing its mission to provide effective job-training programs.
A Path Forward
Given the consistent evidence of the failure of DOL job-training programs and the perpetual false promise that the next set of reforms will finally produce results, policymakers need to recognize that these programs offer little real opportunity to job searchers and waste taxpayer dollars. A more effective policy to boost the incomes of the economically disadvantaged would be to embrace the voluntary cooperation of individuals in a free-market agenda that will help the economy grow and provide all Americans with better-paying jobs.
- The federal government should reduce taxes on individuals and corporations to reward work and entrepreneurship. By adopting a pro-growth tax code, the federal government will boost savings and investment. In return for lowering the tax burden, the economy will prosper, thus leading to more and better-paying jobs.
- Congress and the Trump Administration need to work together to eliminate regulations that stifle the economy. President Donald J. Trump’s executive order to reduce regulations and the revival of the Congressional Review Act—a law that allows Congress to nullify regulations not properly submitted for review to the legislative branch—has the potential to remove unnecessary barriers to economic growth.
- State and local governments need to reform occupational licensing policies that suppress competition and innovation, in addition to stifling the upward mobility of low-income and unemployed individuals.
A free-market agenda that encourages opportunity for all Americans, not ineffective programs administered by the DOL, is a better course of action to create more and better-paying jobs.
—David B. Muhlhausen, PhD, is a Research Fellow for Empirical Policy Analysis in the Thomas A. Roe Institute for Economic Policy Studies, of the Institute for Economic Freedom, at The Heritage Foundation.