Should the federal government spend more on workforce development?

workforce gathering
Editor's note:

This article is based on Professor Holzer’s testimony before the US House of Representatives Subcommittee on Higher Education and Workforce Development, Committee on Education and Labor on May 11, 2023.

Federal funding for workforce development—which includes job training and other workforce services to help workers find and keep better-paying jobs—is controversial in the U.S..

On the one hand, the U.S. has recently had the highest job vacancy rate ever recorded—roughly twice the unemployment rate, at its peak.[1] One reason that many jobs remain vacant, especially some which pay quite well, is that so many American workers do not have sufficiently high skills to fill them—while many of these workers no longer want to accept low-wage jobs which pay less.[2] For those who do not (and likely will not) have college degrees, supporting other forms of education and job training with federal funds makes sense.

On the other hand, many Americans—including some employers—feel that our federal job training programs are a waste of taxpayer dollars and are relatively ineffective. There seems to be broader political support for other forms of raising funding for higher education and training, e.g., expanding Pell grants (which help pay for college for low-income people) to cover certificate programs where enrollments are not now eligible for Pell funding.[3] But support for investing more money in the Workforce Innovation and Opportunity Act (WIOA) run by the U.S. Department of Labor—which distributes several billion dollars to states and local workforce boards for training and workforce services—is much weaker and generates more partisan division.

Below I make the following points about workforce development in the U.S.:

  • Many millions of Americans without postsecondary education or credentials would benefit from additional skills training, especially low-income workers in very low-wage jobs.
  • Relative to the great need for skill enhancement in the U.S., and to other industrialized countries now and earlier decades in the U.S., our levels of federal public funding for workforce training are extremely low.
  • The impacts of such training in raising participant earnings are dramatically limited by the small funding levels available.
  • Besides increasing funding for training, our top priorities should be to: a) expand highly effective sectoral training models to scale, as well as research and evaluation into what makes them effective; b) support state-led efforts to expand work-based learning models like apprenticeship; c) expand guidance and support services for students/trainees; and d) improve labor market data and those on training outcomes. Passing some version of “short-term Pell”—with appropriate safeguards against funding for weak programs—would help more workers gain funding as well.

1. Millions of U.S. Workers Would Benefit from Additional Skills Training

Economists continue to debate how to define and measure “skills shortages” or “skills gaps” in the U.S.[4] What cannot be denied is that overall educational attainment and worker skills in the U.S. need to be improved, especially among Americans with low earnings.

Nearly half of all Americans over the age of 25 still have no postsecondary education or training credential.[5] These workers, on average, earn low hourly wages and have low weekly and annual earnings. For instance, median hourly wages for high school graduates and dropouts working year-round and full time are approximately $18 and $15  per hour respectively, and those working less than year-round or full time have even lower hourly wages.[6]

But earnings are lowest for millions of workers now working in lowest-wage sectors of the U.S. service economy. Workers of color are heavily over-represented in this group. Median hourly wages of workers in personal care and services, building maintenance, food services or healthcare support are $14-17 per hour, while those at the 25th percentile of these occupations earn just $12-15 per hour (generating $24,000-30,000 in annual earnings for full-time workers).[7] Their earnings would no doubt rise with significantly more education or job training.

Nearly half of all Americans over the age of 25 still have no postsecondary education or training credential.

2. Federal Funding for Workforce Education and Training in the U.S. is Extremely Low

Federal WIOA dollars are distributed to state and local workforce boards around the country to fund a range of services for workers. WIOA Title I programs include a stream for adult workers (primarily low-income), displaced workers, and out-of-school youth. Title I also funds the Job Corps program for youth and a number of smaller programs. Title II funds Adult Basic Education in the U.S. Title III helps pay for the American Job Centers (formerly One-Stop centers) where workers can get either core services (staff-assisted job search), intensive services (testing or career counseling), or vouchers for training (known as Individual Training Accounts or ITAs) that are paid for by the adult and displaced streams of Title I. Title IV covers vocational rehabilitation expenditures for disabled students and is distributed by the U.S. Department of Education.[8]

The core programs in Titles I and II —the formula funding streams for adult, displaced worker, and youth services plus adult basic education—receive just $4 billion in federal funding per year. Total funding levels for Titles I and II are now about $6 billion. In an economy with Gross Domestic Product (GDP) of $25 trillion and over 150 million adult workers, these levels are extremely low.

While there are many other federal employment and training programs—43 in all, according to the General Accounting Office—their total funding levels remain at nearly $20 billion, or less than 0.1% of GDP.[9] If one were to include other federal funding for training—such as the share of Pell grants spent on certificate programs at community or for-profit colleges—that fraction would remain about the same;[10] and Pell funding is not now available for short-term and/or not-for-credit training programs that can be important sources of training for working adults. In contrast, most other industrialized countries spend up to 0.5% of GDP  on such “active labor market policies.” [11]

Over time, our federal spending has become much more stringent. Spending on the Comprehensive Employment and Training Act (CETA) at the U.S. Department of Labor peaked at over $17 billion in 1979, which, in current dollars, would be about $60 billion today.[12] Comparing such expenditures to all those tabulated by GAO, federal government spending for workforce services and training have dropped by two-thirds despite the fact that the labor force has grown by about 50% since that time.

3. The Effectiveness of Job Training is Limited by Low Funding Levels

Of the $4 billion in Titles I and II of WIOA, only about $0.5 billion is actually spent on training for about 220,000 people per year, or just over $2,200 per trainee from ITAs.[13] In contrast, Pell grants provide up to nearly $7,000 per year for 6 million students in higher education.

How effective is WIOA training? The rigorous evidence is somewhat mixed, but the most plausible estimates suggest training impacts ultimately reach $1,500-2,000 per year for adult workers but less for displaced workers.[14] Per dollar of federal spending, this is a strong return if it persists over time. Core and especially intensive services for workers at American Job Centers are inexpensive and appear quite cost-effective at raising wages.[15] Many workers who complete WIOA services and training continue to have relatively low annual earnings—often at $25,000  per year or less—though they are likely somewhat higher than they would be absent any workforce services and training.[16]

Certificate programs at community colleges, whether for-credit or not, also generate significant returns in terms of higher earnings on average, though their value varies considerably across fields.[17] Generally, credentials requiring larger spending investments generate larger earnings gains (longer-term certificates earn more than shorter-term, for-credit earns more than non-credit, etc.).

The very best training programs in the U.S. generate substantially higher average impacts on earnings. Sectoral training programs—based on partnerships between industry and workforce training providers to fill high-demand and well-paying jobs in key economic sectors (like health care, advanced manufacturing and information technology)—can generate large earnings gains for workers while helping employers find qualified applicants for hard-to-fill jobs. Adult programs like Per Scholas and Project Quest raise earnings by $5,000  or more per year without any sign of fadeout over time.; Year Up—a program for young high school graduates—generates even larger impacts.[18] These programs cost $10,000  or more per year. Generally, they also require trainees to enter training with higher basic skills and no other major barriers to employment.[19] Unfortunately, these programs remain relatively small  – though each is being replicated across a number of sites in the U.S..

4. Top Priorities for WIOA Reauthorization (or Other Legislative Vehicles)

Whatever the levels at which WIOA and other federal training programs are funded going forward, I would identify the following as top priorities for the program:[20]

  • We should expand funding for the most effective sector partnerships and help them reach scale, while also establishing more partnerships to follow their examples. Currently, local workforce boards are mandated to establish such partnerships but receive no funding or technical assistance to do so. Support for building effective partnerships would likely improve the quality of the jobs for which workers are trained with federal funding, as would dedicated funding for programs that train participants for good jobs (as in the Commerce Department’s Good Jobs Strategy). We also need research and rigorous evaluation into what makes some such training programs so effective. Local and state efforts to scale effective sectoral programs should be supported and evaluated.
  • Apprenticeship and other forms of work-based learning and incumbent worker training appear promising (though rigorous evidence is limited), and their use by local workforce boards and employers with WIOA funding should be expanded. States are innovating a lot in this regard; supports for and evaluations of their efforts should be a priority.[21]
  • More funds should be provided for career guidance and other support services for trainees, both at American Job Centers within WIOA but also for other training providers like community colleges.[22]
  • Data on WIOA training providers and outcomes—and on labor market skill needs more broadly—could and should be greatly improved. The Employer Training Provider Lists provide data on individual training providers and trainee outcomes compiled by the Labor Department. These lists are supposed to help individual trainees (sometimes with help from local staff at American Job Centers) choose where to spend their training vouchers. But these lists are plagued with many shortcomings and are almost impossible for trainees to navigate at this point. Further, the data provided are not very useful for judging the effectiveness of individual programs.[23] Labor market data improvements at the state and local levels have surpassed those of the Bureau of Labor Statistics, and to education and training program performance data are also being improved.[24]

The very best training programs in the U.S. generate substantially higher average impacts on earnings.

In addition to these priorities, passing some version of “short-term Pell” to expand Pell coverage to shorter certificate and some not-for-credit programs would be an additional way to finance more training. But it is important that safeguards be included to protect workers from ineffective programs when doing so.


A strong case can be made that the U.S. needs to inject more federal funds into job training around the country and that WIOA is currently underfunded. At the same time, we need to ensure that such funding is spent on cost-effective programs and services, and many analysts are not confident that WIOA is an effective mechanism for doing so right now.

I believe that additional funds can and should be spent on WIOA, but they should be targeted on programs and services with proven records of cost-effectiveness and where there is great need. The data that are needed to help potential trainees find the right training providers and which can be used to demonstrate whether or not individual programs are cost-effective (and therefore deserving of ongoing funding) should also be reformed as well.


[1]: In early 2022, the job vacancy rate hit 7.4% while the unemployment rate was 3.6%.

[2]: Job vacancy rates have been high in professional services and health care, where higher education or specific skills are needed, as well as in leisure and hospitality, where wages are relatively low.

[3]: See Baum et al. (2020).

[4]: Barnow et al., Occupational Labor Shortages, Upjohn Institute for Employment Research, 2013.

[5]: US Census Bureau: See also Holzer and Baum, 2017: The latter shows that, while postsecondary enrollments are quite high, low completion rates and low labor market value of many credentials earned limits the earnings of many students who enroll, especially at community colleges.

[6]: See the National Center for Education Statistics: To get hourly wages for high school graduates and dropouts, I divide median annual earnings by 2000 hours for year-round full-time workers.

[7]: Bureau of Labor Statistics, Occupational Employment and Wage Statistics:

[8]: Congressional Research Service, 2022:

[9]:Government Accountability Office, 2019: These data don’t include state subsidies for higher education, including certificate programs. See also Holtz-Eakin and Lee, An Analysis of Federal Training Programs, American Action Forum, 2019.

[10]: The federal government now spends $27B each year on Pell grants, but the vast majority goes to degree students at 2-year and especially 4-year colleges (Congressional Budget Office, 2022:

[11]: OECD:

[12]: Holzer, Harry J. 2013. “Workforce development as an antipoverty strategy: What do we know? What should we do?”

[13]: Deming et al., 2023:

[14]: Holzer et al., 2022:; Heinrich et al., 2013:; Fortson et al., 2017: Holzer et al. and Heinrich et al. use rigorous nonexperimental methods to estimate training impacts; Fortson et al. was based on a randomized controlled trial. But, unfortunately, the interpretation of Fortson’s results is made difficult by the fact that workers in the control group got almost as much training on their own as did the treatment group; it is an “intent-to-treat” estimate where the treatment group got little more treatment that the control group. The Holzer et al. and Heinrich et al. estimates suggest larger differences in training between treatment and control groups in the Adult Worker programs. The $1500-2000 estimated impacts are the maximum impacts observed  on an annual basis in the followup periods, with each study’s figures updated for inflation.

[15]: Heinrich et al. and Fortson et al., ibid.

[16]: Deming et al., op. cit.

[17]: Baum et al., 2020: See also Stevens et al., 2015:

[18]: Katz et al., 2020:; Roder and Elliott, 2021:; Fein 2021:

[19]: For instance, Per Scholas requires trainees to read and do math at least at the 10th grade level, in order to be able to complete their IT training.

[20]: Orrell et al., 2023:

[21]: Negoita and Goger 2020: See also Lerman 2021:

[22]: Heinrich et al. and Fortson et al., op. cit. See also Dawson et al. 2020:

[23]: Deming et al., op. cit.

[24]: Credential Engine is one example of an effort to present data on individual training providers and institutions in a more manageable way. See also Julia Lane, 2023:

The Brookings Institution is financed through the support of a diverse array of foundations, corporations, governments, individuals, as well as an endowment. A list of donors can be found in our annual reports published online here. The findings, interpretations, and conclusions in this report are solely those of its author(s) and are not influenced by any donation.