A Vision for the Future of the Workforce Investment System
This paper was originally prepared as a speech for the annual conference of the National Workforce Association in St. Petersburg, Florida, on December 4, 2006.
I would like to offer a vision for the future of the workforce investment system authorized by the Workforce Investment Act (WIA) and its predecessors. I start by going back in time, then address the state of the current workforce system, and then offer a vision for its gradual growth over the next decade. I say “growth” even in this time of deficits and budget cuts because a hard look at the realities of the labor market in the United States shows that a stronger and more robust workforce investment system is needed if we are to meet the many challenges facing both the labor force and employers. I hope that the ideas in this paper will be part of a broader discussion of the purpose and role of the workforce investment system over the coming years.
I address four themes in this paper:
- the target population, or “customers,” that the workforce investment system serves (note that this paper focuses only on adults);
- the relationship of the workforce system to employers;
- the relationship of the workforce system to the welfare system (and to work support programs); and
- the performance standards that drive the workforce system.
In each of these areas, I am suggesting a new approach that could reinvigorate the workforce investment system and ensure its relevance to the labor force, employers, and policymakers.
Observations About the Past
The past has both defined — and confined — our thinking about the future role of the workforce investment system. I’d like to share four observations about the past because they may tell us about a productive future.
First, from the 1960s, the workforce system’s mainstream programs for adults were designed to move unemployed, low-income people — including unemployed, dislocated workers — into work or reemployment through a range of preemployment services. By and large, job retention and job advancement services — including retraining programs for currently working people — were not part of the services menu for workforce development programs. Certainly, incumbent worker training programs were sometimes supported by federal and state discretionary grants, but they were not part of the mainstream adult workforce system; when they have existed, they have tended to target high-level training for somewhat advanced workers.
The workforce system’s target population and services were entirely appropriate — until perhaps ten years ago. As the following figure shows, unemployment from 1975 to 1994 averaged a little over 7 percent, with a peak of 9.7 percent in 1983.
The major job of the workforce system was, appropriately, to prepare unemployed people for work (or, during the toughest times, for public service employment, because so many were unemployed — nearly 10.7 million in 1982). In contrast, the average annual unemployment rate from 1995 through 2005 was about 5 percent, closing in on the lowest level feasible given the normal movement of people in and out of employment. In October 2006, the unemployment rate was only 4.4 percent. The era of significantly lower unemployment rates that began in 1995 is an important factor in thinking about the future of the workforce system.
A second observation about the history of the workforce system: Over time, the system has worked more closely with employers. Beginning with the Private Sector Initiative Program under a single title of the Comprehensive Employment and Training Act (CETA), and then with the far broader role given to private sector-dominated Private Industry Councils under the Job Training Partnership Act (JTPA) and Workforce Investment Boards under WIA, the business community became increasingly engaged in local policy-setting. But, aside from rapid response to big layoffs, the direct link with employers has not existed at the programmatic level for many mainstream workforce programs — and certainly not for all training programs. This may be due, in part, to the system’s gradual increase in its job placement (labor exchange) functions — or “universal services,” as it is now called in the workforce system — with proportionately less skills training.
My third observation about the past that I think is relevant to the future: The workforce system has, by and large, kept its distance from the other major public job preparation program in the country, namely the one run by welfare agencies for welfare recipients. (For sure, it’s a two-way street; the welfare system has also kept its distance from the workforce system.) With some exceptions, this separation has existed even when workforce programs served large number of welfare recipients. In fact, according to the National JTPA Study in the 1990s, the workforce system achieved solid earnings impacts for this population. However, the institutional goals of the two systems were so different that they operated, in most locations, in a customer-referral relationship. Across the country, a few workforce and welfare agencies have merged, but in most cases even the best merged systems have not become integrated institutions.
One result of the mutual distance the workforce and welfare systems kept from each other is that workforce programs gained little first-hand knowledge about — and no first-hand responsibility for — the range of financial and nonfinancial work supports that continue to be administered largely by welfare agencies. These work supports include food stamps, health insurance (Medicaid), and child care subsidies. The fact that the workforce system paid little attention to these programs was not so important as long as these work supports were limited primarily to a welfare population. But, as I’ll explain in a moment, during the 1990s, many work supports became far more widely available to the non-welfare population — precisely the people that fall under the purview of the workforce system.
It is also notable that most workforce agencies played little role in promoting receipt of a major work support at the time administered through the tax system — the Earned Income Tax Credit (EITC).
My fourth and final observation has to do with performance standards. When the main goal of the workforce system was to move unemployed people into work, it was appropriate to have performance standards that focused on the employment outcomes of program participants — such as the proportion of those who had jobs, how their post-program earnings compared with pre-program earnings, and their employment status some time after getting a job. Once fully in place, the performance standards became — and continue to be — powerful drivers of the entire system.
But there is increasing evidence that the current performance standards may well have had the unintended consequence of driving the system in the wrong direction. In order to meet or exceed the standards, programs may have enrolled and placed many people who would have found training or jobs on their own. Looking back, the National JTPA Study in the 1990s showed that many of the men in the 16 study sites would have done about as well on their own, without JTPA services. For women in the study — with the notable exception of welfare recipients — the program made only modest differences in their earnings. The results of the National JTPA study could, at the time, be seen as a natural, almost predictable, result of the way performance standards were set up to drive the system. An inconvenient truth about outcomes-based performance standards has been found in numerous rigorous studies that I will return to near the end of this paper.
One other point relevant to a new vision for the workforce development system: the performance standards for post-program employment rates were and continue to be a reporting requirement — Are program participants still working? — rather than a way to drive the system to provide job retention services.
Recent Changes in Low-Wage Labor
Now I’d like to turn to much more recent times, where we have seen notable changes in two main areas that should have — but have not yet so far — had an impact on the workforce system: the growth in the low-wage labor market and the expansion of work support programs for low-wage workers.
The first notable change is the growth the low-wage worker population in the United States and the resulting employment problems they and their employers face. Although the trend lines show this starting well before the current decade, low-wage work has now become common in this country for reasons including globalization, the bifurcation of the labor market, the loss of industrial jobs that pay relatively high wages, the decline of unionization, and the failures of our public education system.
It almost always surprises people when they learn that the median hourly wage of the 150 million workers in the country was $14.15 an hour in 2005, according to the Bureau of Labor Statistics (BLS). That means that half of the workforce earned more than $14.15 an hour, and half earned less. As the figure below shows, the annual (pre-tax) earnings for a full-time worker at the median wage was about $29,000 a year. If this worker headed a family of four, the family’s earnings would have been a little over 150 percent of the federal poverty level of $19,350 in 2005. The same family’s income would rise to only about 230 percent of the poverty level if this worker’s spouse also was employed half-time at the median wage.
Further, 25 percent — fully one-fourth — of the workforce earned $9.46 or less an hour. With one full-time earner working at this wage, a family of four would be just above the poverty level. With an additional part-time worker earning the same wage, they are at about 150 percent of the poverty level.
What do we know about the low-wage worker population — other than its large size — that is relevant to the workforce investment system? There is a lot of research showing that low-wage workers often encounter serious problems with job retention that end up costing employers — and workers — a great deal. Without going into details, the consensus is that many low-wage workers do not keep their jobs, and that many of those who lose them move into other low-wage jobs. There is also data showing that an increasing proportion of the low-wage workforce consists of immigrants — an issue that I return to below.
It may also be surprising to learn that low-wage work is going to be around for as far into the future as the ten-year projections of the Bureau of Labor Statistics (BLS) show. In 2004, the proportion of jobs filled by people with a high school diploma or less was about 47 percent. BLS projects that this proportion will barely drop — to about 46 percent — in 2014. In terms of sheer numbers, lower-wage jobs will outpace higher-wage jobs for as far out as the eye can see. Yet, at the same time, employers in some sectors — the fast-growing health care industry in particular — have increasing numbers of better-paying second- and third-level jobs open that they cannot fill. All of this suggests a role for both job retention and job advancement services that, by definition, would work directly with employers.
The second major change in the world of low-wage work reached its peak during the 1990s: the substantial expansion, by the federal government and some states, of a variety of financial and nonfinancial assistance for low-income working families that I discussed earlier. Often called “work supports,” “earnings supplements,” or “work incentives,” they support, or “reward,” people in low-wage jobs with additional income beyond their paychecks. The largest of them — the Earned Income Tax Credit (EITC) — can be received only by individuals who work. Because such supports are not seen as “give-aways,” they have continued to draw bipartisan support, becoming integral components of our nation’s policy for dealing with low-wage work and low-income families.
While two of these work supports — the EITC and the Child Tax Credit (CTC) — are run through our tax system, the other three main ones are administered by state and local public agencies: low- or no-cost health insurance for adults and children, child care subsidies, and food stamps.
Together, these tax system-administered and public agency-administered work supports have substantially changed the income calculus for low-wage work in the country. The chart below shows family income with and without work supports for a family of four in Ohio in 2005 — on the left, for a single parent working full time, full year at the minimum wage of $5.15; on the right, for a two-parent household where one parent works full time and the spouse works half time, both at minimum wage. One can easily see the huge difference these work supports make in family income — and the chart includes only the EITC and food stamps, not child care subsidies or health insurance.
It bears repeating: Work supports have changed the income calculus for workers (and their families) who are filling the large and growing number of low-wage jobs that the U.S. economy creates. Indeed, for many families, they bring in over $700 a month in additional income.
Or, rather, work supports could make this difference. But they often don’t because a huge problem with work supports, such as health insurance and food stamps, is that the proportion of eligible people who actually receive them is under 50 percent. And the proportion of people who receive the full package of work supports for which they are eligible is well under 10 percent — in spite of the large difference they can make in the income of working families.
While here are lots of reasons for why these welfare-administered supports are not taken up, two major ones stand out in surveys and interviews of low-wage workers:
- Working people do not want to be associated with welfare or go to a welfare office that they see as demeaning; and
- As working people, they literally do not have the time to go to multiple offices to complete multiple application forms.
Even the “take-up” rate of the tax-administered EITC and Child Tax Credits remained well below 100 percent.
Why should workforce agencies even care about work supports and increasing the proportion of working people who get them? There are at least six reasons.
First, work supports increase employment and earnings. Evidence from four rigorous studies demonstrates that receipt of work supports by low-wage workers raises both employment rates and earnings over what they would have otherwise been. Therefore, work support programs and the workforce system promote the same goals.
Second, work supports increase job retention. In two credible studies, people who received work supports had substantially higher job retention rates than those who did not receive them. Again, work supports promote the same goals that the workforce system has.
Third, the expansion of work supports that occurred over the past decade has meant that, unlike 15 years ago, the large majority of those eligible for work supports are not welfare recipients. They are outside the welfare system. They are low-income families with one or two low-wage workers. And many of them — and their employers — need job retention and advancement services and help.
Fourth, delivering work supports through the workforce system would reduce the “welfare stigma” for low-wage workers. The Workforce Investment Act’s One-Stop Centers can be an ideal way to make work supports available to working people in a single location.
Fifth, although not an explicit goal of workforce programs, work supports improve child outcomes. Studies show that work supports improve the educational outcomes and behavior of elementary school-age children in families that received the financial supports compared to an identical group of families that did not receive them.
Last, at the most fundamental level, work supports support work, and many are conditioned on work — and that is what the workforce system is all about.
All of these findings beg an obvious question: Given that work supports lead to a large range of positive labor market outcomes, not to mention positive social outcomes, why is the workforce system not better engaged with them?
The long-postponed reauthorization of the Workforce Investment Act has been a source of frustration for policymakers and workforce development professionals alike. In addition, the debate to date has tended to look backwards rather than forwards — in many cases, people are left fighting the same old battles over what are essentially administrative, not policy, issues.
In this way, it is not unlike the welfare reauthorization debate that ended in early 2006. The reauthorization of Temporary Assistance for Needy Families (TANF) largely ignored two central issues that the welfare system faces in the coming decade: first, the dramatic increase in welfare recipients who are working, which has gone from under 10 percent to over 30 percent since 1996 and is almost certain to grow; and, second, the need for welfare to focus on the hardest-to-employ who remain on welfare caseloads because they have particularly serious employment barriers related to mental health and depression, substance abuse, and other problems.
TANF reauthorization was, in part, a real missed opportunity because the size of the welfare caseload in our country is its lowest level since the early 1960s — meaning that the system could literally afford to take on the bigger issues. The TANF reauthorization experience may offer an important lesson for WIA reauthorization. There is an opportunity for the new Congress to take up some of the larger issues confronting the labor force now and in the coming years and to lay out an expanded vision for the workforce system that is also mindful of budget limits.
A Vision for the Future
Let me now turn to the “vision thing.” By now, I am sure you will have no problem guessing where I am going.
At the broadest level, I am suggesting that the next phase in the development of the workforce system include aggressive efforts to develop the skills and capacity of the system — to retool it — so that it will be able not only to prepare the unemployed for work, but also to assist low-wage workers and their employers in raising job retention and advancement rates, along with the rate of receipt of work supports.
Let me describe this vision to you by returning to the four themes I mentioned at the start.
Expanding the target population
Unemployment, even at the low rates we have experienced in most parts of the country, will always be around, and the labor exchange and skills training functions for the unemployed and dislocated worker populations will always have an important role in the workforce system. But, as the data I described earlier suggest, the substantially lower unemployment rate over the past 10 years, and the increasing size of the low-wage workforce, suggest that the workforce system must begin to tackle job advancement and retention issues. The strategy to do this will have to take into account changing demographics, particularly the increasing proportion of the low-wage workforce that consists of immigrants. It would have to take into account the reality that the number of part-time jobs without benefits is also increasing. It would also have to confront the bind in which the workforce system staff might find themselves, if this expansion were to happen; that is, while employers might prefer services that increase retention rates among low-wage workers, the workers themselves might benefit more from switching jobs as a route to better-paying work or work with more advancement opportunities. Recent research suggests, in fact, that changing jobs is a major route for workers to increase their wage rates.
Working with employers on job retention and advancement/retraining
Such an expansion into retention and advancement/retraining would provide an opportunity to engage employers at the program level, and I believe employers would, by and large, respond positively. But the story is not simple. Many employers of low-wage workers have limited advancement opportunities, and their main concern is reducing turnover in low-wage work. As I noted above, this could conflict with the goal of moving low-wage workers into higher-paying jobs that might exist with another employer. However, employers in some sectors do have advancement opportunities within their firms, and they see their incumbent low-wage workers as being strong potential candidates. I can say this because MDRC is conducting, with support from the U.S. Department of Labor and several foundations, the Work Advancement and Support Center Demonstration in One-Stops in four cities. As part of this project, demonstration site staff and MDRC staff have interviewed scores of employers in sectors that have large numbers of low-wage workers. With few exceptions, the employers have been keenly interested in working with One-Stop staff to improve the job retention rates of their workers, to raise their “soft” and “hard” skill levels, to train people for advancement opportunities when they exist, and to make work supports available. Based on this experience, my guess is that most employers would welcome this service and relationship with the workforce system.
Becoming an access point for work supports
I may be wrong about this, but, in my experience, very few One-Stops have housed work support eligibility staff from the welfare agency on site, and few One-Stops offer access to work supports for working people. Yet One-Stops are extremely well positioned — far better than welfare agencies — to provide an access point to work supports for working people, who now make up the majority of people eligible for them. Further, the positive and strong labor market outcomes that result from the receipt of work supports strongly suggests there is a logical, even compelling, role for the workforce system here. Doing this would, of course, require developing close collaborative relationships with welfare agencies, which may not be the norm.
Reexamining performance standards
There are two ways in which these drivers of the system should be reexamined during the next phase in the development of the workforce system. First, as the system develops the capacity to engage low-wage workers and their employers, a new set of standards would need to be developed that are appropriate to this population. For instance, since they come in already working, “entered employment rate” just doesn’t fit — something that the new common measures established by the Department of Labor recognizes. But since they would have current earnings when they would be enrolled, the “earnings gain” standard would have to be quite different, and “job retention” would have to be examined given this new population.
Second, as I noted before, performance standards should not drive the system to enroll people who would do as well, or nearly as well, on their own. This is not an easy issue to deal with because measuring program outcomes has broad appeal and seems to be a transparent indicator of program effectiveness. But looks can be deceiving — and, in the case of outcomes-based performance measurement of program interventions such as job training, looks are deceiving. There are no rigorous program evaluations that show a consistent relationship — and many would say any relationship at all — between high outcome rates and high levels of program effectiveness. In other words, a workforce program is not effective when 90 percent of its participants get jobs — if 90 percent of them would have gotten jobs anyway. This should give us all long pause. So, as we look to the future of workforce development, we need to cast a hard, critical eye on performance standards to ensure that they drive the system in the right direction and are appropriate for the populations and employers it engages. After 25 years of the same basic system, we clearly need new thinking and new approaches to performance measurement and standards in the coming years — so that their use promotes and rewards effective programs.
Building the Vision Within Current WIA Proposals
As you look at the final versions of the 2006 WIA reauthorization bills that are likely to be the starting point for reauthorization activity 2007, you will see the germ of the vision I have outlined. Under a section called “Permissible local employment and training activities” are subsections on both “incumbent worker retraining” and “work support activities for low-wage workers.” It is this section that could be strengthened in order to create the foundation for the development of the new capacity and skills of the workforce development system described in this paper.
With budget constraints in mind, this expanded vision might be undertaken through national grants in a number of select Workforce Investment Boards across the country — say, 50 — so that their experiences and the lessons that grow out of their work can inform the next WIA reauthorization debate. Another low-cost alternative would be to take the model used in the 1970s to initiate private sector involvement in the workforce system by establishing a small program, the Private Sector Initiative Program, nationwide. In the late 1970s, this program allowed each local area to develop, slowly, what five years later, under JTPA, came to be the major policy arm for the local system — Private Industry Councils, which are now Workforce Investment Boards. Under either scenario the system would be developing new skills and tools that could address future labor market realities.
Let me close by saying that any change to the workforce investment system, particularly a long-term expansion into a new target population, would be very tough. We know from the experience in the four sites involved in MDRC’s Work Advancement and Support Center Demonstration that the challenges of integrating workforce development, work supports, and retention and advancement services are truly enormous. Sites are learning to work with employees and employers at work sites as their main venue. They are learning how to provide advancement counseling and advice. They are learning how to work with employers to develop career ladders and structure customized training and retraining programs around them. They are learning how to interact with work support eligibility staff from welfare agencies, and how to integrate access to work supports into the normal operations of One-Stops.
But biggest issue has been culture change. All of this retooling and expansion of the skills of workforce development professionals requires, first and foremost, a recognition that One-Stops would be working with working people, not just unemployed people. Working people by definition have limited time availability and cannot be expected to wait (and lose money) for extensive application procedures or repeated call-backs to an office. Efficiency and convenience of service delivery become essential. It means that staff often need to do their jobs at employer sites or other venues outside the One-Stop offices, rather than requiring customers to come in. And offices need to be open longer hours and on weekends, as they are in most of the rest of the service industry.
I believe that this expanded vision for the workforce system — with slow and carefully documented implementation in a limited number of locations — could lay the groundwork for the growth of a reinvigorated, reenergized system with a bright future.
 Larry L. Orr et al., The National JTPA Study: Impacts, Benefits, and Costs of Title II-A (Bethesda, MD: Abt Associates, March 1994).
 Daniel E. Hecker, “Occupational Employment Projections to 2014,” Monthly Labor Review, Vol. 128, No. 11, November, 2005.
 Charles Michalopoulos, Does Making Work Pay Still Pay? An Update on the Effects of Four Earnings Supplement Programs on Employment, Earnings, and Income (New York: MDRC, 2005).
 Charles Michalopoulos, Sustained Employment and Earnings Growth: New Experimental Evidence on Financial Work Incentives and Preemployment Services, (New York: MDRC, 2001).
 Pamela A. Morris, Lisa A. Gennetian, and Greg J. Duncan, “Effects of Welfare and Employment Policies on Young Children: New Findings on Policy Experiments Conducted in the 1990s,” Social Policy Report, Vol. 19, No. 2, 2005.
 Harry J. Holzer, Encouraging Job Advancement Among Low-Wage Workers: A New Approach, Policy Brief #30 (Washington, DC: The Brookings Institution, May 2004).