Originally published in Spotlight on Poverty and Opportunity.

At the height of the Great Recession, when the unemployment rate reached 10 percent, at least 40 states used federal funds under a special emergency program to put more than 250,000 adults and youth to work, often in private businesses that received subsidies to cover their wages. Governors from both parties hailed the program.

At around the same time, MDRC, a nonprofit, nonpartisan research organization, released results from several unusually rigorous evaluations of seemingly similar subsidized employment programs. The studies found that the programs generally failed to improve the long-term employment prospects of the disadvantaged workers they targeted.

What’s going on? Is this one of those cases where policymakers ignored evidence and funded an ineffective program model? Not necessarily. In fact, the emergency programs had very different goals than the programs MDRC studied and should be judged using different criteria. Unfortunately, this distinction is sometimes lost in the public discourse, since different program models with vastly differing objectives are often lumped together under the heading of “subsidized employment.” A bit of parsing may help make this picture clearer.

What is subsidized employment?

At the broadest level, all subsidized employment programs use public funds to create jobs for the unemployed. The granddaddy of subsidized employment programs, the Works Progress Administration (WPA), created public works jobs for 8 million people during the Great Depression and left a lasting legacy: thousands of roads and bridges, schools and parks, even murals and photographs. A program of this scale and scope seems fanciful in today’s context, but the concept did not die when the WPA ended in 1943. The Public Service Employment program in the 1970s employed more than 700,000 people at its peak, and the state programs funded under the TANF Emergency Fund, described earlier, put thousands to work in 2009 and 2010 (Congress declined to extend the program when it expired in late 2010).

These “counter cyclical” programs aim to stimulate the economy and assist the unemployed — by providing wage-paying jobs rather than welfare benefits — during periods of high unemployment. The wages earned circulate through the system, helping to prop up demand for goods and services. Such programs also allow workers to keep exercising their employment muscles rather than letting them atrophy. All of these programs were controversial to varying degrees, but by focusing on the core American value of work they also generated fairly broad political support, at least during hard economic times.

Another type of subsidized employment program has operated sporadically since the 1970s, mostly under the radar. These kinds of models, sometimes called transitional jobs programs (though many other terms are used), are much smaller and they target particular groups who face high rates of joblessness even when the labor market is strong: welfare recipients, people returning to the community from prison, youth who are disconnected from school and work, and others. Typically they offer a few months of subsidized employment at a nonprofit organization or, less commonly, with private businesses that agree to hire the worker. While the workers are in their temporary job, the programs often provide support services and try to help them find more permanent jobs.

Transitional jobs programs put wages into people’s pockets, but that isn’t their main goal. They use subsidized jobs as a skill-building tool, aiming to prepare hard-to-employ individuals for success in the regular labor market. They operate on a simple premise — that the best way to learn to work is by working — though the programs are far from simple to operate.

What have we learned?

Over the past few decades, starting with the National Supported Work Demonstration in the 1970s, a number of transitional jobs programs have been rigorously evaluated. These studies assigned people, at random, either to a program group that had access to transitional jobs and other services, or to a control group that did not. A few of these programs produced lasting improvements in participants’ employment rates and earnings, but most did not. In particular, studies of five TJ programs for people returning to the community from prison, conducted between 2005 and 2012, found that the programs substantially raised employment rates initially, but the gains were driven by the transitional jobs themselves and did not last — though one program, run by the NYC-based Center for Employment Opportunities, was able to reduce recidivism and save money for taxpayers.

These mixed results suggest that innovative, new approaches are needed to help very disadvantaged people prepare for permanent employment. Indeed, two large federal projects, one sponsored by the U.S. Department of Health and Human Services (HHS) and one sponsored by the U.S. Department of Labor (DOL), are currently testing a next generation of transitional jobs models, and very early results show some trends worth watching.

But evaluations of TJ programs will never fully address the question of whether it makes sense to use public funds to create jobs for the unemployed. While TJ programs are rightly judged on their ability to improve participants’ long-term job prospects, subsidized employment programs that are mostly designed to provide work opportunities and income when jobs are scarce must be judged by a different standard – for example, on their ability to create large numbers of meaningful work opportunities quickly and efficiently, and on the value of the work that is produced. Clearly, there are strong arguments for counter-cyclical subsidized employment programs during recessions. But even today, when the national unemployment rate is relatively low, joblessness remains stubbornly high for disadvantaged youth, for African-American workers, and in many distressed urban neighborhoods and rural areas. For some of these groups and geographic areas, the business cycle runs from bad to worse, so the rationale for targeted counter-cyclical jobs programs may be relevant even today.

Transitional jobs programs are an important strategy for helping very disadvantaged people build their skills and gain a foothold in the labor market, and the ongoing federal research projects will soon provide evidence on whether the latest models work better than earlier approaches. Regardless of what we learn from those studies, however, there may be good reasons to create jobs for the jobless when the private labor market fails to provide opportunities for everyone who wants to work. Many believe that work structures people’s lives, connects them to the broader society, and even improves their health. Children may benefit from growing up in households and communities with steady income and working role models. And subsidized workers could be deployed to help address pressing needs in struggling neighborhoods. These are the kinds of potential benefits that should be considered as policymakers decide whether subsidized employment programs are worthy of public investment.