Institutional analysis of labor markets typically focuses on the effects of social policy and industrial relations. For this research, the United States provides a model of market deregulation. U.S. unions are weak, and the welfare state affects only those at the fringes of the job market.
This contrasts with Europe, where employment relations are highly regulated. Unions set wages for entire economies, and welfare states significantly influence the supply and demand for labor. These institutional differences acquired special importance over the last two decades as European unemployment rose in comparison to U.S. unemployment. Recent trends are striking. While unemployment in the European Union averaged 9.5% between 1990 and 1993, the U.S. average was only 6.5% (OECD 1996). Currently, unemployment hovers around 10% in Germany, Italy, and France, while U.S. unemployment averaged less than 6% between 1994 and 1996.
These trends suggest unregulated labor markets yield strong employment performance. Of all the labor markets of the advanced economies, the United States best approximates the competitive model of neoclassical theory. In this model, job seeking is intensified by meager state support for the unemployed, and low unionization allows wages to adjust to market conditions. In Europe, institutions introduce inefficiency: large welfare states and strong unions stifle labor demand and reduce work incentives (Olson 1982; Lindbeck 1985; Giersch 1993; OECD 1994a).
We challenge this analysis by arguing that labor markets are embedded in a wide array of social arrangements that extend beyond the welfare state or industrial relations. In the United States, criminal justice policy provides a significant state intervention with profound effects on employment trends. The magnitude of state intervention is reflected in budget and incarceration figures. In the early 1990s, $91 billion were spent on courts, police, and prisons, dwarfing the $41 billion spent on all unemployment benefits and employment related services (Statistical Abstract of the United States 1995, table 585). By 1996, 1.63 million people were being detained in American prisons and jails—a threefold increase from 1980 (Gilliard and Beck 1997, p. 1). These figures suggest that incarceration generated a sizeable, nonmarket reallocation of labor, overshadowing state intervention through social policy.
This article studies the penal system as a labor market institution and provides evidence for its dynamic effects. Our central argument is that U.S. incarceration lowers conventional measures of unemployment in the short run by concealing joblessness among able-bodied, working-age men, but it raises unemployment in the long run by damaging the job prospects of ex-convicts after release. Incarceration, unlike social welfare policy, deepens inequality because its effects are increasingly detrimental for young black and unskilled men, whose incarceration rates are highest and whose market power is weak. This argument suggests that incarceration has lowered the U.S. unemployment rate, but it also implies that sustained low unemployment in the future will depend on continuing expansion of the penal system.