Unemployment and Crime

Of course, unemployment rates, age, and offense do not operate alone in determining sentences of imprisonment. In 1966 David Jacobs and Ronald E. Helms showed, in their multivariate analysis of historical and geographical fluctuations of incarceration rates in the United States, that unemployment's impact on crime seemed to have a close linkage with other conditions not noted in prior studies. They point out that rates of state plus federal imprisonment of adults per 100,000 population in the United States changed only from 43.4 to 50.9 from 1918 to 1965, dropped to 35.9 by 1968, then nearly doubled in 13 years to 69.7 in 1981, and rose by 84 percent to 127.9 in 1989. There was no such large movement in rates of crime known to the police, which increased only 12.1 percent in 1975–1991. They found that because of this small range in reported crime rates for this long time span, a multivariate regression only yielded very strong results if crime rates were squared. Also, in their regression analysis, out-of-wedlock births—an index of the rate of breakdown in family relations—was one of the best predictors of imprisonment rates, especially when using birthrate data for five years before the prison figures, to allow for the time before such family conditions could promote higher imprisonment rates. Furthermore, rather than unemployment rates, income inequality, and votes for the Republican Party—an index of conservative political trends—were among the strongest predictors of crime rates.

In 1999 Michalowski and Carlson ascribed the diversity of findings in tests of "the Rusche-Kirchheimer hypothesis" to variations in the historical periods covered by these studies. Their analysis is based on theories about stages in what they call "social structures of accumulation," or "SSAs." The first stage is "Exploration," which occurs when new institutional arrangements emerge to cope with high levels of unemployment and with the displacement of both farm and industrial populations. The second is "Consolidation," an effort to maintain whatever new arrangements seem to help capital to preserve its profit margins. The third stage is "Decay," when conditions develop that impede consolidation policies, thus increasing unemployment and preventing children of workers from achieving upward status mobility.

The Michalowski and Carlson analysis of unemployment and imprisonment for crime focuses on what they and some others call the "Fordist" SSA of 1933–1992, for which they differentiate the three phases indicated. During Exploration, from 1933 to 1947, new types of "capital-labor accords" and welfare policies "socialized the costs of labor force displacement." Thus, unemployment insurance and old-age pensions, both adjusted for changes in the cost of living, shifted burdens of dealing with economic distress in this period from businesses to governments, and helped to maintain social order. In Consolidation, from 1948 to 1966, periods of unemployment were shortened by the bargaining strength of organized labor and its central position in the Democratic Party, which transferred government tax income to the poor, the disabled, and the elderly. These "profit-eroding" developments "made it difficult for capital to protect profit margins in the face of growing foreign competition," these authors claim, so that Decay occurred in 1967–1979, during which unemployment and inflation rose while profits declined and the United States lost the Vietnam War despite sending 3 million workers there.

These authors identify a new Exploration phase from 1980 until 1992, in which "cyber-technology" accelerates labor displacement, and also increases earning inequality between workers in digital information systems and those at "hamburger jobs." The latter were disproportionately from minority groups, and in this period there was a "shift away from placative social welfare strategies . . . toward repressive control strategies based on increased use of imprisonment" (p. 226). Consequentially, even with rates of arrest and of unemployment in the 1990s usually below those of prior decades, the rate of imprisonment nearly tripled, confining predominantly young men who had dropped out of school and work. The latter were disproportionately African Americans, whose rates of unemployment exceeded what one would predict for white workers of similar education and work experience. Although their higher rates of joblessness are doubtless due in large part to prejudice against them, it is also because they are below the growing number of Latinos in the labor market in their "reservation wage"–the lowest pay rate for which they would be willing to work (Moss and Tilly, p. 9). Another major factor in their unemployment rates is that they exceed other racial and nationality groups in the proportion who have entered military service, or are enrolled full-time in schools, which probably removes those with the highest earning potential from the labor market (Mare and Winship). But still another influence was their ready opportunity, in the segregated slum areas in which most of the housing available to them was located, to gain much greater income from drug dealing than from the jobs available to them.

In summary, evidence and analysis indicate that unemployment is predictive of crime, but disproportionately for youth, the least educated, those in broken or disorganized families, and those segregated in poor minority residential areas. Also, these relationships of unemployment to crime are likely to continue unless unsegregated housing, special education, family unity, work experience, and appealing career jobs become more readily available to those who are unemployed.