Indirect effects of unemployment and low earnings: crime and children's school performance

This thesis consists of three self-contained essays that consider indirect effects of unemployment and low earnings on crime and children’s schoolperformance.

The first essay, Crime, unemployment and labor marketprograms in turbulent times (joint with Jonas Agell), investigates the effect of unemployment and participation in labor market programs, in general and among youth, on Swedish crime rates using a new panel data set for Swedish municipalities for the period 1996-2000. The exceptional variation in Swedish unemployment in the 1990s provides a remarkable (quasi-) experiment. Between 1996 and 2000 the overall unemployment rate (including those enrolled in labor market programs) decreased from 11.9 to 6.8 percent, and for those most likely to commit crimes, people under the age of 25, unemployment decreased from 21.2 to 8.7 percent. But the decrease in unemployment was far from uniform across the country, and our identification strategy is to use the exceptional variation in the improvement in labor market conditions across municipalities to isolate the relationship between unemployment and crime. We also consider whether placement in labor market programs reduce crime. Such an effect could arise for many reasons. Program participation may imply: (i) that there is less time for other activities, including crime; (ii) social interactions that prevent the participant from adopting the wrong kind of social norms; (iii) a greater ability to earn legal income in the labor market. Unlike most previous studies we identify a statistically and economically significant effect of general unemployment on the incidence of burglary, auto-theft and drug possession. Contrary to much popular wisdom, however, we could notestablish a clear association between youth unemployment and the incidence of youthful crimes and there is no evidence that labor market programs – general ones and those targeted to the young – help to reduce crime.

   The second essay, Earnings and crime: The case of Sweden, analyzes whether low earnings has an effect on Swedish crime rates, considering the overall crime rate and specific property crime categories, using a panel of county-level data for the period 1975–2000. Various measures of the income distribution are considered, based on annual labor earnings as well as annual disposable income. The results indicate that the effect of low earnings on crime in Sweden is at best weak. We estimate a significant effect of low earnings on the number of auto thefts, but the effect is small. Low earnings seem to have no effect on the overall crime rate, the number of burglaries or the robbery rate. The results give, however, further support for an unambiguous link between unemployment and the overall crime rate as well as specific property crime categories. These findings are in contrast with results from, for example, the United States where wages are found to have a stronger impact on crime than unemployment. The differing results could, at least partly, be explained by the fact that during the period investigated, Swedish unemployment has been of a more permanent nature than U.S. unemployment, and that transitory earnings fluctuations appear to dominate the Swedish earnings distribution for young men, a part of the population committing a disproportionate share of many crimes.

   Finally, the third essay, Parental unemployment and children’s school performance, considers another possible indirect effect of unemployment, namely the school performance of the children of the unemployed. I use Swedish data on individual GPA from the completion of primary school at age 16 and final grades from upper secondary school for a majority of all children completing primary school in 1990 directly moving on to three years of upper secondary school, which they complete in 1993. The empirical method builds on the idea that primary school GPA can be used to control for family and individual heterogeneity. The huge variation in Swedish unemployment during the beginning of the 1990s, which can be traced to macroeconomic events, provides an ideal setting for testing the hypothesis that parental unemployment affects children’s school performance. The main results can be summarized as follows. If a mother is subjected to an unemployment spell during the period when one of her children attends upper secondary school, the school performance of the child marginally improves. This implies that, for women, the positive effect of having extra time on your hands exceeds the negative effects of the disadvantages caused by unemployment. This positive effect of having an unemployed mother seems to increase with the length of the unemployment spell. On the opposite, having a short-term unemployed father has a negative effect on a child’s school performance while the effect is insignificant for long-term paternal unemployment. The fact that a long-term unemployment spell of the father has a less clear effect could be interpreted as the shock of unemployment wearing out. One explanation for the differing results across genders could be that women in general cope better with being unemployed and hence are able to use their new extra time doing something productive, such as spending quality time with their children.5