Search Results

The underlying theoretical premise is that strong, stable two-parent families are related to the state of the marriage market. When there are large numbers of marriageable men relative to unmarried women, fewer female-headed families will form. But imprisonment reduces the supply of marriageable men, according to this theory. Thus, the central hypothesis is that when there is a change in sentencing policies that increases imprisonment there will be a corresponding reduction in the supply of marriageable men and an increase in the incidence of female-headed families.

To test our hypothesis we designed three research models to examine the relationship between family structure and incarceration, but using different measures and datasets. The goals was to see if testing for the same impacts using different data would confirm our theory that changing sentencing policies has adversely affected families. The first model (Module A) merged the National Longitudinal Survey on Youth (NLSY) for 1985 and 1994 with the Urban Institute's 1980 and 1990 Underclass Database (UDB) and the 1984 and 1993 National Correctional Report Program (NCRP) data set for counties. It measured the impact of inmate admissions and releases on female-headed families, female family headship, and out of wedlock births. Module B merged data from the Current Population Survey for 1985 and 1995 with state level data to measure the Darity-Myers sex ratio and expected welfare income and to measure their relationship to family structure, sentencing guidelines, and minimum sent e nces for drug related crimes. Finally, Module C used data collected from inmates entering the Minnesota prison system in 1997 and 1998, information from the Minnesota Crime Survey, and the 1990 Census to assess any connections between incarceration and family structure.

While the results of the project support parts of the underlying theory, the conclusion that imprisonment increases female-headed families is not strongly supported. Using the NLSY data set, we find few statistically significant impacts of prison admissions on different measures of family structure. And in the model specifications where we do find statistically significant impacts of admissions on family structure, e.g., in fixed effects models for 1994, the results emerge for states without sentencing guidelines but not for states with guidelines. The lack of strong and robust support for unintended impacts of sentencing guidelines on family structure may be a result of how we have measured the intervening influences of imprisonment. We looked at admissions rates and release rates and the ratio of admissions rates to release rates. It is possible that an examination of the number of admissions or releases will change these conclusions. But the main finding using the NLSY data set is that there are no strong or robust indicators of the adverse impacts of sentencing reforms on family structures.

When we looked at the replication of the Darity-Myers model of sex-ratios and family structures we again find little support for the adverse impacts of sentencing reforms on family structures. We find unequivocal support for the underlying model that links sex-ratios to family structure. But we find ironically that sex-ratios in recent years are slightly higher tor blacks in sentencing guidelines states than there are in non-sentencing guidelines states. In states with mandatory minimum sentences for drug-related crimes the sex-ratios are not much different from those of states without manda to r y minimums. As a result, we do not find consistent evidence that sentencing guidelines or mandatory minimum sentences adversely affect family structure, despite very strong and consistent evidence that lower supplies of marriageable men are associated with higher incidences of female-headed families. There are some minor differences in the responsiveness of family structures to sex-ratios in states with different sentencing policies and there is a small impact of drug enforcement policies on black (but not white) family structures. These impacts should be explored further in future research.

We hoped to examine in greater detail one state with a long history of sentencing reforms and the impacts of local community factors on inmate family structure. We wanted to know whether the aspects of locations that lose marriageable males due to incarceration were more important in explaining the family structures left behind by prisoners than the characteristics of the prisoners themselves. If sentencing reforms adversely affect local neighborhoods then one would expect these effects to translate into differences in family structures faced by inmates who come from these communities. We find, however, that the local impacts are small. Far more of the variance in inmate family structures is explained by individual inmate demographics than location characteristics.

In summary, then, we have looked at three different data sets at different points in time and using different notions of family structure and we find little consistent support for the theoretically plausible hypothesis that there are strong unintended impacts of imprisonment policies on family structures.