| Abstract: | This study examines the impact of model selection on substantive conclusions about racial differences in presentence and sentencing practices. Using a statewide sample of convicted felons, an additive model is estimated to assess the direct and indirect effects of race on various dispositional decisions. Then, separate models for blacks and whites are estimated to evaluate within- and between-race differences. A comparison of the results from each specification illustrates how the standard additive approach used in past studies can mask and suppress racial differences in criminal processing. The findings of race-specific models are then discussed in terms of their implications for further research on racial disparity in the application of criminal sanctions. |
