This paper considers whether the disadvantages that accrue to black Americans with a dark skin hue also induce a transition into criminal activity—an outcome consistent with standard economic models of crime. We also examine whether or not prison terms are conditioned on skin hue. With data on black offenders in the state of Mississippi, we estimate Cox proportional hazard specifications of the transition into criminal activity, and find that it is conditioned on the darkness of skin hue. Our parameter estimates are consistent with a theoretical framework in which being black and having a dark skin hue induces a transition into criminal activity by limiting the set of legitimate opportunities for an individual. Given a conviction, we also find that the severity of punishment for black offenders as measured by the length of sentence is an increasing function of the darkness of skin hue. JEL Classification: J0, J7, K4, Z0 ∗Professor and Chair, Department of Economics, College of Business Administration, 4202 E. Fowler Ave, BSN 3403, Tampa, Florida 33620–5500, email: kgyimah@coba.usf.edu, Tel # (813) 974-6520, ∗∗Corresponding Author: Director, Mississippi Urban Research Center, Jackson State University, P.O. Box 17309, Jackson Mississippi, 39217, email: gprice@murc.org, Tel #: (601), 979-1428. This paper was prepared for presentation at the Annual Meeting of the Allied Social Science Association, joint AEA/NEA paper session on “Skin Tone Discrimination and Economic Outcomes”, January 7, 2006 in Boston, Massachusetts. Conditional on being black there is evidence that skin hue matters for a wide array of socioeconomic outcomes. For example, relative to blacks with a light skin hue, there is evidence that blacks with a dark skin hue fare worse in terms of wages (Goldsmith, Hamilton and Darity, 2005), occupational prestige (Hill, 2000), unemployment ( Hunter, Allen and Telles, 2001), access to health resources (Bodenhorn, 2002), and intergenerational wealth accumulation (Bodenhorn, 2003). Such findings suggest that the distribution of advantage and disadvantage in American life is conditioned not just on being black, but given that one is black, skin hue as well. That skin hue conditions economic outcomes for blacks suggests that profit maximizing firms and utility maximizing individuals optimize across a preference for blacks with light skin hues (Goldsmith, Hamilton and Darity, 2005). If so, the distribution of disadvantage among blacks will fall along a continuum of gradations in skin hue, with disadvantage increasing with respect to the darkness of skin hue. One possible manifestation of disadvantage is crime. In the canonical economic model of crime of Becker (1968) and as extended by Ehrlich (1973), disadvantaged individuals can be viewed as those with constrained opportunities for engaging in legitimate activities, relative to illegitimate activities. To the extent that conditional on being black, opportunities for engaging in legitimate activities are inversely proportional to the darkness of skin hue, the probability of participating in illegitimate activities—crime—may also be conditioned on skin hue. Despite the apparent importance of skin hue in the distribution of advantage and disadvantage, the economics of crime literature tends to view blacks as one homogeneous group. To the extent that skin hue matters, aggregating across black Americans in this manner could lead to biased estimates of the effects of being black on participation in criminal activity, and to inferences that being black is associated with higher stocks of “criminal capital” relative to non-blacks. However, given that race and skin hue determine the distribution of disadvantage among black Americans, the effects of being black on criminal activity may instead reflect the disadvantages that accrue to being black conditional on skin hue. In this paper we consider the effects of skin hue both on the likelihood of participation in criminal activities, and on sentencing for black Americans conditional on being convicted for crime. To the extent that individuals, firms and social institutions optimize across preferences for blacks with a light skin hue, it is plausible that blacks with a dark skin hue face, relative to those with a light skin hue, constraints on opportunities for legitimate activities that motivate illegitimate activities and once arrested, are punished more severely than black offenders with a light skin hue. See for example Ehrlich (1973) and Gyimah-Brempong (1997). The only evidence that we are aware of that links skin hue among blacks with criminal justice issues is that provided by Johnson, Farrell and Stoloff (2000). They found that among black men with prior criminal records in Los Angeles, the jobless rate for those with a dark skin hue was 54 percent—in contrast to 41.7 for those with a light skin hue.