The analyses of the death penalty reported in these pages, while diverse in both subject and method, are part of a distinctive post–McCleskey v. Kemp (1987) generation of studies in which the social science literature on capital punishment has begun to stand separate from ever present constitutional litigation on death penalty issues. This independence of scholarship from litigation has come as a consequence of rejection: A majority on the U.S. Supreme Court has forcefully rebuffed the attempts of social scientists to influence the constitutional law of capital punishment.

How are violence and pain put into legal discourse? How does law distinguish its violence—capital punishment—from other kinds of violence? Do the strategies used to differentiate legal and extralegal violence alleviate anxiety about law and the uses to which law's violence is put? This article addresses these questions through an analysis of a capital trial in which an African-American man is being retried for the murder of a young white woman. It examines the social and cultural resources used to speak about violence and to differentiate legal and extralegal violence, and it suggests that the juxtaposition of narratives about violence in capital trials arouses rather than alleviates anxiety.

Sociolegal research focused on equal protection and the death penalty faces an impasse as the result of recent Supreme Court decisions limiting the admissability of social science research in general and of aggregate data on bias in the application of capital punishment in particular. This impasse, however, provides an opportunity to renew investigation and analysis of the deeper meaning and implications of the death penalty in contemporary societies, not in abandonment of but as a complement to equal protection–oriented research.

This article examines the Supreme Court's use of social science research evidence in 28 capital punishment cases decided between 1986 and 1989. The study describes the frequency and major correlates of the justices' citation of social science authorities in the 1986–89 sequence of cases. Social science evidence figured significantly in several death penalty cases, although a majority of the justices were more eager to discredit and discount research conclusions than to use them as premises for their decisions, and prevailing case opinions generally promoted principles that had little to do with empirical evidence concerning the administration of capital punishment. Social science citation patterns in majority and dissenting opinions, and in the opinions of “liberal” and “conservative” Supreme Court justices, in significant respects parallel the Court's shifting doctrinal premises in capital punishment decisions.

We address capriciousness in decisions by prosecutors to charge homicide defendants with a capital crime. We suggest that it is useful to think of such decisions as a kind of lottery in which one should focus on the distribution of outcomes when considering both the nature of capriciousness and the degree of capriciousness. After our conceptional framework is introduced, we illustrate our ideas with the analysis of a data set from San Francisco.

In their article in this symposium Berk, Weiss, and Boger (1993) have presented a unique approach in the examination of the issue of capriciousness in capital charging decisions. They define capriciousness as one of two forms of arbitrariness. One form of arbitrariness, discrimination, concerns the use of inappropriate factors such as the defendant's social class or victim's race. In the second form of arbitrariness, capriciousness, it is not the use of inappropriate factors that is problematic; rather it is the inability to identify the factors that “meaningfully” differentiate the few defendants given a capital charge from the many who could have been charged with a capital crime but were not. Berk and his colleagues only concern themselves with capriciousness.

We appreciate Professor Paternoster's evenhanded review of our article and consider many of his concerns legitimate differences of opinion. But it is, nevertheless, useful to specify the key points on which there is agreement or disagreement.

In its 5–4 decision in Tison v. Arizona (1987), the Supreme Court put forth disputable empirical assertions about where community sentiment stood on the death penalty for felony-murder accessories. The Tison majority claimed that the community would support the death sentence for an accessory where (1) major participation and (2) reckless indifference to human life were shown. Two experiments were run to test these empirical assertions, and the results indicate quite the opposite. By a wide majority, community sentiment rejects the death penalty for such an accessory and rejects equal treatment of principal and accessory. Even beyond the death penalty question, to the issue of conviction under the felony-murder rule per se, attributions of causal and moral responsibility showed the principle of proportionality reigning supreme. Implications for how the Supreme Court goes about doing social science are discussed.

Twenty years after the Supreme Court struck down existing death penalty statutes (Furman v. Georgia 1972) and a day after Justice Thurgood Marshall's death, the senior Justice of the U.S. Supreme Court, Harry Blackmun, charged the Court with coming “perilously close to murder” in its latest death penalty ruling. He made this charge in an unusual oral dissent from a decision which held that a federal appeals court could not hear newly developed evidence of a death row inmate's possible innocence (Herrera v. Collins 1992). The Court, nine months earlier, had barred a lower federal court from hearing new evidence challenging execution by asphyxiation as cruel punishment (Gomez v. U.S. District Court 1992) and had then taken the unprecedented step of ordering no further stays of execution (Vasquez v. Harris 1992).1 How did we reach this point when 20 years earlier the Court had declared that the death penalty as applied was too arbitrary to be constitutionally acceptable, and when virtually all other modern industrialized nations of the free world either had already, or soon would, do away with capital punishment?

This study tests the impact of OSHA enforcement on workplace injuries. Using data on injuries and OSHA inspections for a panel of 6,842 large manufacturing plants between 1979 and 1985, we find significant specific deterrence effects. Inspections imposing penalties induce a 22% decline in injuries in the inspected plant during the following few years. We suggest that narrow deterrence perspectives have led to unduly pessimistic assumptions about enforcement effectiveness and that a managerial attention model is more consistent with our findings.

In a technical appendix we describe the Chamberlain technique, a powerful analytic approach for panel data that provides tests and corrections for potential biases endemic in enforcement studies, including unmeasured heterogeneity among units, serially correlated dependent variables, and endogeneity of inspections. We argue that more empirical studies of enforcement impacts are necessary to provide an appropriate perspective for descriptive and analytic studies appraising regulatory behavior.