The national debate regarding federal habeas corpus for state prisoners is fueled in the main by ideology. To some, the authority of the federal courts to entertain constitutional challenges to state criminal convictions is the embodiment of all that was right about the Warren Court and the vision that Court offered of a meaningful system of American liberty, underwritten by independent federal tribunals willing and able to check the coercive power of government. By this account, the Bill of Rights is the protean source of safeguards for individual freedom – commanding generous, imaginative, and insightful elaboration by federal courts at all levels. Because the Supreme Court sits atop a large system and accepts only a few dozen cases each year, it can scarcely shoulder sole responsibility for giving effect to constitutional law. The lower federal courts, receiving habeas corpus petitions from prison in- mates, provide the indispensable machinery for maintaining and invigorating individual rights on a daily basis. To others, by contrast, federal habeas is a constant irritant – an expensive, time-consuming, and redundant enterprisethat frustrates law enforcement and needlessly injects the federal courts into matters better left to the states. By this second account, habeas is the paradigm of all that was wrong with the Warren Court – namely that Court’s asserted failure to appreciate the societal threat posed by crime and its palpable distrust of the states and state courts. Between these extremes, there is little common ground. The two camps assign radically different values to the interests at stake in the debate – law enforcement, local authority, individual liberty, federal oversight.

Unfortunately, the ideological debate over habeas has not often been informed by hard evidence about the actual workings of the system under current arrangements. We mean to fill that gap in this Article. We here offer an empirical study of habeas practice, sponsored by the State Justice Institute and undertaken by the Institute of Judicial Administration at New York University. Using conventional social science techniques, we examined actual case files to build a data base from which to generalize. We presented a preliminary report on our findings at a session of the Fortunoff Criminal Justice Colloquium at NYU in April 1990. We present our final results and analysis here.