programs to evaluate was the focus of a set of five guiding principles established in a recent report (NRC, 1998). The use of principles such as these is vital, given the daunting array of DOJ programs that could be evaluated. NIJ needs greater flexibility to identify, in a carefully orchestrated manner, those programs that warrant evaluation.
The expansion of NIJ's role in program evaluation has implications for research training. To meet the growing demand for skilled evaluation researchers, more and more individuals from many disciplines require training at the predoctoral and postdoctoral levels. NIJ provides a modest degree of support for individual trainees, but it does not award full-fledged training grants to institutions. Institutional, as opposed to individual, training grants can create a critical mass of trainees at a given institution, draw young people into a career path, augment the field's research capacity, and sustain the field for the future. The Institutional National Research Service Awards established by the NIH provide a model for an institutional training grant.
NIJ should give its highest priority in crime prevention research and program evaluation to studies bearing on prevention of violence, especially lethal violence. While the terms ''violence prevention" and "crime prevention" are often used interchangeably, Zimring and Hawkins (1997) have shown that thinking of violence simply as a subset of "the U.S. crime problem" fundamentally distorts our understanding of the problem and confuses our responses to it. Rates of death and life-threatening injury from assaultive behavior in the United States are 4–18 times higher than those in other developed nations, even though rates of property crime are about the same. Zimring and Hawkins argue that the uniquely higher rate of lethal violence in America is attributable to differences in the social environment, not differences in either the volume of crime or the number or characteristics of offenders. They also argue that the usual tools of criminal justice are not adequate to the task of reducing lethal violence and that undifferentiated emphasis on the prevention and control of "crime" or even "violence" misallocates resources, diffuses the focus of attention, and fails to address the specific factors most likely to reduce lethal violence. Thus, emphasis should be placed on the prevention of lethal violence (i.e., that subset of violent events that present a risk of serious injury or death).
The committee recommends that NIJ continue to give explicit priority to the prevention of violence, especially lethal violence, within its overall activity in crime prevention research and program evaluation, and that NIJ establish new institutional training grants for violence prevention research at academic institutions.
NIJ also is to be credited with its collaborative approach to research support. As described earlier, NIJ sponsored a joint solicitation with NCIPC for extramural research proposals on the topic of violence against women. It also collaborated with NIH, NCIPC, and other federal agencies on a grant solicitation