Lawrence W. Sherman, "Crime Prevention's Bottom Line," THE WALL STREET JOURNAL, August 6, 1997, p. A 15. Crime rates are going down and Washington is eager to take credit. The Crime Act of 1994 quadrupled funding for state and local crime prevention to more than $3 billion per year, the highest levels ever. Whether that funding increase caused the subsequent decline in violent crime is unclear, primarily because federal funds support hundreds of different programs. Some programs work, some don't, and some may even increase crime. But until Congress spends more to evaluate these programs, we can't tell. That is the major conclusion of a report prepared by the University of Maryland's Department of Criminology and Criminal Justice at the request of the Justice Department and mandated by Congress. By asking for a comprehensive scientific evaluation of the Justice Department's funding for local crime prevention, Congress framed a more basic question: How do we know what works? What's the bottom line? Despite recent efforts to reinvent government with closer analogues to a bottom line, the more appropriate model for evaluating crime prevention programs may be the Food and Drug Administrations approach to evaluating drugs: Do field tests with human beings show this program to be safe and effective? There are already documented examples of how this approach is relevant to crime prevention: l. A University of Southern California evaluation of a gang prevention program found that the program was holding the gang together and sustaining its violent crimes. When the program lost funding, the gang broke up and its crime rate declined. 2. Controlled evaluations of mandatory arrests for misdemeanor domestic assaults show that while arrest deterred some assailants, arrest caused some other assailants to increase their violence against women. Despite these research results, Congress has spent hundreds of millions of dollars a year on domestic violence and gang prevention programs that may actually cause more crime to occur. Whether Congress is aware of these bottom-line results is unclear. What is clear is the absence of a systematic process for measuring the effectiveness of crime prevention programs. Some of the most popular crime prevention programs have already been tested - at great federal expense - and found ineffective. Yet their funding continues unabated. Military-style boot camps for juveniles, Drug Abuse Resistance Education (D.A.R.E.) Classes taught to schoolchildren by police officers, and neighborhood watch programs are but three examples of proven failure to which Congress gives millions of dollars annually. More effective ways of treating juvenile offenders and preventing drug abuse have been demonstrated by careful evaluation research. Teaching juvenile offenders skills like reading works better than boot camps. Making schools more firm and consistent in overall discipline works better than D.A.R.E. But these strategies lack political advocates and lose out in the competition for funding. The biggest danger from ignoring the bottom line of program evaluations is not wasted funding; it is wasted opportunities. Some of the most effective crime prevention strategies known to science are not even on the national agenda for funding, let alone in congressional appropriations for the Justice Department. Two of these focus on a widely recognized cause of crime problems: early childhood development in the home. In early infancy, home visits by nurses can produce major reductions in child abuse, which is a risk factor for later delinquency. Such visits are required in a number of Western nations. Hawaii offers it on a voluntary basis to all new parents, the majority of whom accept. Repeated visits for the first two years of life provide the greatest benefit to children at high risk. This is expensive, but the results are good. A recent RAND study suggests that home visits to infants may be more cost-effective than prison. In preschool years, Head Start programs with home visits by teachers can prevent delinquency up through age 25. This clear effect has been found in several controlled tests started in the 1960s. Yet today, few Head Start programs are budgeted for teacher home visits, and Head Start itself lacks sufficient funds to meet demand. The good news is that there is increasing evidence to support the largest single program in the 1994 Crime Act: putting 100,000 more police on the streets, funded at $1.4 billion in fiscal year 1996. National Institute of Justice studies show that an increased police presence reduces crime in high crime "hot spots," where most crime is concentrated. Full enforcement or "zero tolerance" policies in those hot spots may even reduce gun crime, the leading theory about New York City's free-falling homicide rate - now almost two-thirds lower than in the early 1990s. The bad news is that much remains unknown about strategies for deploying police and the cost-effectiveness of different alternatives. Sudden increases in personnel provide an unprecedented opportunity for controlled tests of policing strategies, especially in high-crime poverty neighborhoods where policing often backfires. Funding this program without a substantial investment in comparing the impact of hundreds of different approaches to community policing is more than a wasted opportunity. It is comparable to the Eisenhower administration giving schoolchildren many different polio vaccines without testing any of them for safety and effectiveness. The University of Maryland report recommends that 10 percent of all federal funding for local crime prevention be set aside for independent evaluations of the programs' impact, commissioned by an independent evaluation agency in the Justice Department. Several bills now pending in Congress use the same formula. Opponents of this idea say that evaluation money could be better spent on effective programs to fight crime. But until Congress invests more money in program evaluations, it will have no way of knowing which programs are effective.