Helping Children in a Downsizing World

"The challenge is to take the array of forces at our disposal and put them to work in a downsizing environment--to actually change welfare systems for the better, not just hack away at benefits and services. In short, we must learn how to do more with less."

--Nick Bollman,

James Irvine Foundation, to researchers at a RAND conference on the new federalism, May 1996.

That imperative--to do more with less--has led RAND to look not only at massive welfare programs like Aid to Families with Dependent Children, but beyond them at a broad array of other less famous (and less studied) efforts. The search has been for programs and strategies that not only help children but that may have outsized payoffs for the dollars invested. These include investigating novel or overlooked approaches to the problems of children and exploring untapped resources in neighborhoods, communities and the family itself.

Among many other child-focused projects, in recent years RAND has carried out major, scientifically rigorous studies of the effectiveness of school-based drug prevention programs and of the influence of the family on children's school performance.

Grandparents as the 'second line of defense'

A multi-site drug prevention experiment funded by the Conrad N. Hilton Foundation proved successful in curbing marijuana and tobacco use among seventh and eighth graders. The two-year program yielded positive effects in highly diverse environments, including urban, suburban, and rural communities and in schools with both high and low minority attendance. After the lessons ended, however, the benefits of the program eroded, suggesting the need for continued prevention efforts during high school. Reinforcing the anti-drug message in high school via "booster" programs could further delay for many the initial use of these substances, the researchers said. And that would be no small victory: Children cannot be protected forever, but postponing initiation until after high school buys them time to develop more mature judgment and skills in resisting the multifarious pressures to use drugs.

The student achievement study found that, contrary to mainstream opinion, student test scores are rising for all students--and rapidly, in the case of minorities. From about 1975 to 1990, the average math and reading scores of students 13 to 17 years old increased 3 percentage points for white students, 11 points for Hispanic students and 19 points for African American students. A major part of the explanation is improvements in the family environment, notably, smaller family size and a dramatic increase in the education levels of African American mothers.

New efforts are planned or under way in other areas: studies of programs that seek to improve parenting skills by involving inner-city parents in their children's schooling; studies of the role of grandparents as the "second line of defense" in protecting and caring for their grandchildren; and a large-scale investigation of every aspect of neighborhoods that may be important in the lives of children--from schools, churches, libraries, recreational programs, clinics and hospitals to the social, economic and cultural characteristics of neighborhood interactions. (Also see the box accompanying this article for an account of RAND's involvement in the "Early Childhood Public Engagement Campaign.")

Perhaps the most promising news comes from a recent study of strategies aimed at keeping high-risk youth out of trouble with the law. A RAND research team found intriguing evidence that crime might be reduced more cost-effectively by these strategies than by the longer prison sentences that are currently in vogue.

Dollar for dollar, the analysts found, programs using financial and other incentives to induce disadvantaged high school students to graduate avert five times as many serious crimes as the stiffer prison terms stipulated by California's "three-strikes" law. Programs that provide parental training and therapy for families whose children have shown aggressive behavior in their early school years avert almost three times as many serious crimes.

RAND's earlier study of the effects of the three-strikes law found that full implementation will produce a 21 percent overall reduction in crime at a cost to the state of an additional $5.5 billion per year in spending on the criminal justice system, notably for prison operation and construction. The new work indicates that a combination of graduation incentives and parent training could achieve a similar amount of crime reduction for less than $1 billion.

"None of this suggests that incarceration is the wrong approach," emphasize authors Peter W. Greenwood, Karyn E. Model, C. Peter Rydell and James Chiesa. The point, they explain, is that policies that allocate huge sums to imprison career criminals and little for targeted efforts to discourage such careers are "lopsided."

Directed by Greenwood, the team analyzed numerous pilot programs employing four different approaches.

• Home visits by child care professionals beginning before birth and extending through the first two years of childhood, followed by four years of day care.

• Training for parents and therapy for families with young children who have shown aggressive behavior in school.

• Four years of cash and other incentives to induce disadvantaged high school students to graduate.

• Monitoring and supervising high-school-age youth who have exhibited delinquent behavior.

A preliminary assessment of these programs found that home visits reduce crime by 50 percent; parent training, by 60 percent; graduation incentives, by 70 percent; and delinquent programs, by 10 to 20 percent.

However, in the "real world" these benefits can be expected to decrease. The RAND team subtracted a percentage for relatively new and untested programs and another percentage for the decay of effects over time. Another factor is how well the programs target the population most likely to eventually commit crime. Not surprisingly, the later the intervention--such as with graduation incentives and programs aimed at delinquents--the better the targeting.

Weighing all the factors, the analysts estimated the number of serious crimes prevented per million dollars spent for each program. The options with the most immediate crime-reduction effects appear to be parent training and graduation incentives.

Greenwood and his colleagues are cautious about their findings, pointing out the need for better data about the results of such programs when they are implemented on a much broader basis. However, their analysis showed that the cost-effectiveness advantage of the best of these programs will hold up even if further study reveals somewhat weaker program results than those reported here.

Large-scale, multimillion-dollar demonstrations of these promising programs "would be an investment worth the cost," the authors conclude.

Diverting Children from a Life of Crime: Measuring Costs and Benefits, Peter W. Greenwood, Karyn E. Model, C. Peter Rydell, and James Chiesa, RAND/MR-699-UCB/RC/IF, 1996, 82 pp., ISBN: 0-8330-2383-7, $15.00. The text appears in its entirety on the World Wide Web (/publications/MR/MR699/). The research is summarized in a one-page brief, RB-4010. "Drug Prevention in Junior High: A Multi-Site Longitudinal Test," Phyllis L. Ellickson and Robert M. Bell, Science, Vol. 247, pp. 1299-1305, 1990. "Preventing Adolescent Drug Use: Long-Term Results of a Junior High Program," Phyllis L. Ellickson, Robert M. Bell, and Kimberly McGuigan, American Journal of Public Health, Vol. 83, No. 6, 1993, pp. 856­861 (RAND reprint, RP-208, no charge). The studies described in this article were supported by the Conrad N. Hilton Foundation, the Lilly Endowment Inc., the James Irvine Foundation, and RAND.