Eight featured speakers offered their views on the role of basic and applied research in formulating and executing government and educational policy and policing strategies, and in furthering foundation objectives. They also looked at linkages to investigative reporting (see box) and advances in experimental methods.
Their bottom line may have best been expressed by Laurence Lynn, at the University of Chicago's School of Social Service Administration: "Given the sheer complexity of social problems and human behavior, research rarely settles anything. It often raises more questions than it answers."
Nonetheless, from all sides of the spectrum, speakers generally agreed that social science research can often effectively describe and interpret social conditions and trends, design and evaluate social interventions, provide useful analytic tools for administrators and practitioners, and positively influence public debate and public policy--if policymakers would only listen.
From a funder's viewpoint, "the strengths of social science lie in its ability to produce accurate and in-depth descriptions of social realities," said Paul Lingenfelter, vice president of the John D. and Catherine T. MacArthur Foundation. He pointed, for example, to recent research sponsored by MacArthur on the effects of domestic violence on women's employment. He said social science evaluations can also deepen understanding of simple replicable interventions, such as the effects of class size on learning, though he conceded that it is difficult to design such interventions with sufficient power to yield strong effects. Of paramount interest to MacArthur is the usefulness of research that helps improve practice.
Richard Nathan, director of the Rockefeller Institute of Government, believes that social science is most effective when it studies conditions and trends, conducts demonstration studies, and does evaluations. He offered as examples work done by the Manpower Demonstration Research Corporation, and the Urban Institute's current study of Family Well-Being.
Yet as the series proved, social scientists may be their own harshest critics.
In his talk, Nathan revisited his 1988 book, Social Science in Government: Uses and Misuses, and said it had an upbeat tone he wasn't sure he would support today.
For starters, he said the "three bad habits" of applied social science he described in his book are still valid criticisms: overspecialization, technical overkill, and a quest for precision about causality that is too much like the natural sciences. "Social science and social behavior are a special environment and we need to be very thoughtful about that," he said.
Though social science can be useful as an input for politicians, Nathan said he worries that "too much of the time social scientists doing relevant research act too much like politicians. This is role confusion." It may be impossible to achieve, he said, but social scientists should strive to be more impartial, objective, and neutral. "I am more pessimistic about social science behavior than the usefulness of it," he explained. "Social scientists do an awful lot of shouting. I don't know why that is our role."
Lynn's perspective was whether social science research can be useful for public management, and whether management even matters to governmental performance.
While there are many positive benefits of scientific knowledge, he said, there is relatively little research on managerial performance. What little exists is often overlooked and many other factors impinge on both the behavior of public managers and the formulation of policy.
Focusing on what he termed "the darker side of social science research usage," Lynn thinks that "beliefs, ideology, self-interest, and the exercise of political property rights drive policy in spite of the availability of good thinking and good data." He said policymakers are strongly influenced by political exigencies, such as the need to cut budgets and reduce deficits, as well as the power of money in a media age and the rise of single issue politics. And many other factors affect governmental performance as well, such as the availability of resources, mandated policy and program designs, and social structures of professional policy networks.
From a foundation standpoint, scientific knowledge is sometimes too abstract and probabilistic to be useful to practitioners in decisionmaking, Lingenfelter said. He suggested that more thought be given to "ways of formulating questions and getting connected to practice that will add more discipline and more analysis to decision making about social policy without at the same time imposing on social reality a mechanistic model that doesn't work in the real world."
A more upbeat and methodological perspective was supplied by William Shadish, professor of psychology at the University of Memphis and a visiting scholar at IPR this spring, who explained how social experimentation has made great strides over the past 30 years.
There has been real progress in dealing with selection bias and attrition, and in estimating treatment effects for dropouts from programs, he said. The development of meta-analysis has improved the ability of social scientists to generalize from experiments and they, in turn, have learned how to design social experiments that reflect the diversity of value positions in policy arenas. On the other hand, such experiments are time-consuming and costly, and their usefulness is frequently long-term and conceptual, rather than short-term and instrumental.
Among the problems still to be solved, Shadish said, are obtaining valid estimates of effects from nonrandomized experiments, and better estimates of treatment effects in randomized experiments where there is attrition from treatment. He also thinks there is still unresolved ambiguity about how the role of social experimentation ought to be structured into the policy sciences and into the governmental process. "
Experiments are no longer on the pedestal on which they once stood, but rather are now standing on the same ground with the rest of the many methods we use in the policy sciences," Shadish concluded. "It's a less ambitious and exalted position, but also a safer, more realistic, and more collaborative role."
Earlier speakers in the faculty seminar series included Wendall Primus, a former deputy assistant secretary in the U.S. Department of Health and Human Services,Wesley Skogan (IPR-Political Science), and James Rosenbaum (IPR-Education).