Empirical Research for Public Policy: With Examples from Family Law

Authors

  • Richard Lempert

    Corresponding author
    1. University of Michigan
      *2433 13th Court N., Arlington, VA 22201; email: rlempert@umich.edu. Richard Lempert is the Eric Stein Distinguished University Professor of Law and Sociology Emeritus at the University of Michigan and Research Deputy in the Human Factors/Behavioral Sciences Division of the Science and Technology Directorate of the Department of Homeland Security (DHS).
    Search for more papers by this author

  • This article was not produced in connection with the author's work at the DHS, and no statements or opinions in the article should be attributed to the DHS.

*2433 13th Court N., Arlington, VA 22201; email: rlempert@umich.edu. Richard Lempert is the Eric Stein Distinguished University Professor of Law and Sociology Emeritus at the University of Michigan and Research Deputy in the Human Factors/Behavioral Sciences Division of the Science and Technology Directorate of the Department of Homeland Security (DHS).

Abstract

Drawing on three family law studies as examples, this article discusses strengths and weaknesses of policy-relevant empirical research. Its main message is that policy-relevant empirical legal research should be encouraged, but the policy relevance of research, especially single studies, should not be oversold. It concludes with five points that consumers of policy-relevant empirical research should keep in mind: (1) do not rest policy change or analysis on a single study, no matter how good it is; (2) when reading the report of an empirical study, look beyond the researcher's bottom line to other relationships revealed in the data; (3) no matter how unversed one is in statistics, commonsense and a close reading of tables, graphs, and methodological narratives can take one a long way; (4) always ask about mechanism: understanding why a situation exists is as important to policy analysis as knowing whether it exists; and (5) if results seem too good to be true, this is often because they are not true.

Ancillary