Bruce Spencer Professor of Statistics

Research Interests

I study the production and use of public statistics, particularly in policy-laden contexts. For example, should an organization (such as the Federal government) spend more money (or less) on statistical programs such as the decennial census? Such a question requires an interdisciplinary attack, and I have focused on questions such as (i) how are the data used, (ii) what is the quality (including accuracy, relevance, timeliness, etc.) of the data, (iii) how would changes in the quality affect uses, and (iv) how can we assign measures of value to those effects. I work actively with government agencies on major statistical programs and conduct related research in sampling theory and methods and in demographic estimates. For many years I have worked with the Census Bureau on how to estimate population and how to evaluate the accuracy of their estimates. More recently I have worked on justice statistics, developing methods to estimate the accuracy of verdicts in criminal trials when the truth is unknown; I served on the recent National Academy of Sciences Panel to Review the Programs of the Bureau of Justice Statistics.

Recent Publications

  • Estimating the Accuracy of Jury Verdicts”, Journal of Empirical Legal Studies 4, 305-329, 2007.
  • Statistical Demography and Forecasting. (with Juha M. Alho). New York: Springer, 2005.
  • Statistics and Public Policy. (editor) Oxford: Oxford University Press, 1997.
  • Benefit-Cost Analysis of Data Used to Allocate Funds. New York: Springer, 1980.