IN THIS AGE OF SCIENCE, SCIENCE SHOULD EXPECT TO find a warm welcome, perhaps a permanent home, in our courtrooms. The reason is a simple one. The legal disputes before us increasingly involve the principles and tools of science. Proper resolution of those disputes matters not just to the litigants, but also to the general public—those who live in our technologically complex society and whom the law must serve. Our decisions should reflect a proper scientific and technical understanding so that the law can respond to the needs of the public.

Consider, for example, how often our cases today involve statistics—a tool familiar to social scientists and economists but, until our own generation, not to many judges. In 2007, the U.S. Supreme Court heard Zuni Public Schools District No. 89 v. Department of Education,1 in which we were asked to interpret a statistical formula to be used by the U.S. Secretary of Education when determining whether a state’s public school funding program “equalizes expenditures” among local school districts. The formula directed the Secretary to “disregard” school districts with “per-pupil expenditures…above the 95th percentile or below the 5th percentile of such expenditures…in the State.” The question was whether the Secretary, in identifying the school districts to be disregarded, could look to the number of pupils in a district as well as the district’s expenditures per pupil. Answering that question in the affirmative required us to draw upon technical definitions of the term “percentile” and to consider five different methods by which one might calculate the percentile cutoffs.

In another recent Term, the Supreme Court heard two cases involving consideration of statistical evidence. In Hunt v. Cromartie,2 we ruled that summary judgment was not appropriate in an action brought against various state officials, challenging a congressional redistricting plan as racially motivated in violation of the Equal Protection Clause. In determining that disputed material facts existed regarding the motive of the state legislature in redrawing the redistricting plan, we placed great weight on a statistical analysis that offered a plausible alternative interpretation that did not involve an improper racial motive. Assessing the plausibility of this alternative explanation required knowledge of the strength of the statistical correlation between race and partisanship, understanding of the consequences of restricting the analysis to a subset of precincts, and understanding of the relationships among alternative measures of partisan support.

In Department of Commerce v. United States House of Representatives,3 residents of a number of states challenged the constitutionality of a plan to use two forms of statistical sampling in the upcoming decennial census to adjust for expected “undercounting” of certain identifiable groups. Before examining the constitutional issue, we had to determine if the residents challenging the plan had standing to sue because of injuries they would be likely to suffer as a result of the sampling

1. 127 S. Ct. 1534 (2007).

2. 119 S. Ct. 1545 (1999).

3. 119 S. Ct. 765 (1999).



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement