Measuring Election System Performance

26 Pages Posted: 19 Dec 2010

See all articles by Stephen Ansolabehere

Stephen Ansolabehere

Harvard University - Department of Government

Nathaniel Persily

Stanford Law School

Date Written: December 17, 2010


The controversy surrounding the 2000 presidential election focused the country’s attention on questions concerning the quality of American democracy. That saga, however, is remembered in popular consciousness as one preoccupied with the question of voting technology. This is due in no small measure to the Supreme Court’s focus on the counting and recounting of punch card ballots, with their dangling and pregnant chads open to varied interpretations. For practitioners and scholars of election administration, however, the 2000 election represented a watershed event that exposed problems in the electoral system as a whole. Although we have made great strides as a nation in addressing the technological problems endemic to the 2000 election, we have made very little effort to evaluate the administration of elections in a systematic way. This article attempts to lay out what a system-based evaluation would entail.

The first Part of this Article discusses the difficulties in evaluating both election systems’ inputs and their outputs. Inputs are the components of election administration that will vary based on state and local law and practices, such as methods of vote tabulation and voter registration. A system’s output is the success with which it accurately records voters’ intentions and encourages voter turnout. Part II examines some of the sources from which we derive data concerning elections, such as surveys, audits, and recounts. It discusses some problems encountered when collecting data, such as discrepancies between different sources, and identifies a need for more systematic collection of data. The third Part goes into more depth about systems of measurement used to analyze today’s elections. It discusses the problems of how to calculate the size of the electorate, and how to determine “lost votes,” or those votes that were not counted due to tabulation error or lack of access to the election system. Part III also reports and analyzes 2008 election data from the Cooperative Congressional Election Study (CCES). The Article concludes by positing that while the country has seen substantial recent improvement in election administration, there still exists a need for higher quality data in order to further assess the election system.

Suggested Citation

Ansolabehere, Stephen and Persily, Nathaniel, Measuring Election System Performance (December 17, 2010). New York University Journal of Legislation and Public Policy, Vol. 13, No. 3, 2010, Columbia Public Law Research Paper No. 10-260, Available at SSRN:

Stephen Ansolabehere

Harvard University - Department of Government ( email )

1737 Cambridge Street
Cambridge, MA 02138
United States

Nathaniel Persily (Contact Author)

Stanford Law School ( email )

559 Nathan Abbott Way
Stanford, CA 94305
United States
9175703223 (Phone)

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Abstract Views
PlumX Metrics