Meta-Analysis in Domains Where Many Studies Are Proprietary: Voter Turnout

• In Event: The Problem of Knowledge Cumulation

Abstract

In principle, voter turnout research is well suited to meta-analysis and other statistical methods that aim to extract insights from research literatures. Most studies of voter turnout in the past decade randomly assign interventions and rely on administrative data to measure outcomes; thus, the prevailing research design is far more standardized than in other substantive domains, and the resulting estimates are expressed in comparable units. Moreover, academic researchers who study voter turnout have been early and earnest adopters of pre-registration, which lessens the threat of publication bias. However, an increasing share of the research literature in this domain is conducted by consultants or staff working within organizations that do not register and disclose all of the studies they conduct. This paper reflects on the special challenges that arise when researchers attempt to summarize research literatures that are only partly visible due to selective reporting by proprietary research operations.

Authors

• Donald P. Green, Columbia University

• Thomas Leavitt, Harvard University