Key points are not available for this paper at this time.
Our research uses principal component analysis and information on 33 different state election laws, assembled in seven different issue areas, to create a Cost of Voting Index (COVI) for each of the 50 American states in each presidential election year from 1996 through 2016. In addition to providing detailed description of measurement and coding decisions used in index construction, we conduct sensitivity analyses to test relevant assumptions made during the course of index construction. The COVI reported in the article is the one that is the most theoretically sound and empirically indistinct from the other index construction options considered. We also test the construct validity of the COVI using both state-level and individual-level voter turnout. After controlling for other considerations, we find aggregate voter turnout is lower in states with higher index values and self-reported turnout also drops in states with larger index values.
Li et al. (Fri,) studied this question.