I just finished reading the Hall, Monson, and Patterson piece cited below by Mike.  It’s a nice, multi-state, multi-time point treatment of poll worker performance and how it affects voter confidence, but the paucity of the results at the end raises a concern for me.  What is “voter confidence”?

Hall et al. show that perceptions of poll worker performance predict voter confidence, and also display an array of demographic predictors of poll worker performance.   Lower educated, lower income, and younger voters all give their poll workers lower job ratings.

Here’s what causes me concern–these same categories predict lower trust in government and lower levels of interpersonal trust.  What if voter confidence is not related to the polling place or to election technology at all–but is instead part of a constellation of attitudes related to social and interpersonal trust?

If this is true, then we should see precisely the patterns that we’ve seen in many studies, a relatively paltry relationship between voter confidence and the things we think should matter (voting machines, long lines, poll workers).

What I’d like to see is one of these studies include some of these broader indicators and set these head to head.  Then we might discover that repairing election administration–while important for many reasons–is unlikely to translate into high levels of voter confidence.  Instead, voter confidence will rise and fall along with trust on government, trust in the presidency and congress, and the generalized national mood.