There is a nice article in the new Political Research Quarterly (gated) by Shanna Reilly and Sean Richey on the relationship between “roll off” (voters not completing a ballot) and complex ballot questions.

The takeaway quote should be of interest to our readers in state legislatures and others working on state and local referenda and initiatives:

Our research suggests that an obvious policy solution to decrease roll-off is requiring ballot propositions to be written in a way that is easier to understand. It is already governmental policy for institutional review boards to require researchers to write informed consent forms in an easy-to-read manner. In a similar way, state election offices could require those writing ballot questions to write them in an easy-to-read fashion.

The paper has some very nice policy prescriptions that should be of interest to our readers in state legislatures.  The authors include examples of “easy” and “difficult” propositions, and provide a technology (the Flesch-Kincaid Grade Score) that can be easily tested using Microsoft Word.

For our academic readers, I had a few questions after reading the article, most importantly, what impact do complex ballot questions have on the composition of the electorate who do not roll off (e.g. do we end up with wealthier and more educated “initiative electorates”) and what impact might that have on what kinds of initiatives pass or fail?  The authors may already be investigating these questions.