CBA Record November 2018

One of the stranger aspects of our state’s electoral system takes place in November of every odd-numbered year. It begins on the first day of the candidate filing period, when candidates fromacross the state file their paperwork immediately upon the opening of the state elections office. Then roughly a week later, a smaller group of different candidates arrives at the office and presents their filings in the final hour before the filing period closes. Their purpose? To seek a specific position on the ballot.

C ANDIDATES WHO FILE AT THE BEGINNING AND end of their filing periods are placed into random drawings for the opportunity to be listed first or last, respectively, in their primary election contests. The first and last ballot positions have long been believed advantageous in nonpartisan and party primary contests, especially for lower-visibility offices like judgeships. And both are highly sought by judicial candidates here in Cook County—particularly the first position. Over the past ten years, according to state elec- tion records, more than seventy percent of Cook County judicial candidates filed at the very beginning of the filing period, and another ten percent filed in the final hour. Is there any point to this ritual? Do these ballot positions have any actual effect on the results of elections, and if so, is it substantial enough to justify their pursuit? And should all of this be accepted as part of our electoral system in the first place? A close look at the evidence provides some revealing answers. Past Research Ballot position was an under-researched subject for many years among political scientists. Today the subject is still not fully under- stood, but gradually a wider variety of research has accumulated spanning many different jurisdictions and types of elections. Nearly all of the published studies on the subject have found evidence of a first-position advantage of up to several percentage points. As would be expected, the size of the advantage varies; it tends to be larger in nonpartisan contests and in lower-visibility contests for which voters typically have little or no specific infor- mation about the candidates. The last ballot position has received less research attention; no significant evidence yet exists to support the idea of a last-position advantage. My own research on Cook County judicial primaries examined

the first ballot position, and to my knowledge was the first to do so for any sort of judicial election. In a study summarized here in the CBA Record in 2010, and published in full in the DePaul Law Review the following year, I measured the impact of ballot position and many other factors on the results of countywide and subcircuit primary contests for Cook County judicial vacancies. For countywide Democratic contests between 1986 and 2010 I calculated an overall advantage of roughly 4.5 percentage points for first-position candidates, which is one of the highest figures found in any ballot position research. I also found evidence sug- gesting that the advantage was increasing over time, although the timespan of the data set was not long enough to be certain. Among subcircuit contests I found generally similar numbers but also evidence that the first position is less valuable in areas with high average educational levels. Unfortunately I could not con- clude much more than that, because the subcircuits are so diverse politically and demographically that they cannot be aggregated for analysis, and because individual subcircuits have had too few contests to produce sufficiently large data sets. A Closer Look Prior to the March 2017 primary election, the subject of ballot position was raised in the well-known For What It’s Worth legal blog. The blog material prompted me to take a more focused look at ballot position effects, and in particular to try to determine whether or not there was anything more to be discovered beyond the previously known first-position advantage. I took two approaches to the task. First, I reran my previous analysis using a larger, updated data set and added a separate series of calculations to evaluate the last ballot position as well as the first. The expanded data set covers all twenty primary elections between 1980 and 2018 and more than 500 individual contests.

CBA RECORD 27