Monday, April 04, 2016

Guest post: Dr. Klumpp analyzes the 2016 judicial primary

by Albert J. Klumpp

First and foremost, last month's Democratic primary set a new record for judicial votes cast. Unofficially, the median number of votes cast in the eight countywide contests was 836,222. As illustrated below, this total outdistanced the previous high of 795,418, set in 2008, and marked only the third primary with more than 600,000 judicial votes since partisan judicial primaries debuted in Cook County in 1974. Not surprisingly, the three--1984, 2008 and 2016--were headlined by three of the most spirited and competitive Democratic presidential primaries over that time span. Last month's high voter turnout and the extra judicial votes it generated had a huge impact on the results of both countywide and subcircuit contests.

Using the procedure that I developed for studying groups of nonpartisan or party-primary judicial contests, I analyzed the results of the eight countywide contests. The small size of the data set -- only 23 candidates in eight contests (and one contest distorted by a late withdrawal) -- was a potential problem. But for the most part my procedure produced useful estimates of the impact sizes of the factors that historically have the strongest influences on judicial voting. I must emphasize that the individual estimates are just that -- estimates -- and have margins of error of various sizes. However, they are very consistent with estimates from previous primaries having larger data sets, and as a group they "fit" the data set quite well.

Generally the impact sizes correspond closely to those of the 2008 primary. Above all, the impact of gender, on two primary ballots topped with a female presidential candidate (and the same candidate at that). In 2008 gender conveyed an advantage of roughly 23 percentage points per contest, to a single female or divided among multiple females. The advantage this year was in the 20-22 point range. Gender typically is the most influential factor in the county's judicial contests, but it also is the most variable and has been as low as 6 percentage points. Its impact this year was one of the largest impacts ever of any factor in any single primary.

Of course, when one factor increases in influence, at least one other must decrease. Last month the losers were bar/newspaper recommendations and party slating. Recommendations are typically used by voters who turn out reliably and who have the greatest impact when turnout is low. True to form, a candidate in last month's primary who held both a Tribune endorsement and the highest bar ratings in his or her contest gained only a 6 point advantage. This is only around half of the long-term average value. (The refusal of the Sun-Times to endorse judicial candidates also played a role, as it did two years ago.)

As for slating, it conveyed an advantage of only 4-5 points. Notably, this is the one major difference between last month's primary and 2008's; in the latter, slating was roughly twice as valuable. A common misconception about slating is that without exception, its impact varies inversely with voter turnout. That is not the case; turnout does affect it but so do other important factors such as the mood of the electorate and the popularity of the party and its leaders on election day.

Another common misconception is that the Irish-name advantage long observed in the county's judicial contests behaves like gender, rising and falling in lockstep with turnout. In fact, it is surprisingly stable from year to year and much less variable than gender. Its long-term average has been 9-10 percentage points; its value in 2008 was roughly 11 points; and last month was roughly 13 points.

The one historically important factor that I could not estimate satisfactorily is ballot position. The data set had too few contests to allow it. However, the clearest evidence of its impact is seen in the two-person contest for the Walsh vacancy. Fredrick Bates, the party-slated and Tribune-endorsed candidate, was narrowly defeated by Patrick Powers, who held not a single advantage over Bates besides the first ballot position. In recent years the value of the first ballot position had increased to the range of 7-8 points, and in eyeballing last month's result a similar impact is very plausible, if not provable.

As for the subcircuit contests, a single group of them cannot be analyzed in the same manner as above because of significant differences in impact sizes among different subcircuits. (Also, full information on campaign spending is not yet available, and unlike in the countywide contests campaign spending has a significant effect on subcircuit contests.) However, a simple accounting shows the primacy of gender across the entire county. Including the subcircuits, there were eighteen contests in which at least one male candidate faced at least one female candidate. Fourteen of those contests were won by females. Two others had a lone female with a surname of an ethnicity that has been historically disadvantageous. The remaining two were in the 5th Subcircuit, where gender has been slightly less influential than elsewhere in the county and where last month's surge in judicial votes was the smallest of all of the subcircuits holding judicial contests.

As a basis for comparison, slated candidates and Tribune-endorsed candidates had success rates of less than 50 percent in all contests. In short, this was a very good year to be a woman running for judge in Cook County.

Albert J. Klumpp has been a generous and frequent contributor to this blog over the years. A research analyst with a public policy PhD, Klumpp is the author of several scholarly works analyzing judicial elections including Judicial Primary Elections in Cook County, Illinois: Fear the Irish Women!, 60 DePaul L. Rev. 821 (2011); "Voter Information and Judicial Retention Elections in Illinois," 94 Ill. B.J. 538 (October 2006); and "Cook County Judicial Elections: Partisanship, Campaign Spending, & Voter Information," CBA Record, January 2007 (p. 34).

No comments: