Saturday, April 21, 2012

Dr. Klumpp replies to Mr. Strom

For those who came in late: Today's guest post, by Albert J. Klumpp, PhD, is the third in a series that began with Dr. Klumpp reviews judicial countywide results. Next followed Mr. Strom responds to Dr. Klumpp, a guest post from reader Michael A. Strom. I am grateful to both contributors. I think their discussion has been stimulating and enlightening and I hope readers may agree.

Having performed both qualitative and quantitative research on judicial elections, I can attest that both approaches have their strengths and produce useful insights. So I'm always sorry to hear of anyone shunning one or the other. But many people do, and I respect Mr. Strom's being honest and upfront about his views on the matter. With regard to his comments about my work and my conclusions, I'll gladly address them.

One general comment first. Mr. Strom correctly points out that every primary has unique characteristics. Unquestionably, those characteristics do affect the importance of some variables from one primary to another. For instance, a summary of my analyses for the 2008 and 2010 primaries was posted in this blog two years ago and is still accessible; it shows the substantial difference in the pro-female vote that led to the different success rates for female candidates that Mr. Strom mentions. The same can be said for the values of slating in 2010 and 2012, and in many other situations. As I wrote in the DePaul Law Review last year,* "evidence scattered throughout the thirteen elections [that I analyzed] supports the notion that the idiosyncrasies of both the electorate and the top-of-ballot contests in each individual election can affect the results of judicial contests."

The longer-term measurements in the table that Mr. Leyhane posted last week (for 1986-2010 and 2000-2010) are simply for bigger-picture comparison to the 2012 results and aren't intended to compare one specific year to another specific year. I suppose a good idea in the future is to put a footnote to that effect below those kinds of tables. The alternative is to turn the table into a long string of individual-year measurements, which solves that problem but creates others. In any event, my intention was definitely not to suggest that all primaries are identical.

On the specific topics Mr. Strom addresses:

Dilution: Dilution occurs regardless of whether or not a given candidate is a "ringer." A second legitimate female candidate in a contest will divide the pro-female vote just as a ringer female will. The fact that more female candidates won in 2008 than in 2010 was due to a larger pro-female vote in 2008, as noted above, and was not because any of the females in either year were or were not ringers.

As for ringers specifically, at least in countywide contests, I haven't seen solid evidence of their use since the year 2000. There may have been a few scattered throughout the subcircuits, but again, this is beside the point. The subdividing of advantages like gender or ethnicity is clearly established in our electoral history. It predates the use of ringer candidates and is why ringer candidates were employed in the first place.

Faithful committeemen: With regard to Epstein and Patti in 2010, Epstein had advantages that Patti did not. Namely, the Tribune's endorsement and his half-million-dollar campaign that ran TV ads. And so Epstein got more votes than Patti. Remember, all of the variables present in each contest have their own effects, so all must be considered together.

It's true that slating did not do a lot for Judge Hill-Veal in 2010. It didn’t do a lot for any slated countywide candidate. Again, looking at my posted analysis for that year, the low value of the slating variable stands out. Actually, though, Hill-Veal's 2010 contest is not a good choice for any sort of comparison because it was overwhelmingly decided by one very unusual factor: name recognition. This was the contest that put Aurelia Pucinski on the Appellate Court and was the third contest that she had won strictly based on a large name-recognition boost that drew many votes away from her opponents. A politically well-known name in a judicial contest is an uncommon event, but when it happens, it usually overwhelms the other variables. (Unless a Mary Jane Theis comes along with a seven-figure campaign.)

Educated voters: I was puzzled by the comments pertaining to voter education. Possibly my use of the word "unless" in the original post was a bad choice, since it may have suggested that I advocate educational programs or a greater use of one source or another of voter information. Whatever the reason, I can assure Mr. Strom that nothing could be further from the truth. I do not seek and have never sought to perpetuate the views of "good government types" or anyone else in the judicial selection debate. I have never "lauded" any group of voters and have always been scrupulously neutral about comparing voting strategies or, for that matter, judicial selection systems. I present the evidence and let it speak for itself. To whatever extent certain individual findings support one viewpoint over another, so be it. But I approach the topic with complete neutrality.

On the size of the informed-voter group: Yes, the better-informed judicial voters are more regular voters than others, but even in small-turnout years, it's simply not a large group. It never has reached more than around twenty percent or so in any primary or general election since 1980—turnout level notwithstanding. (The statement that I have "entirely ignored the salutary effect of an exceptionally low vote total" is incorrect; see my published material.) The larger implications of those small numbers can and should be debated; that they are in fact small numbers is well-established.

Ballot position: Other published studies have demonstrated that the first ballot position has positive value and can affect the outcome of electoral contests. The finding in my own research is hardly groundbreaking. Nor is it news to the many political candidates who line up outside of elections offices as soon as a filing period begins simply for a chance at the first position.

With regard to the two contests I cited: Karen O'Malley, who held the first position, defeated Stanley Hill by 1.4% of the vote. If Hill had won the ballot lottery instead of O'Malley, his vote percentage would have been higher and hers would have been lower. Very simple. Same with Jesse Reyes; his margin of victory was small enough that if the first-position advantage had shifted to Boyd or Flannigan, his percentage would have decreased, and his opponent's would have increased, by enough that he would have lost.

The fact that some first-position candidates are defeated doesn’t mean that ballot position is worthless; it merely means that their first-position advantages were trumped by other variables. Pamela Hill-Veal, listed first, was edged out by Mat Delort, who was slated, endorsed by the Tribune, and had top bar ratings. Patrick Sherlock, listed first, ceded other advantages to his two opponents -- gender, bar ratings, Tribune endorsement, slating, even had to share his Irish-name advantage with one opponent -- and so he lost to both of them. Without the first position, he would have lost by a greater margin. To reiterate, all variables must be considered together and not simply in isolation.

As far as why the first position draws extra votes, theories have been offered but the question is still unanswered. It's a fascinating question, no doubt about it. But whether or not the underlying mechanism is understood yet, the evidence clearly shows that the first position is an advantage for the candidates who possess it. And as such, it must be accounted for.

One final comment. I’m very grateful for all of the work that Mr. Leyhane does to cast light on judicial elections and for his interest in my research. And I always appreciate people like Mr. Strom who take enough of an interest in the subject to look for potential problems with my work. Public review and critical evaluation is what makes research better in the long run. But claiming that I "hate" the votes cast by some voters, and implying that I have some sort of anti-democracy desire to see certain people not vote at all... that’s going a little overboard, to put it kindly. My respect for the electorate is part of the reason that I dedicated myself to the judicial-selection topic in the first place; otherwise it isn’t the issue at all and it shouldn’t have been questioned. Let’s keep things civil and stick to the matters at hand.

Now, on to retention season!

* Klumpp, "Judicial primary elections in Cook County, Illinois: fear the Irish women!" DePaul Law Review 60:3 (Spring 2011), pp. 821-847.

No comments: