Tuesday, November 20, 2012

How well does the last election predict the next election on Guam?

As I am awaiting the certified election results for the 2012 Guam General Election, I have given some thought to what one could do to get a benchmark for expected performance (at least of incumbents and returning senatorial hopefuls). The easiest thing would be to assume stationarity: if you did well last time, you might expect to do well this time, all else equal.  I am very eager to get older data and the latest data to work with, but for now, I'll look at the previous three general elections: 2010, 2008 and 2006. So the question is how well does stationarity hold up?

For each of the following graphs, I constructed a line which simply takes the slope as given (1 because we are assuming stationarity) and an average of the deviation of each candidate from their position in the previous election as the y-intercept. Here is the result for the 2008 v. 2006 election:

What we see is that most of the candidates align pretty closely to the line that I constructed, with the biggest outperformers being Telo Taitague, Tina Muna Barnes and Ben Pangelinan. The biggest underperformers in 2008 were Mark Forbes and Jesse Lujan. Keep in mind, all I am saying is that they out- or under-performed where one would expect them to be if they had done roughly the same as in the 2006 election. If you notice, some Republicans fared worse and some Democrats fared better, which makes me think this might be about a Camacho-Cruz bump for Republican candidates in 2006. And Telo's standing is pretty easily explained by noting that she ran for Senator once before, in 2006, so she would have had greater recognition in 2008 relative to 2006. These are just hypotheses, though, and I wouldn't take them as definitive, by any means. More discussion will follow below. What does the 2010 election look like compared to the 2008 election?

Again, most candidates conform to the line I constructed (with less variability!), but the biggest outperformers were Tony Ada and Tina Muna Barnes. The biggest underperformers were Telo Taitague and BJ Cruz. Perhaps the evenly-matched performance of the two gubernatorial teams has something to do with how tight the line is, except the major outliers. There are a few things about BJ Cruz' performance that makes me want to consider him separately (off-island and Apuron). Just for fun, I'll look at how 2010 compared to 2006:

If you look casually at these and other years, it seems that, generally, candidates who run again for one of the next legislatures do better than in their initial performance. If you were to compare 2006 and 2010, you will find that Chris M. Duenas did much better than in 2006. It also looks like first-term senators have a lot more variability in their performance in their first re-election. Some do considerably worse, like Telo Taitague, some do considerably better, like Tony Ada. But most of the longer-term incumbents are more stable. It looks like gubernatorial elections might have an impact on performance. Keep in mind that my comparisons are just comparisons, they are not explanations, nor are they normative. There are many factors that can play a role in generating a certain outcome for a candidate. A lot of it is probably campaign logistics, like fund-raising, effectiveness of advertising, the amount of time and energy the candidate and volunteers spend on the campaign. But there is also a lot of other stuff that might matter: like the rumor-mill, media coverage, negative campaigning (it happens, as we recall from the 2004 election and the ads against then-Speaker Ben Pangelinan), etc. And don't forget more structural things, like demographics, some of which can be affected by get-out-the-vote efforts, but can also be affected by changes in the age, sex, ethnic, employment, income, religious characteristics of voters (and probably more I'm neglecting).

I became curious about how closely a previous election predicts a current election earlier this year, when I tried to develop a casual model. At the time, I came to the conclusion that BJ Cruz' 2008 General Election performance was an aberration, unlikely to be repeated in the 2012 General Election. That seems right. I felt unsure how to treat first-term senators, like Dennis Rodriguez, Aline Yamashita, Chris Duenas, Mana Silva Taijeron and Sam Mabini. Historically, it appeared that some first-term senators get a "bump" the next term and some get a "punishment".  I made assumptions about how much incumbents would be hurt by the fab 5 narrative, as unconscionably dishonest as it was. I was at a loss for how to treat non-incumbents who had never run before. Looking back, my numbers might have been wrong (I haven't checked, yet), but my view about the relative direction of individual incumbents seemed about right (although I still have to quantify it). And, it appears that performing at previous election-winning levels did not guarantee re-election this time. I'll follow up with results from the 2012 v. 2010 election soon.

Sort-of related posts:
Guam General Election Results & Exit Polls
Guam general election 2012
Memo about polling and Guam's election
A few thoughts on elections from the demand-side
Guam Legislature from a labor perspective and a rallying call
Old hands, new hands and flashbulbs
Comments on the Calvo administration's 'spending cuts' and the debt ceiling
Possible response from McNinch (different topic)
Lee Webber admits why he wants a part-time legislature
Functions of the Guam Legislature
A view of Guam's Primary election

No comments:

Post a Comment