Friday, November 2, 2012

Memo about polling and Guam's election

I am somewhat amused by today's editorial in the Marianas Variety (11-01-2012). Polls are quite useful in telling the current level of support of candidates and political issues, like Proposition A, the "for-profit" bingo monopoly initiative (not that it would disallow non-profit bingo). Dr. Ron McNinch indicated that he expects the Democrats to sustain a strong majority in the Guam Legislature (9-6 or 10-5), the re-election of Congressional Delegate Madeleine Z. Bordallo and the defeat of Proposition A.

The editorial cited its own online "poll" results on the congressional race. I know all three of these candidates. When I was a student at John F. Kennedy High School, Lieutenant Governor Bordallo helped an organization I was involved in to get permission for an off-island conference to Seoul, Korea, despite unreasonable opposition by a school administrator. I have had several very positive interactions with Senator Frank F. Blas, Jr., and like him a great deal. I have met Jonathan Diaz a few times and we agree on quite a few things (and he is a blogger, too). What I am about to say is not about the candidates, just the online polling.  I am sure the editors of the Marianas Variety staff are as suspicious as I am with the results, although I would not suggest that a candidate had recruited volunteers to flood the poll. As reported in today's editorial: "Sen. Frank Blas Jr., the Republican candidate, is way ahead. He’s received 688 votes, or 63.5 percent, to Delegate Bordallo's 374, or 35.6 percent. Ten voters said they will vote for someone else, presumably Jonathan Blas Diaz, the write-in candidate." The biggest problem with online polls is that they are self-selecting. I recall one case very clearly where in 2010 there was an online poll on about the gubernatorial race, where Calvo-Tenorio, Cruz-Espaldon and Gutierrez-Aguon were listed and the Calvo-Tenorio camp got over 200,000 votes! That was pretty hilarious. On a less contentious poll the result is usually like 200-300 votes, total. This one had more people voting than live on Guam (even by the old estimate of 180,000). That was pretty impressive, but somewhat undermines the purpose of skewing a poll by organized voting, since the PDN reports the number of votes cast in their polls. Generally speaking, that level of manipulation does not occur. But even if there isn't manipulation, it can still end up with a skewed result without the candidate getting involved. After all, there is already a profile of the readers of an online newspaper which probably does not match the population of voters. Add to that the fact that readers have to take interest in looking at the poll question and voting. When I go to the PDN or Mariana's Variety website, I go for the news and the opinion pages, not the online surveys.  Then there is the chance that those who are interested in the polls may "recruit" others to vote and that can skew the results even more.

I have considerably more confidence in the polling by Dr. McNinch, although any method is bound to have its limitations. The sample size might be an issue, as well as the weighting of the results and accounting for different demographic characteristics. I've seen the results of a professional poll conducted on Guam in 2006 which I was somewhat disappointed by. The sample size was 444 and it was conducted as a telephone poll. The poll results had about 20% more females than males, and over-represented higher income groups. I think pretty much everyone can remember the 2006 poll which published results for the race between Underwood and Camacho. Underwood was going to win! Except that he didn't, by which I was very disappointed.

Unfortunately, there are not a vast number of polls conducted on Guam shortly before the election, to give us a nice snapshot of what different methods would predict. One might misinterpret to think that I am criticizing the pollsters. On the contrary, I think what they do is important and serious work and can have real consequences, depending upon how polling is used. Thank you Dr. McNinch for your consistent efforts to keep public polling alive on Guam. As far as the uses of polling, I would like to see considerably more on the support or opposition of the voters to particular policies or the constellation of policies which we consider liberal/progressive or conservative.

Best of luck to all those who have put themselves forward to serve the people of Guam in the 32nd Guam Legislature and in the halls of Congress. And I sincerely hope that Dr. McNinch is correct about the untimely demise of Proposition A.

Sort-of related posts:
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