My Thoughts on the Election
Well now. Did that just happen?
My involvement in several races kept me from blogging regularly during the campaign season. I'll try to get caught up on several aspects of the race, but for now, let me take just one: polling.
The media's fascination with political polls, any political poll, drove me nuts this campaign season. The polls dominated the coverage and clearly affected voter behavior. My wife was actually in the polling booth last Tuesday and heard a woman in the booth next to her tell her kid that she wanted to vote for Shawn Moody but didn't because the polls said he was way behind. And the curious 11th-hour retraction by Equality Maine of its endorsement of Libby Mitchell was the ultimate in strategic voting - reading the polls and deciding who had the best shot of beating Paul LePage.
All the stories about the various political polls became like experiments in quantum physics, where the very act of observing the experiment affects the outcome. But the polls were all over the place. Was there really a surge for Cutler, or was it mostly manufactured by the candidate, which then was reflected in later polls as voters made their decision on who had the best chance of winning, or of beating the front-runner? Who knows?
Editors and reporters need some schooling in Polling 101. They need to know what's a real poll and what isn't. And the media should adopt some basic standards for their use. For one thing, a poll conducted and released by a candidate should be ignored. Second, they shouldn't give coverage to any poll unless they have the complete poll - the questionnaire, the cross-tabs, the methodology, etc. There are a lot of little tricks you can do with polling to achieve the desired results, and unless you can see the way the questions were asked and who they were asked of, just reading the top line results in a press release can be misleading.
Comparing polls by different pollsters is also tricky since the methodology of each is different. One poll released in the final weeks of the campaign had Cutler trailing by just six points. Another, released a few days later, had him down by 20 points. Was the electorate really that volatile? Of course not. The difference was more a function of the methodology of each poll, which is why different polls can't be compared to detect a trend.
Instead of just using polls to cover the horse race, the media should rely more on polling to figure out if candidates are really talking about the issues that people care about. What do voters want from their candidates, and did they get it with this crop? And maybe they could spend more time on post-election polls to figure out why the candidate that nearly every major newspaper endorsed lost to the one that received no endorsements. Or maybe they don't want to know the answer to that one.
My involvement in several races kept me from blogging regularly during the campaign season. I'll try to get caught up on several aspects of the race, but for now, let me take just one: polling.
The media's fascination with political polls, any political poll, drove me nuts this campaign season. The polls dominated the coverage and clearly affected voter behavior. My wife was actually in the polling booth last Tuesday and heard a woman in the booth next to her tell her kid that she wanted to vote for Shawn Moody but didn't because the polls said he was way behind. And the curious 11th-hour retraction by Equality Maine of its endorsement of Libby Mitchell was the ultimate in strategic voting - reading the polls and deciding who had the best shot of beating Paul LePage.
All the stories about the various political polls became like experiments in quantum physics, where the very act of observing the experiment affects the outcome. But the polls were all over the place. Was there really a surge for Cutler, or was it mostly manufactured by the candidate, which then was reflected in later polls as voters made their decision on who had the best chance of winning, or of beating the front-runner? Who knows?
Editors and reporters need some schooling in Polling 101. They need to know what's a real poll and what isn't. And the media should adopt some basic standards for their use. For one thing, a poll conducted and released by a candidate should be ignored. Second, they shouldn't give coverage to any poll unless they have the complete poll - the questionnaire, the cross-tabs, the methodology, etc. There are a lot of little tricks you can do with polling to achieve the desired results, and unless you can see the way the questions were asked and who they were asked of, just reading the top line results in a press release can be misleading.
Comparing polls by different pollsters is also tricky since the methodology of each is different. One poll released in the final weeks of the campaign had Cutler trailing by just six points. Another, released a few days later, had him down by 20 points. Was the electorate really that volatile? Of course not. The difference was more a function of the methodology of each poll, which is why different polls can't be compared to detect a trend.
Instead of just using polls to cover the horse race, the media should rely more on polling to figure out if candidates are really talking about the issues that people care about. What do voters want from their candidates, and did they get it with this crop? And maybe they could spend more time on post-election polls to figure out why the candidate that nearly every major newspaper endorsed lost to the one that received no endorsements. Or maybe they don't want to know the answer to that one.
Posted by Dennis Bailey at 11/5/2010 1:09 PM 
Categories: uncategorized
Tags: Paul LePage polling elections Maine governor Shawn Moody
Categories: uncategorized
Tags: Paul LePage polling elections Maine governor Shawn Moody
Very good assessment. I agree.
Reply to this