International bloggers tackle polling

For many of us, scanning poll results have become a fixture of politics. The media has long been criticized for covering elections much like a horse race instead of concentrating on issues and probing the mind of voters. Candidates often complain about the polls, too, especially when those results show them trailing. Senator John McCain, who is behind in many polls, told Meet The Press that he trusts his senses and instincts as well as quantitative data. Obama supporters worry that their candidate’s leads in battleground state polling could result in voter complacency and keep many potential people from braving long lines to get out and vote come election day.

There’s also the old adage that polling is an art, not a science. Some social scientists argue that polling is now skewed by technology, where a cellphone dependent population are often not contacted by poll takers.

With all these negative impressions of polling, it has a long history that has become an ever present part of U.S. elections. If polls are not to be trusted, or worse, prove to skew votes and disenfranchise voting blocks, why do we digest their results so ravenously? One theory, I suggest, is that many people aren’t undecided voters and are tired of hearing last minute policy appeals from either candidate. Polls give people who have a horse in the race a sense to look at an election as sort of a game, or — dare I say? — a war with its own series of battles, fronts and retreats.

Here’s a general condemnation of polls from Paul Cotterill in West Lancashire, England who blogs at the Bickerstaffe Record:

…polls are supposed to reflect views, not drive them, aren’t they?  So from the perspective of one (ie. me) who actually values democracy, I’m always a tadge wary about extolling their validity, let along their virtues.

From Russia, here is a discussion on the issues facing pollsters this election cycle by Serge Markov at I Putin:

Polling has always mixed art and science: at least one of those polls this week, and perhaps both, is wrong. During the primary season, especially on the Democratic side, many pollsters missed their targets by a mile, for example by predicting that Mr Obama would win comfortably in New Hampshire, only for Hillary Clinton to triumph.

This year has presented unusual challenges, largely because of Mr Obama. One difficulty is turnout. Traditionally, pollsters know that youngsters are less likely to turn out, and blacks too. But the youthful and black Mr Obama has brought both groups out in large numbers, getting them to register, to attend rallies and, during the primary season, to cast votes. Pollsters struggled to keep up. For example, in the black-heavy states of the south, Mr Obama was expected to win comfortably. Instead, he won overwhelmingly, and the big margins brought him crucial delegates.

Then there is the much-discussed “Bradley effect.” Many experts still think that white voters will tell pollsters that they might vote for Mr Obama, but secret racism, concealed for embarrassment, will have them voting for Mr McCain instead. No such across-the-board effect could be seen in the primaries, as, in most states, he outperformed the polls. Where he struggled in the primaries, however, could be important for the presidential election: he did worse than his polls in Rust Belt and Appalachian states. A few of those remain in the balance; Ohio, in particular, broke heavily away from Mr Obama in the primaries, even more so than polls predicted.

Pollsters are also flummoxed by the question of how to count voters who use only mobile phones. Most polling is done over fixed-line phones, but a growing number of Americans have only a mobile telephone. These tend to be younger, technologically savvy and urban types, who are probably inclined to favour Mr Obama. Although pollsters try to correct for this in their models, it is impossible to know how significant the problem is, as phone-owning habits are changing fast. More polls do include a sampling of mobile-phone respondents and these seem to show a marked leaning towards Mr Obama. But this does not mean that those pollsters have the right mix of phone-owners in their sample.

A last factor is turnout operations. A candidate may lag in the polls, but have a crack organisation that will be more successful at getting its supporters to the polling station on election day. George Bush is widely credited with a powerful machine that helped him to win a close race in 2004, when many factors (including a worsening war in Iraq) weighed him down.

(And here is a post on the role race and the Bradley Effect may play in this year’s Presidential election.)

Another issue is that all polls are not created equally. Roger Darlington, from the United Kingdom, shows why we must be a bit leery of most polls.

First, because these polls relate to how those who say they are likely to vote say how they are likely to vote – which is not the same as actually going to the polls and and actually voting for the candidate in question. More than in previous elections, there has been a record number of new registrations (especially by Democrats) and there is likely to be a record turnout (especially by young and black voters). Obama – the former community organiser – has a formidable organisation on the ground and this should make a crucial difference that might even mean his position is understated in the national polls.

Second, because the President is not chosen by a national vote of the people but by an Electoral College of the states. Each state is represented on the basis of a combination of the number of members in the Senate (two for each state regardless of size) and the number of members in the House of Representatives (roughly proportional to population). The states with the largest number of votes are California (55), Texas (34) and New York (31). The states with the smallest number of votes – there are six of them – have only three votes. The District of Columbia, which has no voting representation in Congress, has three electoral votes. In effect, therefore, the Presidential election is not one election but 51.

The Poll Bludger in Australia addresses bits of unconventional wisdom that may help shape the presidential election. For example, the Obama campaign could counter the Bradley Effect by getting a very high turnout of African American voters to show up to polls and vote for him. In fact, U.S. pollsters may be underestimating this demographic.

Whatever methods are being used to account for the certainty of higher black turnout, I’m pretty confident they are overly conservative. When a pollster is required to explain inaccuracy after the event, “I was going on past experience” makes for a more professional sounding excuse than “I made a wrong guess”. I haven’t studied this systematically, but the one example I have looked at has proved to be an eyebrow-raiser: the most recent SurveyUSA poll of Pennsylvania has 10 per cent of black voters among its overall sample, whereas this paper from the Joint Center for Political and Economic Studies tells us it was 13 per cent in 2004. Nate Silver at FiveThirtyEight provides support for this and related impressions in taking to task pollsters who have gaps of 4 to 6 per cent between results for “registered” and “likely” voters.

Secondly, there is something called the “Late Republican Surge.”

I recently heard it said that Republican candidates tend to come home strong in the last week or two of campaigning. Remembering how much of Bill Clinton’s lead vanished shortly before the 1992 election, I thought this sounded plausible and went burrowing through the archives for evidence.

He found that in 1992, George H.W. Bush did make up lost ground before losing to Bill Clinton. So did Bob Dole in 1996. However, in both 2000 and to a lessor extent in 2004, George W. Bush’s numbers dipped in the final days before the election, yet still ended up winning.

Then there is “front runner decline,” which shows that underdogs could be underestimated by two percent and front runners could be overestimated by the same amount.

Finally, advertising plays an important role in the final days of the election, and money is an area where Obama clearly enjoys a significant lead.

Political affinity plays a role in how people view polls. For McCain supporters, the polls are mostly hogwash. The Jewish blogger Life of Rubin claims that the mainstream medias’ portrayal of McCain having little chance of winning came about primarily because the press had long ago anointed Obama as President.

The media, showing who they want to win, is helping Obama get elected. They don’t do any of the same investigating on Obama as they do for Palin, or McCain, or heck even Joe the Plumber. They don’t report on polls that show McCain closer to Obama, only on the “huge leads” from the most extreme, unbalanced polls. (Most of these polls take a sampling of at least 10 percent more Democrats)

Even more amazing is the coverage now, the narrative is all about the inevitable and future Obama Administration and how anything McCain does now is to overcome Obama’s “insurmountable lead in the polls.”

On CNN yesterday that had a bunch of foreign media blowhards all discussing Obama’s first 120 days. Its a wide spread campaign to demoralize voters into thinking this election is already over. Even though for as many polls that say he’s up by 12, there are those that also show he’s only up by 3 or 4. Yet they never report on those polls.

Here’s his thesis:

This is an effort to convince people to stay home because its a lost cause. When the media is powerful enough to override a candidates message and affect the outcome of an election we have a serious problem.

It goes the other way, too. This just in from the Kenyan blogger What an African Woman Thinks.

In view of recent developments in the political arena in the US, I’ve calibrated my view of the average American into more positive territory. I now concede that I may have been too quick to judge the more extreme rhetoric on race and creed from certain parts of the US as mainstream.



No more.



I do not hesitate to offer credit where I believe credit is due. That both national polls and the polls of various battleground states are where they are at this point in the process gives me to understand that I’ve been unfair, to a large extent, to the average American on account of a few individuals who’ve been off the normal curve.



I will not go so far as to say that racism is no longer or will no longer be an issue in America.

But I do think it has suffered a major blow. Hopefully, the injuries sustained from said blow will send it into ICU, and someday, in the not so distant future, it will die an unremarkable and lonely death, in a dark and dingy room.


It turns out that it served the world well for Martin Luther King to have had a dream.

I salute the average American citizen today.

1 comment

  • The mainstream media coverage of polling is utter crap. Those people who would like a clear picture of what all of the many polls really add up to should visit http://www.fivethirtyeight.com , a website that cobbles together the numerous polls at both the national and state levels to illustrate the current state of the US Presidential and Senate races. Currently Obama has a 96.7% chance to win to McCain’s 3.3%. Those who do visit the site should be aware that this is a statistical model which interprets polls, not a media blowhard, as is the case in the US mainstream media. In other words, it’s about the most trustworthy polling data you can find on the Internet today.

    – A U.S. voter whose vote counts about as much as a foreigner’s

Join the conversation

Authors, please log in »

Guidelines

  • All comments are reviewed by a moderator. Do not submit your comment more than once or it may be identified as spam.
  • Please treat others with respect. Comments containing hate speech, obscenity, and personal attacks will not be approved.