Tuesday, March 9, 2010

Adimark poll: Bachelet remains popular after earthquake

Today the polling company Adimark reports that president Bachelet's approval rating remains at an impressively high 84% percent, despite cries that the government mismanaged the emergency relief effort. This poll provides a unique opportunity to determine the impact of government handling of the crisis on its approval rating since it was originally completed on Feb. 24, right before the earthquake. Instead of releasing the report on March 1 as originally scheduled, Adimark decided to a second round of polling to capture the public's response to the earthquake.

The report provides a breakdown of the data by social strata, showing that Bachelet's approval rating has taken a hit among the wealthy (group ABC1, below) even though this is the group presumably less affected by the earthquake--all those images of looting and populous rage following the earthquake must have really done a number on them.

[Independently of your political orientation, do you approve or disapprove of the way Michelle Bachelet is conducting her government?]

Interestingly, Bachelet's approval was the highest among the least wealthy, C3 and D/E, where her approval rating actually increased. Adimark also reports that, predictably, that Bachelet's approval rating was significantly lower in the areas most heavily affected by the earthquake (although the sample size for each area was too small to be worth reporting independently).

Significantly, Bachelet's handling of the emergency, though lower than her general approval rating, was also quite high at 75%. Now this takes me to my next point...

Who's this above you might ask? That's Alf Landon, the 33rd president of the United States, or at least according to the now defunct Literary Digest, whose infamous poll wrongly predicted Roosevelt's electoral defeat in 1936. Instead of losing the 1936 election, as Literary Digest predicted, Roosevelt defeated Landon by a landslide. The reason for this epic polling fail? Literary Digest conducted its poll by consulting mostly it's own readers, who were primarily wealthy, and through telephone interviews, which also biased the results.

But what does this classic textbook example of sample selection bias have to do with Chile's earthquake and Bachelet's handling of the emergency response?

Adimark's poll is based on phone interviews. Now, although Adimark is a reputable polling company and assures us that the results were corrected for "the level of telephone penetration in each region," I can't help but wonder if it properly accounts for the views of the most affected by the earthquake who, presumably, are in no position to complete a phone survey.

Of course, we shouldn't completely disregard the results of this poll and overstate the potential effect of this bias. It's just that a one sentence explanation of the methodology used in the report isn't exactly reassuring...just saying. In my opinion Bachelet has done a decent job and has exuded confidence throughout both the financial crisis and now the earthquake. It is also possible that there are widespread negative reactions to the government response but that these are limited to other members of the administration and simply haven't stuck to Bachelet herself. Whatever the case, it seems the voices on the Right predicting that the emergency response will forever stain Bachelet's legacy have been silenced for now.

No comments:

Post a Comment