Just don't mention politics...

The New York Times has an interesting blog on the Princeton study of Americans' happiness levels throughout the crisis which I wrote about two weeks ago:

One of the intriguing findings is that the order in which Gallup posed questions affected the way people reported these various measures of well-being. When the pollsters asked questions about respondents’ plans to vote, their preferred candidates, their approval ratings of the president’s performance and whether the country was headed in the right direction, the respondents reported increased levels of stress and anger.

“People appear to dislike politics and politicians so much that prompting them to think about them has a very large downward effect on their assessment of their own lives,” Mr. Deaton reported. In fact, he wrote, “the effect of asking the political questions on well-being is only a little less than the effect of someone becoming unemployed.”

On the other hand, there were some unexplained spikes in people’s assessment of their well-being. One of the largest increases came on April 6, 2009, a day that did not stand out for its obvious contribution to happiness or standard of living. Perhaps you can find a rationale in the main events of the day: there was an earthquake in L’Aquila.

In fact, the country's happiness levels spiked on April 6th because that was the day Gallup simply asked people how they were feeling, rather than prefacing those questions with any questions about politics. Dropping the politics questions had a far bigger effect on Americans' happiness than any actual changes in their external circumstances (such as rises or falls in unemployment).
At a conference on well-being measurements I attended last year, a young civil servant from the Treasury worried that, if the government is doing the measuring, any measurements would turn into a 'referendum on the government' rather than a genuine assessment of people's well-being. Turns out he may be right...