Monday, July 6, 2009

Pet Peeve

Sean Trende (Real Clear Politics–TIME.com, What Emerging Progressive Majority?, July 6, 2009) addresses one of my pet peeves.
The only problem with this theory is that, to claim the country is shifting to the left, you first need to establish a baseline. Teixeira's study doesn't have any baseline off of which to operate. It merely asks its questions, then concludes that America is now center-left. Of course, without such a baseline (i.e., a study of American attitudes in, say, the 1980s, at conservatism's height), it is impossible to say with any certainty whether America has moved right, left, or stayed the same.

Unfortunately, Mr. Trende goes on to cite numerous polls that show that the country is trending Right, not Left, but the data is still not conclusive.

Let me be clear: Mr. Trende is correct to impugn the results of the Americans For Progress study, because their study was also attitudinal (as are the Pew and Rasmussen studies referenced to refute the claim). It is an appropriate apples to apples refutation.

Mr. Trende is also correct to require a baseline for comparison, but that is still not enough.

It's all garbage.

Why? Because asking a person to rate themselves on a scale, or to choose from a list of political descriptions doesn't tell us anything other than what the person thinks or hopes they are.

If you asked people to rate themselves according to honesty, the study is going to show that the majority of people think themselves reasonably honest (or greater). The majority could think that cheating on their taxes doesn't make them dishonest, or bringing home a few pens or pencils from work is "OK" because "everyone else does it." Have they or do they routinely download pirated software, copyrighted materials, or ever participated in sending along an email chain letter? Have they or do they routinely cheat on their spouse or "lusted for another in their heart"? Do they tell small lies to avoid embarassment or save face? Have they or do they routinely call in "sick" for work, when they're not? If the answer to any of those conjoint questions is "yes," their honesty quotient is wanting, and doesn't jive with their response.

Anyone who works with this kind of data knows that people are going to pick the rating or the item that they most want to be, but has no actual reality.

What is needed, that isn't out there, is quantitative/conjoint analysis of the results.

The reason I know that all the studies are bunk is because it doesn't jive with other studies and facts.

If, for example, a majority of Americans consider themselves conservative or conservative-leaning, then how did Obama get elected, why does Congress have a Democratic majority, how did Obama's initial approval rating get so high, why do a majority of Americans support some form of socialized medicine (even if titled "reform"), a desire to keep the status quo with respect to Social Security, like the Prescription Drug Program, like farm subsidies, approve of limited forms of trade protections, support regulation of "evil" businesses and corporations, or can't/couldn't understand the Bush Doctrine to save their lives?

There is no fixed definition in the public square to define what "conservative" means, especially defined for them when people respond to attitudinal polls.

Their responses to what they think they are needs to be validated based on how they feel about specific issues. Until that is done, and we have a baseline to establish the error rate of these types of polls, it's all feel-good nonsense.

If the country really is "majority conservative" then why hasn't the majority donned pitchforks and dragged Congress into the streets for a little tar and feathering?

The only data we can rely on with any certainty is that the majority of Americans are getting dumber/apathetic (see above studies/facts) and poorer (see everything Obama and this Democratic Congress has done so far).