<< News Scan | Main | News Scan >>

Ideology Ratings

This report by Jeffrey Jones of Gallup is only marginally on-topic, but ideology and politics are intertwined with criminal justice policy.  The report also implicates the limitations of models and labels.
A one-dimensional left-right model of politics is, like all models, a simplification and not reality.  Bearing in mind that "all models are wrong but some are useful," Americans' ratings of their own ideology and the presidential candidates' are interesting.  Gallup asked the survey respondents to rate themselves on a 5-point Likert scale from 1 as very liberal to 5 as very conservative and then rate the candidates.

"Americans' mean score on this scale is 3.3, meaning the average American is slightly to the right of center ideologically."  How do we define the "center"?  In my view, the average American is the center by definition.  We might quibble whether "average" should be the median or the mean, but if the average American perceives himself to be off-center, then the perception of the center is wrong.  The reason for the misperception, I suspect, is that the constant bombardment from left-leaning sources in the media and academia causes people to believe that the center is somewhere to the left of where it really is.

The word "extreme" gets tossed around a lot.  How do we define "extreme"?  In my view, extremeness or moderation is defined by the absolute value of ideological distance from the American median.  Gallup gives us a nice table of that value: Jon Huntsman on the dead center, Michele Bachmann furthest to the right of the median, and Barack Obama furthest out of all, on the left of course.  That strikes me as quite correct.

But now look right below Huntsman.  Mitt Romney and Ron Paul are tied for second in moderation.  Huh?  How do two such different candidates tie on this measure?  And did I just use the word "moderation" in the same sentence with "Ron Paul"?

That brings us back to the deficiency of the one-dimensional model.  Libertarianism is a different dimension than liberal-conservative, arguably orthogonal to it.  People high on the libertarian scale take some positions that tend to be associated with the left side of the left-right scale, such as legalizing drugs or favoring the Fourth Amendment exclusionary rule, and others that tend to be associated with the conservative side, such as deregulation of business.  Trying to rate a libertarian such as Paul on a liberal-conservative scale is difficult, and it involves voter-specific choices on which issues to emphasize.  Going to the actual data, we see that Paul is rated "very liberal" or "liberal" by more voters than Romney is, and he is also rated "very conservative" by more.

So once more, with feeling, models can tell us some interesting things, but we must always keep their limitations in mind and not confuse them with reality.  This model works reasonably well with Romney, Bachmann, and Obama, but it breaks down with Paul.

Leave a comment

Monthly Archives