This topic does not die easily:
Dr Wilson and Dr Storm restricted their study to white, Protestant
teenagers, in order to eliminate confounding variables. However, their
volunteers came from two different traditions–Pentecostal, which tends
to the conservative, and Episcopalian, which tends to the liberal.
The researchers conducted the study by giving each volunteer a
beeper that went off every two hours or so. When it beeped, the
volunteer answered a questionnaire about what he was doing at that
moment, and how he felt about it.
Dr Wilson and Dr Storm found several unexpected differences between the
groups. Liberal teenagers always felt more stress than conservatives,
but were particularly stressed if they could not decide for themselves
whom they spent time with. Such choice, or the lack of it, did not
change conservative stress levels. Liberals were also loners, spending
a quarter of their time on their own. Conservatives were alone for a
sixth of the time. That may have been related to the fact that liberals
were equally bored by their own company and that of others.
Conservatives were far less bored when with other people. They also
preferred the company of relatives to non-relatives. Liberals were
indifferent. Perhaps most intriguingly, the more religious a liberal
teenager claimed to be, the more he was willing to confront his parents
with dissenting beliefs. The opposite was true for conservatives.
Here is Storm’s doctoral thesis, a source for some of the material. This Wilson blog post is a useful summary; nonetheless it emphasizes the weaker part of the results, namely the claim that conservatives are more conformist.
Surely someday we will make progress on this question; it is unlikely that any two non-random groups will have exactly the same personality traits. But what is the correct comparison? By looking only at religious Protestants, we are holding some factors constant, but perhaps choosing atypical members of each political ideology. For instance maybe all we are doing is comparing Pentecostals to Episcopalians.