Males are more libertarian

Share on FacebookShare on Google+Email this to someoneTweet about this on Twitter

It is a rather robust cross-cultural finding that if there is a sex difference in religiosity, males will be less religious than females. Bryan Caplan has a theory about this. In any case, likely less surprising to readers is the generalization that males are more libertarian than females. Just as any random group of atheists is going to exhibit male surplus, meetings of self-identified libertarians usually seem to exhibit the same imbalance. Atheists and libertarians are both extreme cases of the distribution, and so it stands to reason that any mean difference would result in radically different representations several standard deviations away from the norm.

To explore this is a cross-culturally I looked at the WVS wave 5. Broke it down by sex & country, and looked at the following question:

People should take more responsibility to provide for themselves vs The government should take more responsibility to ensure that everyone is provided for.

The responses exhibit the range from 1-10, with one being “government should take more responsibility” and 10 being “people should take more responsibility.” I computed the mean by weighting these values and the frequencies in each class of values. So if a category had a mean value of 6, that would indicate on average a more libertarian sentiment than a mean value of 4. In the total WVS sample the mean value for males is 5.7, and for females 5.6. In other words, men are somewhat more libertarian than women, but only slightly. Contrastingly there are almost 3 units in the range across countries. Below the fold are pairs of charts. The first simply displays the between sex difference for each country. If males are more libertarian in a given country the data are to the right of the chart, and if females are, the data are to the left. The scatter plots show the strong correlation in attitudes between sexes internationally.

Broken down by sex:

Broken down by sex and limited to those over the age of 40:

Broken down by sex and limited to those under the age of 40:

Broken down by sex and limited to those on the political Right:

Broken down by sex and limited to those on the political Left:

It is interesting that the sex difference seems to diminish the political Left, but less so on the Right. Perhaps this is due to the fact that in much of the world the “Right” political party is really not that libertarian in any case, but more focused on social conservatism (e.g., Christian Democratic parties).



  1. I find it more interesting that highly capitalistic countries like Hong Kong and South Korea are less libertarian than Sweden. Probably that’s because their baselines are different; self-reliance in Sweden is very different than self-reliance in Hong Kong.

  2. So I guess some people should have been a little more suspicious at the original Libertarian Girl blog. I think that an analysis at the extremes i.e. % of people who answered 8-10 might show the variation a little better in terms of ‘true’ libertarian people.

  3. I think this question is kind of misleading, since the two answers are not necessarily opposed. Someone could quite conceivably agree with both statements. I know I do.

  4. which ? would be more informative?

  5. My Baloney Detection Kit begins grumbling whenever I read a study where they asked people to choose between conceptually-overlapping alternatives. 
    It’s not *that* difficult to ask good questions. When people don’t, that’s a good reason to inquire about their motives.

  6. YES, BUT ARE ANY OF THE QUESTIONS GOOD? i don’t provide links nominally.

  7. Actually, I take back what I said. I think I misunderstood at first. I take it that they are two seperate questions and a score of 6 means that the ratio of agreement with the statements is 6:4? It’s probably a useful measure. I just wonder how many people agreed with both. It could be measuring something else, like people’s tendency to criticize others which I think men are more likely to do… I’d probably say the first question, about taking responsibility is the more ambiguous one.  
    perhaps splitting the results into 1-10 scales for agree / disagree for each question would reveal more.

  8. To answer your question, razib: After taking a look at the root surveys, it seems to me that many of the binary-choice questions do not present options that are logically distinct. Ideally one option should be the negation of the other, stated in such a way as to split up the response space into two parts. 
    There may be a few good questions, hidden somewhere in the surveys, but as a whole I’d say they’re of low quality and extremely limited usefulness.

  9. I think a good analogy would be surveys where they ask the two questions “should government spend more to help people” and “should taxes be lowered”. People will reply “yes” to both questions even though they are somewhat opposed to each other.