There’s been a political psychology firestorm happening recently: several studies have emerged which seem to suggest that conservative attitudes are linked to lower cognitive ability, prejudice, and low-effort thinking. This isn’t a new debate; a study from 2009 showed that higher childhood intelligence predicts a tendency to vote for left of centre parties and to be more politically engaged in adulthood. Several studies from the late 90s and early 00s link the personality trait openness to experience with both intelligence and left-wing attitudes, which means that left-wing people tend to be more tolerant and open to different lifestyles and attitudes, and this element of personality is correlated with higher intelligence.
All this research points towards the conclusion that people with low intelligence and cognitive ability tend towards conservatism, but this obviously does not then mean that all conservatives are of low intelligence and cognitive ability (that would be the fallacy of affirming the consequent). The research actually indicates this:
I never meant to say that the Conservatives are generally stupid. I meant to say that stupid people are generally Conservative.
John Stuart Mill, in a Parliamentary debate with the Conservative MP, John Pakington (May 31, 1866)
Taking a step back
As a left-wing person myself, I know how tempting it is to jump for joy at the data in these studies and allow oneself to feel smug and self-righteous. I mean, I must be more intelligent, right? As with most things in life, however, things aren’t quite that simple. Let’s take a step back for a second and think about the data and what conclusions we should draw from it.
Education level and political affiliation
Education level correlates with intelligence; IQ levels can be used to predict educational achievement. As such, there appears to be a causal link between intelligence and educational achievement. If intelligence can predict educational attainment, and intelligence can predict political orientation, we should expect, based on the aforementioned studies, for the more educated to be more left-wing. When we look at educational achievement and voter behaviour, however, the picture is extremely complicated.
Rural voting and educational achievement
It is a well known fact that rural voters in both the UK and the US tend to vote for right-wing candidates more so than their urban counterparts, who tend to vote for left-wing candidates. Figure 1 and figure 2 demonstrate this fact nicely (note: in the UK map, red = Labour, blue = Conservative, yellow and green = a variety of parties, including Liberal Democrats, Green Party, Plaid Cymru, and the Scottish National Party). If less intelligent people tend to be right-wing, and intelligence can predict educational achievement, then we should expect those living in urban centres to be better educated.
Data from the Statistical Digest of Rural England, however, indicate that people in rural England get better results at GCSE and are more likely to go into Higher Education than their urban counterparts. The areas that tend to vote conservatively are more likely to be better educated. This looks like a simple case of educational attainment correlating with a tendency to conservatism. As I shall explain later, however, there are caveats to this ignored by this analysis of this data.
US voting and level of education
How about the US? This data, taken from the 2004 election, is very telling (data taken from CNN):
|Level of education||Bush||Kerry|
|No High School||49%||50%|
|High School Graduate||52%||47%|
Look at how many college students voted for Bush. It seems that more education (until you reach postgrad level) correlates (or, at least, correlated during this election) with more right-wing votes. An analysis of the data from the 2008 election showed that there was no correlation between education and actual vote cast. Interestingly, however, during this election, polls indicated that those with a university degree significantly favoured Obama to McCain. This is more in line with the stereotypical image of university campuses flooded with ultra left-wing hippie types. The voter data, however, suggests that this picture is not as simple as it seems. It looks like those with higher levels of education are more likely to say they are liberal, but this does not necessarily translate into actually voting for the left-wing candidate.
Putting up a liberal pretense?
It is the responsibility of the government to reduce the differences in income between people with high incomes and those with low incomes
The government in Washington ought to reduce the income differences between the rich and the poor, perhaps by raising the taxes of wealthy families or by giving income assistance to the poor
and less likely to agree with the statement
The government should not concern itself with reducing this income difference between the rich and the poor.
However, even though the more intelligent respondents were more likely to identify themselves as liberals, they were also less likely to agree with the first two statements. It seems they identify themselves as liberal, but don’t actually agree with specific elements of left-wing political ideology.
This is where it gets even more complicated. As Kanazawa notes, nonblack liberal respondents are significantly more likely to agree with the statement
Blacks have been discriminated against for so long that the government has a special obligation to help improve their living standards
and less likely to agree with the statement
The government should not be giving special treatment to blacks.
Interestingly, given the previous results, more intelligent nonblack GSS respondents are more likely to agree that the government has a special obligation to help blacks. Furthermore, more intelligent GSS respondents of all races are more likely to state that the amount of their federal income tax is too low. These are the same people who disagreed that the government should concern itself with reducing the income disparity between rich and poor. Clearly, these results are conflicting and difficult to read. This study would seem to suggest that intelligence has some relation to political liberalism, but it’s not obvious exactly what that is.
Too broad, wrong variables
So many different sources of data show contradictory results about political identity and actual political beliefs, and educational level and political affiliation. The most likely explanation for this is that we’re wither measuring the wrong things, or taking a broad-brush approach to variables that should really be broken down into more specific variables. To give a concrete example, the type of degree studied at university might be relevant here. Whilst writing this post I couldn’t find any data to support this, but there is a chance that those studying certain kinds of arts and humanities degrees are more likely to be liberal than those studying vocational courses. Simply talking about levels of education might be too broad, and could be blurring important distinctions. If the kinds of courses studied were used in these data sets, then we might have quite different results. Note: this is an intuition based on anecdotal evidence rather than hard data, but it seems worthy of mention.
Regarding the urban vs rural voting difference, this study focussed on more variables than just a simple, broad definition of rural vs urban in England and found strikingly different results to the previously mentioned study. They argued that a broad-brush approach missed out several important distinctions otherwise ignored; for example, 1) there are large regional differences in rural areas, the South East being particularly affluent in comparison to the North and 2) some rural areas are very well connected to the nearest schools whilst some are much more distant. They also noticed that two pupils attending the same school are more likely to be attaining more similarly to each other than if they went to different schools because 1) they likely to be in the same classes and so might have the same teachers and 2) the school is likely to have the same policies in a number of areas, and 3) likewise schools that are within the same local authority are going to have more similar average attainment than they would if they were in different areas because of common local authority education. They factored these (and a few other observations) into data gathered from the Longitudinal Study of Young People in England which is a nationally representative study of young people, started in 2004. They got a rather surprising result: the difference between educational attainment in rural and urban areas disappeared entirely. They also found that for certain ethnic minorities, they performed worse in rural areas than in urban areas, which is likely because there are far fewer ethnic minorities in rural areas and so the minorities that do live there are more likely to feel a sense of cultural isolation and lack a support network.
There are many more things I could discuss here, but for the sake of space I won’t. I could, for example, discuss how intelligence is measured, and the theories of multiple intelligences that encourage us to move away from a single, standard IQ measure. I could also discuss the important distinction between social conservatism/liberalism and economic conservatism/liberalism. As it is, I’ll leave these for another time. They’re a blog post in themselves.
To conclude, whilst I’m well aware of the temptation to leap at any scientific result that seems to add credence to your political beliefs, we should all strive to be more careful with how we interpret these results. As I have been discussing, things are a lot more complicated than they seem.