Liberals smarter than conservatives, Republicans smarter than Democrats

The General Social Survey includes a ten item vocabulary test, coded WORDSUM, which is likely strongly correlated with intelligence. This is an opportunity for us to determine whether liberals and Democrats are smarter than conservatives and Republicans.

For the purpose of this comparison, I consider a high vocabulary score to be 9 or 10 correct answers on the vocabulary test (WORDSUM 9-10).

LIBERALS ARE SMARTER THAN CONSERVATIVES

Of those who consider themselves “liberal” or “extremely liberal” (POLVIEWS 1-2), 20.8% had high vocabulary scores.

Of whose who consider themselves “conservative” or “extremely conservative” (POLVIEWS 6-7), only 13.8% had high vocabulary scores.

REPUBLICANS ARE SMARTER THAN DEMOCRATS

Of those who identify as Republican (PARTYID 5-6), 15.6% had high vocabulary scores.

Of those who identify as Democrats (PARTYID 0-1), only 11.9% had high vocabulary scores.

ANALYSIS

This is an extremely surprising result. Because there is a strong correlation between liberal and Democrat, and between conservative and Republican, I would have assumed that if something were true for liberals it would also be true for Democrats, and the same for conservatives and Republicans.

My hypothesis is that people consider many factors when determining if they are liberal or conservative, including leftist or rightist orientation (leftist being liberal), religiosity (religious being conservative), and attitudes corresponding to what’s considered “Openness” in the Five Factor personality model (with Openness being liberal).

But when people decide which party to vote for, a major factor is how much money they make and/or what social class they consider themselves a part of. People who are of a higher social class or make more money (both of these factors correlating with higher intelligence) are more likely to vote Republican.

UPDATE

I may have been wrong. See my post: Democrats may now be the more intelligent party.

Comments

[SciFiGeek]

Nice job Half, I think you hit the nail on the head. I'd like to add in the correlations, which measure the trends:

correlation between WORDSUM and POLVIEWS: minus .03
correlation between WORDSUM and PARTYID:
.11
correlation between PARTYID AND POLVIEWS:
.29

So, yes, you can have two variables correlated with each other which correlate in different directions with the same variable. Weird huh?

Now look more closely at the frequency tables for WORDSUM and POLVIEWS, however, and you see something really interesting. There are SEVERAL big blocks of red (indicating more likelihood than chance) here. In fact, I see not two, not three, not four, but FIVE, in a twisted X-shape:

WORDSUM 0-3, POLVIEWS 1-2. Dumb and really liberal. These are your urban poor. The Democratic proletariat.

WORDSUM 9-10, POLVIEWS 1-3. Smart and really liberal. Latte-sipping urban yuppies. The Democratic elite.

WORDSUM 3-6, POLVIEWS 4. Average in every way. Middle America.

WORDSUM 7-10, POLVIEWS 5. Pretty smart and slightly conservative. Management and fiscal conservatives. The Republican elite.

WORDSUM 1-4, POLVIEWS 7. Pretty dumb and very conservative. The rural poor. The Republican base.

Just goes to show you that you can't make statements like 'liberals are smarter than conservatives'. Which liberals are you talking about?

[SciFiGeek]

Does anyone know how to manipulate the analyze section so you can, for example, look at people with a Republican PARTYID and a liberal POLVIEWS and see how they did on WORDSUM?

Oh, and Half, I'd like to apologize for encouraging you to become a liberal. It's pretty clear your political views lie elsewhere and I misinterpreted you based on a few posts. At any rate, I really am enjoying your blog.

[SciFiGeek]

Ah, you can do it! Enter the third variable as a control variable! This will give you, for example, POLVIEWS versus WORDSUM for every value, from most Democratic to most conservative, of PARTYID. What this means is that you'll see a vocabulary-versus-views table for strong Democrats, then a vocab-versus-views table for weak Democrats, and so on to the vocab-versus-views table for strong Republicans.

And what do we find?

For Democrats, and third party people, conservatives are dumber.
But among Republicans, conservatives are smarter.
Among independents, moderates are smarter.

Hmmm, what do we make of this?

[Randall Parker]

I would expect the ratio of verbal to non-verbal IQ to be higher on the Left than on the Right. So the use of vocabulary as an IQ test is problematic.

[Half Sigma]

Randall, high verbal IQ actually correlates with "conservative" positions on economics issues, which explains why high verbal IQ is correlated with being Republican.

Obviously the problem with "liberal" and "conservative" is that they have come to include both economic, religious, and cultural indicators. I am going to limit the use of the term "liberal" in my blog because after studying the GSS I think its meaning is too amorphous. Someone might call themselves "liberal" simply because they're not religious even though they don't have any "liberal" economics views.

[Randall Parker]

Half Sigma,

Engineers are less verbal than they are spatial and quantitative. Engineers are also more conservative than humanities types who are more verbal than they are spatial. My guess is there is a cause and effect going on here.

Granted, you can find highly verbal people who are highly spatial as well. But my guess is that people who can spin words but not model the world well using quantitative and spatial reasoning are going to trend leftward and embrace all sorts of wrong-headed ideas.

But there also must exist inter-racial and inter-ethnic group differences in other cognitive qualities that play big roles in determining whether one tends to be conservative or not. So comparisons of verbal abilities and political leanings that do not adjust for race and ethnicity will suffer from other differences messing up the relationships between cognitive ability and political leanings.

[Dr Katerina Vaugier Oppenheimer]

I have to say that lying is not particularly acceptable. Firstly, conservatives or republicans are generally more intelligent than liberal people, with a higher income level, more educational backround and something I like to call, common sense. Liberals tend to focus on trivial inconsquentialities, republicans look at the larger picture. Harvard did a general study taking nearly 30000 people from 48 of the 50 states, it showed a significant difference from your findings, Those who admitted being republican scored 13% higher than those who admitted they were democrat. This of course was not a tainted study like your normal web poll. The General IQ of the democrats who had taken the IQ test, was 99, the average of the republican 112. The samples were not tainted, nor told the goal of the survey prior to taking it. Though I will admit, a small percentile of liberals are extremely intelligent. Being that Pathological liars would need to be intelligent to come up with the lies that they do. Harvard did a study on pathological liars, with correlated liberals and conservatives, guess who had the highest ratio of pathological liars... 13 to 1. Republicans do no lie; at least as often.

[Joel Waxman]

The so called Dr. Katerina is surely not a doctor of any reputable source or her writing would not be so poor and she would not source bogus studies. The Harvard studies she cites do not exist. The most definitive studies in liberal and conservative idealogy in relation to intelligence and religion were done by Thomas Symington in the 30's. Many others were done with similar conclusions drawn around religion and intelligence: http://www.huppi.com/kangaroo/L-thinkingchristians.htm

Other studies have clearly demonstrated that when income is controlled for the relationship beween liberalism and higher intelligence becomes even more pronounced.

Education is a poor judge of intelligence. IQ is not perfect either but is better than education. What education has demonstrated is that there are a lot of average intelligent college graduate republicans that long to be in the upper class so they vote republican so they can feel like they are already there. The Republican party is the party of mediocrity. While it is true that many low intelligence people vote for Democrats, the one splitting atoms are doing the same.

The great minds of the world have overwhelmingly been liberal. Socrates, Jesus, DaVinci and Einstein were all Liberals. The list goes on and on.

While there are many in the lower class that may vote democrat because of social justice reasons they are not necessarily Liberals on a variety of issues. The same is true for many wealthy Republicans that only vote that way because of one issue...taxes.

The bottom line is if you voted for George W. Bush you are a big moron no matter what your intelligence is. If you are one of the 29% that still say he is doing a good job, you are either in denial or truly stupid. At least the kool-aid drinkers are finally waking up in this country. They are realizing that the Republicans are the party of the utterly stupid and greedy. Now you can add pedophile protectors to their resume.

David Kuo's book substantiates the obvious; people of religion have been duped. Those dumb enough to believe that Republicans were the party of faith are getting what they deserve. Now we all have to live with the disastrous consequences of this administration.

[Jackie Jones]

I like how, the so called "Dr Katerina Vaugier" at Oppenheimer does not really exist. Much like her made up study. Conservatives will never post a link or source a comment.

[Martini Drinking Atheist]

I suspect the discrepancies can be explained somewhat like the following.

Overall, people who self-identify as liberal will score higher than conservatives. However, in terms of party-identification, well-educated liberals only make up a minority portion of the Democratic Party. The vast bulk of the party are working poor, blue collar, and middle class voters who may not necessarily be highly-educated.

The Republican Party is primarily made up of middle class white males, who have a slightly higher education quotient than your average working class male, hence by party, Republicans will score slightly higher, but by political identity, liberals will score higher.

[Anonymous]

Why do we have to identify ourselves so narrowly in terms of our political views? I am not Democratic or Republican, conservative or liberal. I am no independent party follower.
Why would anyone even bother studying this kind of thing? It doesn't matter. We are all Americans, and our role as voters is not to elect the richest politician in our party, it is to elect the best person for the job.

In summary: who cares about this kind of thing? Does it matter?

[Josh]

WORDSUM is a horrible choice to test intelligence. According to the theory of multiple intelligences this would only apply to those that possess high verbal linguistic intelligence. Here this is pasted from wikipedia, "
Verbal-Linguistic
Verbal-linguistic intelligence has to do with words, spoken or written. People with verbal-linguistic intelligence display a facility with words and languages. They are typically good at reading, writing, telling stories and memorizing words and dates. They tend to learn best by reading, taking notes, listening to lectures, and via discussion and debate. They are also frequently skilled at explaining, teaching and oration or persuasive speaking. Those with verbal-linguistic intelligence learn foreign languages very easily as they have high verbal memory and recall, and an ability to understand and manipulate syntax and structure.

Careers which suit those with this intelligence include writers, lawyers, philosophers, journalists, politicians and teachers.

[edit] Logical-Mathematical
This area has to do with logic, abstractions, inductive and deductive reasoning, and numbers. While it is often assumed that those with this intelligence naturally excel in mathematics, chess, computer programming and other logical or numerical activities, a more accurate definition places emphasis less on traditional mathematical ability and more reasoning capabilities, abstract pattern recognition, scientific thinking and investigation, and the ability to perform complex calculations.

Those who automatically correlate this intelligence with skill in mathematics criticize this intelligence by arguing that logical ability is often more strongly correlated with verbal rather than mathematical ability: for example, in the Graduate Record Examination, a test often used in the United States to decide who will be admitted to graduate school, the old Analytic section correlated more strongly with the Verbal section than the Mathematical. One possibility is that formal, symbolic logic and strict logic games are under the command of mathematical intelligence, while skills such as fallacy hunting, argument construction, etc. are under the command of verbal intelligence.

Careers which suit those with this intelligence include scientists, mathematicians, engineers, doctors and economists.

These are two entirely different forms of intelligence and many scientists, philosophers, etc. where more mathmeticians and reasoners. This wordsum quiz would undermine their intelligence. If I had to take a guess I would claim that liberals tend to be more verbal-linguistic while conservatives are more logical-mathematical based on the views of the two sectors. I think if this study or poll proves anything it is that my view of that is correct.