Thursday, September 18, 2008

IQ and voter turnout

++Addition++Half Sigma weighs in, pointing out that voting is largely not a self-interested action, but a fulfilling of civic responsibility. This is especially true at the Presidential level, with roughly half the country living in electorally-safe states. Relatedly, he shows that more intelligent people are more likely to attend church services than less intelligent people are, even though the smarter folks are less likely to profess religious beliefs, the point being that smarter people are more responsible and committed than less intelligent people are. The consequences of this are wide-ranging, similar to what we see in the relationship between livability and intelligence--smarter places are better places to be, no matter how it's looked at (with the glaring exception of lots of kids running around, if that's considered better than just a few of them).


The hunt for some socially desirable attribute that correlates inversely with average IQ continues. Electoral participation didn't strike me as a particularly promising candidate, but one entrusted with such an important task must leave no stone unturned, and so I continue to kick the rocks.

Using a great data table on '04 turnout numbers from GMU Professor Michael McDonald, the percentage of the voting-eligible population by state is correlated with its estimated average IQ. McDonald eschews two flawed methods of gauging turnout that I came across when looking for a reliable data source.

Several sites simply report actual votes cast as a percentage of the voting-age population within a state. In the case of felons or the mentally incapacitated, that's not too problematic--their numbers proxy well for the size of the double-digit IQ population, and they are distributed across the country roughly accordingly. But it is flawed because it doesn't account for the presence of non-citizens. In a state like California, where more than one-fourth of the population is foreign-born, the state is artificially dropped near the very bottom in terms of voter turnout, which tells us little about the relationship between intelligence and how likely people who are able to vote actually are to vote.

Others look at the percentage of registered voters who go to the polls. But states are slow to adjust those numbers when people move or register to vote elsewhere, which is why some states show more than 100% of the voting-age population being registered to vote! Voter fraud, anyone?

McDonald gets around these problems by coming up with the size of each state's voter-eligible population. He pulls the foreign-born, the incarcerated, the mentally ill (contingent upon law, as the rules vary from state to state), and half of those on parole (an estimate arrived at through reports from the Department of Justice) from a state's total 18 and over population, then compares that figure to actual votes cast for the highest office in contention on the ballot.

Crunching the numbers confirms the expected. Voter turnout as a percentage of the voter-eligible population and average IQ correlate at a statistically significant .65 (p<.000001) at the state level. Average IQ alone thus 'explains' 42% of a state's voter turnout. That is a strong relationship for the social sciences. Put in another way, it suggests that for every one point increase a state's average IQ, voter turnout should increase by nearly 4%. This brings to mind the words of Fredo Arias-King, adviser to former Mexican President Vicente Fox:
Several [Democratic legislators] tended to see Latin American immigrants and even Latino constituents as both more dependent on and accepting of active government programs and the political class guaranteeing those programs, a point they emphasized more than the voting per se. Moreover, they saw Latinos as more loyal and "dependable" in supporting a patron-client system and in building reliable patronage networks to circumvent the exigencies of political life as devised by the Founding Fathers and expected daily by the average American.
Last year, Pew reported that voter-eligible Hispanics are less likely than other Americans to cast ballots:
Hispanics who are eligible to vote are less likely to register and less likely to cast a vote than either whites or blacks.

About 54% of Hispanics who were eligible to vote [thus illegal immigrants do not factor into ineptness] registered in November 2006. Among whites and blacks, the figure was 71% and 61%, respectively.
The '08 Democratic primaries brought record turnouts in several states across the country. But the race was competitive longer than any Democratic contest has been in several decades.

As Hispanics come to account for one-third of the total US population, which the Census now estimates will happen before mid-century, the country's average IQ will decline from an estimated 97 or 98 today to between 93 and 95, similar to that of contemporary Argentina or Romania. So there will be two forces working against high voter turnout--a less intelligent, less civically-minded public, and a larger Hispanic population more resigned to a prodigious, corrupt government. The linear equation created from the simple regression presented above suggests that nationally voter-eligible turnout will fall from 61% today to just under half (49%) by 2050.

An uninformed, apathetic public doesn't bode well for the future of the United States, which will simultaneously be beset with a rapidly growing elderly population (the number of people 85 and older is projected to increase by more than 300% over the next four decades). As the ratio of dependents to workers continues to grow, those who do vote will increasingly find themselves on the receiving end of entitlement programs and will thus be reluctant to vote against the expanding governmental leviathan that feeds them by eating up the productive assets of the proportionally shrinking productive classes.


BGC said...

Increasing IQ is beneficial in just about every way except the one that really matters: fertility. So, in modern society, IQ is fitness-reducing.

Which leads to the strange situation of increasing IQ being positively correlated with increasing social status and wealth (etc) in all modern societies studied - but still being biologically maladaptive.

The other significnat negative correlation with IQ is religiousness - which is a morally-bad or -good thing according to individual perspective; but is certainly another fact contibuting to lower fertility (because fertility is boosted by religiousness).

BTW - I think your theory of economic dependency linked to voting is correct. I first came across it in fleshed-out form in the writings of Arnold Kling of Econlog.

Kling regards the future as a kind of race between increasing numbers of economic dependants and technology-driven improvements in productivity - alternatively there need to be some serious reforms to reduce the dependency ratio such as a large raising of the retirement age.

Or else dependency will grow faster than the economy and it could be very difficult to break out from the downward economic cycle.

JBS said...

"About 54% of Hispanics who were eligible to vote [thus illegal immigrants do not factor into ineptness] registered in November 2006."

These numbers probably *understate* how low Hispanic voter turnout will be in the future.

The Hispanics most likely to vote are the, comparatively, best educated and highest earning fraction of the Hispanic population, and, therefore, the portion of the Hispanic population with the highest levels of European admixture.

The Hispanics who do not vote are, on average, the least educated and the least European in appearance.

Because the less European looking Hispanics presumably have much higher fertility levels than the better earning/educated, whiter Hispanics, it is probable that future Hispanic voter turnout will be *lower* than the current 54%.

Also remember that these trends are not set in stone.

At least 40% of all Hispanic births since the early 90's up to today have come from illegal alien mothers.

If Hispanics illegals of childbearing age are removed from the US and we switch our immigration system to a points system where we would import more high IQ and more European admixed Hispanics from South America as oppossed to Central America, the growth rate in uneducated Hispanics would slow considerably.

Audacious Epigone said...


At least in the developed world, that appears to be the case. Well stated.

Re: dependency, there is room for improvement on both ends.

Allow smart students to take college classes online or via video-recorded lecture during their early teenage years. Make degrees contingent entirely upon testing, so the most enterprising students need not stop for three months every year, or share classroom space with kids who are disruptive or simply not able to handle the same material. Put students with double-digit IQs on vocational schools at a young age rather than having them waste their time, energy, and morale in a university setting they'll derive no benefit from.

Raising the retirement age, tying social security benefits to an inflationary/poverty measure rather than to wages, and removing legal obstacles to voluntary euthanasia are all ways to reduce dependency on the other end.


Amen. As it's been argued here since the blog's inception, we need a restricted merit immigration system. The 14th Amendment needs to be reinterpreted as not granting birthright citizenship to those the US should have no jurisdiction over in the first place (ie, illegal immigrants).

Stopped Clock said...

I think that at least some gifted students attend summer school and learn advanced things there. I remember going to a nearby college campus when I was seven years old to learn advanced (for my age) math, and then back again in a couple years to learn creative writing. I think that some of the other kids in those summer classes attended private schools with advanced curriculum. My grades in school were never more than a little above average, though; I think that I wasn't so much a child prodigy as just someone who enjoyed learning. I was in normal public school during the school year.

I like your distance learning idea; it might already be in practice, but I havent heard of it. I remember a few young kids mixing with the normal-age students at my previous university. I think that happens everywhere.

What do you mean by "make degrees contingent entirely upon testing"?

Audacious Epigone said...

Stopped Clock,

Randall Parker has several posts dealing with educational reform, specifically video recording lectures. You might check some of his writings out.

Re: degrees contingent upon testing, I'm sharing in Charles Murray's idea that various college degrees should be awarded not by completion of X hours in X courses, but by passing something similar to the CPA exam. To determine proficiency based on hours in class epitomizes the blank slatism paradigm in American education. Since everyone is the same, only the amount of instruction they've received differentiates them, thus after four years of being talked at, you're qualified to receive a degree.

Fat Knowledge said...


I am working on a post on a similar theme and I was wondering if you can help me out with one implication of your analysis.

What is the likelihood that a person 1 STD below and above the average will vote?

Lets say it is 40% and 80% respectively. Can you then say that per capita intelligent people have twice the influence in government? That if you have 1 mil low and high IQ populations, that the high IQ population will have twice the amount of voters?

Oh, and any correlation between IQ and the party they vote for? My guess would be that Democrats are more represented at the low and high ends and Republicans are more represented in the center, but I don't really know.

Audacious Epigone said...


I'm going to take a look at the GSS to get some idea of that to see how it relates to this post once I finish the FF10 review I'm currently having fun with, but I can't give you an answer based on state-level comparisons (unless I look at state's as individual people to get some idea, which would be interesting to do in comparison to actual individuals).

I suspect your presumption is right, especially for Democrats at the low end (that's what community organizers are for, after all ;)

I'll be interested to see what you come up with in your post!