Intelligence & violence

Two of the stranger trends of recent decades have been the Flynn effect (the tendency for IQ to rise over time) and the decline of violence: yes, including terrorism*. A new paper (pdf) by Eugenio Proto and colleagues suggests that these two trends might be related.

They show that in repeated prisoners' dilemma games, people with higher IQs are more likely to cooperate with each other and less likely to defect, with the result that they end up with higher pay-offs.

This is not because they have an innate tendency to cooperate more: in the early rounds of the games, there was no difference in the strategies chosen by low or high-IQ subjects. Nor was it because IQ is correlated with other personality: conscientiousness, they found, had no effect upon whether subjects cooperated or defected.

Instead, it seems that high-IQ subjects learned quicker that reciprocation and cooperation were the best strategies. They were more likely to play tit-for-tat and less likely to always defect; this might or might not be related to the tendency for high-IQ people to have lower time discount rates.

It's easy to see a possible link here with the decline of violence. Violence is a non-cooperative strategy and so would decline as IQ rises.

This might be reinforced by another mechanism - peer effects. As violence declines, it becomes aberrant rather than normal and so carries a stigma, which disincentivizes violence**.

This might also explain why there tends to be more violence in poor areas. It's not just because the poor have less to lose from criminal acts, but because the poor tend to have lower intelligence; this is not (just?) because low intelligence causes poverty, but because poverty reduces (pdf) intelligence***.

I don't offer this as the only, or even best, reason for the decline of violence; it's probably just one of many. And I certainly don't want to join those rather rum coves who think that IQ is enormously important. Instead, I say it for another reason. Perhaps this reminds us that social change happens unintentionally and even without our knowing it at the time. To paraphrase a noted purveyor of violence, progress is what happens when you are busy making other plans.

* Despite Diego Costa's efforts, scenes like these are rarer today.

** This first struck me at the 1999 Arsenal-Manyoo semi-final. After the match, a few people ran onto the pitch to fight. The response of most other fans was to laugh derisively at them as their bald pates glistened under the floodlights and their haymaker punches missed by miles.

*** I'm thinking here of low-level street scraps; it is not (pdf) the case that poverty causes terrorism.

Comments

[Luis Enrique]

have you looked at this?
http://www.sup.org/books/title/?id=23082
[I haven't, and my prior is unsympathetic]

[Matt Moore]

Discount rate is so so important. I would not be surprised if it turned out to be the true casual factor here (not IQ, which would be a spurious co-correlate.

And that rate very much isn't given ex deus at birth. It is developed, particularly in infancy and early childhood. It is essentially the old-fashioned virtue of patience.

To the extent that happiness / satisfaction is generated by having time-consistent preferences (I believe, to a great degree) this study means that both (i) personal satisfaction and (ii) societal harmony are strongly shaped by early experiences.

This is why I worry about the erosion of bourgeois values in the family home.

[Blissex]

For another and somewhat similar take, B Schneier, a wise (computer, but not just) security expert argued that the best security-enhancing strategy is gentrification.

As to intelligence, intelligent people are bound to realize that exploitation rather than cooperation also works, that is violence works, as over a thousand years of feudalism (the protection racket by another name) in Europe demonstrated, but it has high volatility.

Gentrification is less volatile.

[Matt Moore]

@blissex. Violence is typically a bad strategy if you can avoid it. In what sense did the exploitative system of feudalism "work" compared to free market capitalism?

[Blissex]

«Violence is typically a bad strategy if you can avoid it.»

That may be a pious illusion. Violence works pretty well for the winners. Crime, even violent crime, usually pays if done without utter stupidity. The problem with that is the volatility. Few winners, many losers, and winner one day, in a coffin the next.

«In what sense did the exploitative system of feudalism "work"»

First of all it has endured in some variations for arguably 12 centuries or more. Not much else has. If it did not work in some fundamental sense it wouldn't have lasted so long and adapted so much to circumstances. That's a burkean style argument, and indeed in a perverse way that's how Burke was making an apology of the feudal system.

Also it has worked very well for the winners. Many extended families have lived in splendid luxury for over a thousand years on the proceeds of the protection racket. We'll see whether the families of the Mozillos, ONeils, Fulds, Caynes, and other "heroes of wealth creation" in our late stage decaying asset stripping crony financialism will last that long.

Again the problem with winning by violence is not average winnings; it is the volatility in the very skewed distribution of winnings, and for the winners the high volatility across time. It is like the protection rackets in every time and place: a "winner takes all" system, where "all" includes the lives of the losers, and every winner is constantly fighting some pretenders. Reading medieval cronicles in particular is like reading the stories of any gangster war, and in effect most wars in Europe until the 19th century were gangster wars.

«compared to free market capitalism?»

It is hard to compare anything to that because it is a pious illusion. Something like some free markets or a degree of capitalism perhaps existed for a few to several decades in and around Manchester in the 19th century.
JM Keynes still wrote a few decades later of the "Manchester system" when it was already largely gone.

Perhaps this delusionary talk about «free market capitalism» arises from the usual and often wilful confusion between the industrial system of production and the currently-imaginary political system called «free market capitalism».

If you believe there is «free market capitalism» around us try to borrow 50 billion dollars at 0.2% from the Fed by discounting toxic paper, without being a bankrupt financial corporation "sponsoring" a number of senators and congresspeople, or try to get bank holding company status in less than a week :-).

[Blissex]

«without being a bankrupt financial corporation»

As to this enduring state, two delightful quotes that I found recently in a 1986 paper by the always perceptive H Minsky:

«No matter how exalted a bank may have been, we all know that if assets were marked to market, the net worth of many of the giants of international banking would dissappear. Nevertheless these banks are able to sell their liabilities in financial markets, because the buyers believe that they will be protected against losses by the central bank.»
«Formal deposit insurance, as introduced in the 1930's, is not responsible for this change. It can be demonstrated that if banks and other financial institutions were marked to market the claims on the reserves of the deposit insurance agencies would far exceed their resources. Deposit insurance is viable because it is supported by central bank and government commitments to validate the liabilities of protected institutions.»

Funny that :-).

[Deviation From The Mean]

These experiments don't tell us anything much about human behaviour, they simply tell us how humans behave in the artificial world created by these experimenters.

But why bother with these experiments when we have the actual world? They bother because mathematicians can only handle a few variables! Once variables get out of hand mathematics is as useful as reading tea leaves.

The danger with these experiments is that wider assumptions are made. The danger is that these experiments are taken seriously.

So my conclusion is there is no proven link at all.

[Kris]

Chris, check this article out - there is very strong evidence that lead in petrol was the single most significant factor in the dramatic increase in violent crime from the 50s and 60s through to the 90s and its removal from petrol is the cause of the equally significant fall in violent crime. Check Kevin Drum's original article at the link below. I'm fully convinced and think that this is one of the great scandals of the 20th century with a continuing impact on society today.

http://www.motherjones.com/environment/2013/01/lead-crime-link-gasoline

[Kris]

I should have added that the article explicitly links IQ levels and violent crime - lead decreases IQ and increased violent tendencies, inability to delay gratification, reduced concentrations spans, etc.

[Theophrastus]

Feudalism didn't last for 12 centuries. The scholarly consensus is seven centuries - 9th -15th C.

[Deviation From The Mean]

"reduced concentrations spans"

reduced concentration spans seems to have been a government policy rather than lead induced. We see that everywhere in culture, TV reflects an era of reduced attention span.

It seems to me that we have had an increase in reduced attention spans while seeing a decrease in violence (so where is the link).

How anyone can ever believe one variable can be plucked out to explain something like violence is beyond my reasoning. But the limits of mathematics and the esteem mathematics is held in mean we get ridiculous meta theories taken seriously, such as the ones discussed in this article.