6th March 2011
Review of Geoffrey Miller’s Spent
As usual, I seem to get these world best-sellers at the fag end of their season, when they are in the clearance bin. The most "entertaining" places (to me!) are these clearance book shops which temporarily sprout across the Melbourne city, and "Opshops" (used goods shops). A couple of weeks ago I found three great books and want to report on this one which is an excellent eye-opener: Spent: Sex, Evolution, and Consumer Behavior by Geoffrey Miller.
A new personality test (GOCASE)
There is now apparently a growing consensus re: personality types. A 6-factor personality assessment tool called GOCASE has now been developed and has superseded the Myer-Briggs inventory (on which I wrote this blog post some time ago). The six factors, outlined by Prof. Geoffrey Miller, are:
a) G or general intelligence (usually called IQ)
b) O or openness
c) C or conscientiousness
d) A or agreeableness
e) S or stability
f) E or extraversion
Apparently the last five of these are totally independent, and a) and b) are mildly correlated. More importantly, these can be tested through just 10 questions (excluding IQ which is best tested independently). I quote directly from the book:
Measuring Your Big Five
The Big Five can now be measured with moderate accuracy, using a self-rating scale, in about one minute. The psychologists Beatrice Rammstein and Oliver John published a Big Five scale in 2007 called the BFI-10 that uses just ten questions. They found that people's scores on this very short scale were very reliable across a two-month period (with test-retest correlations about .84), and correlated very highly (about .82) with their scores on much longer personality scales.
Their BFI-10 scale is reprinted below in a slightly clearer form; try it and see how you score.
After each statement below, write down a number from 1 to 5 to represent how well the statement describes your personality, where
1 = disagree strongly,
2 = disagree a little,
3 = neither agree nor disagree,
4 = agree a little,
5 = agree strongly.
I See Myself as Someone Who
1. has an active imagination ________
2. has few artistic interests _________
3. does a thorough job _____________
4. tends to be lazy ________________
5. is generally trusting _____________
6. tends to find fault with others ______
7. is relaxed, handles stress well _____
8. gets nervous easily _____________
9. is outgoing, sociable ____________
10. is reserved ___________________.
Here's how to score yourself. Items 1 and 2 concern openness; 3 and 4 concern conscientiousness, 5 and 6 concern agreeableness, 7 and 8 concern emotional stability, and 9 and 10 concern extraversion. For each successive pair of items, subtract the number you wrote for the even-numbered item from the number you wrote for the odd-numbered item, and that gives your score for the corresponding Big Five trait. Scores can range from -4 (very low on the trait) to +4 (very high on the trait), with 0 being about average.
For example, if you "agree a little" with "I see myself as someone who has an active imagination," you should have written a 4 for item 1. If you "disagree strongly" with "I see myself as someone who has few artistic interests," you should have written a 1 for item 2. Then you'd subtract your response to the even-numbered item (1) from your response to the odd-numbered item (4), yielding a score of 3 which would mean you are quite high on openness, given that the average is 0 and the maximum is 4. (p.162-3)
[You can get yourself tested on a slightly different scale here
. Also see this
: OCASE = 2 : 1 : 2 : 0 : 0. In other words, I am pretty much a 'normal' person (closer to 0 represents the statistical mean). Compare this with my MBT inventory here
. The closest fit between the two relates to extraversion, and I come out as neither extroverted nor introverted in both tests (the fact that I'm disclosing my results means I'm slightly extroverted?). Judging and conscientiousness seem to be closely related, but I score somewhat more moderately on that with OCASE. Overall I'm more a 'common man' on OCASE than I am with MBTI.
What's the validity of these results? Don't know. All I know is that Miller states somewhere that those who have all 4's are strongly likely to be super-successful!
Bell curve of human nature
I like the idea that the GOCASE model confirms the bell-shaped model of human nature I've outlined in Discovery of Freedom
. I will now cite Miller in support of 'my model' – noting that I have a more extensive model than GOCASE, and focus more on the tail.
In particular, the GOCASE model doesn't show what Hitler's personality type would be, or that of a Gandhi. To be truly successful a human nature (personality) model should be able to ALWAYS identify a Hitler, or a sociopath. And what is the relationship between this and the 5-level Jim Collins model that I believe best represents human leadership?
I'd say that the jury is out on whether GOCASE is the ultimate word on human personality testing. But it is nice to learn about this advancement.
Further additions to human bias and heuristics
This book, which focuses on the impacts of sexual selection on human nature, provides fascinating insights on how men and women can be biased in their decisions and actions, based on their state in the natural cycle of fertility.
Men who want to impress women are more likely to be non-confirmists "as long as the anticonfirmity doesn't make them look more negative and closed-minded than their rivals, and doesn't lead them to make an embarrassing factual error." (p.245) This, and a wide range of experimental conclusions cited in this book demonstrate that our preferences (even political views) can be influenced strongly by the state of our hormones! This is very provocative conclusion, but likely to be true.
By discussing such biases in human nature, this book advances the direction of work initiated by Kahneman and Tversky. He cites Robert Frank quite a bit, and that is a good thing. Frank is, generally speaking, a very good economist.
Most interestingly, I now better understand gaps in our rationality that I've sometimes struggled with. There is a lot of seeming irrationality in the world.
For instance, it is OBVIOUS to even the most stupid student of economics that free trade generates wealth, that political freedom is beneficial. But then why do millions of people refuse to implement these basic ideas in practice? The caste system is another classic case. It might have been rational in the past. Why is it still in place?
The reason he offers is very potent: that we promote irrational ideas because OTHERS whom we desire (not admire intellectually!), believe in them. If you want to marry a bright young thing and her parents are devout Hindus who believe in the caste system, you are better off by believing in it (and by belonging to the right caste!). One fool with a beautiful daughter creates another fool. Irrationality survives because it is INDIVIDUALLY rational to be irrational!
Miller provides a number of illustrations to demonstrate why people have "righteous contempt for any empirical evidence that would undermine them". You clearly can't out-argue anyone in this world on the basis of facts. I've never succeeded in doing so, at least. So why is it that people are so irrational that they reject evidence? Because we benefit by belonging to a 'biradari' that promotes our personal (reproductive) interests! – A very challenging idea, but possibly true.
Here is Miller's most profound sentence
(which closely matches the social contract I've outlined in DOF
): "More subtly, because mating is a social game in which the attractiveness of a behaviour depends on how many other people are already producing that behaviour, political ideology evolves under the unstable dynamics of social imitation and strategizing, not just as a process of simple optimization, given a particular set of self-interests."
My Nash Equilibrium social contract model would actually endogenise these mating strategies. But this is a profound conclusion, and makes Miller's book a worthy contribution to the literature.
Miller is often very, very wrong
Miller then moves on to areas where he not only has no specialisation but no theory. I refer particularly to his advocacy of the consumption tax. His concern is to reduce "consumerism". But sorry, Geoff, you are wrong on this. Your job to DESCRIBE the reasons why people behave the way they do; not to try to change them. That idea of intervening in people's choices – regardless of what their reasons are – is a huge problem! Is Miller or any economist a God that he/she sits separately from the 'people'? What gives us the right to change people's consumption choices? (I don't have any issues with charging a Pigovian tax for negative externalities, though.)
The theory of the state must be founded on a set of clear and consistent principles. A consistent theory of the state requires INCOME taxation
– not taxation of consumption (see details here
I don't care how many "economists" Miller cites in favour of his views about the consumption tax. The idea of asking a government to deliberately act to reduce 'consumerism' and increase savings is wrong. This whole idea about dabbling with people's incentives is a real issue. Already the central banks have made a mess. You don't want to confound things totally. These so-called "economists" have no underlying model of the state, and so they superficially prescribe "solutions" by meddling with, and disrupting, the clear link between the citizen and state. To Miller, I'd say: "Please build a CONSISTENT theory of the state before you prescribe a tax model."
Miller is also very wrong in his comments regarding Hayek and the classical liberals (which are peppered throughout the book). I won't go into that in detail, since knowledge about such things is not expected of him, anyway.
Is Miller a racist?
Miller has great "faith" in the idea of IQ, as a kind of fixed number (which it is NOT – it continuously varies over time), and that 'groups' have different IQs or other personality traits. Thus, at p. 165 he notes, "Nation, region, langauge, culture, …. may predict consumer behaviour mainly because they are directly correlated with some of the Central Six traits". But how? The traits are supposed to be normalised. ALL nations must have the same mean. The law of large numbers must prevail.
I believe that IQ is largely (not entirely!) determined by culture and by the level of freedom (see my model in Chapter 2 of BFN
). Miller himself notes this, at p.198: "my Latin teachers … were open about why we had to learn to read Virgil: familiarity with Latin roots, prefixes, and suffixes would boost our SAT vocabulary test scores". And he admits that SAT is a proxy for IQ. Clearly a PARTICULAR type of education is influencing "IQ" scores!
Miller needs to read and think more widely. His understanding about IQ and the causes of its variations (and that of other traits) is poor. He does talk about how parasitic infection risk might influence openness, and that is not a 'racial' determinant of group variation. But overall I get the strong sense that Miller is a racist.
The book is far more long-winded than it should be. I had to skip solid chunks of idle chatter. The book should be shrunk into half.
However, Miller does come with a unique perspective on human nature that I've learnt a lot from: the elaboration of how sexual selection might work in humans. Overall, I'd say this book is a GOOD READ. Worth reading – but WITH AN OPEN EYE!
I believe that economists and marketing professionals will particularly benefit from this book. But they should never let down their guard. Much of what Miller writes is untested or not fully proven..