The Intelligence Paradox: Why the Intelligent Choice Isn't Always the Smart One - Satoshi Kanazawa 2012
This book challenges some of the common misconceptions that people have about the nature of intelligence: what it is, what it does, and what it is good for (if anything). People have a tendency to equate intelligence with character, and to believe that intelligence is the ultimate gauge of an individual's worth. They believe that people are not worthy human beings unless they are intelligent, at least in some way. Somehow, intelligence is regarded as the hallmark of human worth and character, the most important trait that any human can have.
Conversely, people claim that, because all humans are equally worthy, they must all hold the potential to be equally intelligent as well. They balk at scientific discoveries that suggest that some groups of individuals on average may not be as intelligent as others, as if that is tantamount to saying that such groups of people are somehow less worthy humans because of their lower intelligence. When they realize that some people do not score high on standard IQ tests, they automatically assume that IQ tests must therefore be biased against some groups of people. They take group differences in performance on the IQ tests as prima facie evidence of such bias.
At the same time, they also maintain that there are “multiple intelligences”1 and fabricate other types of “intelligences” such as “bodily-kinetic intelligence” for people who are athletic, and “interpersonal intelligence” for extroverted people who have social skills. No longer does it suffice to say that some people are athletic and others are social; they must all possess different kinds of “intelligences.” They have to ensure that everybody is intelligent in some way because everybody is an equally worthy human being.
In this book, I want to break this equation of intelligence with human worth, by pointing out that intelligence (and intelligent people) may not be what you think. While more intelligent people can do many things better and more efficiently than less intelligent people, there are many things that they cannot, and intelligent people tend to fail at the most important things in life from a purely biological perspective. The list of what intelligent people are not good at may surprise you. Intelligent people are only good at doing things that are relatively new in the course of human evolution. They are not necessarily good at doing things that our ancestors have always done, like finding and keeping a mate, being a parent, and making friends. Intelligent people tend not to be good at doing things that are most important in life.
There is no question that intelligence is a positive trait, but then so are beauty, height, and health. Yet we don't equate beauty, height, and health with human worth. (However, we do a little bit when it comes to beauty, by maintaining that people who are not physically attractive nonetheless have “inner beauty.” “Inner beauty” is to physical attractiveness what “multiple intelligences” are to intelligence.) We don't necessarily think that beautiful, tall, or healthy people are better, more worthy humans than ugly, short, or unhealthy people. Nor do we claim that everyone is equally beautiful, equally tall, or equally healthy. But we seem to believe that more intelligent people are more worthy human beings. Or, conversely, because all humans ought to be equally worthy, they must all be equally intelligent. In this book, I want to convince you to stop having this assumption and break the equation of intelligence with human worth. Intelligence is just another quantitative trait of an individual like height or weight.
What Do People Want?
This book is also about what people want—their preferences and values. What do people want? Why do people want what they want? Where do individual preferences and values come from? These are some of the questions that I address in this book.
These questions are known as the “problem of values,” and it is one of the central questions in social and behavioral sciences. Mathematically elegant models of human behavior, especially in microeconomics, can explain human behavior very well, if we know what people want in the first place. These microeconomic models essentially say, “Human actors do the best they can within their constraints and circumstances to achieve what they want to achieve,” and their mathematical models are about how these constraints and circumstances affect their choices. However, these models cannot tell us what human actors actually do and what actual choices they make, unless we know “what they want to achieve,” namely, their goals, preferences, and values.
Without the knowledge of individual preferences and values, any choice can be used to support the microeconomic model. Let's say two individuals, A and B, are facing identical constraints and circumstances; they both find themselves at a local grocery store, with limited choices of items to buy and only $10 to spend. If, under these conditions, Individual A purchases apples and Individual B purchases oranges, both choices can be used to support the rational actor model, by positing, ex ante, that Individual A has a preference for apples while Individual B has a preference for oranges. If, under the same conditions, Individual C chooses not to buy anything at all, that, too, supports the theory because Individual C has a preference for savings.
In order to provide a full explanation of human behavior, we must endogenize individual preferences and values, by making them part of the theoretical explanations. We must first explain individual preferences and values before we can explain the behavioral choices. We must be able to explain why Individual A wants apples, not oranges, and why Individual B wants oranges, and not apples, and why Individual C wants neither.
However, economists themselves have given up on the attempt to explain preferences and values. In 1977, two University of Chicago economists (and future Nobel prizewinners), George J. Stigler and Gary S. Becker, published a very influential paper in the American Economic Review entitled “De Gustibus Non Est Disputandum.”2 The Latin title means “There's no accounting for tastes.” In this paper, Stigler and Becker argued that economics should not be concerned with explaining preferences and values, but instead with how constraints affect choice. In other words, they wanted to leave preferences and values exogenous to the economic models, rather than endogenous to them. (Endogenous means within the model of explanation whereas exogenous means outside of such models. Endogenous factors are explained within the model whereas exogenous factors are left unexplained.)
Economists since then have faithfully followed Stigler and Becker's dictum and left preferences and values unexplained. They focus on money—how individuals seek to maximize income and wealth—because money is fungible3; people can use money to buy virtually anything they want (except, of course, for love). Economists reason that they don't need to know people's preferences and values because they would all want to maximize monetary incomes regardless of their idiosyncratic preferences and values. Whatever they want, they all want more money because more money will allow them to have more of whatever they want. With more money, you can buy more apples, more oranges, or increase the level of your savings. So, in the above example, Individuals A, B, and C—in fact, all individuals—would want to maximize their monetary income because it would allow each to purchase whatever they want. The focus on the fungible resource of money obviates the need to know what individuals really want. This is where microeconomics stands today.
I want to endogenize preferences and values to models of human behavior, by introducing evolutionary psychology. I want to explain what people want and why they want it, and I believe evolutionary psychology is key to such an explanation. So this book is also about explaining the origin of individual preferences and values.
Given the main focus on intelligence in this book, I will emphasize the role that intelligence plays in influencing people's preferences and values. How does intelligence influence people's preferences and values? What do more intelligent people want? What do less intelligent people want? And why are there differences?
• • •
So this book is at the intersection of two scientific fields: evolutionary psychology and intelligence research. In Chapter 1, I briefly introduce you to the field of evolutionary psychology. Those of you who have read my previous book Why Beautiful People Have More Daughters4 can probably skip Chapter 1 without missing much.
In Chapter 2, I discuss one important implication of evolutionary psychology for the functions and constraints on the human brain. I introduce an important idea called the Savanna Principle about what the human brain can easily do or has difficulty doing. Readers of my earlier book may recall reading about the Savanna Principle in it.
In Chapter 3, I discuss the concept of intelligence and introduce you to the field of intelligence research. I will explain what intelligence is, what it is not, and what it does (and doesn't do). There are so many misconceptions about intelligence, and I will try to dispel many of them in this chapter.
In Chapter 4, I introduce you to the central idea in this book, which I call “the Intelligence Paradox.” It is about how more intelligent people may differ from less intelligent people in their preferences and values, often in quite counterintuitive ways. Readers of my Scientific Fundamentalist blog at Psychology Today will see that “the Intelligence Paradox” captures the same essential idea as what I call “the Hypothesis” in my blog, but applied specifically to individual preferences and values.
The rest of the book (Chapters 5—12) details various applications and manifestations of the Intelligence Paradox. Chapter 5 is about political attitudes on the liberal-conservative continuum, and explains why liberals in the United States are on average more intelligent than conservatives. I also explain why liberals are stupider than conservatives. (No, that's not a typo. Confused? Read on!)
Chapter 6 is about religiosity and belief in God. I will explain where religion came from and why atheists are more intelligent than people who believe in God. It's not because religion is false or there is no God. It's because religion is deeply human. We are designed by evolution to believe in God, and that is why more intelligent people are atheists.
Chapter 7 is about preferences about sexual exclusivity. I will explain why more intelligent men prefer being sexually exclusive more than less intelligent men, but more intelligent women do not value sexual exclusivity more than less intelligent women. People think that evolutionary psychology is all about sex differences between men and women, but, in this book, this is the only area in life where the Intelligence Paradox makes different predictions for men and women. The chapter also explains why, despite their greater preference for sexual exclusivity, more intelligent men are nonetheless more likely to cheat on their wives and girlfriends than less intelligent men. Once again, that's not a typo. More intelligent men are simultaneously more likely to value being sexually faithful and more likely to cheat.
Chapter 8 is about circadian rhythms, whether you are a morning person or a night person. I will explain why night owls are more intelligent than morning larks. Believe it or not, more intelligent people go to bed and wake up later than less intelligent people, and it's because it's unnatural for humans to stay up late at night.
Chapter 9 is about sexual orientation, where I will explain why homosexuals are on average more intelligent than heterosexuals. Sexual orientation, at least for men, is largely genetically and hormonally determined before birth. But, within their genetic constraints, more intelligent individuals choose to engage in homosexual behavior more than less intelligent individuals. Genetic influence and individual choice are not necessarily incompatible.
Chapter 10 deals with people's musical preferences. It is generally assumed that more intelligent people prefer certain types of music, like classical music. But why? Why should intelligence affect people's musical preferences? And what other types of music do intelligent people prefer? The answer may surprise you. It turns out that intelligent people also like elevator music.
Chapter 11 is about substance use: alcohol, tobacco, and drugs. Consistent with the Intelligence Paradox, there appears to be some evidence that more intelligent people drink more alcohol, smoke more cigarettes, and use more illegal drugs than less intelligent people. Not only are more intelligent people more likely to consume alcohol more frequently and in greater quantities, they are also more likely to binge drink and get drunk. Yes, more intelligent people are more likely to do stupid things. I will explain why in this chapter.
Chapter 12 is about how intelligent people fail at the most important task in life, which is reproductive success. This is where I make the point that intelligent people are the ultimate losers in life. (Yes, that's an evolutionary pun.) They are really not that great when it comes to important things that truly matter in life. There is some evidence that more intelligent people—at least more intelligent women—have fewer children and are altogether less likely to become parents. I will also discuss what this finding likely means for the average intelligence of the population of the future generations in western industrial societies.
In Chapter 13, I explore some of the other possible implications of the Intelligence Paradox, and wonder what other values may be affected by intelligence. For example, it has been known that criminals on average have lower intelligence than law-abiding citizens. But why? And why are vegetarians more intelligent than meat eaters?
In this book, I continue the system of references that I borrowed in a modified form from Charles Murray5 and that I began using in Why Beautiful People Have More Daughters. For endnotes which simply give citation information, I use the standard endnote reference numbers, like this.1 References which give greater information not contained in the text have reference numbers with brackets, for example.
A Brief Word on the Data
Most of the empirical analyses that are summarized in this book use three different data sets: General Social Surveys (GSS) in the US, National Longitudinal Study of Adolescent Health (Add Health) in the US, and the National Child Development Study (NCDS) in the UK. They are all large, nationally representative, high-quality samples. Two of them are prospectively longitudinal data which have followed a cohort of individuals for many years, in one case, since birth for more than half a century. Here I will briefly describe the data sets and how each survey measures intelligence, which is the central concept in this book. Interested readers who want to know more about the actual empirical analyses can consult the academic papers cited in the chapters.
General Social Surveys (GSS)
The GSS is widely considered to be the best source of data in the world on a wide range of social attitudes and social trends. Its success has spawned counterparts in other nations (such as Canada, the UK, the Netherlands, and Germany), which now conduct their own periodic surveys of social attitudes which are modeled after the GSS in the US.
The National Opinion Research Center at the University of Chicago has administered the GSS, either annually or, more recently, biennially, since 1972. Personal interviews are conducted with a nationally representative sample of non-institutionalized adults over the age of 18 in the US. The sample size is about 1,500 individuals for each annual survey, and about 3,000 individuals for each biennial one. The exact questions asked in the survey vary by the year. The 1972—2008 cumulative data file contains 53,043 individuals and 5,364 variables (although not all variables are measured of all individuals).6
The GSS measures the respondent's verbal intelligence with a 10-item synonyms test. The survey asks the respondent to select a synonym for a word out of five candidates. Half of the respondents in each GSS sample answer 10 of these questions. This is a measure of verbal intelligence, not strictly general intelligence, which is the focus of this book. However, verbal intelligence is known to be very highly correlated with general intelligence.7 In fact, as I explain in Chapter 3, all measures of intelligence are highly correlated with each other.
National Longitudinal Study of Adolescent Health (Add Health)
The National Longitudinal Study of Adolescent Health (Add Health) is a prospectively longitudinal study of a nationally representative sample of more than 20,000 adolescents in junior high and high school (Grades 7—12 in the 1994—1995 school year) in the United States. The Add Health cohort has been followed into young adulthood with four in-home interviews (in 1994—1995, 1996, 2001—2002, and 2008).8
Add Health measures the respondent's verbal intelligence with an abbreviated version of the Peabody Picture Vocabulary Test (PPVT). The PPVT asks the respondent to identify a picture that corresponds to a word out of four candidate pictures. While properly a measure of verbal intelligence, the PPVT has been shown to be a good measure of general intelligence as well.9 For all analyses of Add Health in this book, the intelligence is measured during adolescence (at Wave I) whereas all the outcome measures are taken in early adulthood in Wave III.
National Child Development Study (NCDS)
The National Child Development Study (NCDS) is one of the world's oldest prospectively longitudinal studies. It has followed a population of respondents in the United Kingdom since birth for more than half a century. The study includes all babies (more than 17,000) born in Great Britain (England, Wales, and Scotland) during one week (03—09 March 1958). The respondents have subsequently been reinterviewed eight times (at ages 7, 11, 16, 23, 33, 42, 47, and 51). In each survey, personal interviews and questionnaires are administered to the respondents, to their mothers, teachers, and doctors during childhood, and to their partners and children in adulthood.
The NCDS has one of the best measures of general intelligence of all large-scale, population-based surveys. The NCDS respondents take multiple intelligence tests at ages 7, 11, and 16. At age 7, the respondents take four cognitive tests (Copying Designs Test, Draw-a-Man Test, Southgate Group Reading Test, and Problem Arithmetic Test). At age 11, they take five cognitive tests (Verbal General Ability Test, Nonverbal General Ability Test, Reading Comprehension Test, Mathematical Test, and Copying Designs Test). At age 16, they take two cognitive tests (Reading Comprehension Test, and Mathematical Comprehension Test). From these 11 cognitive tests, a measure of childhood general intelligence is calculated by a statistical technique called factor analysis. The technique measures the underlying general intelligence that explains the individual's performance on all of these varied cognitive tests at three different ages. It also eliminates all random measurement errors which are inherent in any measurement of human traits.
1. Gardner (1983)
2. Stigler and Becker (1977)
3. Friedman, Hechter and Kanazawa (1994)
4. Miller and Kanazawa (2007)
5. Herrnstein and Murray (1994); Murray (2003)
6. Davis, Smith, and Marsden (2009)
7. Huang and Hauser (1998); Miner (1957); Wolfle (1980)
9. tanovich, Cunningham, and Feeman (1984); Zagar and Mead (1983)