FREAKONOMICS: A ROGUE ECONOMIST EXPLORES THE HIDDEN SIDE OF EVERYTHING
By Steven D. Levitt and Stephen J. Dubner,
Penguin, $25.95

Imagine you treated all kinds of things that you take for granted as mysteries. What determines what names parents choose for their children? When will people whom one might describe as possessing average honesty ? school-teachers, sumo wrestlers, perhaps even cricketers ? cheat? What explains variations in crime rates? What are the similarities between the organizational cultures of seemingly disparate organizations like street drug gangs and Fortune 500 companies? What determines when people feel generous? What makes for better parents? Imagine further, that you have a whole set of conventional answers to these questions. But what if those answers are completely mistaken? Freakonomics, though not as original and earth-shattering as claimed, is a wonderful entr?e into social science. It will induce a sense of wonder about all kinds of quotidian and not-so-quotidian aspects of our social life. And it will prompt you to look deeper for patterns that might help resolve your sense of puzzlement.

Steven D. Levitt, a professor at Chicago and one of the most felicitated young economists in America, caught the public?s attention with a paper explaining why crime rates in the United States of America had dropped dramatically in the Nineties. Levitt went through all the explanations ? tougher policing, improved economies, tougher gun control laws, and so on ? and came to a rather startling conclusion. The drop in crime had something to do with the legalization of abortion after Roe vs Wade. Levitt, following many other studies, argued that a disproportionate number of crimes were committed by children of broken homes and, what might be termed, unwanted children. The legalization of abortion reduced the number of unwanted children, and hence crime. But what made this politically incorrect and explosive conclusion harder to dismiss as armchair theorizing was the impressive data Levitt put together, including the rather startling finding that crime rates had dropped first in the five states that were also the first to legalize abortion. The ensuing debate over the data shows no sign of abating, but Levitt was able to put this forward as a serious social-science hypothesis.

Whether or not one is convinced by Levitt?s data, the manner in which he argued his case exemplifies what I think is the central methodological theme running through Freakonomics. First, Levitt is very good at exposing what social scientists conventionally call selection bias. This is the phenomenon where we come to explain a particular outcome only by looking for a specific set of explanations, without considering and testing the full range of hypotheses that might explain it. Indeed, the limitations of previous crime studies were their selection bias. What makes Levitt inventive is that he manages to expose selection bias in the most unexpected places.

Second, much of the charm of the book lies in its ability to marshal data patterns. The book is full of quirky data. For instance, a bagel-seller who kept meticulous records of how patterns of voluntary contributions to defray the expenses of bagels in different offices varied, allows Levitt to establish quirky facts ? employees in smaller offices were contributing more than in bigger ones, weather affected the mood of donors, in the period immediately after 9/11, people in this group were more honest. Or he has data to show that parents who, by and large, have little hope and no prospects for advancement, tend to give their children names that are, to put it mildly, not classy. Again the charm of the chapter turns on accessibly presenting huge amounts of data to explain as quotidian a phenomenon as naming.

Much of the interest of the other stories turns on data patterns as well. Levitt manages to get access to the accounts ledger of a street gang dealing in drugs, which allows him to conclude that in terms of incentives internal to the gang, it does not behave differently from Fortune 500 companies. In some cases, inventive probing of data allows Levitt to detect cheating. He was instrumental in showing that many Chicago school-teachers were fiddling with tests of their students in order to achieve learning improvement targets. If this sounds obvious, the means by which Levitt arrives at this conclusion are anything but. Indeed, Levitt?s analysis of data will remind you that discovering patterns is an extraordinary art.

Other interesting cases in the book turn on the old cornerstone of economics: incentives. But Levitt?s distinctiveness consists in not assuming how incentives operate. Instead he takes an empirical approach, what is now known as behavioural economics, to determine how incentives actually operate. And it turns out that both, designing proper incentive structures and the range of incentives that people operate under, are not quite simple. Take a small example from the book. A day-care centre in Israel decided to impose fines on parents who were coming late to pick up their children. And oddly enough, the incidence of late pick-ups increased! This could partly be explained by the fact that the fine may not have been high enough, but Levitt suggests another possibility. In some instances, substitution, an economic motive for a moral motive, actually has perverse consequences: a surprising conclusion for an economist.

At a theoretical level Freakonomics is not particularly startling. In a way, its central themes are now obvious: look harder for patterns in data, momentous outcomes can be caused by very remote and seemingly trivial causes, avoid selection bias, conventional wisdom is often wrong, information hoarding is a key mechanism through which power is exercised and finally, incentives matter. But Sherlock Holmes said to Watson, ?Your grasp of the obvious amazes me.? The virtue of Freakonomics is to remind us that the obvious may not be so obvious after all.