Superfreakonomics, By Steven D Levitt & Stephen J Dubner

In Freakonomics, Steven Levitt and Stephen Dubner pulled off an extraordinary trick. They took behavioural microeconomics and the intricate mathematical analysis of very large databases and made a bestselling book of it. Levitt's work is genuinely original, pleasingly rooted in empirical research. It deploys an interesting take on models of incentives and disincentives that recognises how diverse and interactive those currencies are. Dubner gave human form, narrative structure and a good dose of gee-whizzery to the unexpected counter-intuitive results that Levitt's analysis revealed.

Now they're back with more of the same. An inebriated friend decides to walk home after partying rather than drive over the limit. Good idea? Not at all; analysis of driving and pedestrian fatalities in the US reveals that you are many times more likely to be killed walking drunk than driving drunk. Similarly, despite regular media panics around the dangers of sharks, the likelihood of being killed or injured by one is considerably less than of being killed by an elephant. But in media markets elephants means Dumbo and sharks mean Jaws - and Jaws sells papers.

People, like editors, respond to those kinds of incentives. Smart authors of economics books who want to reach a bigger audience and know that there's a market for tales of sex will act accordingly. Thus Superfreakonomics opens with the microeconomics of prostitution in Chicago. Employing prostitutes to pay and interview other prostitutes after seeing a client allowed the authors to amass an enormous and reliable data asset on everything from how much prostitutes stole from clients, to what sex act at what price to a client of what race was transacted.

Compared to the early 20th century, an era for which they have some comparable data, the price of sex and the income of prostitutes have fallen; in the case of oral sex, the fall has been precipitous. Demand overall has fallen, because of the transformation in sexual mores. The normalisation of oral sex has removed a taboo tax.

Pimps, it appears, add real value. Prostitutes who work for them receive higher wages for less sex, and suffer lower levels of arrest and client violence. They are more likely to have sex with a police officer than to be arrested by one. There are sharp spikes in demand for sex around the 4 July holidays; this raises prices, which mobilises a part-time army of workers.

Cute, but for me the serious point of the research was obscured by this kind of aside. The incomes available to predominantly working-class African American women in Chicago remain so poor that prostitution, given all its disbenefits, remains a viable option for many. How and why the nexus of inequality should be so concentrated on these women, and so unamenable to change, should not be left unexamined.

I was left similarly engaged and queasy by their work on Islamic terrorists. Suicide bombers should buy life insurance because it offers powerful disguise when the authorities come trawling through their financial transactions. Extensive profiling by banks reveals that a cluster of around 20 specific actions (big opening deposits, many small withdrawals, not buying life insurance) are together a very high indicator of subversive activity.

Contrary to the thesis that poverty and desperation creates willing martyrs, suicide bombers do not generally come from poor families or lack education. Their deeds might be better thought of as "civic action on steroids". But does that mean that the grinding poverty and inequalities of the Middle East and Western Asia are not factors in the emergence of jihadism? This reliance on statistical correlation, alongside the bracketing of structural and cultural questions, is the strength and weakness of the book. It leads to some brilliant questions – such as why, given that kids occupy rear seats in cars most of the time, do we not redesign seat belts to fit them rather than use child seats, so rarely fitted correctly? On the other hand, there is a kind of glibness at work. The arrival of oil certainly saved Western cities from the environmental crisis caused by horse-drawn transportation and its mountains of organic waste, and the switch to mineral lubrication oils saved the whale from extinction. But to suggest that "oil was a cheap and simple fix" is to bracket so much that my incentive to agree with them collapsed.