The sequel (or freaquel) was not as sharp, incisive and surprising as Freakonomics but it’s still a good read. Like Freakonomics, it avoids theory and goes for unusual and riveting yarns. Per mile, drunk walkers run higher risks than drunk drivers. Elephants kill many times more people per year than sharks. More soldiers died per annum in the years before the US went to war in Iraq and Afghanistan, astonishing until you realise that the army was much bigger. You get the idea. Data contains nuggets that need unearthing by smart data miners.
Challenges liberal orthodoxies
The authors search a little too hard for sensational examples and the follow up to the drug dealing in Freakonomics with the economics of prostitution this time round, seems a little forced. Another rather odd undercurrent is that they love to dig up results that challenge liberal orthodoxy. Feminism is blamed for poor schooling, Ramadan (fasting during pregnancy) for a spike in disabilities. Car seats are seen as killers. At times it also moves towards behavioural economics and there’s worrying example of how the Americans with Disabilities Act (ADA), designed to protect disabled workers, led to fewer jobs for the disabled. Why? Employers, scared of the Act, hired fewer disabled staff.
Some of the examples are downright depressing. In a son worshiping country like India, where a boy is a profitable, retirement plan and a girl a dowry fund liability, there are 35 million missing women (aborted and murdered),and, they claim, endemic domestic abuse. The unlikely, partial solution to these problems turns out to be the introduction of TV, which lowers both birth rate and domestic violence. This is fine but it tends to both present the problem rather crudely and then say ‘problem solved!’
For me, the book springs into life when it describes how an algorithm that looks at the financial affairs of bank customers, combined with other variables, can be used to identify potential terrorists. You start to get a feel for how analytics experts approach data sets and fine-tune their algorithms to narrow don on their goal. This is interesting, but frightening, as it has led the NSA in the US to massive surveillance, not only of its own citizens but foreigners also.
The book ends with an extraordinary experiment. What happens when you teach a monkey to use money? Turns out they’re just as miserly and subject to ‘loss aversion’. They even exchanged money for sex! They had to stop the experiment as monkey capitalism had quickly led to prostitution. Who knows where it would have gone next?
In learning Levitt and Dubner have a habit of uncovering uncomfortable truths. In their first book it was cheating teachers and parenting myths. Here it’s the impact of feminism on the quality of women teachers. Before you scoff – read the piece as it’s convincingly there in the data. There is the well known Eriksson finding that ‘talent’ is overrated and that practice that makes perfect but on the whole it’s not as rich in educational examples as its predecessor.
Although the book is really an endless series of fascinating examples, beneath the fun is serious analysis. That’s the point, that some sets of data have stories to tell, unexpected stories, that can be unlocked through digging deep and applying the maths. There’s more to all this than fodder for dinner-party anecdotes. It’s all about data and the power of good mathematicians and economists to discover interesting correlations, especially in large amounts of data. This is a book for the beach rather than background research but once you’ve gorged yourself on amusing tales and stocked up on tales to tell, you’re likely to have an appetite for more theory.