There’s loads of surprising stuff contained in Freakonomics, Steven D Levitt and Stephen J Dubner’s romp through the application of economic principles to sociological data. The thesis that legalised abortion leads directly to lowered crime rates – that’s pretty shocking, even if you’re someone (like me) who thinks that abortion should be legal and that one of the strongest arguments for its legality is the unhappy circumstance of unwanted children. Or the analysis of cheating by sumo, which cleverly presses the data from sumo wrestlers’ championship bouts to discover the circumstances under which a wrestler seems to be willing to hand the win to his opponent. The data are structured to account for variables, interrogated for controls, analysed – and the outcomes are often revelatory, not so much for the conclusions, but for the implicit argument that something as seemingly ineffable as human behaviour can be measured in this way.
There’s an argument, pushed by Meghan Falvey for n+1 magazine, that the book is too tightly focussed on a cost-benefit analysis and incapable of accommodating outcomes which fall outside of the financial, but I don’t think that’s a fair representation of Levitt and Dubner’s work. As a commentary on their method, it’s roughly acceptable: analysis requires that an approximate value can be set on the input and output of a transaction, and some things (money, lifespan, crime per capita) are easier to quantify than others (happiness, love, the dappling of sunshine on treetops). But Falvey’s ethical objection – that Freakonomics espouses a narrow view of human behaviour as reward-driven and rational, and so falls into the service of stakeholder-society welfare restriction – seems simply wrong.
After all, Freakonomics says that everyone from most powerful to least is influenced by calculations of self-interest: that’s an argument that recommends putting limits on activities where one person holds an advantage over another. And more than that, a lot of the behaviour described in Freakonomics barely qualifies as rational anyway. The section on baby-naming picks up on the phenomenon of names cascading down the social classes from high to low to total disuse. Levitt and Dubner argue that this demonstrates that parents tend to ‘name up’, picking a name which they associate with wealth or success and giving it to their child in the hope that the glamour of the name will be absorbed by their offspring. This is palpably irrational. The name, they argue, is a symptom of the parents’ socio-economic background, and it’s the background which determines the child’s outcome, not the name. So while the progress of names through society stems from a sort of logic, it’s not a logic that demonstrates any sort of grasp at all on what will be best for the child.
Outside of directly acting rewards and penalties (‘If I throw this fight to an opponent they’ll throw me a fight when I need it’ sort of thing) it’s really, painfully hard to discriminate the best course. The relationship between action and outcome isn’t always immediately obvious, and when you start to deal in probabilities rather than certainties, human rationality begins to strain at the seams. “The risks that scare people and the risks that kill people are very different”, says risk communications consultant Peter Sandman, quoted in Freakonomics. “When hazard is high and outrage is low, people underreact. And when hazard is low and outrage is high, they overreact.”
Statistics are the best tool we have for assessing risk and tempering outrage, and they’re a tool which very few people are able to use effectively or consistently. Why don’t we prize the analysis of numbers as highly as the analysis of words? But terms like ‘control’, ‘risk’ and ‘bell curve’ flit inconsistently through the curriculum while ‘metaphor’, ‘simile’ and ‘superlative’ get close attention from the primary curriculum on. When poor benighted Jeni Barnett says (extract taken from Sciencepunk’s transcript):
We have evidence, however much people say we don’t, we have evidence that if a child’s immune system is weak; my daughter was one of them, she was very asthmatic as a child, she could not have received that triple vaccine, she couldn’t have done it so I made a calculated decision that I didn’t want to go there.
she demonstrates pretty clearly the impossibility of making a “calculation” where you don’t have or deny yourself access to the data. And of course, it’s not just Jeni: she just happens to have stated particularly beligerantly what plenty of far more influential people (including, disappointingly, Ian Hislop and Paul Foot, as well as the usual idiots) either believe, or used to believe, or believed to be plausible. The argument of Freakonomics isn’t that people are capable of being rational and acting consistently for the good: it’s that people might make slightly more rational choices, given the right information and possibilities of action.