As a politics student I never much liked the term ‘political science’. It seemed to imply that those of us doing the studying could somehow put our own prejudices aside, achieve a pure state of dispassionate reason, and study other people as if they were rats in a laboratory, rather than people just like us. In the same way, the problem with ‘evidence-based policymaking’, which is one of the buzzwords of the day, is that the certainty of the sciences can never be fully applied to the social sciences, because it’s never possible to run fully controlled experiments on people’s lives. Correlation, in trying to analyse the effect of one influence among millions on how people act, can rarely prove causation: at best it can give us an educated guess, which falls some way short of pure science.
Economics is a good example: the mountains of data it has accumulated still refuse to point to a clear conclusion. The more the discipline progresses, the more disagreement there is between the Keynesians and the Austrians, both of whom can torture the statistics as they see fit and produce apparently convincing arguments. The more I read from either, the less clear I am about which side is right: in theory the Keynesian position is appealing, but in practice the amount of responsibility it places in relatively few hands is a problem too. Neither Ed Balls nor George Osborne really has a clue what will happen if we fiscally retrench, stimulate demand, put up interest rates, print more money, or whatever. Each can quote roughly appropriate and superficially attractive examples from history to support his case, but neither can possibly quote examples that cover all of the variables that are in play in what is today an enormous, unique and uncontrollable experiment. Beyond the very short term, our economic outlook is simply unknowable: unfortunately, politicians and commentators are respected for making confident predictions (even if those turn out to be completely wrong) so you won’t catch many of them admitting it. Science and social science are, as it were, overlapping magisteria: the bigger and more complicated the political question, the weaker our ability to deal with it by scientific method.
The phrase ‘evidence-based’, then, when applied to politics, where there is rarely a definitively correct answer, is one we ought to be wary of. Indeed, since governments so often debase evidence-based policymaking into policy-based evidencemaking, we have to be even more sceptical than usual when dealing with any politician, journalist or public figure who purports to be operating on the basis of ‘the evidence’. Politics is about how we run society, what we do with our and other people’s money, how we share our time on earth with other people, how we order the fundamentals of our lives. Unlike crunching data from radio telescopes or splicing plant genomes, these are emotive issues, and it is impossible to be fully dispassionate and to disregard our own experiences of our lives and of society. Anyone who claims to be able to do so should warrant our suspicion. Anyone who can genuinely do so is a sociopath.
If we are being honest, we all — scientists and academics included — at some point or other fall into one of the following:
- anchoring trap — giving disproportionate weight to the first item of information received
- status quo trap — bias towards maintaining the current situation
- sunk cost trap — making decisions which justify previous decisions, even if these are no longer sound
- confirming-evidence trap — concentrating on finding evidence to support a favoured option
- framing trap — posing the question for the decision in a particular way
- overconfidence trap — overestimating the accuracy of estimates and forecasts
- prudence trap — being over-cautious
- recallability trap — giving more weight to recent and dramatic events
(Hammond, J S, Keeney, R L and Raiffa, H (1998), ‘The Hidden Traps in Decision Making’, Harvard Business Review, September-October)
As ever, Orwell is ahead of me:
Implied in the demand for more scientific education is the claim that if one has been scientifically trained one’s approach to all subjects will be more intelligent than if one had had no such training. A scientist’s political opinions, it is assumed, his opinions on sociological questions, on morals, on philosophy, perhaps even on the arts, will be more valuable than those of a layman. The world, in other words, would be a better place if the scientists were in control of it. … The fact is that a mere training in one or more of the exact sciences, even combined with very high gifts, is no guarantee of a humane or sceptical outlook. The physicists of half a dozen great nations, all feverishly and secretly working away at the atomic bomb, are a demonstration of this.
Go and read his short piece in full, then come back to this one.
The problem is at its most acute when, as with the Keynes/Hayek question, an issue presents us with binary choices. The drugs debate is a good example, and one where the liberal side (including the BBC’s Mark Easton, who no longer makes any pretence of impartial journalism) claims to have monopolised the evidence, citing Britain before prohibition, contemporary Portugal, and elsewhere as examples of successful legalisation and harm reduction. Many of these arguments are convincing, but then so is the argument that giving visitors boarding passes that read WARNING: MANDATORY DEATH SENTENCE FOR DRUG TRAFFICKERS UNDER SINGAPORE LAW, and carrying out the threat, has resulted in a low drugs problem there. Though they are clearly incompatible, essentially I agree with both sides: either extreme would be an improvement on the status quo. But there is evidence on both sides, and neither side can conclusively demonstrate that their position, when applied to all of the other circumstances that are unique to our socio-economic situation, will work as they intend it to (for example, on the one hand the relaxation of late-night licensing laws failed to turn us into a continental-style cafe culture and merely extended the existing problems later into the night in the areas that adopted it, but on the other we had the legality and availability of ‘legal highs’ like mephedrone to thank for a drop in the use of other, nastier drugs like cocaine — at least until we made them illegal highs). Both sides are advocating a leap in the dark, to a greater or lesser extent: but both are on firm ground to say that it is impossible to call the compromise that is our present policy a success. What we ideally need is a global referendum in which the status quo is not on the ballot. Much the same applies to the debate on crime and penal policy, where again the government seem to be advocating a mixed policy of community-based rehabilitation rather than short-term imprisonment, but without the vast amounts of money that such a policy would clearly need to have any chance of working — this seems to be the same sort of compromise as we have in drugs policy, and seems to me to have about as much chance of working.
The best way forward for politics and political science is to try to apply the methods of the information revolution: by acknowledging the limitations of social science as a discipline and the limitations on how much any individual or small group of people can know, by admitting the role of individual and social experiences in shaping people’s opinions, and by aggregating as much information as possible when formulating policy. Policy should operate in the free market of ideas as price does in the free market of exchange: in the modern parlance, we should crowdsource it. This is the opposite approach to fascism and communism (both of which, Orwell tells us, were supported by plenty of scientists and academics in their time), which both ultimately failed because too many decisions were being made by too few people with too little information. The representative democracy we currently have — whereby we are ruled by an elite making what it thinks are the best decisions on our behalf, with a degree of our consent — is half way between the two positions. The ultimate, revolutionary, goal — in synthesis with the revolution in access to information, which inclines me towards the rational optimist camp — is for as many people as possible to have as much of a say in as many decisions that affect them as possible. Direct democracy may be some way off yet, but it is the direction in which we should be agitating to travel.