One of the most important aspects of the financial crisis, but one that is often at risk over getting over-looked, is its consequences in relation to research and knowledge.
To try and explain what I mean by this, we need to start by thinking about economics. Economics has presented itself as the ‘hardest’ of the social sciences, favouring quantification above qualitative data, and prizing its version of scientific rigour that is organised around statistical modelling in various forms. It claims to follow a hypothetico-deductive model in that it tries to test theories through the use of empirical data. It follows the positivistic mantra of variables not being real unless they can be measured.
Economics, as a discipline, receives very large amounts of funding from research councils, and even now is marked as a priority area for further research. This seems to me to be a huge mistake.
This isn’t because we don’t need to understand economics better, its because the discipline of economic has failed on a colossal level to meet its own demands for methodological rigour and relevance, and the wider societal need to contribute to our understanding of the economy.
Economics is still largely based on Newtonian, normal distribution-driven models of mathematical modelling. Writers such as Ormerod, Mandelbrot and most recently Taleb have shown the mess this results in. Economists confidently predict outcomes based on models that bear little or no resemblance to anything anyone experiences in the world, are shown to be wrong again and again, but continue as if their methodology remains sound and sensible. Every time there is a financial crisis, you might have noticed how economists label it as a one in a hundred year event. But these events keep coming. Economic models assume away the reflexivity of humans (which Soros damns them for in his work), presents financial outcomes in a Gaussian, normal distribution (which Mandelbrot has shown is entirely inappropriate) and gives predictions that are wrong over and over again.
In a time when the economy is booming, it doesn’t tend to matter that the predictions of economists aren’t entirely accurate, as basing what you think will happen tomorrow on what happened yesterday won’t result in a problem. The problem occurs when the economy goes through a turning point – the most important time to be able to predict – and economists fail to predict it, or to be able to provide any sense of why it happened or how long the particular downturn will last. And yet we keep going back to them, asking for more spurious predictions of when things will get better, as if they are able to get it right this time.
This has huge implications for the way we carry out social policy research. There are strong movements within social policy to utilise economic approaches because of the supposed rigour they will generate. I think this is absolutely bonkers. Economic approaches don’t work in economics, so why on earth should they work anywhere else? I have no problem with quantitative modelling provided it is based on sensible assumptions and it is careful in its recommendations. Economics has not done any of these things. It has failed us, and its about time that we started addressing that problem rather than trying to copy its methods in academic contexts where it is even less suited than its home area.
Social policy does need theoretical roots. But there is no reason why these can’t be pluralistic, so long as we can understand each other’s languages, and we make explicit our assumptions about the world when carrying out research. Social policy based on economics often fails to do any of these things – it presumes that, because it is based on economic principles, that it should be accepted on those terms. It’s about time we challenged that. The city economist Roger Bootle has just published a book showing the failure of academic economics to address real world problems. Economics (the discipline) doesn’t work when addressing the problems of economics. Advocates of its methods in other disciplines need to be challenged far more often than has so far been the case.