In praise of econometrics…
Shirley Williams once said, “Economics are all right if you don’t inhale”. She also helped push through LibDem support for HASCA12 (the NHS Privatisation Act) so shredding any respect I might once have had for her.
Actually, we do need to inhale economics. We need to study it more, and better.
A friend of mine and I were discussing the recent decision of a mutual friend’s daughter to read economics at university. He blamed the discipline for the present state of the economy, and thought that actually studying it was the problem. The existence of an academic discipline of economics isn’t the problem; it’s the way it is now taught and practised that is. It should be about working out how we can make everyone better off, eliminating destitution instead of making a few people much richer. It always used to be the point: Alfred Marshall always kept a picture of a destitute street urchin in his rooms, to remind why he did what he did.
Students at Manchester University, followed by others around the world, have already realised this. The Post-Crash Economics Society was established earlier this year to challenge the neo-classical curriculum that predominates in most university economics departments. Andy Haldane, the much-respected chief economist of the Bank of England, wrote a positive foreword to the report they published in April, in which he outlined – and criticised – the way economics has developed, seeking to lay firm theoretical foundations on which to build mathematical models, just like the physicists were doing. He thinks that the problem is that the economists wanted to be too like the physicists, instead of the historians. I think the problem is that perhaps they didn’t want to be enough like the physicists; although the history matters too.
The big difference between economists and physicists is that the theoretical physicists share the same departments, the same journals, the same laboratories, the same canteens as the experimental physicists, a small fraction of the budget and, ideally, equality of prestige. The experimentalists probably beat them in prestige, because they have bigger budgets and better toys. Where are the experimental economists? (Partially, in the banks). In physics, experiment and theory go together. Experimental results guide the theoreticians’ theorising; the theoreticians’ predictions guide the experimenters’ experiments. Classical economics operates without the experimental division; it is trying to do science without the empiricism. So it becomes just another branch of theology; the fact that the language it uses is advanced mathematics rather than Latin or Greek makes it no more reliable than theology, because it has no empirical base. Econometricians, the nearest thing to experimental economists, aren’t given parity of esteem in the department; they’re considered a rather weird bunch and a whole school of economics considers them irrelevant. They, and their discipline, are the very foundation stones of positive economics. They should have the big offices and the big budgets, and feed lots of lovely data to the theoreticians.
Gradually, however, we are beginning to realise that there is a lot of data out there. We can be critical about it. Experimental physicists are very critical of the data they collect. They understand that there is never an exact answer; they know that the best they can say is that “the probability that the answer lies between these two values is such-and-such”. They analyse and allow for the errors in their instruments, the most important and error-prone of which is the experimenter, or observer, himself. Economics needs to develop the physicists’ critical analysis of bias in measurement. Double-blind clinical trials are structured so as to eliminate the possibility of the experimenter’s bias affecting the results. Is there anything like such rigour, anywhere near the heart of the discipline of economics? Economists still think along the lines of “the data will say anything you want if you torture them enough”.
There are signs, around the discipline, that things are beginning to change. Piketty’s work has been data-driven; he isn’t the only one. The Manchester students are arguing for a discipline based on a critical analysis of historic episodes, the policy responses and the results; and how these events drove the development of economic thought. There is much less historic data than there is current data; and a lot of current data is hidden because it’s about people’s money; and it’s proprietary because it can be used to make money. Surely better that it should be used to inform the policies on which the nation is governed? Yet another argument for transparency: if current financial data is available for analysis, anonymised, there’ll be lots of empirical data from which to make economics a sound academic discipline.
The idea of evidence-based economics, without political bias, seems to be a pipe-dream today. On the left and on the right, academic economics is partisan. There are schools of economic thought: the Chicago school, the Cambridge school, the Austrian School. They’re selective about the data they choose to support their arguments. They’re also hand-in-glove with Government and the financial services industry; their decisions affect everyone’s lives. It’s irresponsible not to want to try to find the right answers, and you can’t get the right answers without experiments. Lots and lots of them.
No economic paper should ever be published without asking the question: how do I test this hypothesis? Is it testable? if not, it’s theology, not published in this journal. If so, let’s test it. Don’t let anyone publish another paper in the same vein without some further research, data driven, empirical research to support it.