Monday 29 January 2018

Economics, the evidence-free discipline – from the horse’s mouth

If economists wished to study the horse, they wouldn’t go and look at horses. They’d sit in their studies and say to themselves, “What would I do if I were a horse?”
Ely Devons, British economist

Ever since I started working in a job that periodically puts me in economics lectures, I’ve noticed that economists have a very different idea of what constitutes evidence for their statements than what scientists do. And when I say “noticed”, I mean it’s been thunderingly obvious. The health sciences, in particular, spend hours upon hours drumming into their students’ heads how much it takes to call your practice “evidence-based”. In economics lectures I’ve heard lecturers say outright, “If your analysis of the data disagrees with economic theory, trust economic theory.” Economics is at about the stage medicine was at in the mid-nineteenth century, when blood-letting and cold showers were the go-to treatment for every ill because physicians knew how the body worked, damn it, and didn’t need jumped-up empiricists coming in telling them how to do their job thank you very much.

A nineteenth-century physician practising bloodletting

But I don’t think nineteenth-century physicians ever proudly declared that their theories were evidence-free and thought it a mark of superiority. Yet I encountered economists saying exactly that, in an article from only a year ago, in a debate with a libertarian on Tumblr recently. I’m referring to the Mises Institute’s “Ten Fundamental Laws of Economics”. The list includes some uncontroversial items, but also contentious ones such as “Productivity determines the wage rate”, “Labour does not create value”, and “Profit is the entrepreneurial bonus”. It ends with Law 10: “All genuine laws of economics are logical laws.” This is explicated as

Economic laws are synthetic a priori reasoning. One cannot falsify such laws empirically because they are true in themselves. As such, the fundamental economic laws do not require empirical verification.

Which basically translates to “Anyone who disagrees with us is wrong by definition.” This is not about economics being a “soft science” rather than a “hard science”. This statement makes economics as practised by the Mises Institute not a science at all.

The phrase “synthetic a priori” is, in this context, pure bafflegab, but unfortunately it’s going to take a bit of unpacking. The philosopher Immanuel Kant divided truths along two lines. First, they can be “synthetic” or “analytic”. An analytic truth is basically simply a definition of a word: the usual go-to example is “All bachelors are unmarried,” which is true because being unmarried is part of the definition of being a bachelor. A synthetic truth is one that can’t be derived from the definitions of words alone, such as “I have two cats.” Second, truths can be a priori or a posteriori – Latin for “from before” and “from after”, respectively. An a priori truth has to be true in any conceivable universe; you know it is true before you go investigating. “One plus one equals two” is an a priori truth. An a posteriori truth is one that might or might not be true, and that you therefore can’t know is true until somebody investigates, such as “I have eaten the last of the cheese.”

Two lines of distinction potentially divide a set into four subsets, in this case analytic a priori, analytic a posteriori, synthetic a posteriori, and the one we’re interested in, synthetic a priori. Three of these are uncontroversial. All philosophers agree there is no such thing as an analytic a posteriori truth, and most philosophers agree there are analytic a priori truths and synthetic a posteriori truths. The big disagreement over Kantian philosophy is over whether there is such a thing as a synthetic a priori truth – whether there is anything that has to be true in any conceivable universe, but that can’t be reduced to definitions of terms and logical deductions from such definitions. The Mises Institute puts economic principles in this category. What would this mean?

Kant himself populated the synthetic a priori category with mathematical truths like “One plus one equals two”. Personally I’m inclined to the school of thought that mathematics is in fact analytic. There are arguments to be had on both sides, and I won’t go into them. I suppose the four-colour map theorem might count as synthetic a priori – no-one has ever created a map that needed more than four colours to fill it without any two areas of the same colour touching, and somebody has proved that this is indeed impossible via a computer program, but the proof is too complex for a human mind to comprehend. The point is that no-one can even imagine such a map. It’s inconceivable. For the Mises Institute’s “laws” to be synthetic a priori, it would have to be the case that no-one could even imagine a world in which productivity did not determine the wage rate.

The Mises Institute might respond that someone who thinks they’re imagining a world in which productivity doesn’t determine the wage rate is kidding themselves, just as someone who thinks they’re imagining a five-colour map is kidding themselves (your mental picture is just a vague squiggle; you aren’t filling in the details). But it is not at all difficult to imagine a shareholder-profit-maximizing corporation deciding to funnel 100% of the profit margin from a productivity boost into shareholder dividends instead of wages. Nor is it hard to imagine every other firm in the market doing the same, thus leaving no competing employer to whom the employees could defect. The Mises Institute is committed to the claim that this scenario is not merely implausible but unimaginable. Either that, or the phrase “synthetic a priori” in their statement is bafflegab.

Of course, if economic principles are not synthetic a priori, then what the Mises Institute is doing is making up excuses for why their beliefs shouldn’t be exposed to empirical testing, and if you find someone doing that then they’re up to something dodgy. My Tumblr correspondent claimed that societies organized according to these theories could produce and distribute goods and services “way more [efficiently] than any Marxist or Keynesian society can.” Well, that’s an empirical question. I have sat through more than enough economics lectures to be well aware of the theories as to why the free market is meant to be the most efficient means possible of maximizing production and optimizing distribution of goods and services. But only empirical data can tell us whether it is the most efficient means possible. Any attempt to put it beyond the reach of empirical investigation, such as the Mises Institute is here guilty of, suggests that its proponents fear it would disappoint them.

And before someone asks, yes, for all the respect I have for the Marxist community for the work they do in activism towards social change, I do have to concede that Marxists are equally prone to shielding their theories from the possibility of being refuted by reality. I have written about that elsewhere on this blog, if you care to go looking. But Marxist theories don’t at present dominate the global economic system. Capitalist ones do. The point is that only reality can tell you what’s true. Nineteenth-century medicine wasn’t dislodged by some other theory-driven approach; it was corrected by recourse to empirical evidence. In economics as in medicine, when the “experts” cling to their pet theories over reality, people die. We need evidence-based economics and we need it yesterday.

No comments:

Post a Comment