After reviewing our debate (which I’m reposting below), I realized there were a few points where I misunderstood what Friedman was saying. It will not surprise you to hear that I think there were also several points where he misunderstood me. (For one obvious example, of course I know that modern physicists think the universe is non-Euclidean. That has nothing to do with my point: Even in light of that fact, if a student in 7th grade tried to “prove” theorems in geometry by appeal to experiments, he should be marked down heavily for totally misunderstanding how geometry works.)
Rather than re-litigating every twist and turn in the debate, let me try a new way of motivating my position (and I’ll send this post to David to see if he’s interesting in pursuing the discussion). I claim that the people who today consider themselves free-market economists got that way not because of empirical research, but because they read ingenious thought experiments from the likes of Bastiat, Say, Henry Hazlitt, Milton Friedman, Ludwig von Mises, Friedrich Hayek, Murray Rothbard, David Friedman, and Steve Landsburg.
Take free trade in particular. Nobody believes in free trade because of regression analyses, or of “natural experiments” in which some country dropped tariff barriers as an offshoot of some other, unrelated event. Rather, people believe in free trade because of brilliant explanations such as David’s own “Iowa Car Crop” (which I first read in Landsburg’s Armchair Economist, and which Steve correctly described as “a thing of beauty” that can “change the way we see the world”).
Going further, to the extent that you might abandon your teenage belief in free trade, it would again be due to a thought experiment, a simplified tale showing how (say) a big country could affect the world price of cotton and so maybe a small tariff could, in principle, be welfare-enhancing. Even here, the disputes would be more philosophical rather than empirical. Nobody would convince someone to change his mind on free trade, by pointing to empirical evidence. For example, I remember when Pat Buchanan and some others pointed to the strength of the US economy when there were high tariff barriers. That didn’t faze me a bit (and still doesn’t), because there are a million moving parts in these historical analyses. I think the US would have been even richer had they adopted free trade back then.
Another consideration: Intro textbooks or even pop books on economics often start with a section about, “How to Think Like an Economist.” These principles or “laws” of economics are almost always non-falsifiable. They are things like, “There are always opportunity costs,” or, “There’s no such thing as a free lunch,” or, “Decisions are made on the margin.” You don’t go out and “test” whether these statements are true, instead you decide a priori that you are going to use these principles to parse the evidence.
Yes, if we’re trying to assess whether the Great Recession was more due to supply-side or demand-side “shocks,” then that is an empirical question. But it’s not a “law of economics” to ask whether the recession from 2007 onward was due to the supply or demand side.
In contrast, the “laws of physics” are entirely empirical. In principle, they might be different in 200 years, as physicists learn that Nature obeys a different set of rules from what we currently believe. In contrast, when the “marginal revolution” occurred, it wasn’t because of new data, it was because economists came up with a better mental way of organizing and explaining what humans had known for centuries.