Rodrik supports the mathematical nature of economics as bringing clarity of meaning, and argues that the subject is far more applied and empirical than its detractors realise. But he criticises large-scale macro models and time series regressions. “I cannot think of an important economic insight that has come out of such models,” he writes. He also flags up the lack of testability of many economic models: they purport to be deductions from theoretical principles, but as they are ‘deduced’ to explain a particular phenomenon (credit rationing, say), then that phenomenon cannot be used to test the model. “Very few of the models that economists work with have ever been rejected so decisively that the profession discarded them as clearly false.”
Pointer from Mark Thoma.
My line is that economists deal in non-falsifiable interpretive frameworks. Read Coyle’s entire post. She makes Rodrik sound like someone I would agree with, although not everything I have read of him would indicate that.
The conversation between Tyler Cowen and Dani Rodrik keeps circling back to methodological issues. For example, Rodrik is wary of overrating randomized control trials. Rodrik suggests that graduate students should spend more time in the real world.
I keep thinking of the quote of Minsky writing that economists are well trained but not well educated. You are trained to solve equations. Nowadays you are trained to do the sort of narrow empirical studies that Rodrik thinks are overrated. But you are not educated in history or financial institutions or secular changes in the economy.
Also, Noah Smith has more to say.
philosophical empiricism is far more frightening for economists than for natural scientists. Living in a world of theoryderp is easy and comforting. Moving from that world into a Popperian void of uncertainty and frustration is a daunting prospect. But that is exactly what the credibility revolution demands.
Read the whole post. As I read it, he thinks that economists will have to reconcile themselves to less theory and more empirical work. I do not really agree:
1. I think that economists rely a lot on what I call interpretive frameworks. These do not have standing in philosophical empiricism, because they are not falsifiable.
2. Philosophical empiricism does not provide a guide to evaluating interpretive frameworks. Unfortunately, economists have not thought about this question. Frameworks become popular because they are tractable or interesting, and they stay popular without ever being evaluated for usefulness.
3. I think that an interpretive framework is strong if it offers explanations in many contexts, if it does not encounter too many anomalies (phenomena that seem to run counter to the framework), and if it is reinforced by other beliefs.
4. Supply and demand is an example of an interpretive framework that is very strong. That is, it seems to explain a lot, one rarely encounters anomalies, and it is consistent with other beliefs that we tend to hold.
5. Keynesian macro is an example of an interpretive framework that is not very strong. Many anomalies have cropped up over the decades: the ability of the U.S. economy to rebound after World War II in spite of the staggering drop in government spending; the breakdown of the Phillips Curve in the 1970s; the failure of many Keynesian stimulus policies in many countries, including the U.S. And Keynesian macro is notoriously inconsistent with many other beliefs that economists tend to hold.
“My line is that economists deal in non-falsifiable interpretive frameworks.”
“Keynesian macro is an example of an interpretive framework that is not very strong. Many anomalies have cropped up over the decades”
Wouldn’t those anomalies mean it’s been falsified?
Dang it, Steve, you beat me to this by ONE MINUTE. 🙂
That’s what you get for taking the time to edit those quotes to make your post more compelling than mine. As every tool who’s ever written “First!” at the top of a Marginal Revolution comment thread knows, timing always trumps substance.
Statement 1: “[I]nterpretive frameworks…are not falsifiable…Philosophical empiricism does not provide a guide to evaluating interpretive frameworks….Frameworks…stay popular without ever being evaluated for usefulness.”
Statement 2: ” I think that an interpretive framework is strong…if it does not encounter too many anomalies…Keynesian macro is an example of an interpretive framework that is not very strong. Many anomalies have cropped up over the decades: the ability of the U.S. economy to rebound after World War II in spite of the staggering drop in government spending; the breakdown of the Phillips Curve in the 1970s; the failure of many Keynesian stimulus policies in many countries, including the U.S.”
Aren’t you contradicting yourself? Calling a framework “not strong” because it has “anomalies” = rejecting a theory because it conflicts with data.
No, that is not how it works, for two reasons.
Firstly, calling something an anomaly is a way of saying you won’t let the data upend your theory. Newtonian gravitation couldn’t explain the precession of Mercury. That didn’t cause Newtonian gravitation to be rejected, because it was the best available theory, and it explained lots of other stuff. It just meant that this anomaly was marked off and metaphorically “put to one side” as something we didn’t understand. At any given time, a field may have any number of “anomalies” – indeed, a field with no anomalies is dead. Over time, the anomalies will either be resolved, or they will keep mounting up, at which point the explanatory power of your theory will be sufficiently diminshed that some other theory is likely to take over. See Kuhn generally.
Secondly, a framework is a much looser thing than a theory. They are ways of looking at the world, rather than specific predictions. Both “supply and demand” and “Keynesian macro,” the original examples, are “all else being equal” ways of looking at the world, rather than specific predictions about prices, unemployment, etc. Over the next ten years, the price and quantity sold of oil could follow just about any path, and it could be consistent with a story of supply and demand. After all, “demand” is something not observed, but rather inferred from the price and quantity. And so on.
Maybe it’s worth pointing out, because I have a feeling that this post could at least be misunderstood: w/r/t the Popperian framework, providing a falsifiable theory is actually a strength. Also, a falsified theory is not ‘failed’ as I read this post to imply. You won’t hear people laughing off Newton, or calling his law of gravity ‘weak’ because it has been falsified. A theory that cannot be falsified is worthless, ‘not even wrong’, indeed.
Well, that’s not true, actually. A demarcation criterion is just that, not a statement about what something is worth.
In my view, rejecting a theory because it does not fit the data is something that happens when the evidence is clear and convincing to everyone. In the case of Keynesian economics, this is obviously not the case. There are plenty of economists who are not as troubled as I am by the anomalies that I mentioned.
My point about calling these things “interpretive frameworks” rather than theories is that theories are falsifiable and interpretive frameworks are not. You cannot stuff economics into the empiricist boxes.
“Rodrik suggests that graduate students should spend more time in the real world.”
Advice like this creeps me out. Graduate students are not in a position to save academia.
Anomalies that only exist in the minds of those who don’t understand? As opposed to classical theory that has no place or explanation for recessions at all? Or austrianism that has no place for facts at all?
What is the Keynesian explanation of recessions? They have an explanation for recovery.
Animal spirits.
The idea that empiricism is the solution to a credibility problem is way off base. Putting aside Noah’s flippant use of a term intended to insult people of diminished capacity (derp), he clearly doesn’t understand empiricism and Popperian epistemology.
Popperian empiricism isn’t what most people think it is. Here’s Popper in “The Logic of Scientific Discovery”
“In point of fact, no conclusive disproof of a theory can ever be produced; for it is always possible to say that the experimental results are not reliable, or that the discrepancies which are asserted to exist between the experimental results and the theory are only apparent and that they will disappear with the advance of our understanding. If you insist on strict proof [or strict disproof] in the empirical sciences, you will never benefit from experience, and never learn from it how wrong you are.”
Noah’s criticism doesn’t apply to what Kling calls “interpretive frameworks” or what Pete Boettke calls “theory.” A priori, deductive theory has received a bad rap, mostly because it is highly misunderstood.
Here’s an interesting paper on the subject: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2229570