I once talked to a theorist (not RBC, micro) who said that his criterion for serious economics was stuff that you can’t explain to your mother. I would say that if you can’t explain it to your mother, or at least to your non-economist friends, there’s a good chance that you yourself don’t really know what you’re doing.
Right, but I think the spirit of the Krugman quote is that complication may be unavoidable, but shouldn’t be made into a goal or a badge of honour the way the theorist did. Also that complicatedness is ceteris paribus weak evidence of incorrectness, because of the logic I stated earlier.
What Krugman is implicitly arguing is that macroeconomics is different. The idea seems to be that since academic macro theory is not (yet) good for making quantitative predictions, we should focus more on communicating ideas. Communicating ideas—or “storytelling”, as some call it—requires simplicity and clarity.
Um. We can’t forecast anything so let’s construct some narratives..? :-/
The quote basically says that there must be a dumbed-down version of describing whatever you are doing and you must know it—otherwise you’re clueless. And that just ain’t true.
Specifically, it’s not true in most hard sciences (an in math, too, of course).
Krugman, however, used to do research in economics where there is not much hard stuff and even that spectacularly doesn’t work. Accordingly, there is nothing which is really complicated but does produce real results (as in, again, hard sciences). Given this, he thinks that really complicated things are just pointless and one must construct narratives—because that’s how economics, basically, exerts its influence. It makes up stories (about growth rates and money and productivity and… ) and if a story is too complicated it’s no good.
That’s fine for economics but a really bad path to take for disciplines which are actually grounded in reality.
OTOH it was Feynman who said something like ‘we don’t know how to explain [something] to freshmen, therefore we don’t really understand it yet’, and Einstein and Rutherford are alleged to have said similar things about explaining stuff to grandmothers and bartenders respectively.
He probably was using the word “understand” in a relatively narrow sense (after all he was the same person who said that no-one understood QM), and I agree with your general point, but certain people do overestimate how impossible it is to explain certain things in a way that can be understood by intelligent laymen (as done e.g. in Feynman’s QED).
I think according to Feynman the fact that nobody understands QM is the reason why we can’t easily teach it to freshman.
In some sense I think modern physics even dropped the goal of understanding. It got replaced by the mantra of “Shut up and calculate.”
but certain people do overestimate how impossible it is to explain certain things in a way that can be understood by intelligent laymen (as done e.g. in Feynman’s QED).
I also often underrate it. I once tried to teach a first year student in informatics A the principle of recursion. The whole course uses Haskel to make a point of teaching recursion. I don’t think why was stupid but the new phenomenological primitive of recursion was really hard to get into her brain. I think I spent 2-3 hours in one-on-one tutoring.
There no way to explain a concept that requires 3 new phenomenological primitives that a layman doesn’t have to that layman to make him really understand. You might find substitutions and explain the concept in a way that reduces to primitives he already has, but then you aren’t really explaining the full concept.
Yes, I think Krugman was just repeating a well-known observation without meaning all that much by it. However I think my point still stands: people in hard sciences (where results are checked against reality) can afford to make such observations, people in soft sciences (where what matters is sounding convincing) can not.
--Paul Krugman, “The Trouble With Being Abstruse”
big inferential distances usually --> long chain of reasoning --> at least one step is more likely to be wrong
On the other hand some things really are complicated.
Right, but I think the spirit of the Krugman quote is that complication may be unavoidable, but shouldn’t be made into a goal or a badge of honour the way the theorist did. Also that complicatedness is ceteris paribus weak evidence of incorrectness, because of the logic I stated earlier.
The implication seems anti-rationality.
As Noah Smith points out
Um. We can’t forecast anything so let’s construct some narratives..? :-/
I think the point is more “good forecasting requires keeping an eye on what your models are actually saying about the real world.”
That’s not what the quote expresses.
The quote basically says that there must be a dumbed-down version of describing whatever you are doing and you must know it—otherwise you’re clueless. And that just ain’t true.
Specifically, it’s not true in most hard sciences (an in math, too, of course).
Krugman, however, used to do research in economics where there is not much hard stuff and even that spectacularly doesn’t work. Accordingly, there is nothing which is really complicated but does produce real results (as in, again, hard sciences). Given this, he thinks that really complicated things are just pointless and one must construct narratives—because that’s how economics, basically, exerts its influence. It makes up stories (about growth rates and money and productivity and… ) and if a story is too complicated it’s no good.
That’s fine for economics but a really bad path to take for disciplines which are actually grounded in reality.
OTOH it was Feynman who said something like ‘we don’t know how to explain [something] to freshmen, therefore we don’t really understand it yet’, and Einstein and Rutherford are alleged to have said similar things about explaining stuff to grandmothers and bartenders respectively.
He probably was using the word “understand” in a relatively narrow sense (after all he was the same person who said that no-one understood QM), and I agree with your general point, but certain people do overestimate how impossible it is to explain certain things in a way that can be understood by intelligent laymen (as done e.g. in Feynman’s QED).
I think according to Feynman the fact that nobody understands QM is the reason why we can’t easily teach it to freshman.
In some sense I think modern physics even dropped the goal of understanding. It got replaced by the mantra of “Shut up and calculate.”
I also often underrate it. I once tried to teach a first year student in informatics A the principle of recursion. The whole course uses Haskel to make a point of teaching recursion. I don’t think why was stupid but the new phenomenological primitive of recursion was really hard to get into her brain. I think I spent 2-3 hours in one-on-one tutoring.
There no way to explain a concept that requires 3 new phenomenological primitives that a layman doesn’t have to that layman to make him really understand. You might find substitutions and explain the concept in a way that reduces to primitives he already has, but then you aren’t really explaining the full concept.
You might enjoy the book How the Hippies Saved Physics by David Kaiser, if you haven’t already read it.
Thanks, I will put it on my reading list.
Yes, I think Krugman was just repeating a well-known observation without meaning all that much by it. However I think my point still stands: people in hard sciences (where results are checked against reality) can afford to make such observations, people in soft sciences (where what matters is sounding convincing) can not.