Sunday, December 30, 2018

Sumner on teaching economics

Scott Sumner has a terrific post on teaching economics. (HT Marginal Revolution)
The core ideas of economics are extremely counterintuitive and are not accepted by most people....
Non-economists also tend to reject the central ideas of basic economics, and for reasons that are not well justified. [In particular these central ideas do not rest on hyper-rationality.] For the economics profession, our “value added” comes not from spoon feeding behavioral theories that the public is already inclined to accept, rather it is in teaching well-established basic principles of which the public is highly skeptical.  Thus we should try to discourage people from believing in the following popular myths: 
1.    People don’t respond very strongly to economic incentives.  (I.e., the demand for life-saving drugs is very inelastic.) 
2.    Imported goods, immigrant labor, and automation all tend to increase the unemployment rate. 
3.    Most companies have a lot of control over prices.  (I.e. oil companies set prices, not “the market”.) 
4.    Policy disputes over taxes and regulations are best thought of in terms of who gains and who loses. 
5.    Experts are smarter than the crowd. 
6.    Speculators make market prices more unstable. 
7.    Price gouging hurts consumers. 
8.   Rent controls help tenants. 
These myths are all widely believed by the general public.
Our primary goal should not be to add new information, it should be to have people unlearn false ideas about the world.
My emphasis.

One is tempted to add to the list. (An invitation to comments.)  Many of them stem from a basic principle -- "find the supply response" or ("demand response") that the fallacy ignores. "State the budget constraint" is another good habit.  Look for competition, entry, and choice among alternatives -- a market is not just bilateral negotiation. I might add reverse causality and selection bias -- empirical economics has stories to tell as well.

Scott frames the essay as a reaction to an Atlantic story advocating more teaching of behavioral economics. Scott is very clear: he is not opposed to behavioral economics. (He will likely be misquoted on this. Some behaviorists are very touchy. I know this from painful experience.) He is merely opining that our profession has more value added in teaching regular economics first. Regular economics is harder, less intuitive, less known, and therefore more valuable. To really understand behavioral economics, you have to understand what it is behaviorists object to -- and the vast amount of regular economics that good behaviorists agree with. Art schools might do better teaching people to draw, music schools to teach classical before atonal, physics programs newtonian before quantum mechanics, and so forth.
Most people find the key ideas of behavioral economics to be more accessible than classical economic theory. If you tell students that some people have addictive personalities and buy things that are bad for them, they’ll nod their heads.  And it’s certainly not difficult to explain procrastination to college students. [Dave Henderson's nomination for best sentence in the essay!] Ditto for the claim that investors might be driven by emotion, and that asset prices might soar on waves of “irrational exuberance.”  ... One should spend more time on subjects that need more time, not things that people already believe. 
I.e. let us not indulge in our own quest for teaching ratings via confirmation bias.

Yes, people do nutty things. But if you approach rent control, and all you have in the back of your head is behavioral stories, you will miss the clear prediction, borne out time and time again, that within a decade there will be a massive shortage of rental housing.

Scott does not neglect how awful most economics courses are
That doesn’t mean that I agree with the way that economics majors are currently being taught.  Our intermediate level courses are far too theoretical; they waste students’ time on lots of minor theories that would only be useful for people planning to do graduate work in economics.  (Most students do not.)  Too many homework problems with Cobb-Douglas utility, Hicksian demand, marginal rates of substitution, Giffen goods, gross substitutes, indifference curves, etc.  Some of that is appropriate, but all economics courses should focus heavily on applied economics. 
Most students come out of such courses still unable to coherently judge Scott's nice list of fallacies. Most of our courses are histories of thought, "greatest hits" of past theoretical contributions, passed on rather mindlessly. We teach many harmful parables. For example, natural monopoly due to increasing returns to scale, and the need for resulting regulation is a staple, passed down from about 1910. It has little to do with modern industrial organization in a global economy.

In part, it's easy to get through an hour by moving the curves around. Teaching real applied cases is much harder.

Macroeconomics teaching is in worse shape. Keynesian macro, like behavioral economics, enshrines most people's intuitive fallacies. Consuming more will increase output - forgetting the budget constraint. Breaking windows is good as it gives employment to window repair people. Good Keynesian macro justifies these apparent fallacies with carefully described "frictions," by which classical economic results fail. But you have to understand those classical results first to arrive at a correct economics that recognizes frictions (like behavioral biases) but doesn't violate budget constraints and accounting identities. Most macro teaching consists of young professors pushing IS-LM curves around, though such curves appear nowhere in their own research, nor anyone else's since the time they were born. Well, it passes the time easily.

An important point is implicit. Economics is not hard because of math. The math in even graduate level economics is no greater than in sophomore physics. Classical economics is hard because it can attack social problems in a value-free, cause-and-effect way, and upends the little morality stories that most people use to think about those problems -- rents are high because landlords are greedy. "Learning to think like an economist" is indeed best learned by application. And "learning to think like a behavioral economist" requires learning to think like an economist first.


  1. This may seem an odd suggestion based on whose post you're riffing on, but how about the tenets of monetary theory and, by extension, mainstream macro?

    To wit: banks are mere intermediaries, bank lending is no different than other forms of lending, money creation can be modeled as independent of the loanable funds market, incremental borrowing requires incremental saving from prior income, incremental investment requires incremental saving from prior income, etc.

    Macro theory doesn't exist without these assumptions, empirical "conclusions" throughout mainstream research require them, and yet they don't match the real world.

    Maybe someone with the clout to really change the way economics is taught will eventually study real-world banking practices, update Friedman and Schwartz's Monetary History with the money measure F&S studied (not the measures they helped popularize), realize that the original measure - unlike M1 or M2 - has retained its predictive power to the present day because it correlates almost perfectly with bank investment portfolios (the transmission mechanism!), and then start rooting out the monetary fallacies that virtually "blow up" mainstream macro.

  2. Thank you for your thoughtful insights on this topic - please keep this up! We need more long(fish) form expositions on how to remedy the ignorance that is plaguing our society...

  3. Hey, I enjoyed all my econ classes. Ha. I like the math and what the models are pointing to. They're not perfect by any means. I myself have some bones to pick with the assumption that consumers are rational when it comes to finding optimal bundles with utility functions and budget constraints. But, I find the math fun. Maybe that's just me.

    You want to see some weird stuff? Check out intertemporal choice in the long term versus the short term. Even Thaler quipped in a paper, "If you have enough utility functions, you can explain anything." Made me laugh.


  4. Very nice article. One nitpick: the article sometimes used "i.e." (that is) where it should have used "e.g." (for example).

  5. The most depressing aspect about free-market acolytes is that they constantly refer to the minimum wage and rent control, but never to routine criminalization of pushcart or streetside vending or property zoning.

    It is as if free-market types are trying to convince the public by the argument that they are for free markets when they benefit wealth and privilege, but when free markets cut the other way...well, not so much.

    Ironically, property zoning is probably the largest structural impediments in the US economy today. Free-marketeers should ascend to the very panicles of moral indignation and outrage and condemn property zoning on a daily basis.

    Instead we are told it is those people making minimum wage who are holding America back.

    Oh gee, the public is not convinced?

    1. This is false. The only question is whether it reflects astonishing ignorance and therefore feebleminded willingness to go along with left-wing demonization stereotypes, or, if you know anything about free market "acolytes", a deliberate falsehood. Go spend 10 minutes on the Cato, or Hoover websites, or any other free-market, libertarian organization. You will see a steady stream of outrage about zoning, other building restrictions, occupational licensing, BS legal harassement of the poor, and so on. When you see criticism of rent control and minimum wages, you will see that it is entirely that these policies hurt the poor, by reducing the supply of housing and jobs, not that they hurt anyone who makes over 30k a year. Completely false, twice in a row.

    2. John Cochrane:

      I read Econlib and your excellent blog regularly.

      Perhaps I exaggerated to make a point.

      But in fact Sumner's piece that you cite specifically mentions rent control butbut mute on property zoning.

      It is exemplary of the deficiencies I cite.

      As soon as we end property zoning and the routine criminalization of push-cart vending, I agree on wiping out the minum wage and rent control.


    3. Ben,

      You're just another pedantic douchebag upset someone didn't mention your pet topic. You're why the comments sections everywhere are generally terrible and lacking insight.

  6. A good portion of behavioral economics comes from actual calculations and decisions being costly while most of economics assumes them costless. So people put things off. They follow what they see others doing etc. Serious work would build such costs into models. Herb Simon's 1947 book makes some progress but this stuff is hard and usually ignored.

  7. I wrote a textbook that addresses almost all these issues. The ancillary material can be freely downloaded from:

    1. Interesting article. My degrees are all in criminal justice. A concept that was bantered about was " The skinny man wants what the fat man has." So if one thinks in simplistic terms, that would also apply to economics, which I always thought was overly complicated by those who taught it. Hmmmm

  8. Excellent stuff. Sometimes, insights come from throw away quips. A trader on the Merc mused, "Why is the market going down?" Someone replied, "More sellers than buyers."

  9. It's not hard to miss out on finding folks who don't quite misunderstand what is meant to be avoided when not using positives. Maybe avoiding double & triple negatives would make things more clear:
    "Thus we should try to discourage people from believing in the following popular myths: " << Econ teachers should try to teach truths, which can then be compared to the false myths that are often believed.

  10. I teach introductory economics at an inner-city community college. I have found success by teaching it as moral philosophy, rather than as "science". I also believe that my students are instinctive economists: they make choices, after all. What they lack is the vocabulary and basic structure to systematize what they do every day.
    I also try to disabuse them of the notion that suppliers and consumers are in opposition. Suppliers compete with other suppliers; consumers with consumers. And: when they go out to get a job, *they* are the suppliers, and behave just like all other suppliers, seeking the most they can receive for what they supply.
    Economics need not be difficult and obscure.

  11. Amazing post.

    I am not sure how it happens, but I’ve met way too many people with economics training who believe, at least implicitly, in nearly all the myths listed here (and many more). Not long ago, I was having a drink with a few colleagues after work and someone brought a young government policy analyst with an MA in environmental economics and public administration. He said that he and his group were helping to set government subsidies on green energy investments just right so that society could optimally fight climate change. I asked him, “If you can do that and really make it work, could you not also use that type of analysis to make optimal capital allocations for everything across the entire economy?” He thought about it for a second and without the least bit of irony he said, “Yeah, I think with enough effort it could be made to do that.” I nearly spit out my beer!

    Somehow there seems to be a deep disconnect between the stylized facts that students learn in school and important real world implications. It’s like a physicist who has been drilled over and over again on ideal heat engines, yet misses the principles behind the laws of thermodynamics, and then tries to build a perpetual motion machine in his first job out of college!

  12. It is time for Thomas Sowell to write a new " economic facts and fallacies " book

  13. I would add one piece to Sumner's list:
    Read, absorb and study Thomas Sargent's commencement speech at Berkeley 2007. Excellent stuff.
    This is basically economics condensed into 335 words.

    I cannot find a good link, here is one:

  14. PhD training ought to make sure students know basic economics, but usually it skips it. Chicago used to be famous as the only top place that taught the basic, non-math, econ. I don't know if it's still like that.

    We really need to think about updating the fallacies we try to dispel. Rent control is out of date. Probably the biggest fallacy now is conservation: that markets will use up natural resources too fast, with the related fallacy that natural resources are especially valuable compared to labor and capital. I still haven't figured out how to kill that idea. We really should teach the touchy implications of the free market for discrimination by race and sex, but one of the senior lecturers here at Indiana got fired for trying that not long ago. The ideas students most need to learn are the ones most dangerous to teach them.

  15. I'm teaching principles of macroeconomics this coming semester (for the first time ever). What are the 5 main things you think the students should be tested on in that course?

    1. That's a great challenge. I'd start with growth -- how fabulously better off we are than our great great grandparents, how productivity alone is responsible. Growth comes from ideas. Then cross country -- the immense gap in the levels of GDP per capita across the world; catch-up vs. frontier growth, and the catch-up growth miracles and controlled experiments -- hong kong, Japan, Korea; N vs. S Korea, E vs. West Germany, just how poor china was and how fast it caught 1/3 of the way up. Shake them out of the idea that growth rates rather than levels matter, and that recessions and distribution of income are first order worries on that scale. Then introduce money via hyperinflations. The first thing to learn about money is neutrality, and the fiscal foundations of really big inflation. Sargent's ends of inflations is great. You can only do non-neutrality once you understand neutrality. Finally, the basic model of a bank run.

  16. Good essay, good blog treatment. A few quibbles.

    "For example, natural monopoly due to increasing returns to scale, and the need for resulting regulation is a staple, passed down from about 1910. It has little to do with modern industrial organization in a global economy."

    Perhaps but it does reflect actual regulatory practice. Oligopolies and natural monopolies do tend to be heavily regulated (e.g., open-pit mines, telecom companies).

    "Economics is not hard because of math." Yes, it is. The math is simple, no argument there, but look at how the vast majority of educated adults use 'math'. They do not. They mostly use arithmetic.

    Take the most basic prisoner and social dilemma models -- all ripe with potential policy insights. The math is simple but the models are hard to understand unless you work through them.

    Economics will remain relevant in part because it is 'hard'.

    Its relevance will continue to elude many at the undergraduate level in part because it essentially ignores hierarchy and blithely assumes that everybody is a free agent with equal rights and no budget constraint for endless, costly litigation.

    The whole notion of the crucial importance of well-defined, secure economic property rights has lots of teaching moment potential but as long as both economists and fellow citizens appear to believe that significant ethnic or sectarian exceptions should be made then the material becomes too controversial for the typical North American economics class room.

    In other words, NA economic undergraduate teachers avoid 'boot heel economics' by simply talking around the subject or ignoring it all together. That may not always be the best approach if appealing to entry-level economics students.


  17. I teach a number of MBA business strategy courses and one of the things I usually explain to my students is that while 90% of the x-axes in an economics textbook have "quantity" labels, most of the x-axes in strategy have to be labeled with "quality" or "intensity of attribute A." And when econ textbooks do discuss "quantity" they usually fail to maintain critical distinctions between scale (rate) and volume (see Alchian) and between capacity and output.

    I somewhat disagree about natural monopoly if the assumption of regulation is removed. An interesting fact about many modern businesses is that we have at least some competition even though there would be at least a static efficiency gain to monopoly as marginal costs are very low relative to recurring fixed costs.


Comments are welcome. Keep it short, polite, and on topic.

Thanks to a few abusers I am now moderating comments. I welcome thoughtful disagreement. I will block comments with insulting or abusive language. I'm also blocking totally inane comments. Try to make some sense. I am much more likely to allow critical comments if you have the honesty and courage to use your real name.