Sunday, November 29, 2015

Why teach kids macro at the intro level?


I've been writing some posts about Econ 101 (post 1, post 2, post 3). The basic theme is that kids need to learn a lot more evidence at the intro level, including methods. One response has been that it's actually very hard to teach kids even the simplest methods like OLS. At best, it would take a lot of time, and a one-semester intro course is just not long enough.

That got me thinking: What are the kids doing in the second semester of a year of intro economics? At a lot of schools, they're taking intro macro, often called Econ 102. I used to teach that class at Michigan, actually. And to be honest, I'm starting to think that this class is not really necessary. I think econ departments should think seriously about turning 102 from a macro class into a data/econometrics class.

Why should undergrads learn macro in their first year of econ? If they go on to be econ majors they can easily start out with intermediate macro and not miss anything important. If they just take the first-year econ sequence and then go into the business world, what do they really need to know?

In terms of growth theory, they might as well know the Solow model, so they can understand that capital accumulation by itself won't be a sustainable source of economic growth. Really, the Solow model is just a convenient way of teaching growth accounting, introducing ideas like capital, labor, human capital, and total factor productivity. And it's the one time in intro econ where you use a differential equation, so that could be useful for general math skill. Other than that, there's not really much growth theory an intro student needs.

How about business cycle theory? When kids go into the business world, it will probably help to know the standard Milton Friedman, New Keynesian, AD-AS, accelerationist Phillips Curve theory of monetary policy. That doesn't mean the standard theory is right (maybe Neo-Fisherian theory is better!), but since most people in the business world and at central banks sort of think it's right, kids will benefit from knowing it. It's easy and quick to teach to 101 students, since AD-AS fits right into the supply-and-demand graph stuff they're doing anyway. And if you want to mention RBC, you can just take the AD-AS graph and draw a vertical AS curve.

That's all I can really think of, and it's only about two weeks of material.

We have a lot of macro theories, but none that really work well unless you pick your data set very carefully and squint very hard. We have some methods for gathering evidence, but none that are very satisfying, and none that are simple enough for most intro level kids to understand.

And to make it worse, most of the macro theories that economists take halfway seriously are too hard for intro kids, so they end up learning silly stuff like Mundell-Fleming and Keynesian Cross that no one even halfway believes. Do we want kids going out in the business world and making deals as if interest rates will eventually equalize across all countries? God, I hope not.

So devoting 50% of the first-year econ sequence to macro just seems like a giant waste to me. We tend to think of macro and micro as symmetric things, but really "micro" is a lot bigger and more general than "macro". It affects a lot more policy questions, it has a lot more abundant and reliable evidence, it has a lot more interesting theoretical methodologies (game theory!), and it is more directly relevant to what students will eventually encounter in the business world.

So here's my new proposal for the first-year intro econ sequence. Replace useless macro with useful micro empirics. Either:

A) make 101 about theory and 102 about micro evidence, and stick Solow and AD-AS into 101, or

B) make 101 and 102 both predominantly micro (with Solow and AD-AS thrown in), and intersperse theory and evidence while stretching the sequence over 2 semesters.

And just kill the all-macro 102 course. It's not really doing anyone any good. It's kind of a barbarous relic.


Updates

1. On Twitter, Matt Yglesias says that econ probably has too much prestige tied up in macro to ditch the intro macro class. Maybe that's true. But I wonder if intro macro might now be generating negative prestige for the profession. Since the crisis there have been millions of words written in the media about how macro is bunk, macroeconomists don't know anything, etc. Killing intro macro might actually be saving the field some embarrassment, because a lot of those 102 theories really are obvious bunk, and most of the rest have taken a beating since 2008.

2. Some people (in the comments and on Twitter) have suggested that simply teaching kids the meaning of GDP, inflation, etc., and how these things are measured, could fill a whole semester. My response is: Maybe, but why go looking for ways to fill a whole semester? It sounds to me like status quo bias - because we've always taught an intro semester called "macro", we had better go find some "macro"-y stuff to teach. Sure, learning about growth accounting and NIPA calculations and whatever is cool. But there's a whole bunch more super duper useful stuff they could be learning instead: data analysis! Data analysis is becoming more and more important in econ, and more and more important in the business world. We're doing students a disservice by not teaching it in intro classes, and if we're going to correct that, something will have to make way. And I think the best "something" is macro.

33 comments:

  1. I think it is important for students to understand some of the "language of macro" in their early classes. What does GDP mean as a concept? How is it measured? What are the national accounts and why (and how) do we compile them? What is the big picture of long run macro history - the turbulence of 19th century capitalism, etc.

    The way I teach the MBA economics course, which is a one-shot economics course, is to focus on the "measurement of macro" - focussing on the national accounts, labour force survey, international accounts, money and banking etc - so that students know what all the macro data seems to mean. This is a really important foundation for any further work, which is often done poorly in order to get to the theory stuff.

    ReplyDelete
    Replies
    1. I see that there are at least two of us...

      Delete
    2. In response to Noah's update. I don't think data analysis is that useful unless you know what the data is! I may teach a whole course on "macro measurement". Or instead I may embed these important points in some other course, say one called "Economic Data", which would go into survey methods and basic data analysis for all types of economic data.

      Delete
  2. Having done my damnedest for the last 25 years to avoid teaching intro to macro, I find myself in the somewhat absurd position of defending it. A fairly significant part of macro (as I taught it, when I taught it--thanks be for retirement) was not about macro theory. Where I was teaching (a third-tier public), we needed to spend at least 2-3 weeks re-booting micro (at least re-running the basic demand-and-supply framework).

    Then I spent a considerable amount of time trying to get students to understand where the major bits of macro date--GDP and its components. price indexes and the calculation of the rate of inflation, labor force data (employment, unemployment)--come from, what they mean, and what they do not mean. I also spent a considerable amount of time on money and interest rates. In general, I spent only 4 or 5 weeks (of a 15 week semester) on macro theory.

    So while I would mostly agree that spending a lot of time on macro theory is an over-rated use of time, I would actually have been happy to spend that 4-5 weeks doing more with where the numbers come from and what they mean. Because my students, at least, had mostly never been exposed to it (or their exposure was so long ago that they had forgotten it).

    ReplyDelete
  3. Anonymous11:49 PM

    The gulf between modern macroeconomics as embodied in DSGE and the stuff taught to undergraduates is the size of the Mariana Trench. I used to think this reflected poorly on the discipline. After all, someone studying undergraduate chemistry or physics is prepared for study at the graduate level. Not so economics. Now that DSGE has been so thoroughly refuted by realty, however, while based ISLM actually has a good track record for nearing a decade, it turns out that this gulf is a good thing.

    ReplyDelete
    Replies
    1. I am Responding to an Idiot10:56 AM

      Asserting things and posting them on the Internet does not make them true.

      Delete
  4. You do realize that your average college freshman, before taking a macro class, thinks that inflation (whatever that is) is 50%, GDP (whatever that is) is some kind of thing that is bad because it doesn't account for quality of life, interest rates are too high for sure, the government is bankrupting the country through its spending on foreign aid, and the Federal Reserve is some shadowy organization which was organized by the IMF and the WTO to destroy the value of the dollar and ruin social security.

    If nothing else, the main purpose of an into Macro course is just basic economic literacy and citizenship. I'm sure you knew all this stuff when you were a college freshman, but you're not exactly representative.

    ReplyDelete
  5. And don't get me started on what they think of exchange rates, trade deficits, immigration and globalization

    ReplyDelete
  6. Anonymous8:47 AM

    The people who don't take Mundell-Fleming seriously usually fail to understand currency regimes. Please appreciate their importance for smaller countries and trade in general.

    ReplyDelete
    Replies
    1. Anonymous5:05 AM

      People keep referrring to MF as arguments against fixed currency regimes like the Euro. The problem is they often downplay trade finance issues (eg the costs of borrowing in other country's currencies). Paul Krugman does this, and has not gone unnoticed by people in the real world business of banking and trade. You also do not need to learn a whole lot of crap about and derive things like the LM curve to understand the very simple points that MF does make - you can do it in a couple of sentences.

      Delete
  7. To join the chorus, I think you're greatly overestimating the background knowledge of a typical high school grad, even those who go on to attend first tier universities. I teach high school economics, and the macro portion of AP Econ (which lines up pretty close to my intro macro class at a first tier university) is all about what you'd probably consider background knowledge. Take a look at the syllabus on pages 30-31 of the PDF (pages 26-27 of the printed numbers).

    It's mostly vocabulary and measurement building. How to measure/understand GDP. How the CPI works (we bludgeoned this to death in my college intro macro). A basic understanding of the business cycle (not even what causes it (let me know!), just that it exists, the parts have names, and temporary changes =/= long-term changes.) What the Fed is, how banks work, how interest rates work. Basic vocabulary around international trade, currency exchange, and comparative advantage.

    I guarantee you don't want anyone entering intermediate without this knowledge, and I also guarantee you can't count on the students to walk in with it. Even beyond econ majors, this is stuff every college grad should know to be an informed citizen; many high schools have kids take poli sci for one semester, and macro the next.

    Now, we do do some slightly more theoretical/complicated stuff. AS-AD models, a very small amount of time on the interplay between the money market and loanable funds market, the Phillips Curve, and a few basic models of how international trade can impact domestic markets. I think you're right that we could cut some of this and not be totally ruined. But I also doubt we would get nearly as much time back as you seem to think. I also think that if the intro courses are to be a way for students to see whether economics might be the field for them, it would do them a disservice not to expose them to at least some basic macro theory.

    Finally, you may appreciate this Less wrong post.

    ReplyDelete
  8. Yes, by all means dump macroeconomics. But replace it with finance, not econometrics. Students need to understand present value and options much more than they need to understand heteroskedasticity, multicollinearity, etc.

    ReplyDelete
    Replies
    1. Also probably a good idea. But finance is heavily empirical, so these aren't mutually exclusive.

      Delete
  9. In the macro classes, the students need to understand the weaknesses and critiques of the bunk theories, and not just take them as truth. The economics profession is still developing new core theories, like effective demand ¯\_(ツ)_/¯.

    ReplyDelete
    Replies
    1. I am Responding to an Idiot10:58 AM

      You can pretend "effective demand" is a real thing, and post it everywhere, but it will still be nonsense and ignored by the profession, of which you are barely a part.

      Delete
  10. Perhaps along the lines of Peter Dorman's texts? They do deserve an idea of limitations and an inoculation against bunk. Some history can serve empiricism.

    ReplyDelete
  11. Personally I think one way forward would be to admit that social science disciplines (economics, sociology, psychology etc.) do not neatly divide in the way the natural sciences do; so all courses should have modules in the other areas, at least in year 1, and students can then specialize from there. It would, I think, help give greater appreciation for the macro/micro, theory/experiments divide that exists in most areas of social science.

    ReplyDelete
  12. I think you are very naive if you don't realise that even the most simplified and unrealistic theories you teach in an intro macroeconomics course would be a hell of a lot better than the dangerously stupid notions an average student has been taught by the media.

    Also, teaching students to analyse GDP and inflation etc without knowing what they actually are is suicidal.

    ReplyDelete
  13. Anonymous2:27 PM

    You must have AMAZING undergrads if you think AD-AS/IS-LM are simple enough to cover in a week or so. Just drawing a vertical AS curve seems incredibly simplistic. They are smart enough to want to know why something is vertical, horizontal, or in-between. And that question leads you into the bowels of the theory, which takes time.

    I am often baffled by reading economists' blogs where learning is considered a trivial thing. Even inflation is not so simple. Kids know that the iPhone didn't always exist, and they don't quite understand how to account for that transition, at first glance. Price stickiness (if that's the way you go) is a jarring concept, coming from basic micro. To not cover these things is to educate people in a way where they are not able to even understand the debates around Fed actions. With cranks like Ron Paul around, I'd rather take the time.

    I'm all for empirics, and very much in favor of treating 'weird' markets like labor in careful way. But just tossing out macro, with some casual remarks about Solow and aggregate D/S along the way, is not sufficient. Think about Krugman's remarks a while ago, about how even younger faculty at top-notch institutions did not know IS-LM. If this were easy to absorb, this wouldn't be an issue.

    ReplyDelete
    Replies
    1. Anon is moving in the direction I do not see much of in either Noah's original post or most of the comments, which has to do with the idea that for most intro students who will never take any econ beyond those courses, what one is teaching them is "how to be a good citizen," which means making it possible for them to read media discussions of the economy in their future lives and have some idea what is going on and how it all affects them. In this regard, theories are secondary and institutions may be more important, along with all those defitional issues such as what is GDP and unemployment rate and inflation and so on.

      It is not just the Fed but also the budgetary process. So, obviously one has to bning in at least some theory to talk about how monetary and fiscal policies affect (or might affect) the economy, but for the students it is also important to understand who carries out these policies and how they do it. When you put this together with the definitional stuff, that will take up most of the time, especially if one wants to include some international coverage and worry about long run growth and so on besides the more intense focus on shorter run stabilization policies. But getting them to the point where they can simply understand to some degree general discourse about the economy in media may be the most important goal.

      Barkley Rosser

      Delete
    2. I wouild add that indeed the majority of that discourse is about macro issues, which gets frustrating for most professional economists because micro is where we think we have these nicely settled theories that we can teach them, even though at higher levels they may not be as neatly settled as one finds them in the principles textbooks. We all know that macro theory is a mess of all thesse competing theories, but this can be handled with a basic classical versus Keynesian sort of discussion, with the institutional and definitional stuff maybe in the end being more important.

      JBR

      Delete
  14. Anonymous9:33 PM

    Firstly - to suggest that things like Mundell-Fleming are silly ideas is itself silly. If you think Mundell-Fleming is silly because interest rates across the world don't equalize in reality then the standard demand/supply model is also silly because I have yet to see the demand-supply equations for any product. The reason to teach students macro-economics is to allow them to see the proverbial "forests" and not lose it for the "trees". Mundell-Fleming may not be a good descriptor of the world, but taught well, it allows people to think about the interaction between currency pegs, fiscal policy, demand shocks, etc. A very good a real-time example for you Noah: "How long can Saudi Arabia hang on to the Riyal peg?", "What will happen if they abandon the peg?", "Why aren't the Russians hurting as much as (say) Venezuela because of the drop in oil prices?". Mundell-Fleming at least sheds some light or provides a useful way to think about these issues. Maybe it is too simplistic - but that is your fault, not M/s Mundell or Fleming's. Come up with a better model if you can.

    ReplyDelete
    Replies
    1. Anonymous4:22 AM

      "Mundell-Fleming at least sheds some light or provides a useful way to think about these issues."

      No it doesn't. You would understand far more about the things you mention by understanding something about Russia and something about Venezuela. There is nothing in the ISLM gadget that will give you an informed insight into these things. These things distract, not inform. NK

      Delete
  15. Anonymous8:04 AM

    Hmm, I think a compromise between what you and Don and Cameron are arguing. Understanding headline macro numbers is generally useful in business, good thing to teach in first year. A whole semester of intro macro not really necessary.

    Btw Tom W here, too lazy to log in on not my comp.

    ReplyDelete
  16. The biggest problem with data analysis is that properly done it is actually hard for most of the average students. In one semester they will not learn enough to use it usefully and mostly not even to understand the errors they are doing. Person adequately educated in data analysis should have at least 1 year of statistics worth and 1 year of data/programming/machine learning worth of teaching.

    ReplyDelete
  17. Anonymous10:50 PM

    teach data analysis using data from FRED. easy.

    ReplyDelete
  18. Anonymous3:17 PM

    The problem is not only with Macro it is also with Intro Micro. Economics is sort of like teaching Newtonian Physics as if there is nothing more to learn and then discovering in later courses that the theory in foundational terms is just not adequate to discovering the world. Actually Micro is more insidious because it doesn't really answer the Institutionalist challenge. You really have to go deep to understand not only market imperfections in any level of detail but also the behavioral complexity of markets. Macro is useful is taught right because of the nature of national discourse. So understanding at least the basics behind Fiscal and Monetary Policy wrt Price Levels and Employment not to mention growth will make for better citizens much more than leaving them with some silly notion that supply and demand are the answers to any question. Remember what Alfred Marshall said about any short statement about Economics. And also remember what Keynes said about Madmen.

    ReplyDelete
  19. Anonymous7:19 PM

    Macro is far more interesting than micro, and far more useful to understanding public policy and politics. As a person that reads the news, I care about macro, but not micro.

    If you learn data analysis, go take a statistics course.

    ReplyDelete
  20. Anonymous11:47 PM

    As someone who works in policy, I have to say one of the most common difficulties dealing with other people with an economics background is that they know micro, but they are clueless about macro - or they know something, but they don't have any comprehension of how it links into micro, or why you can't apply micro theory to economy-wide analysis.

    You can't really participate in a sensible discussion on a lot of issues in labour market policy (for example) without understanding macro and how it links into micro. People who know micro very well often say quite ridiculous things about macro, basically confusing micro concepts that only apply at a firm or individual level with macroeconomics. In some areas you really need to have both.

    Econometrics is becoming increasingly important, and I'm all for teaching more of it. But sidelining macro would be exacerbating an existing problem, which is the weakness of macro knowledge among many people with economics qualifications. I would be more inclined to ease back on micro theory in favour of microeconometrics.

    ReplyDelete
  21. Anonymous4:15 AM

    Interesting views Noah, and I think you are on to something. Macro-theory is a waste of precious time for most people. But I think in any social science critical reasoning skills are central. I did a PHD in economics, but I am familiar with other disciplines through my work as an academic. It is critical reasoning which is taught at a very low level in the economics profession. I agree people need to learn data techniques, but there is a hell of a lot more to investigation that running data through computers. And I really think this is not properly taught in economics: how to use primary material, how to build up a coherent narrative from information that is messy and contradictory.

    ReplyDelete
  22. Anonymous5:13 AM

    You are right. Strip macro right down. ADAS, maybe Solow. Basic Keynesian insights - the aggregation problem, fallacies of composition - Say's Law etc. This will set people up well for whatever path they choose later.

    Keep rational expectations and DSGE out.

    ReplyDelete
  23. Do they teach the concept of money being neutral in economics courses? Do they reference Ben S. Bernanke et al's 2002 FAVAR paper that found the US Fed from 1959-2001 largely had little or no influence over a variety of economic variables (3.2% to 13.2% out of 100% of an variable change, statistically significant but not that overwhelming)? Do they reference that perhaps economics is non-linear, and at best you can only estimate with 60% confidence the envelope that some economic variable may fall within? I sure hope so, but somehow I doubt it.

    ReplyDelete
  24. Noah, your series of posts came up in a discussion on Nick Rowe's blog about intro economics courses. Frances Woolley (Nick's colleague) brought your posts up in the 1st comment.
    http://worthwhile.typepad.com/worthwhile_canadian_initi/2015/12/just-a-little-bit-of-macro-first.html#more

    ReplyDelete