Sunday, January 10, 2010

Initial Thoughts on Greg Mankiws' "Macroeconomics"

So, first I'll have to give Gregory props -- he's a good businessman. One of his innovations is to sell advertising space in his textbook (and, I suppose, on his blog -- see my post on his Steven Landsburg review). I have no problem with this (especially if this could be used to reduce the price, $178, of the textbook), but when you do so in the context of listing what good resources are for students, to only list resources who will pay is poor scholarship. For instance, he lists Aplia, Mankiw WebCT, the Wall Street Journal, and he tells his readers that "Faculty receive a complimentary 15-week subscription when 10 or more students purchase a subscription." If we go to the Economist website, we can get access to: "Blogs. Blogs cover economics as well as U.S. and European politics." He doesn't mention that you can get access to Econ blogs on the world wide web, nor does he provide a list of the most influential Econ bloggers. Why not? Because they won't pay for advertising space in Mankiw's book.

The students lose out.

Second advice to Gregory: Donate 10% of the proceeds of your $178 book to a charity which makes college education more affordable for low-income students and feel free to include this in the cover. You'll come across as less of a blood-sucking leech who just uses his position in society and the Harvard name to extract every dollar out of students he can (while dispensing less-than-worthless economic advice), which is unfortunately my impression at this juncture.

As one would expect from N. Gregory Mankiw, the book includes some whoppers...

P. 3: "Voters... know that government policy can affect the economy in powerful ways. As a result, the popularity of the incumbent president often rises when the economy is doing well and falls when it is doing poorly."

Except this isn't true. When the economy is doing poorly, voters take it out on incumbents regardless of whether the incumbents have anything to do with the poor state of the economy or not. Usually, the incumbent party has at best a very modest impact on the business cycle, and voters are generally unjustified in thinking otherwise (some rare cases excluded). It would be nice if Econ textbooks highlighted this so that voters would stop acting so irrational. (Greg: It's comments like this that got your Mensa application denied...)

On P. 60 -- Gregory buys into the "Great Productivity Slowdown" of 1973-1995. Of course, this 22 year period begins and essentially ended with low-productivity growth periods, while the shorter periods before and after it contained only brief recessions. We can make much of the "Great Productivity Slowdown" go away by slight arbitrary shifting of the dates. Greg's first period is chosen as 1959-1973, when productivity grew at 2.8% vs. 1.4% in the slowdown period. But of course, this is a questionable start date since average productivity growth from 1956-58 was only 1.3%, and a questionable end date since productivity growth in 1974 was -1.6% (according to the BLS ). Absorbing these four data points into the first period, which is eight years shorter than the second, would make productivity in the first period less than 2.4%. If we then make the second period 1975-1992, the average productivity growth was 1.7%, which cuts the magnitude of the "Great Productivity Slowdown" in half. If we exclude the three recession years during that time, then we get average productivity growth of 2.2% -- within the MoE of the first period (although this is kind of cheating -- it does help us "explain" why there was a productivity drop). The third period, from 1993-2008, is 2.2% as well (compared to 2.4% from 1995-2007).

I still think there was a productivity slowdown, it was just a minor blip (less than 1%) which, after accounting for the oil shocks and Volcker's battle with inflation, is probably within the MoE. Part of the reason the myth of the Great Productivity Slowdown took hold was that Europe and Japan grew like wildfire rebuilding after WWII, which may have helped US productivity as well by increasing their demand for American goods. Also, bias in the CPI is likely increasing over time, so that, in reality, we'd have to guess that productivity was roughly flat in the first two periods and increased in the third, but this increases our error estimates, so that it's actually tough to say. I don't put much faith in government bean-counters...

I'll have to admit, though, that the "The Great Productivity Slowdown" was a such a big, serious issue for Macroeconomists says more about the sorry state of Macro and Econ than the sorry state of Greg Mankiw, as I'm fairly certain every other Macro book says the same damn thing.

More to come...

UPDATE: Just opened Chapter 7. Greg starts the chapter off with a list of a few select countries by GDP with the title "International Differences in the Standard of Living". Greg has chosen the countries (leaving out the likes of Norway, Ireland, the Netherlands, or Switzerland) so that the US looks like far-and-away the country with the highest living standard in the world. Students will draw from this that the US therefore must have the greatest economic system for its citizens, and will begin to wonder how to explain the US's preeminence. What I recommend Greg do in the next version is to also list each countries rankings in terms of life expectancy, where the US ranks 38th in the world, and infant mortality rate, where the US ranks 33rd, just to give students a little perspective on how GDP translates to "Standard of Living". Also, as Greg knows, it makes more sense to include median GDP per person, where the US doesn't match up as well, or, even better, median income per hour worked, where the US would not longer even rank in the top 10, and would lag even France by a wide margin...

One of the countries Greg compares the US to is Germany, which lags the US by about 25%. Except, of course, Germans work about 1/3rd less than do Americans. Throw in the fact that Eastern Germany is still about 25% poorer than West, and it's clear that Greg Mankiw is leading his students to a faulty conclusion.

UPDATE II: The growth section is just AWFUL! Almost unspeakably bad. On page 215, Greg plots income per person against population growth, and sees a clear downward trend, about which he says "high population growth tends to impoverish a country because it is hard to mainatin a high level of capital per worker when the number of workers is growing quickly". Basically, he pitches it as confirmation of the Solow model. Except, there are two other theories it confirms -- the first, and more obvious one being the Demographic transition -- the (i had thought) well-known fact that when societies get rich, they have fewer children (so that income per person predicts population growth rather than the other way 'round), and the second being the Malthusian Model -- African countries which depend heavily on agriculture and resources have fast population growth.

Greg does go on to discuss the Malthusian Model, but there are some crazy comments in this section too, as Manks discusses the Malthusian Model as a relic of the distant past, having nothing to do with the world today. "There is now little reason to think that an ever expanding population will overwhelm food production and doom mankind to poverty." Greg, this should be changed to: "There is now little reason to think that an ever-expanding population will overwhelm food production and doom white countries to poverty." What about Sub-Saharan Africa? Much of Africa is quite likely still at least weakly Malthusian, definitely not completely out of the Malthusian woods yet...

Re: The Kremer Model -- Jared Diamond is the proper citation there, Diamond came first(I suspect Kremer meant to cite him). Diamond's theory is also more interesting than Kremer (even though Kremer came later) so you need to explain it.

Also Greg, a tip: AJR (2001) is fatally flawed research and, as such, has no place in an undergraduate text. First, you should go read Crosby and the original Diamond (3rd Chimpanzee, 1992). The basic problem with AJR (2001) is not that the authors made up their data, although they did. It's that for settler mortality to be a good proxy for institutions, it needs to be uncorrelated with geography. But, by construction, it's almost perfectly correlated with geography. The europeans could also bring their human capital and agricultural technology to other temperate areas -- not just their institutions.

Somewhere, in your section on growth, you've got to show a frigging map of the world w/ per capita GDP so students can see the latitude connection...

Fortunately, I got this book for free. (Sometimes you get what you pay for...)


  1. The Mankiw text was one of the recommended readings for my macroeconomics courses last year.

    Waste of my money and time. And waste of an entire semester's worth of macroeconomics lectures too, come to that.

  2. I can imagine that... the book is just garbage... expensive garbage, but garbage...