Macroeconomics Textbook Manifesto

-- Bradford DeLong (delong@econ.berkeley.edu), May 04, 2000

Answers

I find much of your analysis of the problems with existing textbooks very persuasive. I fully support your views on the treatment of the money stock, AS-AD analysis (typically confusing and badly thought out), and the approach to growth theory. I would like to suggest though that there are a number of other, less theory based, things that require more emphasis in texts. Professors tend to forget that they cause confusion because they have long since thought the issues through....

Firstly, the distinction between stocks and flows, and stock equilibria and flow equilibria. This distinction is typically very poorly handled. A simple example - often it is not clear whether the term "savings" refers to a stock (financial assets) or a flow (unspent income). This may seem an obvious thing, but in my experience even very clever people miss tricks here.

Secondly, causality. The problem of presenting lots of nice pictures in text that show equilibria is that one sometimes loses sight of the causal mechanisms that drive the economy. And sometimes causation is disputed (eg do savings cause investment or the other way round?). Explaining that the role of social science is to uncover the mechanisms that drive things is therefore an important task - particularly for a field where disputeds about methodology have often been heated - macroeconomics.

Thirdly, implicit assumptions. To use the example above, whether one believes that additional savings frees up resources for growth, or causes recessions, depends largely on your starting point - full employment or underemployment. Much confusion could be avoided by pointing out that a lot of "debate" in economics really boils down to the implicit assumptions one starts from, and that a change in assumptions can radically affect the presumed relationship between variables.

Fourthly, the difference between action by individuals and action at the level of the system. (Fallacy of composition or decomposition). Macroeconomists have to think at both levels. Much of the trend towards optimization models, representative agents etc tends to blur this difference, and people can easily assume that aggregation problems in macroeconomics don't exist, because the textbook writers have chosen functional forms that assume them away. Nothing wrong with such simplifying devices as long as they are flagged and not used to hide conceptual problems.

Fifthly, and a bit more esoteric...As a Cambridge educated economist I still remember arguments surrounding the use of aggregate production functions and aggregate capital stock in growth models. I think that this issue is poorly handled (or ignored) in just about every text I have seen. No doubt this is because most textbook writers believe it is not an issue - and they may be right - but by not even addressing the capital theory issues (or trying to explain why they do not affect the core results) textbook writers tend to give an over-consensual view of the development and hisory of growth theory.

Finally, and in sum, I feel that the major hole in a number of texts is a lack of methodological sophistication. This runs through my comments above. This is nothing new in economics. For example, the acceptance of the Friedman line that assumptions don't really matter (I accept this may be a misinterpretation of his position) was a dogma that, incredibly, and to the shame of the profession, caught on. Just read Arjo Klamer's book of interviews to see how this view pervaded early new classical work.

Fortunately the world moves on. However, the desire to present economics as a hard "science" leads writers to try to present a view of economics that is too pure, mathematicised and data driven. But economic methodology is not settled - different approaches give different insights. The trick is to know what is useful and when. When do the simplfying assumptions affect the robustness of results and when don't they? This type of thing requires critical judgement - and hence a degree of scepticism of scientific method. But the rules of debate are that we have to discuss rationally. That is what students should come away with.

As a parting shot, it is my view that most economists would benefit from reading Jon Elster's "Nuts and Bolts" for the social sciences. This sets out a methodologically indivuidualist approach - hence congenial to economists - but one that focus on causal mechanisms rather than laws, and hence forces us to be humble in our ambitions. Mechanisms can change, as do economies. In my view any economic textbook that left a reader feeling that macroeconomics is a way of thinking constructively about economic relationships, and not just a bodied of codified "knowledge", would be well worth the price.

My conclusion would be that a short introductory chapter on methodological issues would help any text. But you are the author....

Best wishes for your book.

-- Dave Livingston (dbl@ibm.us.com), May 04, 2000.

I think very little emphasis is given in the long term relationship of a country's real exchange rate policy with economic growth. The book by Charles Wyplosz and Michael Burda "Macroeconomics: A European Text" had some partial success in this area. I think the branding of competitive devaluations as a "mercantilist" strategy is merely a protectionist ploy.

I also think that very few texts tell the students how heated the debates in monetary economics have been all these centuries. When students read the orthodox, mainstream monetarist approach which most textbooks present, students end up thinking that it is "the" monetary theory. What about the debate on the endogeneity or exogeneity of money? What about the money and credit views? I think these are more important than theories of growth accounting that end up settling the puzzles of growth with a "catch all" term called total factor productivity. . . . . . . . . . .

-- Emilio S. Neri Jr. (juneri@skyinet.net), May 04, 2000.