Graduate Macroeconomics @ Berkeley

Updated due to "unclarity":

The retirement of George Akerlof is tremendously distressing--as is Chad Jones's decision to go to visit Stanford next fall. (Memo to Chad: if you stay at The Farm, you (a) get $, but (b) you lose the best office view this side of Beta Lyrae, (c) you must teach mbas, and (d) you confirm the Stanford administration's belief that it was right to nuke you for tenure when you came up internally because you had already given them a call option. Your marginal utility of wealth can't be so high that (a) outweighs (b), (c), and (d), can it?)

Note: The real reason for Chad Jones to think about changing from Berkeley to Stanford is, of course, the presence of Pete Klenow and Paul Romer at The Farm. Add Chad Jones to that and you have the strongest endogenous growth group in economics.

At any event, we now have to figure out how to teach graduate macro with two holes ripped in our fabric...

There are seven slots in the curriculum--first come four eight-week slots that are taken by everybody in the Ph.D. program in their first year, and second come three sixteen-week slots that are taken in their second and third years by Ph.D. candidates especially interested in the field. As I understand macro, the slots are:

• Moral hazard, near-rationality, and coordination failures
• Neoclassical growth theory
• Keynesian and new-Keynesian models
• Dynamic stochastic general equilibrium models
• Monetary theory, policy, and history
• Capital markets and macroeconomics
• Endogenous growth theories

Plus there are all the "international finance" topics...

The question is: are these the right seven slots to teach, and is this the right ordering of them?

Comments

[Lorenz]

Brad,

Thanks for taking up this issue. I would at least change the order and follow closer the structure of Romer's textbook. (I would probably also add more details and more tools than the textbook.) :

1. growth (neoclassical and endogenous)
2. RBC and DSGE in general
3. short history of thought: from IS/LM to modern macro
4. Keynesian, New-Keynesian economics
5. moral hazard, near-rationality, coordination failure, animal spirit, ...
6. monetary policy, financial markets, asset pricing, intermediation, ...
7. fiscal policy, political economy, social security, ...

Of course, there is much more to add, but I think given the limited time this would already be quite an ambitious first year course in macro.

I added point 3 since the transition from undergraduate to graduate macro is often difficult. For many graduate students it is hard to understand why the typical macro graduate course tends to look so different from the undergraduate material. The problem is not so much the additional math (as in micro or metrics), but in terms of content. (Graduate students at top programs usually do not struggle too much with the math anyway.) Sometimes it seems like there is not much of a connection between what one learns in intermediate macro and the syllabus of a standard macro first-year sequence. I think a historical approach might be useful in providing a bridge.

I haven't found a good graduate textbook that provides this transition, but I'm looking forward for suggestions.

[JS]

Is (C) really a problem ? I would hate to think that I aspire to be a burden to higher-calibur economists in the next 1.5 years.

Brad, perhaps you could outline the unique losses experienced by teaching mbas.

[Mu-Jeung]

Something George Akerlof taught generations of Berkeley graduates is how to overcome the "Neutrality Results" in Neoclassical Macroeconomics (Permanent Income, Ricardian Equivalence, MM-Theorems...), but I always felt that his part was actually much too advanced to be the first quarter-course of Core Macro, but it was very worthwhile...

So I agree with Lorenz on at least two points:

1. It gives "Macro" more of a "unifying structure" to start with the Neoclasscial Growth Model and develop various questions (taxation, microdistortions, business cycles), afterwards (for example in contrast to it...) - I always felt that Blanchard and Fisher (1989) for example had a very good structure...

2. there has to be a transition that connects undergraduate IS-LM to graduate macro (the "microfounded IS-LM")

What maybe seems somewhat troubling with 1. is that unifying macroeconomics under a neoclassical benchmark seems more or less like an admission that "RBC Macro" triumphed in the end over various Keynesian ideas...

And I think here a cautious transition from undergrad IS-LM to "Graduate Macro" is crucial. I would suggest one or two lectures, maybe at the end of math-camp (BEFORE the official start) to connect IS-LM to modern macro. Something starting with: "Every Business Cycle Theory has real interest rates, inflation rates and real output; these we typically connect with an IS-Curve, a Taylor-Rule (or an LM Curve) and an expectations-augmented Phillips-Curve." Then one could go on to spread uneasyness with the IS-LM graphs holding expectations fixed (Lucas Critique) and emphasize the need for "micro-foundations" ... The key point though is, that this "consolidated version" of IS-LM gives essentially rise to the same structure. In between we learn something about the importance of various rigidities in dynamically determining propagations of shocks, but the representative agent Ramsey model is the start, not the end of the journey...

I did something like this for the incoming first years last summer and did it almost exclusively with graphs that reminded them of "good old textbook IS-LM"; on the other hand, I could easily already discuss the role of real rigidities in monetary transmission, the "Taylor Principle", the policy response of monetary policy to supply shocks under inflation targeting, Krugman's "Liquidity Trap ideas"... (for the motivation...)

A last thing: I think the strength of Berkeley Macroeconomics is indeed our focus on policy issues and empirical methods early on in the Core. But it might also be worth looking at the topics in the Core from a perspective of tools to teach people (at least to a limited extend). Topic I always felt were great to connect evidence with theory, but were a bit scattered or rushed in the Berkeley Core are Consumption (see for instance Deaton's "Understanding Consumption") and Lumpy Investments (good example: Doms-Dunne (1998, RED)) and one get a whole lot of of these topics, how dynamic optimization and analysis of micro-evidence go together...

[UBC]

"you lose the best office view this side of Beta Lyrae"

Clearly you haven't remembered to include the economics department at UBC in your comparison group.

[thorstein veblen]

I concur with the "History of thought" although I would emphasize the attempts of policy makers to implement economic theory, making case studies out of: 1) Malthusian Theory, and its application (i.e., the reform of the Poor Laws), 2) Smith, and laissez faire policy in Britain during Irish Potato Famine, (vs. the economics of the Federalist Papers, and the rational for central policymaking...) (fast forward to) 3) Hoover & his damned balanced budget & staying on Gold during the height of the Great Depression vs. Keynes (fast forward to) 4) tweedle dee & tweedle dum, i.e., Robert Lucas & Ed Prescott, doing hoover proud and writing that business cycles are rational and engender almost no welfare loss... and then to 5) Stiglitz, and his critique of Washington Consensus...

This would be a nice introduction for the first three weeks, while you could have the TA focus on solving models...

Every PhD Economist, much less macroeconomist, should be able to crank out a 10 page paper on any of the above case studies... Most *standard* phd programs skip 1-3 altogether, do (4) from a vastly different angle, and try hard to pretend 5) doesn't exist. Macro exams and grades are mere algebra marathons, with 10 pages of mechanical algebra (replacing any actual thought) the norm for a 90 minute exam... The typical Macro professor just copies his own notes on the board and calls it teaching. The TAs generally mark the exams, and there is no discussion at all in class. Nobody mentions that they haven't learned anything and that the teaching is horrid on the semester evaluations, b/c they what to flatter the profs, who will subsequently mark the preliminary exams...

[Andrew Samwick]

[Yes: the real attraction of Stanford for Chad Jones is Paul Romer and Pete Kleniw--both awesomely smart, very good company, and working along exactly the same lines as Chad is. Our chances of keeping him are not high...]

Brad,

Perhaps you overestimate the lure of having colleagues who will impugn your motives and integrity publicly based on your professional choices. I hope you will come up with something classier to say about Chad at his bon voyage party.

Andrew

[thorstein veblen]

well put, andrew...

Chad cares nothing whatsoever about any potential large pay increase; it's solely his devotion to research for the betterment of mankind that will lead him to Stanford. (On the other hand, if Berkeley can match, it is his profound dedication to the Berkeley community that will keep him at UCB, it has NOTHING whatsoever to do with the raise...)

[Barkley Rosser]

Face it. There is no replacing George Akerlof.

[Gautam]

Thoughts from an incoming Berkeley grad student:

1. I would very much enjoy a "history of thought" section / approach to the macro courses. But isn't ECON 210 supposed to cover this?

2. I have had no intermediate macro myself, so a gradual transition from undergrad to grad macro shouldn't matter much to me. Perhaps I am an outlier.

3. I'm all for a focus on connecting theory to data in the core. This will be the only time many of us will study macro, so it would be good to know what the data says. I've taken one grad macro class, and found the growth models elegant but entirely implausible.

4. Akerlof seemed like a fantasitc person to talk with at the flyout. I am very disappointed that he's leaving, and do wish we had been told about this earlier. I have heard great things about his class, and was particularly looking forward to his "research idea" assignments. Perhaps his replacement could continue this? Will Akerlof be around at all as an emeritus professor?