D-Temp

Gabriel asks why Micro textbooks don't have as much of a data/empirical focus as Macro textbooks. Which is only partly true. "Micro" textbooks such as Mas-Collel-Whinston and Green are very theoretical as is Varian, and so on. "Macro" like Romer, or ... uh ... uh ... Carlin and Soskice (it's an undergrad book but grad programs currently using Romer could benefit from adopting it instead even though there's much less fancy math in it. Or is there?) have a lot of data/empirical work in them. At EI I speculated that part of it had to do with the insecurity of macro theory vs. micro. If in the back of your mind you know your theory, or in this context a lump of theories, a methodology, is schizophrenic (which Macro very much is) you're gonna throw up some numbers as barricades against obvious accusations. But the main - although related - reason for this phenomenon is that micro is simply much more specialized than macro. You pick up a Health (Micro)Economics textbook or a Labor(Micro)Economics textbook or a Sports(Micro)Economics textbook, it will be chuck full of data, I promise. But what are the specializations in Macro? (And how does that reflect on the status of this sub-discipline) Well... uh ... there used to be something called "Monetary Economics". Which got absorbed into "Money and Banking". Or maybe it was the other way around with M&B getting absorbed into ME once the finance people got purged and macroeconomists decided that somehow Central Banks have "preferences", i.e. "Monetary Rules"

Digression are these cardinal? ordinal? ordered subjective evaluations merely represented by a monotonic function invariant to affine transformations? Does BB wake up in the morning and say "I'd like to consume a little less inflation today" or where does this stuff come from?... It comes from people who insist on "Microfoundations" that's where it comes form. I did mention Schizophrenia, didn't I?

There's also the "growth and development" or "growth" and "development" specializations but even with quite impressive advances in both theory and data in these areas over the past twenty years they are still treated as "something we have to cover in ye ol' textbook but let's get it out of the way as quick as we can so we can start talking about business cycles" kinda way. Maybe it's the nature of the subject matter - and I don't think it is - but there's very little specialization in sub-disciplines in Macro (of course there's a good bit of specialization in the sense of running the same particular regression (or recalibratin') with minor tweaks over and over and over again) and well ... as Adam Smith told us, no specialization means low growth. From what I understand, these days fluctuations-Macro people specialize in explaining "puzzles".

And note that I haven't even brought up financial markets.

Consider another anomaly. Umm ... let's call it the "Data Theory Excess Methodology Puzzle" (D-Temp). At this point in time, Micro data is way way way better than Macro data. This is true even if you ignore all the aggregation issues and quibbles. To get at causation rather than correlation you need some really specific, well designed, and well collected, information and Micro is much closer to that than Macro which is still in the Torquemada stage of data gathering (i.e., torture and false confessions). Yet the standard Micro textbooks - assuming Gabriel's right here - focus on theory whereas the macro textbooks spend a lot more time on the data. Or, what I think I already said above, the determination of theory vs. data emphasis is simultaneous here. What gives? If you got good data then (assuming more data is a normal good) you should be more empirical. But if you look at the textbooks at first glance it looks like it's the bad-data-having-macro that has the more empirical approach.

This is the "D-Temp Puzzle".


Also note that most of the people who make the argument that "economics is too abstract", or "economics is too mathematical" or "economics as taught in grad school has nothing to do with the real world" or some convex combination of those three arguments, got their economics degrees like 15+ years ago. Give it about five more years and we'll start seeing (and I've been trying to do my part to start, resurrect, or more precisely reincarncate the trend) people talking about ad-hoc empirical estimations, causality, Lucas critique, structural econometrics and just the plain fact that the so called "Applied Micro" people are mostly making shit up and regressing the annual depreciation of your mother's slippers against your dog's fleas until it's significant (No. Really. That's really what they do. In fancy, statistical way but it's what they do. And then they call it "Letting the data speak for itself")

And when the pendulum swings I will feel a little better about being a Macroeconomist, even though a reluctant one.