Doug Campbell: Sitting in on a growth theory lecture:
Doug Campbell @TradeandMoney: Sitting in on a growth theory lecture. What is clear is that this field of economics is largely a scam. Top growth theorists know nothing at about topic they study. It’s remarkable to behold. Some of the misguided beliefs in yesterday's talk: (1) that property rights institutions are a key to growth. (Actually, it looks like a wide-variety of institutions are consistent with economic growth...)
Brad DeLong @de1ong: As somebody who has been predicting for 30 years that China’s lack of property rights institutions will cause it’s growth to crash in five years or so, you feel my pain...
Doug Campbell @TradeandMoney: This guy's take was that China was at risk of a middle-income trap because of its property rights institutions. I wonder how many years he has been saying this for. He also insinuated that China is mostly just copying and not innovating. An idea which is out-of-date by 15 years at least, if it was ever accurate.
Brad DeLong @de1ong: Let him go to Shenzhen...
Doug Campbell @TradeandMoney: (2) that there is a causal relationship between financial development and growth. In fact, there is no close relationship, and what relationship there is is unlikely to be causal. (3) that the number of patents across US states has a causal impact on state GDP growth. (Likely 3rd factor causality here...) (4) That Japan started growing slowly after 1990 because it didn't reform while Sweden did. Actually, what reforms the speaker wanted were not stated, but it appeared to be reform of the tax system. But, taxes in Japan are still much lower than Sweden, so taxes on innovation (5) are not going to explain Japan's poor performance since 1990 (and, yes, it has been poor). The liquidity trap, and the timidity of the central bank are the key explanatory variables. (6) the implicit assumption that if a paper is written by people from prestigious universities, inevitably very, very close friends of the speaker, and published in a top 5 journals, it must be correct. Even despite relatively obvious flaws, there for all to see. (7) At one point the speaker responded to a question about whether the correlation between two clearly endogenous variables was really causal by saying. "Yes, because the author had an IV." He didn't explain what it was. But, in reality, few IVs are convincing.
Brad DeLong @de1ong: Truly: the specification search for instruments and the pretense that geographical autocorrelations do not exist or the twin banes of the growth literature... ..."
Doug Campbell @TradeandMoney: (8) One piece of research the speaker cited at face value was the Bloom, Draca, Van Reenen paper on China. Indeed, this is a paper published by titans from MIT and Stanford, in the REStud, one of Econ's top journals. And it is also a comedy of sloppy errors. (9) But, hey, I'm not at Harvard, so what could I possibly know?
Epic Factory Owner @GurdCarm: What does this say about these top journals?...
Doug Campbell @TradeandMoney: Haha, I'm not the best person to ask. I have a rather cynical view about what it is that top economic journals actually do. What isn't appreciated enough is the "least common denominator" feature of publications in top journals. To publish, you need to write something the median referees in your field agrees with. In other words, you're probably going to need several people of rather middling intellect to sign off on a paper to get into a top 5 or top field. Combine this w/ no downward mobility in academia, the cliquesh nature of the field, the under-appreciated inherent difficulty of doing science, and of judging science, and it is easy to explain some real pseudoscientists at top departments...
#shouldread
