Wednesday, August 16, 2017

Alfred Crosby Deserves a Nobel Prize in Economics...

for his absolutely brilliant insights into Economic History and Development. Arguably, he should share it with Jared Diamond, who really succeeded in popularizing Crosby's ideas and also added quite a bit himself. Although some might say that he could have cited Crosby a bit more than he did. 

This came up recently when Anton Howes started a nice thread on twitter asking for recommendations for items to teach undergrads in economic history. I suggested teaching Crosby. Pseudoerasmus suggested it's easier to just give them Diamond, but agreed that, if there was no Crosby, there would have been no Diamond.

For it was Crosby, a historian by trade who decided he would look into the little-researched field of ecological history. Any grad student in Economic History/Development, or, scratch that, any grad student in economics, should go read at least one of Ecological Imperialism or The Columbian Exchange. The latter term, coined by Crosby, highlights the very unequal exchange of diseases and technologies between peoples of the old world and new. Europe got potatoes, corn, tomatoes, chocolate, and plastic. Native Americans got grains, livestock, bacon, and a whole host of vicious diseases which leveled their populations. The only disease to come from the Americas was perhaps syphilis, which is still debated. 
Suppose you wanted to believe that Europeans had better institutions, and thus colonized the Americas rather than vice-versa. However, Crosby tells the story of when the first American settlers coming over the Appalachian mountains arrived in Kentucky, and thought that Kentucky blue-grass was a native plant. In fact, Crosby notes that Kentucky bluegrass actually is native to France. It arrived in the Americas aboard a ship, and just happened to colonize the continent even faster than humans. The same was true of European dogs, mice, rabbits, and bees. It would be bizarre to argue that rabbits had better property rights institutions, and thus colonized the Americas. And, given that human's colonization of the Americas was not unique, it seems wrong to focus on unique aspects of humans. Weeds, bugs, and various other animals pulled the same feat. 

Here's an except from a Crosby essay: 
Europeans in North America, especially those with an interest in gardening and botany, are often stricken with fits of homesickness at the sight of certain plants which, like themselves, have somehow strayed thousands of miles eastward across the Atlantic. Vladimir Nabokov, the Russian exile, had such an experience on the mountain slopes of Oregon:
      Do you recognize that clover?
      Dandelions, l’or du pauvre?
      (Europe, nonetheless, is over.)
A century earlier the success of European weeds in American inspired Charles Darwin to goad the American botanist Asa Gray: “Does it not hurt your Yankee pride that we thrash you so confoundly? I am sure Mrs. Gray will stick up for your own weeds. Ask her whether they are not more honest, downright good sort of weeds.”
     Why was the Columbian Exchange so unequal? Essentially it comes down to the large Eurasian landmass being a far more competitive ecosystem than the north-to-south oriented Americas (admittedly, Diamond deserves credit for noticing the difference in the axes of the continents). Any type of grass in the large Eurasian landmass that adopted a new, useful trait would come to dominate the entire continent, as much of it shared similar climate. By contrast, if a type of grass in New York has some new beneficial trait, it will be unlikely to spread to Mexico or northern Canada given the vastly different climates there.  Thus, evolution would play out a little faster in Eurasia, as would technological progress among human societies. On the other hand, the distribution of useful large domesticable animals was probably just random. And this spurred the pandemic diseases. 

    Crosby notes that most European settlers ended up dominating the "Neo-Europes" -- areas of the world similar to Europe in terms of climate -- but much less so in the tropics. Many economists believe that institutions are the reasons the tropics are poor, or genetics.  However, Crosby lays out another reason:
The reasons for the relative failure of the European demographic takeover in the tropics are clear. In tropical Africa, until recently, Europeans died in droves of the fevers; in tropical America they died almost as fast of the same diseases, plus a few native American additions. Furthermore, in neither region did European agricultural techniques, crops, and animals prosper. Europeans did try to found colonies for settlement, rather than merely exploitation, but they failed or achieved only partial success in the hot lands, The Scots left their bones as monument to their short-lived colony at Darien at the turn of the eighteenth century. The English Puritans who skipped Massachusetts Bay Colony to go to Providence Island in the Caribbean Sea did not even achieve a permanent settlement, much less a Commonwealth of God. The Portuguese who went to northeastern Brazil created viable settlements, but only by perching themselves on top of first a population of native Indian laborers and then, when these faded away, a population of laborers imported from Africa, They did achieve a demographic takeover, but only by interbreeding with their servants. The Portuguese in Angola, who helped supply those servants, never had a breath of a chance to achieve a demographic takeover. There was much to repel and little to attract the mass of Europeans to the tropics, and so they stayed home or went to the lands where life was healthier, labor more rewarding, and where white immigrants, by their very number, encouraged more immigration.
One of the big problems with both Acemoglu, Johnson, and Robinson's (AJR) work on development, and also with Spolaore and Wacziarg's QJE paper on genetic distance, is that they hadn't actually read, or properly internalized, the teachings of Crosby and Diamond. AJR argued that disease climate was a proxy for institutions and not geography, whereas clearly one might make a case that it's also a good proxy for climatic similarity. Areas of the world where European peoples died are also areas of the world where European crops and cattle also died. This was an insight that AJR missed. If Acemoglu got a John Bate's Clark for his work on institutions, then Crosby certainly deserves a Nobel. 

    Of course, there's more here. The insights of Crosby/Diamond don't end in 1500. First, history casts long shadows. But, aside from that, the world was largely agrarian even long after the Industrial Revolution in 1800. In Malthusian societies, agricultural technologies are very important. If you are a farmer in Angola, those new varieties of wheat, and farming technologies discovered in the American midwest in 1900, or even 1950, are not going to help you. If you live in Australia, the Southern Cone countries, or Europe, they will. 

    One thing that is fascinating is that, while the US is similar to Europe climatically, it's not 1-for-1. The US is much hotter and more humid in summer. The US south actually does share some similarities with Africa. And it was in the US where air-conditioning was invented. This technology almost certainly was more helpful than the tropics. (Note: I haven't turned on my AC all summer in Moscow, Russia!...)  But, again, I see this as a Crosby/Diamond type of affect.  

   In any case, here's a nice interview of Crosby. Here's a short reading you can give to students. Here's Crosby's wikipedia page

   Alfred Crosby is now 86 years old! He deserves a real Nobel, much less the fake one awarded in Economics, and he deserves it sooner rather than later. 

10 comments:

  1. Maybe we should read AJR again:
    "Many social scientists, including Montesquieu [1784](1989), Diamond (1997) and Sachs and coauthors have argued for a direct effect of climate on performance, and Gallup, Mellinger, and Sachs (1998) and Hall and Jones (1999) document the correlation between distance from the equator (latitude) and economic performance. To control for this, in columns 5-8, we add the absolute value of the distance from the equator in degrees (latitude) as a regressor (we follow the literature in using a measure of latitude that is scaled between 0 and 1). This changes the coefficient of the index of institutions little."

    ReplyDelete
    Replies
    1. Nice try, but no. Two main problems with AJR:

      1. Death rates of European settlers could itself be a proxy for climatic similarity to Europe. Indeed, it could be better than simply just latitude. This is because, for example, much of Europe and the US have similar average temperatures even though Europe is much further north. Latitude is more different than climate. Good settler mortality data would probably pick this up.

      2. Go read Albouy. The IV they construct for "institutions" has unreasonably low settler mortality rates for Singapore and Hong Kong; they mistranslated the name of an African country, and they assigned settler mortality to most of Latin America based on 18 deaths of clergy in Mexico. Nobody takes that data seriously anymore. https://www.researchgate.net/profile/David_Albouy/publication/268365609_The_Colonial_Origins_of_Comparative_Development_An_Empirical_Investigation_Comment/links/5544e9980cf234bdb21d4e42.pdf

      Delete
    2. Like there was no reply from AJR to Albouy?
      https://economics.mit.edu/files/8620

      Delete
    3. There was a reply, yes. But a serious reply, no. It was merely intended to try and cloud the issue as much as possible.

      First, their reply says nothing about point 1 above. Or a related problem that settlers could also bring human capital and/or their technologies with them, and not just institutions.

      Second, to see how disingenuous AJR can be, check out this version of the response: https://pdfs.semanticscholar.org/17b7/d3b6bc1d42d3044229c93c38f2d47f78c93a.pdf

      In it, they write "The substantive disagreement between AJR (2001) and Albouy's current and previous comments revolve around Africa."

      That statement by AJR is patently false, and clearly intended to mislead. The substantive disagreement also revolved around the American and Asian data points. Yes, they became more sophisticated in their rebuttal as time went on, but it doesn't change the fact that their original data was sloppy, and making the Albouy changes makes their results go away. Or that their IV was invalid for various reasons in the first place, so the debate over their data isn't a productive use of our time in any case. The profession has moved away from the Settler Mortality IV in any case.

      Delete
    4. 1. If a better variable besides latitude is available, then why not to use it and show how it changes AJR results? Why Albouy did not use it?

      Hypothesis of different human capital endowments was studied in AGR 2014 and it did not work.

      2. Albouy changes in data were sloppy as well. And not only that. Albouys result was driven entirely by one outlier (within 28 observations). After dropping the outlier, AJR result is back there.

      Sure, the research goes on (I like the Dutch-Indonesia case), but mortality IV is used from time to time (see Gorodnichenko and Roland recently)



      Delete
    5. Re: "If a better variable besides latitude is available, then why not to use it and show how it changes AJR results? Why Albouy did not use it?"

      Bro, re-read the post and my responses above more carefully. "1. Death rates of European settlers could itself be a proxy for climatic similarity to Europe." Thus, the other variable I choose to proxy geography is settler mortality.

      Albouy showed a few changes to the series killed the results. Go check out his tables. Why should he introduce a new IV?

      As I wrote above, it's a waste of time to try to get down in the weeds of individual points. The AJR response is designed to fool exactly people like you. But it's not true that just one outlier drives the Albouy results. This was more AJR nonsense.

      However, a reason why I don't go and write a paper about this is because no one really takes that paper seriously anymore. Oh, sure, you probably are not the only dead-ender, and Acemoglu certainly has a fan club, but outside of that club, the paper is openly derided for good reason.

      Delete
    6. Man, you've said that latitude is more different than climate and I responded that go ahead and replace it (the latitude) with smth better. I am not asking for a new IV.

      Second, it is very easy (or so cheap) to write that mortality *could be* itself a proxy for climate without providing any evidence for such claim.

      But now it seems like you have not read AJR at all. In fact, AJR used numerous proxies for climate: various tempertures, humidity and soil quality.

      Albouy's paper is not about few changes in data. Just the contrary. He examines majority of datapoints to argue that the IV is weak.

      You are fooling yourself.

      Delete
    7. To reply to the first bullet: yes, if you use the settler mortality data that assigns very low levels of mortality to Singapore and Hong Kong, both in Malaria-central, then it's going to be hard to find a geographic variable that will beat it. But, if you use the revised data, which assign higher rates of mortality to places like Singapore and Hong Kong, then settler mortality is no longer robust.

      On your 2nd point, go read Alfred Crosby. Immediately. The evidence here is quite overwhelming.

      On your third point, I never wrote that AJR don't control for any geographic variables. If you read AJR, you'll realize they didn't consider the possibility that settler mortality could be a proxy for climatic similarity with Europe. And thus that there is a good argument for using it as an IV for climate rather than for institutions.

      I agree that Albouy finds flaws in a majority of the AJR data. However, you only need to make a handful of changes to kill the result.



      Delete
  2. This comment has been removed by a blog administrator.

    ReplyDelete

Note: Only a member of this blog may post a comment.