Created 11/9/1997
Go to
Brad DeLong's Home Page

Review of:

Kevin H. O'Rourke and Jeffrey G. Williamson, Globalization and History: The Evolution of a Nineteenth Century Atlantic Economy (9/5/97 version).

J. Bradford DeLong
U.C. Berkeley

Marginal notes on O'Rouke and Williamson's forthcoming book, Globalization and History .

First, this is going to be an absolutely wonderful book--a magnificent book. The synthesis of material that is pulled together is most impressive. The topic is extremely important. And the gap that it fills has yawned very, very large.

If you do not want to publish this book, other university presses will snatch it from your hands with the greatest possible speed. I think this will become the standard reference on the real side for the next decade or two for anyone interested in the growth of the pre-1914 world economy.

Some more detailed reactions:

  • Page 1-1: On the first page, where the authors state that their convergence club "excluded, of course, most of the Third World and eastern Europe." Why the "of course"? Arthur Lewis--whose Evolution of the International Economic Order plowed somewhat similar ground--thought that the exclusion of the periphery from the wave of industrialization was very, very interesting and not at all obvious from first principles. Say that the convergence club excludes eastern Europe and the Third World, and say that this book is not going to deal with processes taking place therein, but don't say "of course,"
  • Page 1-1: One of the central issues that any reader of this book will think about will be why was pre-World War I convergence so globalization-driven when globalization is almost always assigned a secondary place in studies of post-WWII convergence and distribution trends. Is the character of international trade fundamentally different now? Are today's economists blind as bats? I think it is important to foreshadow applications to contemporary debates, and I do not think that page 1 is too early to foreshadow them. I do think that more needs to be said than that "these facts are directly relevant to debates over globalization today."
  • Page 1-3: when the authors write that "many poor countries in the Third World only exhibited catching-up when the researcher controlled for the fact that they were hampered by high population growth, low investment rates and human capital shortfalls," they raise the question: should researchers control for these? Aren't they as much consequence as cause of poverty--if you are poor your real exchange rate is low for Balassa-Samuelson reasons, and you can't afford to import anything. If you are poor your households' calculations of the benefits of more children are different than if you are rich.
  • Page 1-4: I'm not sure I like the classification of Abramovitz and Baumol as "less formal, historical." I always have a hard time saying what I learn from the "formal" cross-country growth literature (even from my own contributions to it) because it seems that there is just too much left out--that the theories and null hypotheses are just too implausible, and the tests of hypotheses too unconvincing.
  • Page 1-5: Would the pattern still be the same if your net included more well-positioned Latin Americans? Uruguay, Chile, Venezuela, Brazil, and Mexico all should in some sense have done much better than they in fact did.
  • Page 1-6: I thought the points that it is important to get the history of the OECD right before one ventures outside, and that factor prices have at least as large a claim on our attention as per capita incomes, were both very powerful. I would urge you to strengthen them.
  • Page 1-7: Can you do more to quantify the reduction in tariff barriers in the mid-nineteenth century?
  • Page 1-8: C(8) rises from .143 in 1830 to .402 in 1846? Can this possibly be correct?
  • Page 1-8: The US increased her real wage advantage from 45% in 1830 to 89% in 1846--implying a rate of real wage growth in the U.S. of some 4% per year or more? Can this possibly be correct?
  • Page 1-9: If C(13) shows a later turning point than C(8) for the shift from divergence to convergence, won't the C(large) that you really wish you had data for show an even later turning point?
  • Page 1-10: Is it the Argentinians who call 1870-1913 the "Belle Epoque", or is it the French?
  • Page 1-14: Why aren't trends in the wage-rental ratio an indicator of performance per se--shifts in the factor-price frontier are very interesting measures of technological change, and movements along the factor-price frontier are very interesting measures of structural change.
  • Page 1-14: What is "Austria" before 1918 to Angus Maddison?
  • Page 1-17: The US's sudden leap ahead in productivity between 1890 and 1913 appears more remarkable than ever.
  • Page 1-22: I missed the fact that the authors' "real wage rates do not take account of unemployment". How do the authors' real wage rates fail to take account of unemployment? How should they take account of unemployment?
  • Page 2-2: I'm not sure whether a shift in demand toward (nontradeable) services is a decline in globalization or not. I suppose it is a philosophical question at bottom.
  • Page 2-4: It would be nice if the mention of the Erie Canal carried with it an estimate of how much of the value of bulk cargo the price of shipping it along the Erie Canal amounted to.
  • Page 2-5: But there were no coaling stations between Port Said and Bombay? Even that distance seems a little long for the steamships of 1870 (but maybe I'm wrong).
  • Page 2-9: From what base had averaged duties on manufactured products imported to the European continent declined? How much of a reduction was the reduction in average tariffs to 10%?
  • Page 2-11: "It cost 6 shillings and 11 pence to ship a quarter of wheat by lake and rail from Chicago to New York in 1868." How much does a quarter of wheat weigh? What is the value of a quarter of wheat in Chicago in 1868?
  • Page 3-2: My impression is that Adrian Wood is losing his debate--Richard Freeman, at least, has said that the way to understand Wood is that the effects of globalization on demand for U.S. unskilled labor is large if and only if the U.S. would under autarky be using Third World technologies to produce goods now produced in the Third World, but it wouldn't. I think Freeman is correct. After all, Paul Krugman first made his reputation by pointing out that the net embodied factor content of post-WWII trade flows appears to be surprisingly small...
  • Page 3-10: I worry about treating the U.S. as a point in any CGE...
  • Page 3-16: What (if anything) is Angus Maddison's estimate that Japanese GDP per capita increased by 17% from 1820 to 1870 based on?
  • Page 3-17: How much of Punjab agricultural product was exported to Europe in 1906? Is the export boom large enough to plausibly drive such a 68% fall in the wage-rental ratio?
  • Page 4-3: I think that Irwin deserves at least a paragraph more on why he reached the conclusions that he did.
  • Page 4-10: Surely we can make guesses about international trade elasticities? Surely the authors are better-placed than anyone else in the world to make such guesses. They should not leave us hanging in mid-air. Was Torrens right?
  • Page 4-14: I don't understand the "price shock" view of the effect of Repeal on Irish emigration. Surely food and rent were major components of Irish workers' consumption. Even if Repeal sharply lowered the price of Irish agricultural output and the wage of Irish workers, it also lowered the price of their consumption bundle. Even a large fall in the real wage (deflated by the GDP deflator) would have been only a small fall in the real wage deflated by Irish farmworkers' consumption. Have I missed something?
  • Page 4-15: We live in degenerate times. How many of the readers of this book will know who Phileas Fogg was?
  • Page 5-1: As a whole, the globalization backlash chapter is very nice.
  • Page 5-10: Why are the trends in price gaps not distorted by problems of comparability?
  • Page 5-14: How much evidence is there that the "marriage of iron and rye" made sense for iron-masters and rye-growers? I have always had a suspicion that one (maybe both) of the participants was a dupe. And I would like to see more about what would have had to be true about the structure of trade and production for this set of matched tariffs to have made any sense.
  • Page 6-13: I sense that the atuhors may not fully believe their findings that tariffs aided growth. I find myself skeptical as well: it seems to me that the costs of tariffs are first order--and that the reductions in investment from the lower GDP caused by higher tariffs are first order--while the gains from concentrating investment in externality-heavy sectors are second-order. What else might be going on in the authors' regressions?
  • Page 7-1. How about a scale estimate of the African "migration"?
  • Page 7-7: "Spain where the destination wage is represented by Argentina alone..." The pattern of outmigration from Iberia gives at least an estimate of going to a place where the folks spoke your language. Perhaps the authors could use the Iberian experience to reach at least some conclusions about cultural barriers and emigration? Did French have a disproportionate tendency to migrate to Quebec?
  • Page 7-11: The story of finance-constrained emigration at the beginning turning into market wage-difference driven emigration sounds convincing. Is there any more evidence that could back it up--other than its inherent plausibility? How about collections of letters about how now we can migrate because great-uncle Federico has sent the money for the ticket that you could cite? The ticket evidence is very nice, but I have a feeling that more would be good.
  • Page 8-12: Is there good reason to think that Ireland and Sweden are representative cases for Europe as a whole? Anything odd about the countries that would make them unrepresentative?
  • Page 8-15: David Card points out that in recent years immigrants into the U.S. as a whole are not that much more unskilled than native workers (I'm writing this on my laptop at UCLA, where I am a guest of that incredibly articulate and entertaining madman Deepak Lal). Immigrants from Mexico into Texas, Arizona, New Mexico, and California, however, are a different kettle of fish...
  • Page 8-21: The partial-equilibrium migration impact assessments are very, very large indeed. Is it the sheer magnitude of the migrants relative to unskilled workers? Or is it something about the authors' labor demand elasticities? If the second, do they really trust the labor demand elasticities?
  • Page 8-23: "The relative insensitivity of GDP per capita convergence to migration is a result of countervailing effects inherent in the algebra..." Such countervailing effects presumably meant that farmers who owned land, or members of the petite bourgeoisie who owned stores could benefit substantially from the immigration. I find myself wanting a decile-by-decile distributional analysis of the effects of immigration, taking account of the distribution of property across the income deciles.
  • Page 9-5: When I read Borjas's stuff, I get confused. How does inequality fare among the native born as a result of immigration? The convention is to look at slots in the labor market (high school degree/unskiled), but in the presence of immigration more of the native-born will find it worth their while to go to community college--and the prices of goods made by the unskilled fall in any event.
  • Page 9-7: I could be wrong, but I had thought that Adrian Wood argues that there are three types of labor: skilled and semi-skilled in the industrial economies, and semi-skilled and unskilled in developing economies. Thus increased trade increases inequality in both regions: the skilled gain relative to the semi-skilled in the OECD, and the semi-skilled gain realtive to the unskilled in the periphery.
  • Page 9-17: I think it would be worth some effort to calculate the effect of immigration and trade not just on relative inequality but on absolute standards of living of the poor native-born in the United States. Consider public provision of some commodities, the opportunities open to even the poor native-born to "improve themselves," and the effect of increased immigration on the price of services. Taken all in all, were America's native born poor hurt in absolute terms by pre-WWI trade and immigration? Or were they just hurt in relative terms?
  • Page 10-2: In the U.S. tariffs were not on the rise before 1914.
  • Page 10-5: I found myself wanting a little bit more on the California/Australia experience of "premature anti-immigrationism". Do their reactions in the 1880s truly foreshadow what was going to happen throughout the white settler regions a generation later?
  • Page 10-21: I found myself wanting to know about the standard errors of the estimates of policy-index drops in Timmer and Williamson. How tight are these estimates? How variable are the estimates to small changes in the specification? How many degrees of freedom are there in the regressions? What are the influential observations?
  • Page 11-2: I'm not sure that Eichengreen would blame capital flows for the end of the golden age. I think he would say that the failure to figure out how to handle an international economy that included capital flows was partly responsible.
  • Page 11-4: "United States... registerd high levels of foreign capital dependence towards the end of this century." Was 26 percent the net stock or the gross stock of foreign liabilities?
  • Page 11-5: Arthur Lewis would claim that labor was not in surplus in either Malaya, Ceylon, Kenya, or South Africa--and that massive flows from India and China were needed to provide the labor force for the plantations.
  • Page 11-7: May I complain that ex post rates of return are simply too crude a measure?
  • Page 11-10: I'm not sure that Hobson's theory was Marxist. It seems more proto-Keynesian to me. Of course, Lenin later stole it...
  • Page 11-13: " enormous 115 percentage point difference [in the price of capital] between Sweden and Japan"! Where does it come from? Why didn't this diferential cripple Swedish growth?
  • Page 11-14: More details of Obstfeld and Taylor's argument would be appreciated.
  • Page 11-16: I think that an extra paragraph or two giving more details of Taylor's argument about foreign investment and Argentine retardation would be a good thing.
  • Page 11-17: By now, between migration, trade, and capital, the authors have accounted for much more than 100% of convergence before 1914. This suggests that other, presumably technology-related factors were driving countries apart before 1914 very, very rapidly. Why, given that technology looks a lot like a public good?
  • Page 12-8: I liked the discussion of the frontier and trade-migration complementarity.
  • Page 12-12: Is the Kuznets cycle really driven by trade-migration and the expanding frontier? Or is it driven by animal spirits, and endogenous reactions to low unemployment when people in the periphery are very optimistic?
  • Page 13-1: I would drop chapter 13. It seems like a sudden intrusion from some other book.
  • Page 14-3: Can I complain that convergence "controlling for initial educational levels" is no convergence at all? A poor country with a high initital education level is a strong sign that the country isn't really poor--and perhaps the national product is wrong.

If the authors do everything I have asked for, they have an immense amount of work to do. But work is good for the soul.

If they do nothing I have asked for, the book is very much worth publishing. But I think this book stands a good chance of becoming a classic--something that will still be on economic and social history reading lists a generation from now. And since it stands a chance of becoming a classic, it is worth working hard to make it as good a book a possible.


Created 11/9/1997
Go to
Brad DeLong's Home Page

Associate Professor of Economics Brad DeLong, 601 Evans
University of California at Berkeley; Berkeley, CA 94720-3880
(510) 643-4027 phone (510) 642-6615 fax