Brazil was not a country before 1914, economically-speaking

 In comments, Jonathan asks the significance of the finding that Argentina is not alone in experiencing relative decline after 1914. Rather, it spans most of the Conosur. The fact that Uruguay also experienced decline is well known, so I will not comment more. And I will punt on Chile, since I know not sufficient economic history to say anything sensible ... there is evidence that Chile was in fact quite poor at the time but the headline figures for GDP per capita show a similar relative decline. But I will discuss southern Brazil. 

(Short version: southern Brazil was much richer than most think before WW1. It then experienced relative decline much like Argentina. Any explanation for Argentine decline that cannot also explain southern Brazil, or vice versa, is therefore likely to be wrong.)

Longer version: The popular version of Brazilian economic history is that Brazil in the 19th century was an underdeveloped country, like Peru. It then underwent a period of explosive catch-up growth in the 1950s and 1960s, but growth collapsed in the 1970s, only to pick up again recently.

That picture, however, does not jibe with what we now know of the non-economic history of Brazil or what we know of regional economic development in that country. Molly Ball (Rochester) has done some excellent work tracking down wage data in São Paulo. She collected data from textile firms, gas & electric companies, and the tramways. Ball’s textile wage data do not go back before WW1, but her series for the latter two industries go back to 1891.

It is possible to compare Paulista nominal hourly wages with U.S. wages for the same two industries, taken from Historical Statistics of the United States. (I calculated American hourly wages under the assumption that electric and tramway workers had a 48-hour week, which is consistent with data from the U.S. Postal Service; manufacturing workweeks fell from 55 to 50 hours in 1891-1913.)  I could not easily find data on comparable Italian wages, but is also possible to compare them with the estimates of hourly wages for skilled construction workers generated by Stefano Fenoaltea (University of Rome Tor Vergata). The United States was, of course, the second-richest destination country in the New World; Italy was a major country of emigration in the early 19th-century to both America and Brazil. 

Bearing in mind that these are nominal wages, the results look like this:

The basic image is that Brazilian wages were lower than in the United States, but much higher than in Italy. (They were within striking distance of much of northern Europe and beat Sweden.) It would be hard to call São Paulo underdeveloped the way we think of it today. 

(The Brazilian volatility is almost certainly an artifact of the fact that the estimates come from industries dominated by monopolies in one city in  whereas the American results are a national population-weighted average.)

Now, nominal wages are not real wages. It is possible that high prices in São Paulo meant that living standards were closer to Italy than to America. Brazil was highly protectionist at the time and had lousy internal transportation, so there are reasons to believe that many manufactured goods would have been pricier.

Fortunately, Luis Bértola and Carolina Román (Universidad de la República) have done much of the hard work. They find, rather astoundingly, that the cost of living in Brazil was one-third that of the U.K. before WW1, although it should be noted that their data is from Rio de Janeiro rather than São Paulo. Now, I strongly suspect that their estimates are off by a significant amount and the benchmark was Britain, not America ... but I would also bet that correctly accounting for the net effect of differences in the cost of living would still raise Paulista real wages with respect to the United States.

(I would also submit that for many migrants, who intended only to work and save in the new country for a while before returning home, nominal wages would be the relevant draw.) 

In short, southern Brazil was rich before 1914. Not as rich as Argentina, not as rich as the United States, maybe not as rich as the richest parts of Europe, but rich. 

The reason why Brazil as a whole ranked so poor in 1913 was that the United States of Brazil resembled the French Empire much more than it did an integrated nation-state like the Argentine Republic or the United States of America. Its huge northern hinterland, like France’s African territories, were populated by near-subsistence peasants isolated from the national economy and providing little save some tropical products. They were effectively prohibited from migrating south (or, in the case of France, north) by a combination of extreme poverty, poor transport links, and racial prejudice. 

The problem is that a mass southward migration of northerners masks the relative economic failure of the south after 1913. We know much less about the timing of that decline than we should. We also know much less about the reasons for the decline than we should; unlike Argentines, Brazilians not only do not obsess about their fallen position in the world, they’ve forgotten that they ever had a better one. 

The fact that the relative declines spans Argentina, Brazil and Uruguay means that Argentina-specific explanations are almost certainly wrong. Rafael di Tella may be correct that Argentina was “hit with 700,000 bullets,” but it can be asserted that the ones which did not also go to hit the neighbors did not kill the target. 

Kind of long, but I wanted to make the point about Brazilian relative decline. Did I answer the question?

Comments

[Jonathan]

Great explanation. So noted decline is the result of something that affected the entire Southern Cone. Policy unlikely, so some kind of geographical issue, like maybe cost of shipping imports?

[Noel Maurer]

Could still be policy, if it was something adopted by all the countries for some underlying reason common to the three.

[Randy McDonald]

But Australia and New Zealand also experienced relative decline on their parts, if (in Australia's case, at least) nothing quite so dramatic.

Interesting and interestinger.

[shah8]

Okay, my first inclination would be to blame it on the changes in the rubber industry that occurred around this time. And in general, a renewed process of industrial advances and colonial exploitations, like systemic rubber plantations in SE Asia and Africa, such that margins for primary exports fell or became negative. This would have had knock-on effects to local industries that sopped up the incoming money.

As a thought experiment, what do you think would have happened to northern US industry had the UK implemented a colonial scheme around cotton forty years earlier than it did?

Well, that's my lazy theory...

[Noel Maurer]

That was up in the north, around Manaus. It didn't contribute to the Sao Paulo economy: very few linkages. Not trade, not fiscal transfers, not labor migration, not rents going south. The rubber boom might as well have been in a different country; to effect, it was.

Same applies to Argentina. Remember, any explanation has to explain the entire Conosur.

There are other problems with the hypothesis, but that's the big one.

Almost certainly not rubber.

[shah8]

Understood. That's why I mentioned the cotton thing--northern manufacture had economic underpinnings from cotton export. Could be the money sloshing around Amazonas eventually reaches the hub and quietly underpin economic activity. I was also thinking in the back of my mind about Spanish gold and silver driving economic growth in the rest of Europe while undermining Spanish economy...

It's also understood that this was a very facile guess, because it's so obvious and any economist worth her salt would have twined together a tight theory of economic history. The alternative explanation for me was that shipping became a larger component of final goods provision somehow. Or it could be that the reason to be at distant ports declined, say, the whale oil industry made use of many ports and made further trade profitable, but when the whale oil trade declined, lower margin tradestuff went elsewheres.

Of course, we're basically sixty years too early, since whale oil industry declined severely from about 1850 on and there were booms later on. We could check what the really profitable goods (non-obvious stuff) were in the relevant time period, but I suspect that the actual reason is probably some absurd interplay between monetary issues from excessively hard Sterling of that era(austerity after Encilhamento--which popped, mind you, after a minor Argentine political revolution nearly caused widespread default), coffee glut dynamics, rubber?, mining, and industrial mismatch.

Historical mysteries are fun.

[Noel Maurer]

Rubber is not a terrible guess, but there are two things to note beyond the fact that it won't explain why Argentina follows the pattern.

First, cotton wasn't that important to the North in the United States. It wasn't nothing, but a counterfactual north without any links to the South develops about as fast. Maybe faster. See Irwin and Ransom and the rest. (You should read the collected works of Gavin Wright ... every American should. I'll add Leah Boustan when her book comes out next year.)

Second, rubber was even less important to Brazil as a whole.

The lack of links between the north and south is just icing on the cake.

But it isn't a bad first hypothesis.

[shah8]

Looking at Amazon, I appreciate the Gavin Wright stuff. Old South, New South got put on the wish list.

Irwin, Ransom, I'm not finding quickly at 4AM, so I'll find later. Boustan, only way I'll see that one is from a library, I think, and I'm a deep pile under of books I think I should read.

[Noel Maurer]

Leah's book isn't out yet; she's still finishing the manuscript. But some of the papers are here: http://www.econ.ucla.edu/lboustan/research.html

She is indispensable to understanding modern America. It is fair to think of her work as The Warmth of Other Suns, but with systematic evidence.

Her book is going to be seminal.

Ransom is two basic works: with Richard Sutch One Kind of Freedom: The Economic Consequences of Emancipation and the rather less serious The Confederate States of America: What Might Have Been. (Spoiler Alert: what might have been is very very bad until Teddy Roosevelt reconquers the place.)

Doug Irwin's work are mostly papers on the historical impact of protectionism.

With Wright, don't forget Sharing the Prize: The Economics of the Civil Rights Revolution in the American South. That plus Old South, New South may be the best economic histories of the United States ever written, because of their regional focus rather than despite it.

[shah8]

Obviously relevant link to before the crash...

http://www.voxeu.org/article/external-integration-and-development-evidence-argentina