« Memo Question for February 27: Colonialism | Main | Memo Question for March 12: Technology, Investment, and the Industrial Revolution »

February 29, 2008


Monica Deza

Engerman and O’Brien stated that from slavery and trade with the rest of the world contributed to European capital accumulation; however, this contribution was not very significant. Even Europe had obtained a very significant growth in capital accumulation, they growth caused by it could not have been permanent. Thus, to understand European economic growth, we should focus on other factors that could have come from trade besides direct impact on profits and capital accumulation.
Acemoglu finds that Atlantic Trade Potential-the ratio of Atlantic Sea coast to land area- has a significant effect on urbanization- a proxy for GDP. I am a bit skeptical about the effectiveness of this regression. The
regressor Atlantic Trade Potential (ATP) is only, as its name says,indicates only potential trade, not actual trade, and the fact that a city had a vast sea shore does not mean that it was being used efficiently in trade.
Regarding the question for this memo, “How can anyone so confidently assert that the intercontinental trades caused ‘European growth’?” My answer would be that there is no way. Growth cannot be caused by merely on factor. I agree that intercontinental trade played a role on it. However, we cannot ignore trade within the continent through the Mediterranean. I think that trade increases production, which causes multiplier effect. Intercontinental trade also calls for institutions and reliable regulations, which do not only promote trade, but also help citizens feel more safe about investing.
It seems that all the readings for this week have a different theory of what caused the growth of Europe.
Austen and Smith focus on the increase in consumption of non essential goods such as sugar, tea and coffee. The importance of these goods was caused by the rise of domestic production of goods that were not available only for the rich.
All the readings for this week indicated that different indirect effects from trade had an effects on European growth.

Edson R Severnini

Acemoglu, Jonhson and Robinson (2005) [AJR] try to explain the rise of Western Europe with an “institutional change” argument. Their central point is that Atlantic trade (the opening of the sea routes to the New World, Africa, and Asia and the building of colonial empires) contributed to the process of West European growth between 1500 and 1850 not only through direct economic effects, but also indirectly by inducing fundamental institutional change. According to then, in countries with easy access to the Atlantic and without a strong absolutist monarchy, Atlantic trade provided substantial profits and political power for commercial interests outside the royal circle, including various overseas merchants, slave traders, and various colonial planters. This alteration of the balance of political power contributed to the emergence of political institutions protecting merchants against royal power, and with their newly gained power and property rights, those countries took advantage of the growth opportunities offered by Atlantic trade, invested more, traded more, and diverge to a higher level of economic growth (measured by urbanization).

Although the argument makes so much sense, some pieces of empirical evidence do not support the essential point of it: the profits earned in that period were not “substantial: even AJR mentioned that they were modest. So, if this is the case, were them large enough to drive the demand for new institutions by the group of commercial interests? AJR argues that although the profits from Atlantic trade were small relative to GDP, they were still much larger than previous trading profits. Furthermore, they defend that the recipients of them became very rich by the standards of seventeenth and eighteenth-century Europe, and typically politically and socially very powerful. Unfortunately, the authors did only speculations about these points: they did not present convincing evidence for these justifications. How bigger than before the profits became with the Atlantic trade? How richer the recipients became with those profits? I think that the authors should have provide more data about these issues because the profits and the consequent economic power are the main force driving institutional changes in their idea. Based on the evidence that they presented, they cannot so confidently assert that the intercontinental trades caused “European growth”.

Austen and Smith (1992) also argument that some “institutional change” resulting of the Atlantic trade contributed to the rise of Europe. They suggest that the slave and sugar trade was essential to the European industrial development precisely because it simulated and ultimately reshaped the entire pattern, organization and meaning of Western consumer demand. They present some evidence that high levels of imported luxury food consumption appear to be linked to indicators of economic growth pointing toward industrialization, for instance. Neal (1990), on the other hand, emphasize the role of the Atlantic trade in generate the “financial revolution”, another “institutional change”. He mention, for example, that the Portuguese and Spanish discoveries in the East and West Indies at the end of the fifteenth century required merchants throughout northwestern Europe to develop new financial techniques in order to exploit the opportunities of long-distance trade. Therefore, although the argument and the measurement of economic growth differ among the authors, the central point that some institutional change resulting from the Atlantic trade drove the rise of Europe is similar among them. Although without a strong empirical evidence, all these authors seem agree that the colonial relations with the New World and Asia contributed to the European growth, direct or indirectly.

Wayne Feng

Although there is ample evidence of trade was not a significant proportion of GDP (de Vries) (Acemoglu, Johnson, and Robinson aka AJR), we see that with comparisons of states where there was significant Atlantic trade versus states without significant trade that there is evidence that trade did contribute to increased European growth. This contribution did not come about via profits but rather is profound in the economic institutions developed and the redistribution of income/power within these developing countries. AJR argued for the analysis of capitalist institutions which developed in Britain and the Netherlands. de Vries points out, however, the difference between the Dutch and British experience. The definition of growth and period of growth (especially in comparative analysis) is demonstrated as very important. From de Vries we have to question as to why the Dutch did not have as much growth as the British, for de Vries this was the result of differing population growth rates. For us to relate different time periods and different countries is too confining to fully understand the development of the countries according to de Vries. However, there is a clear difference between the growth patterns of countries such as Italy and Spain versus those such as the Dutch and the British. Austen and Smith persuasively argue that the effect of traded commodities was minimal and even proceeded the industrial revolution, pointing to (as they argue) a demonstration effect to gain “respectability”. Thus if the effect of trade was one that was not noticed in the profits and changing supply/demand, then what effect did trade have? With the promotion of trade, certain powers and parties were placed into positions where capitalist institutions could develop. AJR points out clearly the example of Italy where the restrictions upon the amount and type of trade that could develop limited the development institutions. At the same time, we must also keep in mind what de Vries pointed out with the Dutch vs. British example where the bourgeoisie existed (social capacity) but development was slowed compared to later Britain. From this we see that trade is something that is more complicatedly related to growth than initially thought. Trade and growth is related with institutional development, redistribution of income, and consumer taste. At the same time, the profits from trade although themselves not significant in volume are the driving forces which led to economic structural changes. By comparing the trade growth and history of different European nations we see that the role of trade does have a significant correlation to growth despite the fact that the profits from trade itself seemingly did not directly relate to growth (capital accumulation).

James Zuberi

Acemoglu, Johnson and Robinson confidently assert that intercontinental trades caused “European growth” and led to the ‘first great divergence’ by indirectly affecting the institutions of certain countries in such a way that the changes spurred economic growth. They contend that in 1500 and 1600 in western Europe countries were divided between those countries which had absolutist governance structures and those which did not. They take Britain as an example of a society which was not absolutist and Spain as an example of one which was absolutist. The particular channel that trade affected countries which were not absolutist is that trade increased the wealth and so influence of merchants and in those countries they were able to change the balance of power to account for their newfound wealth. In this way, property rights were much better protected in the countries which reformed themselves in this way and the better property rights caused economic growth.

In response to the second question, I think the answer depends on the definition of growth but not necessarily a more complex model of the relationship to growth. In the paper, the authors use urbanization rates as a proxy for growth admittedly because they do not have data for economic growth by state for the large time period which they study. They then use a very incomplete and rough estimate of a dataset to test the results that the get from urbanization regressions. One the few places that the authors did not, in my opinion, provide sufficient historical or other evidence was on the link from urbanization to economic growth. Urbanization definitely requires some level of technical sophistication, but the fact that countries which were engaging in trade had higher levels of urbanization is not in itself compelling. Even if one were to show that nonabsolutist states were growing at a slower rate, there are surely many possible explanations one could give.

As for a more complex model of the relationship of trade to growth, I do not necessarily think that this is the case. Definitely a more complex model exists in reality between trade and growth, but that is not the question. Unless there are large and significant mechanisms at work which are not being addressed by the model, the adding of linearity in the functional form or other elements of structure that OLS invokes do not obviously make me wary of these results. I am not a historian, so I cannot address the lack of any mechanism. These results at the very least show that an explanation for growth using interactions between trade and institutional reform is plausible,


Acemoglu, Johnson, and Robinson (2002) mainly provides two points:

1. Most differential economic growth in Western Europe can be accounted for by the corresponding differential growth of Atlantic trade.

2. While the major impact of trade and the profit associated with it was relatively small, the recipients of such profit became economically and socially powerful and then facilitated the "capitalists institution", which promoted the economic growth in Western Europe.

They run a regression of GDP per capita onto Atlantic trade. The former is measured by urbanization rate and the latter by a dummy variable of belonging to a set of countries or a coastline-to-land ratio. The estimation shows that the Atlantic trade explains the difference in economic growth between Western Europe and Eastern Europe. They also identify other factors that affect the growth, like religion. Those results are quite robust under different specifications and different measurements of data.

Another regression they ran is that of institutional measure onto Atlantic trade and other regressors. They found that the difference in institutional change is significantly correlated with the Atlantic trade measure.

Although their argument is quite persuasive, I'm not very comfortable with those regression analyses. First, the measurements of variables in interest are very rough. In particular, the Atlantic trade potential, the most important explanatory variable, is either a dummy variable with arbitrary choice or the coastline-to-land ratio.

Second, if the measure of capitalist institution is omitted in the first regression, this will cause the severe upward bias in the coefficient estimates. The omitted variable bias on the trade potential is expressed as a product of the true coefficient on the institution and the covariance of the institutional measure and the trade potential divided by the variance of the capitalist intuition measure. If the intuition really matters, the numerator would be large and, as the institutional measure and the trade potential are highly correlated, it would be even larger. On the side of denominator, the variance of trade potential is quite large between Western Europe and Eastern Europe, but that of the Atlantic traders, Britain, Netherlands, Spain, and France, are not and thus leads to bigger bias on sub sample of these.

Third, on theoretical side, I'm not fully convinced why the intercontinental trade over Atlantic ocean had such a decisive effect on institutional evolution. What they identify as a changing force is the profit on merchants and not the trade itself. To me, it rather seems that commerce-friendly nations allow more business freedom on those merchants and thus they expanded trade among other things. This also explains the difference between West and East and between non-absolutists and absolutists as long as there is no unsuccessful free nation in Eastern Europe, which I've never heard of.

Fourth, the comparison with Asia should have been emphasized. Asian countries wildly vary in their attitudes towards commerce and thus the hypothesis should apply here, too. One possibility is that Asian traders cannot have measure to conduct intercontinental trade with Europe and the associated monopoly rent was only extracted by European traders. However, then as the articles in the previous week shows, such profit had been limited most of the time. If relatively small profit did change the institutions in Europe, it could have happened in the Asia as well.

While I listed my rather critical thoughts on the article, the idea seems very legit. What is necessary here is not pretty unreliable data and statistical analysis, but more convincing story how those merchants actually affected the politics in some countries and how they failed to do so in other countries. That would be more useful to show the Atlantic trade did have a causal effect on institutions and thus on economic growth.

Ernie Tedeschi

We tend to think of economic growth as a linear phenomenon that is
purely a function of some measure of income, such as per-capita GDP. Any
enterprise that adds to income contributes to growth, and any period of time
that sees monotonic real increases in output is similarly a time of growth.
There is no need to rethink this side of the definition: it is clearly sufficient
evidence when, say, a new technology increases productivity to then con-
clude that it contributes to growth. In the context of economic history, it is
the necessity of this condition, however, that one ought to reexamine. The
lack of direct growth effects does not diminish an enterprise’s importance in
contributing to economic expansion.
This seeming paradox is a common thread throughout this weeks read-
ings, and is alluded to in the article by Acemoglu, Johnson, and Robinson:
how could the Atlantic trade – with some notable exceptions – return rel-
atively weak profits and yet be considered so crucial in European economic
development? Acemoglu et al tease out the answers in a series of models,
finding that active European trader nations were more likely to develop the
institutions necessary to spur economic growth, such as strong property rights
and checks on executive power. Anticipating causality questions, they use
their time-series data to show that the Atlantic trade was not a significant
proxy for the presence of these institutions before New World colonization
accelerated in the 17th century. So while New World trade itself hardly filled
the coffers of many bank accounts (again, with notable exceptions such as the
sugar trade), it did empower a merchant/trader class to the point where they
had the power to demand property protections and institutional oversight of
royal power.
Austen and Smith also point to a non-institutional benefit of the Atlantic
1trade that indirectly fueled economic growth: the spawn of consumerism that
resulted from the (relative) mass availability of New World products such as
sugar and tobacco, availability which, when limited by supply shocks and
shortages, often caused mass unrest. Without the consumerism forged by
the Atlantic trade, creating the anticipation among many to trade up to new
and exotic goods should they prove within reach, no large-scale market to be
satisfied via mass production would have existed in the first place, and the
Industrial Revolution might have been more limited in scope and effect.
In a final reading, Jan de Vries looks at the growth of the Netherlands
throughout modern history, mostly compared to the United Kingdom, and
finds its pattern of economic expansion unique and unusual, yet no less im-
pressive when viewed in retrospect. He concludes that the technical definition
of economic growth ought to be loosened so as to allow for more organic ebbs
and flows in income. Insofar as some economists may dismiss the Atlantic
trade as a waste of capital without considering the broader political and
economic ramifications, he has a point.

Mark Borgschulte

We know, from our historical vantage point, that long-run economic growth is driven by the accumulation of ideas. Therefore, we know the rise of political institutions that enabled the capture of economic rents by inventors and their investors must be the driving force behind the rise of Europe. (Am I oversimplifying this? Because it seems pretty straightforward to me.) AJR posit, and I would support, a complicated relationship between trade and growth, which includes interactions with the political institutions in the home country. The definition of growth remains change in real GDP per capita, though their mechanisms have their strongest effects on lagged GDP per capita.

What we know about idea-driven growth allows us to reduce our question of "why Western Europe?" to "what about Western Europe led to the development of these institutions?" I'm not sure that AJR answer this question completely, but they certainly give us some food for thought, and point us in the right direction. Their thesis, that European trade gave rise to a merchant class that eventually won political, and hence, property rights in non-absolutist countries, seems entirely plausible, even if the causal pathways between the rise of the merchant class and the emergence of growth promoting institutions is not entirely clear.

Surely property rights were essential, but I'm not sure they are the end of the story; some further questions occur to me. As Neal argues, did the emergence of financial institutions play an essential role, especially as an intermediate stage in the rise of merchant power, or were they merely the consequence of the rise of the merchant class and expanding property rights? How important were corporate organizations that were created to manage trade with the New World? And what about the education and specialization of the work force? Finally, what is the connection between physical, or financial property rights and intellectual property rights- and how did these forces contribute to the take-off period of European growth?

One specific line of questioning that emerges from our readings relates to who it was that drove, and who gained, from the expansion of property rights. Specifically, how much of the change was driven by the rise of small group of merchants pursuing their narrow interests (what we might call an "oligarch" theory) and how much was driven by the removal of restraints on upward mobility, creating a much larger class of potential inventors and investors, which ultimately led to further political changes ("democratic" theory)? Though profits were not large, they were concentrated in a few hands, which may recover their importance, and suggest we focus on the "oligarch" theory. On the other hand, the argument advanced by Austen and Smith, that the rise of consumerism as a result of the Atlantic trade reshaped European economies, lends support to the "democratic" theory. Neal's theory of financial development fits somewhere in the middle, as financial markets created great wealth for the richest merchants, but also democratized access to credit and investment vehicles.

Another set of questions which remain open to me relate to the lack of growth in Spain, France and Portugal. Why didn't the monarchs in these countries see what was happening in England and try to accomplish the same thing, while retaining some portion of their power? Were they unable to aggregate information and identify investment opportunities as well as the market, did they seize too large a portion of assets from successful investments, or was it something else i.e. what exactly about the lack of protection of property rights leads to slower growth? What role did the lack of social mobility play in these countries?

One thing that worries me about these conclusions: we are really looking at the English and the Dutch as the driving forces behind the relationships we are talking about. (I'm not sure if they have as many data points as they claim.) We might be suspicious that there is some unobserved factor driving the associations. I am also curious as to why the industrial revolution proceeded along different lines in England than in the Netherlands, given their similar institutions in this era, but that is a question for another set of readings.

(Sorry for the long post. Hopefully it makes up for the fact that I seem to have more questions than answers.)

I Romem

Upon reading hypothesis of Acemoglu et al I found it quite intuitive and convincing - perhaps too much so. The chain of links between location on the Atlantic and a greater divide between merchants and monarchy were surely not detrimental to the development of Capitalism in England and the Netherlands. The question at hand, however, is not whether this set of circumstances assisted the development of Capitalism, but whether it was pivotal for it.

The evidence brought forth by Acemoglu et al essentially consists of a correlation between distance from London or Amsterdam and an earlier rise in urbanization and GDP per capita levels. The latter two measures are nothing but proxies for economic development. Correlation, however, does not imply causality, and so the correlation observed in the paper is not sufficient evidence for any causal link between the set of circumstances in England and the Netherlands and the development of Capitalism.

There are two scenarios by which the two countries would have increased trade with the New World and would witness the development of Capitalism without having the former cause the latter:
1) The direction of causality could be reversed, i.e. the development of Capitalism in England and the Netherlands could have caused an increase in trade with the New World and promoted institutional restructuring.
2) There could be a third factor that caused both the increase in trade with the New World and the institutional changes in England and the Netherlands and also triggered the development of Capitalism (in as much as the latter can be distinguished from the former). For instance, it is possible that Marx' hypothesis of removing peasantry from the land caused a shift of labor to cities, which promoted the development of industry. This, in turn, could have increased demand (by factories, perhaps) and boosted trade with the New-World and also triggered institutional restructuring. But this particular example need not be the case – any plausible third factor would render Acemoglu et al's implication of causality false.

Consequently, the assertion that trade with the New World is the underlying cause of the development of Capitalism is not convincingly supported by the evidence brought forth in Acemoglu et al. Moreover, even the softer assertion - that trade with the New World merely contributed to Capitalist growth - cannot be supported by this evidence, for the same reasons. So, finally, even the softer assertion remains highly intuitive, but insufficiently founded.

The comments to this entry are closed.

Search Brad DeLong's Website