Spending a Windfall


The paper “Spending a Windfall: American Precious Metals and Euro-Asian Trade 1531-1810”, by André C. Silva and myself, is now available as a GGDC working paper.

Here’s the abstract:

During the early modern period, Asia ran a large current account surplus with Europe. We show that the critical factor to stimulate Euro-Asian trade was not the new trading routes to Asia, but the European access to American precious metals. We use a dynamic general equilibrium model to reproduce historical data and calculate alternative scenarios. We find that European imports of Asian goods were up to thirteen times higher than they would have been without new routes and without precious metals. The effect of American precious metals is six times larger than that of the discoveries of new trading routes.

90% of the European imports of porcelain, silk, tea, and other luxuries from Asia during the early modern period were paid for in silver. The traditional and still currently accepted explanation by most historians for this imbalance of early modern Euro-Asian trade relies on cultural factors. It is said that Asians have always had a penchant for hoarding treasure (this is the view of Hamilton, Keynes, Kindleberger, and Maddison). By contrast, our explanation for the patterns of Euro-Asian trade in the early modern period does not rely on different preferences for Asians. Instead, the observed trade patterns emerge as a natural consequence of rational agents taking decisions in a dynamic general equilibrium context.

Our model of international trade has two “countries”, two goods and money. The demand for money is obtained with a money-in-utility specification. The two agents, denoted Europe and Asia, have identical preferences. Each agent produces a domestic good. All of the silver windfall is given to the European agent. There are also transaction costs, which apply to international transactions only.

We feed the historical data to the model. The windfall is given to the European agent. From 1500 to 1800, 85% of the world’s silver and over 70% of world’s gold came from America. The following figure (taken from Figure 4 in the paper) shows, in blue, the annual amount of “discoveries” of precious metals in America during 1531-1810, expressed in grams of silver per capita. In red, in shows our counterfactual: a world where new routes to Asia were found, but no precious metals existed in America.


The following figures (also taken from Figure 4 in the paper) then decouple the results of the historical and counterfactual scenarios for consumption and trade between Europe and Asia. In this one we can see that thanks to the new routes and the precious metals (mainly silver), European consumption of Asian goods was about 13 times what it would have been by the second quarter of the seventeenth century. However, most of that effect was due to the availability of American precious metals and not to the new trade routes themselves (note that two new routes to Asia were found, one through the Cape of Good Hope and the other through the Pacific). The new routes (and its associated technical and organizational change) would have at most doubled the previous trade levels.


Not only did the new routes have a much smaller economic impact on trade than did American precious metals, but the sign of their effect on trade patters was also different. This is clear if we look at what happens to net exports of European goods to Asia: in the counterfactual world of no precious metals existing, they increase. Historically,  we know that this wasn’t the case.



As the figures suggest, our simulations show that the combined effect of the discovery of precious metals in America plus new trade routes to Asia led to Euro-Asian trade volumes that were up to 13 times what they would have otherwise been. After peaking in the second quarter of the seventeenth century the effect diminishes over time as the early modern period advances but is still over 10 times as late as 1750.

In the paper, we also consider the effect of rising incomes in Europe, and show that its effect was marginal relative to what can be attributed to the windfall of precious metals.

Early modern Euro-Asian trade started off a historical dynamic process with broad consequences for Europe. While foreign trade accounted for a small percentage of European GDP at this time, it provided dynamic expansion opportunities. Early modern trade with Asia led to the emergence of mercantile companies such as the Dutch VOC and the English East India Company, which in many ways were the prototype for modern multinationals.71 It permitted the development of modern financial markets in Amsterdam and London. It induced an industrious revolution which encouraged additional labor input and market participation in Europe, necessary preconditions for the process of modern economic growth and for the industrial revolution itself. It stimulated economic growth and urbanization, through a process of spillovers and agglomeration economies. Trading with Asia (and America), may have also induced a shift in the wealth and political power from the land-owning elite to the hands of a merchant, entrepreneurial class. Positive spillovers also resulted from the increased inter-continental exchange of ideas. Finally, international trade and war always came together, and external warfare was one of the most powerful drivers behind European state-building.



GDP reconstructions for the early modern period

Angus Maddison is justly famous for historical national account reconstructions.

He did much of his work at the University of Groningen, where I currently teach. Work in his tradition continues here, where we have the Groningen Growth and Development Centre, and host the (soon-to- be-updated) Maddison and Clio infra projects.

Despite the importance that Maddison’s GDP and population figures had in stimulating our thinking about economic history and development, it is fair to say that his pre-1820 figures were less than solid. (In one review published in the Journal of Economic History, Gregory Clark said of Maddison’s tendency to “guesstimate” that his figures were as real as medieval relics.)

One reason why most of Maddison’s pre-1820 numbers were indeed questionable is that, while he often relied on the work of others when he could, there just wasn’t much quantitative work available for the period prior to 1820.


But in the last 10 years, economic historians have done a lot work using archival data specific to each country, in ways that are reasonably comparable across space and time (no easy feat). Even taking into account the tremendous uncertainty that will always exist about national account reconstructions for the past (especially about levels, as I explain here), I think it is fair to say we have now made much progress.

Steve Broadberry has been doing an excellent job summarizing what we’ve been learning (and how it was done); see for instance this Journal of Economic Perspectives paper, co-authored with Fouquet. (Ungated version here.)

The following table shows per capita GDP in “international” GK 1990 dollars. (For the sources, see p. 15 of this paper.) To have a rough idea of what this means, keep in mind the World Bank’s definition of poverty as having less than 1 dollar a day, hence about 350 dollars per year. This is a rather arbitrary definition, for sure. It can also be difficult to compare societies which consume very different kinds of goods (see Angus Deaton’s work on PPPs), not to mention that it’s difficult to control for quality improvements. Nevertheless, these problems are mitigated for premodern economies, which consumed reasonably similar baskets, with the exception of different foodstuffs (for instance, for fat, olive oil in South Europe and butter in North Europe).


What is clear from the table is that these societies lived well above physical subsistence. And all of them grew in the 1550-1750 period, though some more than others, and though later there were some reversals. (Note that economic historians conventionally refer to countries using their modern borders which refer to a physical territory even for periods when such “countries” did not yet exist, as was the case here for Germany.)

Maddison claimed that Europe experienced significant levels of real income growth during the early modern (i.e. 1500-1800) period. By contrast, Gregory Clark argues that this was not the case. Getting the timing and magnitude of growth right can help us falsify some theories and clarify others. It may also help new explanations surface. Determining who is right in these matters has important implications for our understanding of the origins of the industrial revolution, for instance.

Though medieval income levels were generally higher than Maddison believed (and hence early modern growth slower, given what we know about nineteenth century levels), the conclusions have overall been closer to the position of Maddison than to that of Clark. Interestingly, the opposite is true if we look at the long term evolution of real wages instead of GDP growth; Clark usually focuses on real wages, though he also has a paper on English GDP; but as it is calculated from the demand side, it rests heavily on those.

Also notice that by 1500 population levels had not yet fully recovered from the Black Death, so when we look at early modern growth we should really be looking at what happens after the 1550s. The most anti-Malthusian implication of the data is the fact that both real income per capita and population levels (the latter not shown above) grew at the same time, for most of these countries, during much of the early modern period. Hence for several of these economies the early modern period was one of both intensive and extensive growth.

Unconventional monetary policy in the past: Lessons for today

I start by sharing a recent vox.eu article written by Patrick O’Brien and myself which relies on our recent work about the Bank of England during the Restriction Period, 1797-1821 (during which gold convertibility of banknotes was suspended).

We are currently revising the paper in preparation to be submitted to a journal and in due time a new working paper will come out, and at that point I’ll post some new thoughts.