Warfare and Economic Growth in the Preindustrial World:
Some speculative thoughts
A consensus among most modern historians is that the frequency and prevalence of war was not an especially important cause of economic stagnation in the preindustrial world. This is evident in Peter Brown’s account of the decline of the Roman economy that I previously criticized. It is reflected in recent scholarly accounts of the Mongol invasions such as those which have influenced recent popular treatments such as The Silk Roads. I tend to think this is mistaken. But the reasons why it is mistaken are not obvious. Here are some admittedly speculative grounds for this skepticism.
The reasons why the economic costs of warfare are usually dismissed fall into a couple of categories:
1. Arguments like: the Romans destroyed Carthage in 146 BC yet by, say, 0 AD Carthage had recovered and was a major economic metropolis.
2. Arguments from analogy: Japan and Germany were devastated by WW2 yet they recovered rapidly and exceeded previous levels of living standards within a decade and a half.
3. Keynesian-style arguments: warfare was necessary to stimulate aggregate demand.
4. A binding technological ceiling on growth in preindustrial economies. Hence in the absence of warfare, growth was limited.
Argument (1) is widely used, but it exhibits Bastiat’s fallacy of the seen and unseen. The fact that Carthage eventually recovered from its destruction by the Romans does not shed light on how large or prosperous it might have become, had it not been destroyed.
Argument (2) is more credible. Everyone is aware of the miraculous recovery of the German and Japanese economies after WW2. The analogy is informative but anachronistic. The destruction of physical capital in modern economies is not necessarily harmful for long-run growth for several reasons: (i) in a world of rapid technological change, capital and infrastructure have to be overhauled and replaced every couple of decades. Therefore destruction in wartime accelerated process of capital replacement which would nevertheless have occurred anyway. This destruction was costly in the short-run but not in the long run.
(ii) In modern growth regimes, human capital matters as much or more than physical capital. Research by Fabian Waldinger shows that German science lost much more through the exodus of scientists than it did from the destruction of laboratories, universities, and physical infrastructure (here). In contrast, in the premodern world, human capital was not particularly relevant for growth and it was physical capital — irrigation, dams, olive and apple trees, cities, that mattered most. Finally (iii), and this is speculative, it is plausible that what matters is the ratio of the ability to destroy, relative to the ability to invest in capital. Total destruction, i.e. reducing the value of physical capital to zero, sets a lower bound. Thus total destruction might arguably be worse in an environment where technology is basic, people are close to subsistence, and the underlying rate of investment is low.
I won’t deal with argument (3) extensively here. I don’t think that the Keynesian model of a liquidity trap or post-Keynesian models of secular stagnation are applicable to the preindustrial world.
Argument (4) is perhaps the most compelling. Tony Wrigley argues in numerous books and papers that “organic economies” had a natural ceiling that limited growth (here, here, and here). This means that prior to the harnessing of fossil fuels, no economy, no matter how commercially developed, could experience sustained economic growth. In this case, there is an upper bound to how warfare could effect living standards.
One obvious drawback to this argument is that it could be that without the destruction associated with frequent warfare, the transition from an organic to a fossil-based economy might have come much sooner.
A second weakness is that evidence suggests that the constraints imposed by the limits of an organic economy were not binding. This is clear from estimates of agricultural productivity in medieval England and France. Average productivity was low. But some farms — those that employed best practice agricultural techniques and were located close to centers of demand such as London or Paris — could often achieve yields equivalent to those obtained by farmers in the 19th century (see the work of George Grantham here).
Now let us consider the reasons for thinking that war had a big impact on living standards in preindustrial economies. Economics distinguishes between short-run and long-run effects. In the short-run, war brings death and the destruction of physical capital. These are the impacts usually considered by historians and we can all acknowledge that they were negative. But in the long-run, frequent warfare also changes incentives and institutions. These factors are sometimes neglected.
Raiding, the destruction of agricultural land, and the disruption of trade networks were a crucial part of warfare from ancient times until the modern era. Jonathan Sumpton’s books on the Hundred Years War describe entire towns being “wiped out” as a matter of course in northern France and Flanders in the 1340s (i.e. Sumption 1999, p. 368). Nor was this just a feature of medieval warfare. The armies of Louis XIV devastated the Palatine on several occasions in the late 17th century. The Duke of Marlborough ravaged Bavaria, burning villages and crops, before the Battle of Blenheim in 1704.
Continuous back and forth raiding and warfare could turn prosperous agricultural lands barren. This occurred, for example, in Anatolia in the 8th-9th centuries. In the words of the Byzantine historian Mark Whittow:
“In the Roman and late Roman periods much of [inland Anatolia] seems to have been settled farming country. Not as rich as the coastlands, but still supporting potential taxpayers. . . .”
But by the 10th century, these villagers no longer engaged in arable farming and population density was much reduced:
. . . the villagers’ main wealth was in animals, and when contemporary Arab sources rejoice over a successful raid what they list apart from human prisoners, are thousands of cattle and sheep. Over the Near East as a whole, in modern times as much as medieval, the standard reaction of peasant farmers to chronic insecurity has been a drift into pastoralism” (Whittow, 1996, 180).
Continuous warfare reduced the incentive to become a farmer or a merchant. It increased the payoffs to becoming a solider or government official. These factors are all but impossible to quantify, but it doesn’t mean that it they were not important.
A recent literature in political economy and economic history emphasizes the positive aspects of war-making in early modern Europe (e.g. here and here). War made the state; and the state made war, in Charles Tilly’s memorable phrasing. This literature suggests that wars in Europe between 1500 and 1800 led to urbanization and the building of state capacity. For Besley and Perrson (2011) external wars constitute an important “common interest” over which alliances could form in favor of building more inclusive institutions. Both Ian Morris and Peter Turchin have recently written enjoyable popular accounts that set war as among the key drivers of human history. This literature is fascinating and important. But it should not obscure the fact that for most of history, warfare was both a blight for the inhabitants of large-scale agrarian societies and a major check on economic growth. I would like to suggest that this is a topic that demands further study.