Optimists about our current civilization tend to ignore the fact that history is full of past civilizations, all of which collapsed except the ones that happened to be around when modern civilization arose and managed to get themselves incorporated into it. For example, in checking the index of Ray Kurzweil’s The Singularity is Near, I find that he has a single one sentence mention (p346) of the issue. Similarly, Julian Simon’s Ultimate Resource II takes almost all its data series from the industrial era and simply ignores the collapse of past civilizations. Clearly, these authors are not thinking very deeply about the data points that might tend to contradict their optimism.
On the other hand, more pessimistic thinkers often subscribe to some flavor of environmental determinism and assume that our current civilization is governed by very similar laws to past civilizations, and therefore is at risk of collapsing along similar lines. For example, Tainter in his book Collapse of Complex Societies argues that all civilizational collapses share a common dynamic: as the society becomes more complex there is a declining marginal return to more complexity, and eventually the civilization becomes overstressed and must collapse. He believes the same law applies to modern civilization and shows a variety of data series arguing that modern civilization is experiencing declining marginal returns along certain dimensions. While he’s careful not to claim specific timing, he clearly thinks the same thing could happen to us that happened to past civilizations.
In this post, I want to explore the idea that there is at least one important class of threats where we might expect modern civilization to be much more resilient than past civilizations. Specifically, modern civilization operates at far higher levels of economic surplus than past civilizations, and this means that it is in a position to devote far higher levels of economic resources on solving certain kinds of problems.
It’s hard to be terribly accurate about this, of course, since modern civilization and pre-industrial civilizations are very different, and past civilizations tend to be poorly documented. Still, the overall pattern is clear. For example, in a post last month, The Net Energy of Pre-Industrial Agriculture, I pointed out that the fact that 75-97% of the population of medieval European countries lived in the country must mean that the overall energetic return of their agricultural sector, taken as an entire system, must have been quite small. By contrast, in modern western countries, only 2-3% of the population is involved in agriculture.
Of course, that comparison overstates the situation; in pre-industrial societies, agriculture was the source of most primary energy, whereas in modern civilization that is fossil fuels. So I present the following graph with rough estimates showing the fraction of global GDP being expended on primary energy. Here I have used methods similar to yesterdays post, in which I am taking global fossil fuel production figures, US price indices, and PPP global GDP estimates from the IMF to estimate the fraction of global world product being expended to obtain fossil fuels. The use of US price indices introduces some inaccuracy, particularly for natural gas and coal which are less globally integrated markets than oil. Further, I have used wellhead/mine-mouth prices – delivered to the customer, you would need to add a few percentage points.
Still, the overall point is clear – we expend less than 10% of our effort as a society in securing our primary energy source, whereas our distant ancestors needed to expend well more than half of theirs.
This means that modern societies can meet a clear existential challenge with very high levels of mobilization. One interesting illustration is response to war. For example, during the second world war, US GDP and the fraction expended on defense was as follows:
A high of 37% of GDP was diverted to defense purposes, taking about five years to increase to this level. In the UK, which faced a more profound existential threat, 55% of GDP was being mobilized by 1943, four years into the war. So these are indications that, faced with a sufficiently great threat, and given a few years to respond, modern societies can reorganize themselves to devote a large fraction of their economic/energy surplus to fending off the threat.
By contrast, consider again the example of medieval Europe – in the Hundred Years War, England and France were able to wield armies of 10,000-20,000 troops. The population of England at the time was a few million, and that of France 15-20 million. So these societies were only able to mobilize of order a few percent of their resources to fight a war with. Presumably, for the elites of the time, this was an existential threat and they were motivated to mobilize whatever resources they could, so the small totals largely reflect lack of ability rather than lack of desire.
It seems clear that a pre-industrial society that can only mobilize a few percent of its resources to meet a challenge has to be a lot less resilient than one that can potentially mobilize of order half of its (much greater) resources.
For example, consider the challenges due to natural disasters. A recent Proceedings of the National Academy of Sciences paper by Buckley et al, discusses the demise of Angkor (in modern Cambodia) in the fourteenth century (Angkor is now believed to have been the largest pre-industrial city with a population approaching one million). Buckley et al argue that drought was probably a serious stressor on the civilization towards the end, but also offer evidence that there were extreme flood years and that large canals, critical to the water distribution required for Khmer agriculture, filled with sand during a flood and were not cleared.
Assuming, just for the purposes of discussion, that Buckley et al are right about this, it seems to me that this is the kind of challenge where we might expect to see a material difference between modern and pre-industrial civilizations. It’s plausible that a society that can only mobilize a few percent of resources to meet a challenge, and has built a water distribution infrastructure over the course of centuries, might be unable to substantially rebuild it in the course of a few years in the face of extensive destruction.
By contrast, if, say, the California Aquaduct system were to somehow get filled with sand in a flood, there seems little doubt that the state, with Federal help if necessary, could wield enough bulldozers to get the thing working again in a short enough time to not destroy California as a society.
This is not to say the modern society has no vulnerabilities. In some ways the high level of resources presents threats of its own. For example, we could near-sterilize the planet with a nuclear war, which is not something our ancestors had to worry about. And there are probably what one might class as Death Star Vulnerabilities. The highly-optimized just-in-time, extremely-specialized nature of our society may create at least some possibilities in which the thing could unravel faster than its ability to repair itself could be mobilized. Many people feared this about Y2K; that particular challenge was either not that serious, or was adequately met in time, depending on point of view, but probably the underlying fear lurks beneath the surface: none of us understands our civilization terribly well, so it’s hard for us to assess what vulnerabilities it might have.
But it does seem to me, on the reasoning above, that the kinds of risks that modern civilization faces are likely to be fundamentally different than the risks faced by pre-industrial civilizations.