World Economic History in One Picture?
I am blogging about a book project on Thomas Malthus’ “An Essay on the Principle of Population,” which was first published in 1798. You can find an overview of all my posts on this topic here that I will keep updated: “Synopsis: What’s Wrong with the Malthusian Argument?”
At first glance, it may seem as if this is old stuff that is only of historical interest. However, Malthus’ book has had a tremendous impact since it first appeared more than two centuries ago. I review this in further posts. My contention is that it has shaped our worldview and in this way our whole culture. And the Malthusian argument is by no means dead even now although I would argue it was clear from the start that it is worthless.
— — —
In 2007, the economic historian Gregory Clark published his book “A Farewell to Alms” with the prestigious Princeton University Press in a series curated by Joel Mokyr who is a renowned economic historian himself. Tyler Cowen supplied a blurb (lifted from an article in the New York Times) that the book would be “the next blockbuster in economics … We may not have cracked the mystery of human progress, but A Farewell to Alms brings us closer than before.” As always with Cowen, you have to read the fine print, though, which puts this in perspective: “From reading his book, I’ve upped my attachment to the Malthusian model from p = 0.2 to p = 0.37; that’s a big shift and it’s to Clark’s credit.” I am impressed it is 0.37 and not only 0.36.
Gregory Clark is not so naive to advocate the Malthusian argument in its original version. He has his own take and also his own terminology, eg. with his fancy usage of the phrase “the Malthusian trap.” Still, Clark makes it abundantly clear that he is indebted to Thomas Malthus. Quotes are liberally strewn over the text, and there are images that harken back to the time.
As I will develop in my analysis of his rhetorical strategy, Malthus uses a clever approach. Instead of proving the theory itself with stringent arguments, he conjures an associated worldview up. I understand a “worldview” here as an intuitive panorama of how the world works. If that is not clear enough, I would point you to my post “Worldviews, Narratives, and Ideologies.”
Malthus anchors his worldview right at the start and presents it as an “obvious truth” (cf. P.3). Gregory Clark owes a lot also in this regard to him. He rams a very similar worldview in already over the first sixteen pages. Then there are some short theoretical arguments that are presented as rather obvious, just like with Malthus, and then the book goes into a lengthy interpretation of events in light of the worldview as if it had been proven. This is then intermingled with lots of digressions that draw all kinds of conclusions which would follow if the argument worked.
Of course, Clark has much more firepower than Malthus had. So there are many interesting tidbits you can find in this book. There are also references and all the footnotes you want. However, the underlying argument remains weak and is based on certain fallacies that I will try to expose in further posts.
— — —
Clark’s worldview, as narrated in the first few pages, is about this: From 100,000 years ago until 1800, there was no progress in living conditions. There were perhaps some fluctuations, but essentially hunter-gatherers in the stone age were just as well off as the population in European countries in 1800. Actually, the claim is starker still: the latter were even worse off. Only afterwards something extraordinary happened and humankind escaped the “Malthusian trap” (not what everybody else calls a “Malthusian trap”).
Gregory Clark’s explanation for this sudden change is original. Although he does not call it that, he thinks that the conditions of a “Malthusian” world effectively operated like an eugenics program in England (!) that bred a population that was able and would then go on to start the Industrial Revolution. I will address this point in later posts, but let me just remark that if that were so, the same conditions should have yielded the result already long before if they had been in place for 100,000 years. The argument proves too much. It should predict the Industrial Revolution perhaps a few centuries after 100,000 BC, or, if you think that agriculture was a necessary precondition, something like 10,000 BC.
My point about a worldview that Gregory Clark conjures up early on is well-supported by a graph he presents already on page 2, which anchors it on an intuitive level. Here it is:
This is impressive. You can easily grasp the eternal stagnation of living standards before 1800. But here is the catch: What are the data behind this graph?
As the caption makes clear, this is “world economic history in one picture.” So “income per person” on the Y-axis and indexed to its level in 1800 should be that for the average person in the world. Now, this may not be obvious. But there are no such data series back to 1,000 BC! Gregory Clark’s arguments mostly rely on a data series just for England from the 13th century AD on. He had to reconstruct it with a painstaking analysis.
However, the graph is supposed to be for the world and back to 1,000 BC. To the best of my knowledge, there are some reconstructions that go back to 0 AD, but only with a much cruder scale. Angus Maddison has tried to figure out what population sizes and incomes per person were across the world in 0 AD, 1000 AD, 1500 AD (data here). That is already courageous. Population estimates for the world are rather reliable only from 1900 on. The farther you go back, the larger and larger the error bars have to become. For 1,000 BC, they are huge.
It is even worse for economic output. Seriously , what do we know about GDP for South America, Africa or Australia in 0 AD? Next to nothing. To a large extent, this has to be informed guesswork, and that is okay if you keep it in mind. But then the graph shows income per person. In other words, we would have to take quotients of these highly unreliable data to obtain the data series.
But if you look at Gregory Clark’s graph, he seems to have data on a rather fine time scale, not just for millennia or even centuries. Eyeballing it, they should at least be for a step size of five or ten years as with Clark’s data for England alone. So, to pose the question again in sharper form: What do we know about how GDP for South America, Africa or Australia changed from 1,000 BC to 990 BC and then to 980 BC and so on? Literally nothing.
Still, there is this graph, and there is only one explanation in my view: Gregory Clark has made the data up or he has just drawn the line without any at all. I can find no explanation in the text or in the index either what we see here. With a certain boldness, Gregory Clark introduces the graph in the very first sentence of his book with these words (cf. p.1):
The basic outline of world history is surprisingly simple. Indeed it can be summarized in one diagram: figure 1.1. Before 1800 income per person — the food, clothing, heat, light, and housing available per head — varied across societies and epochs. But there was no upward trend.
Sure, that’s what you see in the graph. And if you take it for granted that it is based on actual data, it shows the claim on an intuitive level. However, if Gregory Clark has only made the graph up, then it is not as convincing as you are led to believe. If you read closely, Clark keeps a door open for a disclaimer: “Dramatization, don’t try this at home.” But if you don’t pay attention and absorb the general claim about no progress over 100,000 years that is also hammered in in the introduction, you have your worldview ready that then needs little confirmation from the actual arguments. You can see that it was so, it is in the graph!
— — —
Not to be misunderstood: My point is not that there were regular dramatic improvements over the first 100,000 years until 1800. That cannot be so because even with a rather slow rate that would have led to an extraordinary factor for the total increase. Since the level was indeed low in 1800 compared to our times, a huge factor would mean that hunter-gatherers in the stone age must have lived on next to nothing. That cannot be the case.
But there is a subtle trick here that is typical for Malthusians: They treat it as if only regular growth were possible. My own explanation is also that there must have been very long periods of stagnation, but not because of some “Malthusian” mechanism. The reason was that there was no technological progress. That can happen, and then a human population is just stuck at some level and you cannot produce more out of thin air.
However, that does not preclude that there were, for example, a few shifts upwards in the past where living conditions improved materially although still not as massively as over the past two hundred years. That is totally possible even if Clark’s graph were correct. The advent of agriculture, perhaps 10,000 years ago, might have led to such an improvement, and this would not be visible in the graph, even less so with a scale for the Y-axis that accomodates the huge increase at the end.
Clark argues against this and even claims that the move to agriculture was a deterioration. What puzzles him is why hunter-gatherers were so stupid to not stick with their previous lifestyle. I will address this point in further posts. But I can sketch the thrust of my argument already now:
It is entirely possible that people have the same income on average after adopting agriculture, or even less, and are still better off. If you have to suffer from serious starvation every five years that is clearly worse than if you have to suffer from it only every fifty years. You might even exchange this for somewhat lower income because it is a huge improvement.
The average might remain indeed stagnant, and even fall somewhat, but you should reasonably go for agriculture in this case. What’s puzzling here is only how Malthusians are puzzled by it. I would take the deal anytime. And the move to agriculture, often associated with trade, plausibly lowered the volatility of the food supply versus that for hunting and gathering. Just think about what you would do if there is a harsh winter and there is no gathering and you cannot hunt either. With agriculture, you can store food over time or trade it for something else without having to produce it yourself. Hunter-gatherers could mostly not do this. Simple.
— — —
In addition, average income per person is perhaps not the right thing to look at anyway. It is easy to confound it with the average supply of food, but that is not true. Malthus makes this mistake all the time, and Clark seems to be in the same league. How much food we need per person, though, is basically a constant determined by our human nature if you correct for relevant features like height, weight, and the activity level. We need just as much as the people in 1800 or the hunter-gatherers 100,000 years ago. This cannot change at all, like it is silly to assume that technological progress should lead to an upward trend for body temperatures, or else there was no improvement.
If a population survives over the longer run, this proves that they got enough on average. Otherwise they would not be there. The big difference is not that we eat more food now, but that we have perfect food security. We never have to starve involuntarily. That was not as perfect in 1800, but already quite good, and it was much worse 100,000 years ago.
If you think about it, this is a huge improvement that can hardly be overestimated. And the situation in 1800 proves that there had been already a huge improvement over the previous 100,000 years. Large scale famines had become very rare by that time and that had been so already for centuries. Malthus himself struggles with the contradiction that this poses for his theory and hedges his claim about inevitable famine with the words (cf. VII.19):
In every State in Europe, since we have first had accounts of it, millions and millions of human existences have been repressed from this simple cause [ie. general distress]; though perhaps in some of these States, an absolute famine has never been known.
One of Malthus’ pet peeves is actually that because of the “Poor Laws” people suffer too little from starvation, and that could cause problems from population growth in the future. If that sounds perverse, that’s because it is.
— —
Nitpicking about this one graph is certainly not a persuasive argument in and of itself. Yet, it makes a certain technique transparent, and alerts you to the pitfalls of Clark’s reasoning. In this case, it is circular: He already thinks he “knows” what has happened, then draws a graph that illustrates it, and you as a reader are led to believe that the graph supplies the proof.
My critique of Gregory Clark’s “A Farewell to Alms” is only a sideshow to my general critique of the Malthusian argument. But since his book has had quite an impact from what I gather and has again reinforced the Malthusian worldview, it is perhaps appropriate to take it to task as well. This also shows that you cannot ignore the Malthusian argument because it is old. It is still alive and kicking.
