Life expectancy and the dawn of agriculture

Relative to their hunter-gatherer counterparts, early Neolithic farmers were short, had poor dental health due to malnutrition, bone lesions suggestive of disease and stunted spines from the back-breaking labour. This comparison underlies Jared Diamond’s claim that agriculture was the worst mistake in the history of the human race.

However, this decline in health was not permanent. For example, the health of Egyptians 12,000 years ago, shortly after the shift to farming from foraging, was poor. But by 4,000 years ago, height had returned to pre-agricultural levels and evidence of malnutrition in tooth enamel was lower than that of their hunter gatherer comparators. Agriculturalists adapted to their new diet and environment.

This story plays out in a 2007 working paper by Oded Galor and Omer Moav on how the shift to agriculture could have affected life expectancy, and more importantly, still affects life expectancy today. They make a fairly simple argument. When people shifted to agriculture and faced a greater threat of disease and other hazards associated with dense living, those who were predisposed to make larger investments in health (toward, say, higher investment in immune function or, as they model in their paper, resource transfers to children) would have an evolutionary advantage and increase in prevalence. When those threats to the environment later ease, as has now happened in most advanced economies, that higher investment in health translates into higher life expectancy.

The most interesting implication of this argument is the comparison across populations. Those populations that underwent a transition to agriculture earlier will have had more time to adapt to the new high-mortality environment and as a result, will make a larger investment in health. When those same populations transition to a modern low-mortality environment, they have a higher life expectancy than those without the history of agriculture.

Galor and Moav test this idea by examining how the length of time since the Neolithic Revolution is associated with life expectancy in modern populations. Using a range of controls, including latitude, land arability and levels of income in modern times, they found that an extra thousand years of time elapsed since the transition to agriculture results in an extra 1.5 to 2 years of life expectancy for that population. Using those numbers as a rough guide, the extra 3,000 or so years that European populations have experienced agriculture relative to African populations would result in an extra 5 years life expectancy for the Europeans even if gaps in socioeconomic circumstances were eliminated.

There are a couple of implications to this argument. The first is that if mortality risk starts to decline, we would expect investment to start to go the other way, such as focusing on quantity of children. It may be that the couple of hundred years since any population broke the Malthusian shackles is too short for evidence of that rebound. There might also be successive increases in the level of the shock - first with the transition to agriculture, later in the move to high-density disease-ridden cities, with increasing mortality risk playing out right up to the Industrial Revolution.

This paper gets a bit of space in Galor’s book Unified Growth Theory, where it ties in quite nicely. Galor advocates a unified growth theory approach as it captures the patterns in technology, income and population (including a demographic transition) within a single analytical framework. The basic mechanism in Galor’s unified growth models is that investments in education (quality) of offspring feeds technological progress, which in turns feeds the incentives to invest in education. This virtuous feedback loop drives the transition to a modern high-growth economy.

Now take the argument about increasing investment in health when a society has transitioned to agriculture. Those investments form part of the general investment in quality, helping to spur on the growth in human capital. A society with the early transition to agriculture will have higher investment in children, which will in turn drive an earlier transition to a modern growth state than would occur in a society with a short history of agriculture (as is reflected in which populations have made that transition).

Having said all the above, I’m not convinced by this particular argument linking longevity and agriculture (although I could be). Investments in health may trade-off with investments in other types of quality, such as education or brain energy consumption. The transition to agriculture also allowed for higher reproduction rates through the sedentary lifestyle, which is investment in quantity not quality. And lastly, does the difference in longevity simply reflect the maladaptation to modern diets among those with a short history of agriculture? As a test of this, what would be the respective lifespans in a hunter-gatherer context of two groups with short and long histories of agriculture?