Sure, I Can Make a Bet

Bryan Caplan has four main rejoinders to my argument that there may soon be once again a significant negative population externality:

(1) If I am confident there will soon be a negative population externality would I like to bet on when income per capita will start declining? 2020, 2030, 2040?

(2) Land is still superabundant in the United States; most is not used for farming.

(3) Suppose there were a negative population externality. Then, adopting an argument of Steven Landsburg, it would occur only within the family. The other children will be made poorer.

But the future income of the children other families will not be affected. So parents alone can decide on their optimal family size taking this into account. No one else will be affected by their decision.

(4) I have neglected the positive externality from more people through the faster production of knowledge. Growing world populations are a key factor in explaining the final appearance in the world two hundred years ago of consistent technological advance.

My responses are as follows:

(1) Income per capita will not fall in the future, even with continued population growth. Here Caplan has misinterpreted the claim I made. A negative externality associated with population is quite compatible with the continued rise in incomes. I did not mean to imply that the rise in incomes since 1800 need end if population keeps growing.

Suppose technology continues to improve by about 1% per year, as it has done for the past 200 years. That will lead to a growth rate of incomes per capita, absent population externalities, of 1.3% per year because of induced capital accumulation.

The population drag on income growth will be the population growth rate multiplied by the share of resource rents in income, times 1.3% for induced capital decumulation. Currently with natural resource rents a very small share of world income, and population growth rates of 1% of less per year, that drag is trivially small. It could easily be completely countered by economies of scale from larger populations (including economies in knowledge generation).

However if the resource rent share rose to be the 20–25% that was typical of the world before 1800, and population was to grow at 1% per year, then the population drag on income growth would rise to 0.3% per year. Incomes would rise significantly more slowly.

But for the population drag to lead to income declines would require either an impossibly fast population growth rate of more than 3%, or a rise in the resource share in incomes to close to 100%.

So the issue here is not whether incomes will in the future decline because of population growth. That will not happen at current rates of technological advance. It is whether population growth comes to be a significant drag on future income growth.

(2) Farmland is still very cheap in the United States. A recent investor newsletter talked breathlessly of land in Iowa selling for as much as $8,200 per acre. This implies a rental value of only about $240 per acre. That is not a lot of Caramel Frappucinos for a year’s use of the rich soils of Iowa.

The recent rise in farmland prices may not, however, be as transient as most past increases. Perhaps it is the beginning of a new era of land scarcity.

So if Bryan wants to bet even odds that farmland prices will be higher relative to average wages in 30 years time, I am happy to accommodate him. After all with the impending rise in world population to 9 billion people, surely given his views on population and innovation, one of these extra people will end forever the issue of land scarcity!

(3) I do not have a copy of Landsburg’s book with me, but I presume his idea is that the reduced land and resources per child that population growth implies will be felt through the amount of land and resources that each child inherits from their own parents.

The flaw in this argument is that the population externality, if it exists, is a reduction in wages for other members of a cohort which is not fully compensated by a rise in resource rents. That reduction falls almost entirely outside the family. Caplan’s kids are potentially reducing my kids’ earnings.

Since a half of families own no assets in most societies, population externalities would be sure to hurt them. So if there is a significant population negative externality there is no easy fix for your libertarian conscience through the Landsburg defense.

(4) On the claim that more knowledge growth results from having more people, I was including the effect of more people on idea production under the general category of economies of scale, since it is just another aspect of economies of scale. I was just skeptical about how significant these effects are.

Caplan has no such skepticism. “If the Black Death had never happened, there is every reason to think that the modern world would have arrived centuries earlier.”

The theoretical case for some positive effects of population on idea production seems compelling. The strength of this effect, and whether it is great enough to counteract the costs of population growth, is another matter.

It could be that there are only a fixed set of possible technological developments available to society at any time, given the existing knowledge. With this set of ideas, once we attain a certain minimum population size, all the available discoveries will be harvested. Doubling the population will just result in duplicated efforts.

Take a concrete example, Hollywood movies. Suppose we were to double the number of scriptwriters and production crews in LA. Would that double the number of innovations in movie stories and types? Or would it instead result just in Die Hard 6, Die Hard 7, and Die Hard 8 all coming out in one year?

One thing that may matter is not just the number of people, but also their living standards. In that case population growth that simultaneously has negative effects on the rate of growth of income might well also slow the rate of innovation.

Here there are too many theoretical possibilities to make theory much use. It is empirical evidence that matters.

But the empirical evidence population on idea production is weak, especially once we move to population sizes beyond those of remote island communities. If population externalities were such a sure cure for stagnation, why no Industrial Revolution in China by 1 AD, or 1000 AD?

Also from this issue

Lead Essay

  • Bryan Caplan argues that declining world populations aren’t such a good thing after all. While we may have dodged the “population bomb” predicted in the 1970s, the world still benefits from more suppliers — and demanders — of new goods and services. Friends of scientific and cultural advancement should want more people around. All of which leads him to a question: Are there ways to incentivize population growth without sacrificing individual liberty? Caplan suggests several methods, including open immigration, tax incentives for children — and the direct, ultimately private argument that having more kids is more fun than most people realize.

Response Essays

  • Gregory Clark argues that, while Caplan may more or less accurately describe the history of economic development, mineral and other reserves in the West have been significantly depleted. What holds true today will not hold true indefinitely regarding food, energy, and basic mineral commodities. The balance between population growth and resources only recently tipped toward abundance, and it will likely tilt in the other direction soon.

    Clark agrees, however, with Caplan’s suggestion that on the margin, middle-class families stand little to lose and much to gain by having more children; cultural assumptions in the West likely overestimate the importance of nurture, and as a result, middle-class Americans probably do overinvest in their children.

  • Matthew Connelly reviews the history of pro- and anti-natalist policies around the world. He finds them both full of coercion, perverse incentives, misogyny, and — if states try hard enough — atrocity. Population trends, however, are very stubborn, and the payoff to such policies is low. In his words, “economists do not know where babies come from.” Not, anyway, such that they can motivate people without destroying their autonomy. If libertarians are recruited to the pro-natalist side, as Caplan urges, their enlisting could come at the price of individual liberty itself.

  • Betsey Stevenson agrees that parents are often generally less happy than otherwise similarly situated nonparents. But she casts doubt on the theory that excessive parental worry is the source of the unhappiness. The most worrying parents, those who spend the most time on childcare, are also the ones who report relatively higher levels of happiness. There is, moreover, more to life than self-reported happiness, and parents may very well have other desires and values than just this one. Neither self-reported happiness nor economic utility are all that there is to our various ideas of the good life. We still need to learn more about why parents describe themselves as less happy before we can prescribe more children and less worrying as the remedy.