What Does a Declining Population Mean for Economic Growth?

[johnwalker]

The End of Economic Growth? Unintended Consequences of a Declining Population

Professor Charles I. (Chad) Jones of the Stanford Business School, developer of the Jones model for economic growth, has been investigating the consequences of a shrinking population, which now seems likely in many developed countries where the fertility rate has fallen (often dramatically) below the replacement rate. His 2022 paper in American Economic Review, “The End of Economic Growth? Unintended Consequences of a Declining Population” [PDF] presents a model in which the interrelationships of population, innovation, and economic growth are explored. Here is the abstract.

In many models, economic growth is driven by people discovering new ideas. These models typically assume either a constant or growing population. However, in high income countries today, fertility is already below its replacement rate: women are having fewer than two children on average. It is a distinct possibility that global population will decline rather than stabilize in the long run. In standard models, this has profound implications: rather than continued exponential growth, living standards stagnate for a population that vanishes. Moreover, even the optimal allocation can get trapped in this outcome if there are delays in implementing optimal policy.

The video discusses the analysis in the paper and conclusions in a concise 20 minute presentation. If you’d like to be able to flip back and forth through the slides while watching the talk, here they are as a PDF.

This is a highly abstract and mathematical model. If you’re allergic to differential equations, this is probably not the paper for you. What is interesting for policy-oriented readers is that while the fertility rate is slightly above or below the replacement rate does not make a dramatic difference as perceived by families or a country in the short term,

The macroeconomics of the problem, however, make this distinction one of critical importance: it is the difference between an Expanding Cosmos of exponential growth in both population and living standards and an Empty Planet, in which incomes stagnate and the population vanishes.

further:

When the equilibrium fertility rate is negative, the optimal allocation typically features two stable steady states. If the economy adopts the optimal allocation soon enough, it converges to the Expanding Cosmos. But if the economy waits too long to switch, even the optimal allocation can be trapped by the Empty Planet outcome.

This means there is a point of no return after which changing policies has no effect on the outcome. The model shows that once population shrinkage begins, the flow of new ideas drops to zero, so sanguine predictions that “we’ll cope with that when we get there” or utopian fantasies of “degrowth” may not have worked out the consequences of an end to innovation on—everything.

Robin Hanson of George Mason University, who recommended this paper, discusses the consequences of a shrinking population and economy in a 2023-08-21 Substack post, “Shrinking Economies Don’t Innovate”.

This view has a dramatic implication for a world of low fertility, and thus falling population. If the world economy reaches a peak and then falls to a level that is only X% of that peak, but stays within the same growth mode, then at that point the rate of innovation would be no more than X% of the prior peak rate. Fewer than X% of N typical innovations would be found per year.

That is, in a shrinking economy, innovation grinds to a halt! And if after a shrinking economy, economic growth should restart, then a restart of innovation would be delayed until the new economy rose to pre-population-fall levels. If a culture of innovation had withered in the meantime, such a restart might take even longer.

⋮

Thus once we have entered into a shrinking world economy, we cannot put much hope on tech or innovation solutions. If such solutions are not found before the economic decline, they will probably never be found.

This, then, motivated him to ask, since we now appear to be in an age of below-replacement fertility, implying a shrinking population in the future, what are the possible scenarios for the future? Either the human population will continue to decline towards extinction or, at some point, fertility will rise and stabilise the population or resume growth. In a Substack post on 2023-08-24, “16 Fertility Scenarios”, he enumerated sixteen possible futures he could envision and presented the results of a battery of 𝕏 polls he posted, asking respondents to rank scenarios based upon their likelihood, desirability, and story potential of that outcome. Analysis of the results was complicated by the limit on four items in an 𝕏 poll, but the 2732 responses to 24 polls were as summarised below. See the paper cited above for detailed descriptions of each scenario.

---

Now, as I noted above, these are studies in macroeconomics, a subject which is particularly prone to “assume a cow is a sphere”: radical simplification of matters which involve a bewildering amount of detail and complicated and difficult to untangle interactions in the real world. For example, production of ideas is assumed to be proportional to population: the more people, the more ideas. Then, income per person is proportional to the accumulated production of ideas. But, one wonders, what people, and what ideas lie behind the letters in these equations? Certainly that matters, doesn’t it? Can a technological society confronted with a shrinking population solve its problems by importing hordes of low-IQ savages from those countries seemingly trapped in an eternal state of "developing”? Are new ideas for energy sources, manufacturing processes, materials for fabrication, and ways to organise complex endeavours the peer of innovations in deconstruction of literary and artistic works, eliminating embarrassing and distasteful episodes from recorded history, revising language to deny objective reality, or “remxing” the creative output of earlier ages into derivative works?

I present these analyses because they are interesting and pose challenges to commonly-held views of the future and rarely-questioned assumptions about the consequences of trends which now appear to firmly established. These are not my analyses, and I am not here to defend them, but purely to commend them to your examination and independent consideration.

[johnwalker]

One of the 16 scenarios discussed in Robin Hanson’s article, “16 Fertility Scenarios”, mentioned in the main post is:

10. Insular Subculture : Initially small subculture(s) arise that highly value fertility, reliably keep members from leaving, and care little about what other cultures say of them. They probably achieve this by disrespecting outsiders, and defying many cherished outsider values. This tempts mainstream cultures to repress them, but such repression needs to mostly fail. As it wins, this subculture likely discards many elements of tech and culture that we now treasure; it can’t be penalized for being anti-innovation until that restarts. Maybe Africa creates insular culture to keep its fertility high.

This cites an article from 2021-09-11, “The Insular Fertile Future”, which notes:

Today the cultures associated with higher fertility tend to be more “traditional”, and less integrated with the dominant world elite culture. And a few small subcultures, like Mennonites and Amish, or Mormons and Orthodox Jews, even manage to maintain high fertility while staying closely connected to the dominant culture. However, as a big fraction of the youth of such subcultures leave them, it isn’t obvious that these subcultures can long sustain net growth.

But this does point to the plausibly winning strategy: subcultures that are both highly fertile and highly insular, keeping enough youth from wanting to defect from their subculture to join the dominant low fertility culture. Through some combination of genes, culture, and tech, they find a way isolate their members more from outside cultural influence, and thereby to support sustained population growth (or at least less rapid decline).

Here’s an interesting observation in that regard.

[Gavin]

Models can only tell us the consequences of the modeler’s assumptions. The model does not tell us if the assumptions are correct. See Catastrophic Anthropogenic Global Warming.

Might history have some useful lessons? For example, there have been suggestions that the great population reduction following the Black Death actually stimulated positive advances in Western Europe.

[magus]

@johnwalker

But, one wonders, what people, and what ideas lie behind the letters in these equations? Certainly that matters, doesn’t it?

I think this is an important observation.

Ideas, people, and capital are not interchangeable, but macroeconomic models treat them as homogenous aggregates.

Some degree of abstraction is necessary to create a model, but there is a real difference between the discovery of penicillin and the conception of a new type of rain dance. Lumping these ideas into the same variable A would not seem appropriate.

That said, the general schema of more smart people → more innovation → better economic outcomes, is probably sound. Institutions also matter, of course. Smart people born under authoritarian regimes might not have the opportunity to develop their ideas, as despots are afraid of innovations that would threaten their control.

[johnwalker]

Robin Hanson has posted a follow-up to his 2023-08-21 article, “Shrinking Economies Don’t Innovate”, exploring the question “How Much More Innovation Before Pause?”.

One simple first-cut way to estimate this is to simplify our future growth path, and assume that our economy keeps growing exponentially until it suddenly switches to falling exponentially forever. (Compared to a more realistic transitional economic deceleration, this simple model should over-estimate total economic product, and thus innovation, across this period.)

We can then ask: how long would the economy have to have instead kept growing at prior rates to create the same total product as we would have been seen by infinity in the declining economy scenario? Add that duration to the future date at which growth switches to get our key estimate: expect (a bit less than) as much innovation as you would have seen by that future date in a simple keeps-growing-at-same-prior-rate scenario.

⋮

But though I look forward to seeing more careful analyses, these effects just can’t change the overall answer by that much. Expect no more than a century’s worth of further innovation, before the great innovation pause . Maybe only a half century . While I know many are excited by recent AI progress, I think it will actually be quite a challenge to achieve human level AI by this deadline. And of course adopting policies to slow down AI innovation on purpose will only make it harder to make this deadline.

Same goes if you were hoping for life or fertility extension, or artificial wombs. You’ve got less than a century to not only work out the basic concepts, but also all the messy details needed to translate abstract ideas in concrete practice, and then to lower costs through scaling up practice. And to get past regulatory obstacles, which many seem eager to add to your troubles.