Lessons from the latest UN's population projections
On Monday, the United Nations released its world population projections, with some decently eye-popping results. Some highlights that jump to the front:
Global population will breach the 8 billion barrier later this year, and peak around 10.4 billion people in 2080
As soon as 2023, India will surpass China as the world’s most populous country and China will start experiencing absolute population decline
Many countries, especially those in eastern Europe and east Asia, will experience shrinking populations, while most global population growth will come from south Asia and sub-saharan Africa
This data release, coincidentally arriving after weeks of some questionable Elon Musk tweets on fertility rates, provides an occasion to revisit some of the research and implications of these demographic trends.
Major red flags for China
As flagged above, China’s population will start declining starting in 2023. As usual, Our World in Data provides an excellent data visualization of the UN’s projections of national population.
Per Our World in Data, the UN projects the over-65 population in China to top the under-25 population by 2040.
Indeed, the median age in China just surpassed that of the United States, though it still lags well behind much of Europe and countries like Japan.
As we’ll discuss in a second, an aging population can be rather devastating for overall growth perspectives. But it’s worth flagging that there are some measures that can be taken in the meantime. For example, China has a relatively low retirement age (60 for men, 50 for women, 55 for female civil servants). As a result, they start to experience an adverse “dependency ratio” (high ratio of retirees to workers) far earlier than other countries. As National University Singapore’s Bert Hofman (h/t Adam Tooze) estimates, raising the retirement age to 65 would largely forestall the labor supply issue for the next twenty years. But as this report from the Lowy Institute illustrates, since people naturally decrease their work in advance of any official retirement age, these reforms could at best freeze the demographic structure at its current level, and would hardly reverse it.
A shrinking population can harm growth
Why are these population projections (mildly) concerning? After all, wouldn’t a stabilizing population mean less stress on environmental resources? We’ll get into the environmental case in a later section, but the gist is: no. Slowing population generally harms not just aggregate growth, but living standards per person as well through a few key mechanisms.
First, a slowing population means that there are fewer working-age people producing economic output to support retirees. In 1960, there were over 6.5 working age adults per 65+ people in the US, and a 15:1 ratio in China. Today, that has fallen to roughly 4:1 in the US and 6:1 in China. Though China lacks the robust Social Security system that the US has, the transfers needed to support a large non-working population are simply higher when the population skews older, whether that’s via taxes or via voluntary payments or savings.
Second, and perhaps more significantly, there is a large effect on innovation. As Stanford economist Charles Jones writes, the effect of a growing population is implicit in nearly every macro growth model out there. The gist is pretty simple: more people–particularly more working-age people–means more opportunities for those people to have ideas and those ideas build off each other to produce more ideas and translate into economic growth. There is some decently strong academic evidence that ideas are getting harder to find (as low-hanging fruit is plucked), so a growing population with more and more researchers is necessary to produce the kind of increases in living standards that existed throughout the 20th century.
If you’re interested more in that chart and reasons why “ideas might be getting harder to find”, there’s an excellent primer from Matt Clancy here. The tl;dr (though you really should read if you’re interested) is that there’s some suggestive evidence of a “burden of knowledge”: as the state-of-the-art progresses, it takes longer to know enough to advance the frontier of knowledge.
Another paper addressing the issue (available here) from Hugo Hopenhayn, Julian Neira and Rish Singhania discuss the implications of a demographic slowdown on firm entry and exit. In their model, existing firms can absorb new people entering the workforce. But if population growth rises, these legacy firms cannot easily do so, and the remainder of the workforce is absorbed by new firm entrants. As a result, a declining population increases the share of the economy controlled by older institutions. And since these older institutions generally have a lower labor share than younger ones (in part since upfront capital costs are easier to absorb at scale and over time), changes in demographics plays a large explanatory role in the decline in the national labor share (and rise in inequality) over the last few decades.
This brings us to the “trillion dollar bill on the sidewalk”: immigration. A declining population because people are better able to choose their family size is obviously an unmitigated good. As a result, to help improve living standards overall without impinging on those far more important rights, it becomes even more important to accept more people entering the country.
Anti-growth politics are completely backwards
The final point I want to address is the claims of the anti-population growth crowd, who are encouraged by the declining population. Their claims generally fall in three main camps:
A rising population contributes to increasing resource usage, as represented by water shortages, climate change, etc.
Increased population increases crime, disorder, violence and chaos
The fiscal strain of increased population would be devastating
All three claims are not just false, but morally bankrupt.
Let’s first address the eco-pessimistic resource usage claim. The fundamental misunderstanding is that the relationship between population and resource intensive is far from linear.
Take plastic pollution, for instance. According to Our World in Data, despite having roughly 1/3rd the population of the US, the Philippines emitted over 150x more plastic waste into the ocean. They even emit more than twice as much as India, despite having a population roughly 1/13th the size. The simple truth is that effectiveness of trash management systems are far more determinative of the “sustainability” of resource usage than simple metrics like population.
Plastic is far from a cherry-picked stat. India (and China, and most of Europe) have actually gained total forest area since 2000, despite an increase in population of roughly 40%. Total water use in India, China and the US have all been flat in the last decade-plus, despite that growth. Even total land devoted to agriculture has fallen since its peak in 2001.
The basic reality is that the relationship between population growth and resource usage is more complicated than “more people = more resource usage”. If anything, rising living standards can finance more expensive but more environmentally sustainable systems, as represented by the famous “environmental Kuznets curve”. Under those conditions, rising living standards through population and economic growth is not just compatible with environmental protection, but actively supports it.
Air pollution is a clear counterexample: China and India have definitely increased their fossil fuel consumption to provide power to their growing populations. But the fact that (still-growing) countries like the US can see their total pollution decline from 2006 to present suggests that newer power-generating technologies can indeed break the population-resource usage relationship, just perhaps at a slightly higher level of average income. In short, the solution to environmental degradation will come from better technology and better policy, not from enforcing some demographics.
The same story holds true with the other claims. Overall institutions and policies are far stronger determinants of crime, disorder or fiscal status than anything as simple as population growth. The US’s homicide rate is 10x China’s, despite the obvious population gap. Heck, even within the US, the relationship between population density and murder rates are very weak (the highest murder rate in the US is in the hardly hyper-urbanized states of Mississippi and Louisiana).
Key takeaways
I guess the takeaway is pretty simple:
Credible projections show population growth slowing rapidly in much of the world, with global population declining by the end of the century
Declining population as a function of people being able to better choose their family size is obviously an unmitigated good.
At a society-wide level, a slowing or shrinking population has negative implications for economic growth, without many of the silver linings others have suggested
Countries that can should take the “trillion dollar bill on the sidewalk” and bolster their population by substantially expanding immigration
The UN report is a nice reminder that population growth plays an important role in shaping societal trends. But a cursory read of the data suggests that the most simplistic interpretations (“overpopulation!”) are woefully inaccurate. There are plenty of tools governments have (like increasing immigration) to manage slowing population growth while supporting rising living standards.