Bryan Caplan  

Demographic Disaster? What's Wrong With Jonathan Last's What To Expect When No One's Expecting

Prince George County Governmen... Open Borders and Personality B...
Jonathan Last's new What To Expect When No One's Expecting is the best-written, most engaging, and funniest book on the social cost of low birth rates and population decline.  While he focuses on fertility, he breaks with typical conservatives by hailing the massive benefits of immigration too.  Highly recommended. 

Despite my sympathy for Last's pro-human position, though, there is a big gap between his rhetoric and the facts.  Indeed, there is a big gap between his rhetoric and his facts.

The subtitle of Last's book is America's Coming Demographic Disaster.  He tries to tone this down in the intro:
Unlike Ehrlich, I'm not selling doom. (Unless that's good for book-buying.  In which case, we're doomed.)
Yet the text is largely consistent with the subtitle:
[S]ub-replacement fertility rates eventually lead to a shrinking of population - and throughout recorded human history, declining populations have always followed or been followed by Very Bad Things.  Disease.  War.  Economic stagnation or collapse.  And these grim tidings from history may be in our future, since population contraction is where most of the world is headed.
Notice the odd language: "followed or been followed."  The reason is pretty obvious.  History is full of cases where Very Bad Things happen, then population falls as a result: the Black Death, the Mongol invasions, the conquest of the New World.  But history is not full of cases when population falls as a result of low fertility, then Very Bad Things happen.  Indeed, I'm not aware of any clear-cut examples of the latter.  And as far as I can tell, Last doesn't provide any such examples.*

"Very Bad Things" isn't a throwaway line; it's also the title of chapter 5.  After reading this chapter, though, you really have to ask, "Is that all you got?"  When critics of overpopulation cry "disaster," they predict mass famine and social collapse.  The Ehrlichs of the world are silly, but their rhetoric is consistent with their predictions.  Last's list of Very Bad Things is underwhelming by comparison: health care rising as a share of GDP, shrinking cities, closing villages, and unsustainable retirement programs.  He also mentions decreased innovation, but focuses on the debatable effect of median age instead of the undebatable effect of total population.**

Last's Very Bad Things are problems, but they hardly add up to "disaster."  People won't like higher taxes and lower retirement benefits, but they're hardly the end of the world.  If you're really worried, you can protect your future by saving more in the present.  As for the other problems, most people will barely notice them.  At risk of sounding callous: Villages don't shut down due to lack of population until there's almost no one left to lament the shut-down.

To be fair, Last does anticipate blase reactions:
Mind you, these are the concerns of rich, First World countries, like America.  If you're a poor, developing country, the prospects of population aging are much, much worse... A decline in lifestyle for a middle-class American retiree might mean canceling cable, moving to a smaller apartment, and not eating out.  A decline in lifestyle in a place of abject poverty is something altogether different.
He'd be dead right if declining fertility were the only major long-run global trend.  But it's not.  Here's another major long-run global trend: rapid economic growth.  The Great Recession notwithstanding, the global economy continues to swiftly expand - especially in the Third World.  Some countries may get old before they get rich, but as long as demographic and economic trends continue, they'll be rare.

If I disagree with Last on all these points, why do I agree with his natalist conclusions?  Because the greatest costs of declining fertility aren't visible disasters, but missed opportunities.  The millions and billions of people who are never born won't share their ideas with the world.  They won't help spread the fixed costs of idea creation, product variety, and infrastructure.  And they won't enjoy the gift of life.  (If you're already objecting by listing the upsides of non-existence, think again).  Low fertility isn't bad because we'll lose what is.  Low fertility is bad because of we won't gain what could have been.

I'm glad that Last is challenging our demographic complacency.  I hope people take his message seriously.  I want them to buy his book.  My worry, though, is that after they ponder Last's list of Very Bad Things, undecided readers will decide that declining population isn't worth worrying about.

* Last does have a chapter predicting Very Bad effects of declining fertility on foreign policy, but he doesn't present any clear-cut historical examples of such effects.

** Yes, I have debated the undebatable.

COMMENTS (13 to date)
Mark Bahner writes:


We're not going to have an under-population problem. By 2025, the world will be adding 1 billion human brain equivalents every year. And it will be 1 trillion human brain equivalents each year in the year 2033.

A person might say, "Well that will be computers, but not robots." But there could well be half a billion robots added in the year 2025, and tens or hundreds of billions of robots added in the year 2033.

The reason there aren't robots now is because it doesn't make sense to hook $50,000 worth of mechanical equipment to something that has the brainpower of a worm or a mouse.

But by 2025, a petaflop computer (roughly the calculations per second of a human brain) will be a couple thousand dollars. So it will make sense to have $50,000 worth of mechanical equipment hooked up.

The real potential problem is if the tens or hundreds of billions of robots added by 2033 decide they don't like humans.

Tom West writes:

As a programmer, I've always wondered at calculations like Mark Bahner's. Yes, the *hardware* keeps getting better for cheaper, but the software doesn't follow that curve at all.

Software costs in general haven't risen through the roof because we are essentially either doing more of exactly the same thing (processing bank transactions) or where doing the same thing with a nicer interface (word processing).

But look at video games, and suddenly we're looking at budgets that have gone up by a thousand as the hardware has improved.

Human or quasi-sentient brain replication is not bank transactions. The "software" complexity of the brain may not scale steeply exponentially (although on the small scale that's probably does), but it's almost certainly at least mildly exponentially.

Create 10 trillion neurons and you don't have a brain, you have a few pounds of flesh. The electronic equivalent isn't going to be much more useful, and creating the software may take, for all we know, millions of man-years of effort.

MG writes:

I saw Last being interviewed about the book on Book TV by a someone from the Pew Center (who ended up being a good interviewer), and I came across thinking his take on the subject was not all that dissimilar to Bryan's.

Specifically, he was more circumspect about the downside to the fertility bust, but he did a good job of demonstrating why low fertility poses challenges to many things we have become accustomed to: a welfare state, early retirements with tittle private prefunding, innovation in everything. He also stated that there are no historical precendents to what the developed world (and eventually the rest of the world) is experiencing.

Steve Sailer writes:

Most subjects can be profitably discussed in terms of quantity and quality. Unfortunately, when it comes to population, nobody is ever supposed to think about quality (you wouldn't want to have happen to you what happened to Mike Judge with his "Idiocracy," now would you?), so we're only allowed to talk about quantity.

Jeff writes:

On the subject of bad things being preceded by a fall in fertility rates, Phillip Longman noted that both the fall of classical Greece and later the Roman Empire could be traced back to declining births. I would add France to the mix, too. Under Napoleon, they came close to conquering mainland Europe. Half a century later they were getting their clocks cleaned by the Prussians, and of course in the 1930's, Nazi tanks ran them over like a stray dog in the street. The contrast between French and German birth rate changes over that time was not a small factor.

Obviously, none of those examples had ICBMs or carrier groups, but you can imagine the problem manifesting itself in other ways. Maybe China and India will build more and better drone swarms or weaponized satellites or something.

MingoV writes:

The word fertility often is misused by sociologists and economists. Fertility is the ability to create a fertilized egg that develops into a fetus (males: plenty of healthy motile sperm; females: healthy ova that enter and travel through uterine tubes, get fertilized, and implant in the uterus).

Low birth rates can be caused by infertility, but, in first world nations, low birth rates mostly are due to successful use of birth control methods. The latter should not be referred to as a decline in fertility.

Brian writes:


Great post. You've got it exactly right. In general, having kids is better than not, and having more kids is better than stopping. But the results of declining birth rates will likely not be disastrous, just...boring. Low birth rates likely create societies sapped of all vitality, and who wants that? The costs, as you say, are opportunity costs.

Mark Bahner writes:
As a programmer, I've always wondered at calculations like Mark Bahner's. Yes, the *hardware* keeps getting better for cheaper, but the software doesn't follow that curve at all.

Good hardware can power through poor software. Here's a robot that learns like a child:

Robot that learns like a child

And here's an industrial robot that can be trained by a factory worker:

Robot can be taught by workers

I doubt either of those robots is even within a factor of 1000 of the number of operations per second of a human brain. The industrial robot doesn't even have hearing.

Mark V Anderson writes:


You seem to like this book just because you agree with its conclusions. This appears to be the case because you only quoted stupid comments he made in support of his ideas.

I think this is a very bad idea to support books that don't make sense even if they come to the right conclusion. I myself consider such books to be even worse than dumb books that come to incorrect conclusions. All these books do is create the impression these ideas are dumb and work against broader acceptance of the ideas.

Not that I agree that we need more people in this country or the World. I think 7 billion is quite enough. So I don't mind that the guy argues for more people with very bad "facts."

Brian writes:

Mark says: "Not that I agree that we need more people in this country or the World. I think 7 billion is quite enough."

Interesting. And HOW exactly did you determine that 7 billion is the magic number for population sufficiency? Or are you claiming that 7 billion is already way too much and can't quite bring yourself to say where we should start making cuts? I always find it curious when certain people want to argue that we should deny others the chance to exist.

But more to the point, as a general question for everyone, what do you think the carrying capacity of Earth is? When it comes to humans, how much would be too much? I'd be interested to know what people think and why.

For my part, I suspect a carrying capacity that is not well determined, but probably in excess of 100 billion. This is based on currently sustainable population densities and potentially available land mass. In other words, it assumes no fundamental limit due to natural resource use. What do others think?

DougT writes:

What did Malthus (and Dana Meadows) think?

The Malthusians are always and everywhere saying that the lags in population dynamios mean we are in a regime of overshoot and collapse, like Kaibab Squirrels. Since their anti-natalist propoganda is everywhere, it's nice to hear a pro-natalist argument, even if some of the logic isn't as tight as I might like.

Mark Bahner writes:
For my part, I suspect a carrying capacity that is not well determined, but probably in excess of 100 billion. This is based on currently sustainable population densities and potentially available land mass. In other words, it assumes no fundamental limit due to natural resource use. What do others think?

Carrying capacity is determined by technology. For example, there's no way we could feed 7 billion humans with horse-drawn plows. (Or horses to carry the food to cities.)

I agree that the carrying capacity with current technology vastly exceeds 7 billion people...100 billion is a nice round number. But with the progress of technology, the carrying capacity is growing much faster than the approximately 75 million people per year being added.

For example, the progress in desalination technology in recent years is remarkable. And looking to the future, progress seems to be accelerating rather than decelerating. Since 2/3rds of the planet is covered by salt water, the idea that water will be a huge problem in the 21st century seems ridiculous to me.

Water will be "more precious than oil!" Good grief!

Adam writes:

In figuring the social cost or benefit of the marginal birth, one hardly hears discussion of "The gift of life" as Brian described it. It's true that we can't be certain that being born,on average, provides more utility than never having been born in the first place (since we don't know what ante-birth existence consists of). However, we can be certain that once an individual is born, on average, he prefers it to the alternative of ending his life. This is strong evidence that the utility of "the gift of life", for the individual in question is positive, at least relative to not living. In deciding whether the marginal birth is beneficial or costly to society, we can't forget the utility derived from the "gift of life". Guys like Levitt often miss this point when they pronounce that aborted babies are "net losses" to society due to their propensity to be criminals.

Comments for this entry have been closed
Return to top