Population Growth and Technological Change: One Million B.C. to 1990

That is an older paper by the excellent Michael Kremer, worth keeping in mind, here is the abstract:

The nonrivalry of technology, as modeled in the endogenous growth literature, implies that high population spurs technological change. This paper constructs and empirically tests a model of long-run world population growth combining this implication with the Malthusian assumption that technology limits population. The model predicts that over most of history, the growth rate of population will be proportional to its level. Empirical tests support this prediction and show that historically, among societies with no possibility for technological contact, those with larger initial populations have had faster technological change and population growth.

This bears on my earlier Bloomberg column, today cited by Mike Lee, suggesting that having more children is likely to help out on the climate change issue.


"those with larger initial populations have had faster technological change"

Quite: the near-billion population of England in the 18th and 19th centuries guaranteed the coming of the Industrial Revolution.

Well the French had a big population, as did the Chinese, but both of them had too much social convention holding them back from using labor saving technology, which the latter invented, precisely because of this reason (too many people means labor saving inventions--which is most inventions--are mere novelties rather than practical devices). Thus a better start from just "big population" is "better patent laws". But sadly I doubt this will be adopted, so TC's suggested "more people means more inventions since some people--mainly nerds--invent stuff just for the fun of it" is the way to go forward.

Bonus trivia: labor saving devices were used during the late Roman empire, when populations started thinning a bit. And this: https://en.wikipedia.org/wiki/Joseph_Needham ("During his stay in China, Needham was asked to analyse some cattle-cakes that had been scattered by American aircraft in the south of China at the end of World War II, and found they were impregnated with anthrax.") - whoa! USA employed biological weapons? Developed in an army camp in Maryland if memory serves.

I believe that one of the premises above was a society with no contact. Great Britain had enormous contact by and through its sea power, trade, and empire.

But I agree with you in your example that in an open society, population is not as crucial to technological development.

The US surpassed Britain in the 19th because of its larger population and similar technological knowledge. China in the 21st is set to do the same for similar reasons.

You're forgetting another massive reason: natural resources, which in the 18th mattered more than just about anything. China doesn't have that over the US, even though it matters less.

Sorry, 19th

It doesn't matter at all due to trade.

That's why South Korea's stuck behind China then. Population's too small...

China has surpassed South Korea in everything except memory chips, and maybe cars. Coastal China is as rich as South Korea, the interior is still third world and brings the average down. I almost put shipbuilding down for South Korea but China got that too.


Coastal China (even so qualified) has not surpassed South Korea in technological innovation per head. Seriously.

A child born today would not contribute an idea to mitigate any problem CO2 emissions might have until he or she was 25 to 30 years old in 2045 or 2050, well after technology in energy, sequestering, etc. will have ended any concerns.

And who's going to invent the "etc", Einstein?

"Etc" includes virtual reality which might lower travel to work, etc2. But Tyler's view is that there will be little technological innovation until the 2040s.

Tired: we shouldn't leave a broken world for our kids.

Wired: we should totally leave a broken world for our kids .. to fix!

(somewhat off-topic) I know someone who gave birth to 8 children. This person was about 35 years old between the birth of her two middle children. Historically, there have been about 105 males born for every 100 females. The Pyramids of Giza were built around 2500 BC when the total human population was around 5m. If each woman then and afterwards also gave birth to 8 children in the same way, the current human population would be about the following (according to my calculation) -


According to Google, this equals 5.3444188e+82 which exceeds the number of atoms in the known universe (10^82)!

This only works if all of those children are immortal

No infant mortality, no wars, no disease or starvation, no predation. Lot of nos.

And I cant validate his math even though i see all his parameters.

Actually, the world population in 2500 BC was about 15m but maybe about a third (5m) of the population at that time (both male and female) was in the child-bearing years, so 5m might be roughly correct as a starting population.

This calculation doesn't assume that people are immortal (no sigma notation to sum up the generations) but it does assume that all people born survive to adulthood to have offspring themselves.

Unlike the past, a large majority of children now do survive to adulthood. 2500 BC is in the distant past but is recent enough to envision what human populations could be like in the future along a similar span of time, 4500 years.

For instance, if women have 'just' 3 children for the next 500 years, then the population would be about ((3/2.05)^(500/35))*7.5*1000000000 = 1.7276374e+12 (about 1.7 trillion). The numbers get wild after several generations.

The technologies to lower birth rates have mostly been rolled out, but the technologies to raise them have much room for growth. Imagine in the future where procreation is thought to be a human right that's enabled by the government for the infertile, there's more free time to have children, and children are viewed as a status symbol (e.g. everyone already owns an iPhone and a fancy car because of great material prosperity).

"This only works if all of those children are immortal"

Assume a spheroidal chicken in a vacuum....

Even more stupider is the failure to factor infant mortality, other mortality rates, and very short life spans.

Even mas estupido - in the past one million years there were four glaciations and global warmings. Ergo, billions of million-year-old homo sapiens people caused four climate changes catastrophes beginning un milion anos pasados. Just establishing my bi-lingual creds here.

I once wrote a program that proved, using a Monte Carlo simulation, something that Francis Galton and his kind proved back in the late 19th century, and that is, to ensure with a greater than 50% probability of ensuring your DNA will survive in subsequent generations, you need to have at least five offspring. From memory, having two kids I think only gives you a 35% chance of spreading your DNA forever, due to the Poisson distribution nature of human reproduction (as shown by a survey of French solders as I recall). So the nuclear family of two kids is, statistically speaking for those families a futile effort for the individual family (not collectively, since some families will have offspring that have lots of kids and their DNA will 'live forever'). In fact, strictly speaking a pure nuclear family of exactly 2 offspring and no more will not even result in a steady population for society as a whole, but will result in a slowly shrinking population (since the replacement rate is 2.2 or so offspring).

Bonus trivia: Genghis Khan, a murderer, sadist and rapist, was the most successful progenitor of all time, his DNA will live forever. Alexander the Great not so much, since his offspring were all killed by his Greek generals after he died.

Of course, DNA doesn't "live" in any real sense. If making sure that a particular configuration of nucleic bases is still evident in some bodies walking around in a couple hundred years is important to you, I guess everyone needs a hobby.

Adam was the most successful progenitor, as of now. Adam, and not that little Genghis dude, who died in pain and despair long long ago, over the hills and far away. I remember.

There are several people alive today - you may know some of them - who are going to have more children than Adam. Trust me.

"Out of the rocks on the side of the road" was just the beginning of it.

By the way not everybody needs a hobby.
Imagine you are a nobody in today's society, maybe a high school teacher or somebody with an "accounting degree" who jokes about how "you only have to work 40 hours a week", or a woman of genius who home schools her children - and remember that a thousand years from now, you might be a revered ancestor for a reason that only you understand.

That is not going to happen to me, I am just an anonymous guy who tries to write like a bot because I love the downtrodden, including every single person who tries to program a bot to sound like me, because I know I can be trusted, and because I understand the downtrodden.

And I have no hobbies, except having friends, relaxing, and helping people out, and I guess praying a lot, which is easy for me because unlike you I do not have a guardian angel and that makes this world a little scarier but also a lot more vivid and amusing.

Don't try it at home though.

Don't get angry at me get angry at William Arkle, of George McDonald, or the great Carmelite poets, whom I consider my inspiration. We are not talking baseball statistics here.

As of now Adam has maybe 14 billion descendants, poor Genghis who I more or less saw die choking on his own bilious vomit not that long ago has less than a billion, and I can show you in any average shopping mall on any ordinary Saturday two or three people who will have a hundred billion descendants.

All of whom will have at least one guardian angel each.

That is all more predictable, to a greater number of digits, than any prediction you can counter-offer.

Tell me exactly where I am wrong, you probably can't.

1990 was a really good year for many people

(don't bother to look it up -- it is ridiculous to think anyone could get angry at William Arkle or George MacDonald!)

why do i bother

why do i bother - because I remember 1990 so well

and 1974, too ---- and what it was like to love, and be loved, in that blessed year of the Lord

please don't respond but if you do please quote Shakespeare or Li Po or Isaiah




the counter-beat, paul ceylan

the antschels were good people

Alexander was not promiscuous and seems to have preferred men to women. He had just one known child, born after he died. There was a later pretender claiming to be Alexander's bastard son, but since he only surfaced when the legitimate son (and most of Alexander's other relatives) were dead, he was probably a fake put forward by one of the generals disputing Alexander's empire to counter the claim of Alexander's remaining sister and her husband.

"...implies that high population spurs technological change."


how nice -- and just what spurs population growth ??

(of course economic productivity spurs population growth)
(now if only economists knew what sputs economic growth)

It's a spiritual thing, which economists don't like to talk about. It's being found at places like Gobekli Tepe that population centers and vast engineering works preceeded agriculture and state formation. Was agriculture a response to the population growth which was spurred by a spiritual revolution? Because that's what brought them there.

The model predicts that over most of history, the growth rate of population will be proportional to its level.
Meaning that is is Euler exponential and smooth. This is equivalent to saying the error bands about technology are small. In quant theory, it means the tree trunk is round, holograms work. In reality is means that parents are not inordinately busy finding food for the kids and have time to work problems. This is phase minimization, another Euler concept that applies to Gaussian variations, population distributions are bell curves.

I should note, this also means the young and old generations sample each other sufficiently to avoid large sampling errors. There is a large coalescence among various transform theories in math recently.

We need no "help" on the climate change issue. We need nuclear power. Whether you believe in AGW or not (I do not), nukes are the answer. We have a very small technological and political problem of disposing of nuclear waste. The current state of technology already puts to rest most of the other concerns.

The environmental movement has cut its own throat by associating with radical leftism. They don't really want a clean environment. What they want is slower growing Western economies and money funnelled to their Big Green interests.

I do believe in AGW (though not the catastrophic non-scientific predictions). But yes, I agree, nuclear power would, for much lower costs than $200 per ton carbon taxes, solve the issue.

I might come to believe it with some uncompromised data and analytical methods. Every paper I've read has been tragically flawed to the point I really dont want to waste my time anymore. And they rely so often on superficial nonsense that I cant take the people seriously anymore. Personal anecdotes from mountain climbers. Weather events. Missing data. Manipulated data. Trend breaks exactly where two different temperature reading sources are spliced together. Living lifestyles as though they don't believe it themselves.

Personal lifestyle choices are irrelevant in the context of 35 billion tons of annual carbon emissions.

We can't solve climate change through individual action any more than we could solve the smog problem through individual action.

It require coordinated global efforts by governments. It requires passing laws and spending lots of money. There is no alternative.

Unfortunately, large numbers of people are ideologically allergic to the concept of collective action and are forced to come up with every more complicated excuses to explain why the overwhelming scientific consensus on climate change is Fake News.

There wasn't a coordinated effort among the U.S. , the E.U. and Japan to decrease CO2 emissions but lo and behold, they fell around 18% in the U.S. since 2007.

> Living lifestyles as though they don't believe it themselves.

It's a tragedy of the commons. If we insist that people identifying such issues take extreme measures personally to prove their non-hypocrisy, we shift the incentives landscape even further from useful collective action.

No offense intended, but why the heck would any reasonable person expect that in 2019 any other reasonable person would think the opinion of someone who is so fundamentally flawed as to "not believe" in anthropogenic climate change is worth even a millisecond of their time? I certainly won't waste any of mine, since I assume your other opinions are as well grounded (i.e. not at all) in facts and evidence.

+1. He's either a shill for industry or has environmentalist derangement syndrome. I like how he minimizes nuclear waste. The half-life of some of this toxic stuff is longer than from the invention of agriculture to the world today. Small problems, indeed.

Of course the half-life of cadmium is infinite, has a higher LD-50 and is just deep injected to "deal with it".

I mean we are only using thousands of tons of Cadmium to make solar cells ...

In the ancient world, every woman had to have at least six or seven pregnancies just to maintain the population of the tribe, so many mothers and infants died in child birth. Indeed, the life of a woman in the ancient world has been described as sex, birth, death, and decay. How nice. The current administration is promoting populating growth in less developed countries by restricting aid to countries that promote family planning (i.e., birth control and abortion). As population in the developed world stagnates and population in the less developed world explodes, what are the implications for global economic growth? What are the implications for women in the less developed world? [So readers aren't confused about China's position, demographers are projecting the population will continue to grow the first half of this century, and then decline by about 400 million in the second half - that's the mid-range projection].

Curbing population growth does not lead to prosperity, as you ably noted in the beginning. Large families ensure continuity and act as a form of insurance in one's elder years. Therefore encouraging so-called "family planning" switches the arrow of causation, deprives those societies of strength and endurance, and inculcated a decrepit moral philosophy of butchering children rather than raising them.

They dont solve their horrendous social and moral problems by adopting more horrendous socialist and immoral practices.

Except that we have pretty good evidence to the be contrary. As long as a society can maintain its numbers (and most human societies were at rough birth-death balance for most of history) the society will endure, barring external catastrophe.

Tried to resist responding to that inanity but it has to be done. You're simply making shit up. You've no idea what the infant and maternal fatality rates were in the ancient world. Accidents were the major cause of death. Deaths during childbirth and infancy were less common than those among wild animals, since there were midwives, which moose and wolves don't have. Populations in that era were generally healthier than the present obesities, since sweets were a rarity, honey and dates mostly. Practically everyone lived an active, physical lifestyle and epidemic diseases were rare. This all changed in the west when transportation and travel allowed the spread of bacteria and viruses to populations without immunity and slothful inactivity and overeating in women made pregnancy dangerous sometime between the fall of Rome and the Renaissance.

Just google 'Roman child mortality' and read.

chuck's tribe doesn't use the pale face search engine.

The Latina paradox might be of some relevance:

"Latina mothers in the United States enjoy surprisingly favorable birth outcomes despite their social disadvantages. This "Latina paradox" is particularly evident among Mexican-born women. The social and cultural factors that contribute to this paradox are maintained by community networks—informal systems of prenatal care that are composed of family, friends, community members and lay health workers. This informal system confers protective factors that provide a behavioral context for healthy births. U.S.-born Latinas are losing this protection, although it could be maintained with the support of community-based informal care systems."


Early Childhood

Health Disparities

" among societies with no possibility for technological contact"

I'm skeptical of the model. What historical societies actually satisfy that constraint? "No possibility" != infrequent.

That said, I'm sympathetic to the empirical claim, but wonder if it still implies in in our industrial age society. Is manpower still the bottleneck nowadays? Or is it more of a hindrance (as the environment movement claims/implies)?

If it is a model, then it is drawing a conclusion with a simplifying assumption. Clearly there are few societies today with zero contact. This doesn't make the predictions of the model invalid per se. One can and should relax that assumption in later models.

But mankind likely did live out of contact with others for millenia, particularly in Australia, the Americas, Mongolia, Pacific islands, and a few other places. Or their relationships with neighbors were so hostile or standoffish that they traded little information. Those isolated societies clearly lagged the rest of the world in technology and innovation. Nowhere in the world is there a Wakanda or Atlantis. Technology was born in the midst of huge social and economic interactions. Europe learned to make pasta and gunpowder from the Chinese. It learned to make bone pottery not from the Chinese but despite them. America developed the atomic bomb because of the Germans and Japanese. Interactions don't have to be pleasant and cooperative to be technologically fruitful. Thus we can "borrow" the population and practices of others.

I'll give you Australia, but as for the rest:

"the Americas" It's a huge stretch to call all that "a society. " More like hundreds/thousands of different cultures. You might as well call all of Eurasia a society.

"Mongolia" I'm confident they've had contact with other societies at least since the Great Wall was built. Before that? I'm skeptical about data.

"Pacific islands, and a few other places" Maybe Easter Island.... but in general, I think there was contact between various islands. Per Wikipedia (YMMV, of course): "Generally each island maintained a guild of navigators who had very high status; in times of famine or difficulty these navigators could trade for aid or evacuate people to neighboring islands. "

Saying "the Americas" were isolated is pretty damn odd. Why not say "Eurasia was isolated" (from the Americas, Africa)?

"having more children is likely to help out on the climate change issue."

Like General Turgidson asked in another context, , mightn't the solution be "the abandonment of the so-called monogamous sexual relationship, I mean, as far as men were concerned?"

Where on earth did Mr 1993 article do they find these "societies with no possibility for technological contact"? In this non-obscure article which is well known, much cited and has reverberated through the literature on innovation since the time of its publishing.

I think Malthusian pessimism on fast population growth is likely true in the short term, but false in the long run. If average births per woman is 5 then the vast majority of the population would not be in the labor force. Less than 30% of the total population would be working which means each worker is supporting ~3 dependents. This would be a huge stress to those workers and GDP/capita would be relatively low.

However, in the long run the population growth rate would either A) slow down as it hits carrying capacity and the percent of the total population in the workforce would rise to a healthier level or B) technological change as a benefit from more the increase in human capital would make 5 births per woman manageable.

Low fertility rates appear to be a much greater economic stress to society in the long run given the increasing ages of people in society. Older dependents are usually more expensive to care for than younger dependents.

That argument ignores mortality rates, standard of living, public goods, shared subsistence. It also ignores the massive resources of the earth. The entire world population can fit inside the borders of Texas in single family homes. The very resource problems you state was what spurred the expansion of civilization. Scarcity sparked innovation. Agriculture yields have far outpaced population growth. And empirically we see people choosing smaller families when economic conditions are better and risks are smaller. The system is self limiting, rationing and innovative in ways Malthus couldn't imagine.

How long is the short run you imagine? In production, the long run is generally two years. That is a very long time if you are starving but a woman can only pop out 2.67 kids in that period, barring the uncontrollable chance of twins or more. This means the family has sufficient time to adapt. And in extreme privation, the woman is more likely to miscarry and infant mortality rises. Malthus was concerned about the resulting social strife, but that is a normative view.

The world is more at risk from technology than population growth. The question is whether technology will produce solutions faster than it creates them.

I don't disagree with much that you say. I generally favor naturalism and don't think that overpopulation is a threat to humanity. Nor do I think technology is much threat either.

Generally, GDP growth rates are highest in countries with moderate (above replacement) fertility rates and lowest in countries with low fertility rates (well below replacement). Countries with very high fertility rates tend to be somewhere in between. There are exceptions to the rule caused by immigration and economic policy.

I think the main point in my original comment had more to do with GDP/capita and less with GDP. This would be a charitable reading of Malthus and probably not how most of his supporters would defend his views on population growth so I'll concede that.

Edit: second sentence should read natalism, not naturalism

I think Mike Lee (UT-R) stole this idea. He said:

"The solution to so many of our problems at all times and in all places is to fall in love, get married, and have some kids."

He is being widely mocked, but he is seriously propounding the thesis that population growth begets technological advances. He may not be wrong.

Depends on time scale. Today we have computer networks, yesterday we were 20,000 souls at one point. So he has to pick a time scale, how? His experiment naturally selects time scales in which the population growth keeps converging to population level. But those are precisely the populations that match technology and population.

Why does his research naturally select converging populations? Because we have 6 billion today, up from 20,000 people 70,000 years ago, he is going to find the solutions, the two numbers alone verify his hypothesis.

Suggest you read Julian Simon on population growth and economic growth.

growth rate proportional to "level" (which I assume means total population) is exponential growth. As an old microbiology student, I suspect that if we're lucky we'll follow the s-curve like bacteria do (slow start, exponential acceleration and then plateau). It seems few like to talk about it, but the best technology available today does not (!) allow today's western standard of living with energy per capita usage much less than where we (in the USA) are today. There really seems to be a general failure to see that the Emperor has no clothes. What do you call the environmentalists that are to energy as vegans are to diet? I don't know if there even are any. Are there?

I'm currently reading a a very good book by Edward Glaeser entitled "The Triumph of the City" which deals with some of these issues.

Kremer was way ahead on this. Footnote 2 in the 1993 paper, p. 686:
"Ted Baxter of the "Mary Tyler Moore Show" apparently agreed: he planned to have six children in the hope that one would solve the world's population problem. "

Suppose a 2x2 square grid. Left column contains problems that are not exacerbated by larger numbers of people. Right column contains problems that are. Upper row contains problems that people can solve if they are sufficiently clever. Lower row contains problems they cannot. Tyler seems to think that the lower right square is completely empty. I think he is wrong.

i know it's in your job description to occasionally weigh in on trendy topics like how having babies impacts climate change, but my god, if you take such a rationalization seriously, you're as nuts as the women who cooked it up.

Comments for this post are closed