This graph of European urban population over history shows some dramatic declines, suggesting that there may have been historical analogue to our upcoming world population decline. And there have long been rumors that at least elites often had low fertility. Searching for quotes, I’ve
Many of these examples seem to correlate with resource limitations. In an agrarian society, if the amount of land is limited, increasing population leads to privation. This may be expressed in terms of not wanting to divide land to maintain status, but if there are more people on the same amount of land, with static technology, they will all be poorer and eventually starve. Describing this as a fundamentally social process rather than a social response to resource constraints seems like a reach.
Modern technological societies are built on childhood education. This puts economic stress on parents, both because education itself is expensive and because children are no longer a source of income. Such societies also use the production of women, reducing the amount of time women can devote to childrearing, further increasing the downward pressure on fertility. All of this would be true, regardless of social norms, so it seems odd to claim that social norms drive this, in preference to this being the social response to changing conditions.
Beyond that, at least some of the concern over this trend comes from a belief that unless the workforce keeps growing, technology will stagnate and progress will cease. If that is the case, however, removing women from the workforce in order to increase fertility would have the same effect.
My own expectation is that we will find a series of technological improvements (increased longevity, more efficient fertility and childrearing, improved allocation of resources) that will stabilize the population and maintain technological (defined broadly) progress. Whether population increases at that point will probably depend on how well we have addressed the other needs (land, waste, and energy) that constrain human population.
"if there are more people on the same amount of land, with static technology, they will all be poorer and eventually starve." Did you notice that in many places this issue is stronger for those with MORE land? Hard to square that with this all being due to starving risk.
Humans don't have objective access to their future wellbeing of that of their children. There may be large misprediction on that. And as people tend to learn from other people and to anchor on each others predictions and there is little feedback the argument that they "objectively" are better off doesn't have much strength.
I didn’t say it was all due to the risk of starvation. I did some looking for references to some of the examples you gave, and wasn’t able to find Asian references that tied such practices to broad population declines. I obviously may have missed some. In particular, there are papers I don’t have good access to.
"Modern technological societies are built on childhood education."
There's a bit of a problem there, as while it is true there is probably more need for education now than in say 1200, that doesn't mean that the current amount of time spent on education is necessary or optimal. Social norms drive a lot of what "education" looks like, what it consists of, how long it takes, etc. For example, many modern jobs do not require a college degree, but many people go to college, and so many college educated people are working modern jobs that don't require the degree they acquired. Likewise, large swaths of topics could be ignored if we wanted, thus shortening the time and resources spent on education, but social norms prevent that.
You should take a look at Augustus’s Moral Laws (Lex Julia). He’s often described as a devout social conservative (aristocratic, traditional Roman virtue and family structure), and the laws lend credence to that. They offer a glimpse of what Augustus thought had changed amongst his peers. They don’t seem to have worked in the long run (which is informative if we’re thinking about where to make the changes), and many were repealed soon after his death.
Since I do not have access to the original journal articles I would like to drop my questions here:
Question 1: To what extent do these sources represent the attitudes and practices of the societies they describe? Do they primarily reflect the perspectives of the elite and not their society as a whole?
Question 2: What were the gender roles and dynamics involved in decisions about infanticide and fertility? How much autonomy did women have in making these choices?
Question 3: How did religious and ethical beliefs influence these practices? Did they change over time within different cultures?
Question 4: What family planning or birth spacing methods, if any, were used as alternatives to infanticide in resource-constrained situations?
Question 5: What was the demographic impact of widespread infanticide and fertility restriction among elites at the societal level? How did populations stabilize?
Good post, Robin, raising a fascinating idea. I'm very curious about pre-industrial population decline.
However, one question: how many of these stories are about the elite, rather than the total of a given society's population? I know there are serious limitations with sources on this score.
Some quick and lazy research suggests that the only major city that remained inhabited during this period of time was Athens, and that the total population in Greece at this time was 700,000 (I chose 800 BC as my reference year). It doesn't seem that unlikely.
What does it matter the total population in Greece at this time? The chart is about population in cities across all of Europe, not just Greece.
It's very unlikely that European city population could have dropped from close to 1 million to only 10,000 in the span of a couple hundred years. That would be a loss of 99% of previous city dwellers across all of Europe. Some cities were destroyed in the late bronze age collapse, but claims of this were also exaggerated. https://en.wikipedia.org/wiki/Bronze_age_collapse#City_destruction
At its peak (500 years later), Athens alone had a population of 250k.
https://en.wikipedia.org/wiki/List_of_largest_European_cities_in_history appears to list a reasonable set of candidate cities to evaluate as "urban", and a nadir of 10,000 between 1000BC and 700BC is plausible; the four cities of relevancy appear to be Athens, Argos, Tiryns, and Rome. Tiryns stopped existing during that timeframe, so I think it's fair to ignore it. Argos may not have qualified as "urban", possibly contributing fewer than 1,000 people during its nadir, so really we're just talking Athens and Rome, whose population, if indeed this was a low point, likely did amount to around 10,000 people.
The most charitable take of the claims does suggest they're not baseless.
We were looking for a 99% decline in city population between 1300 BC and 900 BC, because that's about what Robin's graph shows.
If we use your method and the chart you linked, we find that for 1300 BC, there's about 90k population listed. Compared to 1000 BC where there's something like 20k population listed, and 700 BC where there's around 75k population listed. So at most your method and chart says there might have been an 80% decline, not a 99% decline.
Surely there are other cities not listed in your chart, which would bump up the 1300 BC number from 90k to something closer to 1 million, and similarly bump up the 900 BC number.
If not, then the anomaly in Robin's graph is the 1300 BC number being an order of magnitude too high rather than the 900 BC number being too low.
20k listed, and a 300 year gap for that to fall another 50% before rising.
And surely there are other settlements - whether or not they qualify as "urban", however, is a relevant consideration. A thousand settlements of 500 people each is a huge population - but would, for most conceptualizations of "urban", contribute 0 people to the urban population.
Many of these examples seem to correlate with resource limitations. In an agrarian society, if the amount of land is limited, increasing population leads to privation. This may be expressed in terms of not wanting to divide land to maintain status, but if there are more people on the same amount of land, with static technology, they will all be poorer and eventually starve. Describing this as a fundamentally social process rather than a social response to resource constraints seems like a reach.
Modern technological societies are built on childhood education. This puts economic stress on parents, both because education itself is expensive and because children are no longer a source of income. Such societies also use the production of women, reducing the amount of time women can devote to childrearing, further increasing the downward pressure on fertility. All of this would be true, regardless of social norms, so it seems odd to claim that social norms drive this, in preference to this being the social response to changing conditions.
Beyond that, at least some of the concern over this trend comes from a belief that unless the workforce keeps growing, technology will stagnate and progress will cease. If that is the case, however, removing women from the workforce in order to increase fertility would have the same effect.
My own expectation is that we will find a series of technological improvements (increased longevity, more efficient fertility and childrearing, improved allocation of resources) that will stabilize the population and maintain technological (defined broadly) progress. Whether population increases at that point will probably depend on how well we have addressed the other needs (land, waste, and energy) that constrain human population.
"if there are more people on the same amount of land, with static technology, they will all be poorer and eventually starve." Did you notice that in many places this issue is stronger for those with MORE land? Hard to square that with this all being due to starving risk.
Humans don't have objective access to their future wellbeing of that of their children. There may be large misprediction on that. And as people tend to learn from other people and to anchor on each others predictions and there is little feedback the argument that they "objectively" are better off doesn't have much strength.
Land near functional people and institutions is the resource limitation of our day.
Fertility decline is a function of pessimism amongst those with the conscientiousness to avoid unplanned pregnancy.
In the Middle Ages/Renaissance these people were full of Christian optimism and had more kids.
I didn’t say it was all due to the risk of starvation. I did some looking for references to some of the examples you gave, and wasn’t able to find Asian references that tied such practices to broad population declines. I obviously may have missed some. In particular, there are papers I don’t have good access to.
You can send your kids to public school for free, and smarter kids can go to community college on scholarships.
There is a lot of signaling around education.
Other costs of raising kids:
Housing size. Putting 6 kids in 2 bedrooms isn't done anymore.
Location - a good neighborhood matters a lot more with kids
Phones, electronics, games, all are basic costs with kids nowadays.
Clubs, little leagues, etc etc etc
Actually raising the kids - you can't hit them, can't yell at them, need to entertain them and give them attention for the sake of attention
I suspect people are less proud of their kids today, and less likely to want a kid when seeing other people's kids
"Modern technological societies are built on childhood education."
There's a bit of a problem there, as while it is true there is probably more need for education now than in say 1200, that doesn't mean that the current amount of time spent on education is necessary or optimal. Social norms drive a lot of what "education" looks like, what it consists of, how long it takes, etc. For example, many modern jobs do not require a college degree, but many people go to college, and so many college educated people are working modern jobs that don't require the degree they acquired. Likewise, large swaths of topics could be ignored if we wanted, thus shortening the time and resources spent on education, but social norms prevent that.
You should take a look at Augustus’s Moral Laws (Lex Julia). He’s often described as a devout social conservative (aristocratic, traditional Roman virtue and family structure), and the laws lend credence to that. They offer a glimpse of what Augustus thought had changed amongst his peers. They don’t seem to have worked in the long run (which is informative if we’re thinking about where to make the changes), and many were repealed soon after his death.
I have looked at those.
Since I do not have access to the original journal articles I would like to drop my questions here:
Question 1: To what extent do these sources represent the attitudes and practices of the societies they describe? Do they primarily reflect the perspectives of the elite and not their society as a whole?
Question 2: What were the gender roles and dynamics involved in decisions about infanticide and fertility? How much autonomy did women have in making these choices?
Question 3: How did religious and ethical beliefs influence these practices? Did they change over time within different cultures?
Question 4: What family planning or birth spacing methods, if any, were used as alternatives to infanticide in resource-constrained situations?
Question 5: What was the demographic impact of widespread infanticide and fertility restriction among elites at the societal level? How did populations stabilize?
Good post, Robin, raising a fascinating idea. I'm very curious about pre-industrial population decline.
However, one question: how many of these stories are about the elite, rather than the total of a given society's population? I know there are serious limitations with sources on this score.
This seems to be more about infanticide than about fertility.
Polytheistic societies just did not have the same attitudes towards shame and murder.
The graph's claimed low of only 10,000 people living in cities in the Greek dark ages seems unlikely. What's the ultimate origin of this claim?
Some quick and lazy research suggests that the only major city that remained inhabited during this period of time was Athens, and that the total population in Greece at this time was 700,000 (I chose 800 BC as my reference year). It doesn't seem that unlikely.
What does it matter the total population in Greece at this time? The chart is about population in cities across all of Europe, not just Greece.
It's very unlikely that European city population could have dropped from close to 1 million to only 10,000 in the span of a couple hundred years. That would be a loss of 99% of previous city dwellers across all of Europe. Some cities were destroyed in the late bronze age collapse, but claims of this were also exaggerated. https://en.wikipedia.org/wiki/Bronze_age_collapse#City_destruction
At its peak (500 years later), Athens alone had a population of 250k.
Ah, misread.
https://en.wikipedia.org/wiki/List_of_largest_European_cities_in_history appears to list a reasonable set of candidate cities to evaluate as "urban", and a nadir of 10,000 between 1000BC and 700BC is plausible; the four cities of relevancy appear to be Athens, Argos, Tiryns, and Rome. Tiryns stopped existing during that timeframe, so I think it's fair to ignore it. Argos may not have qualified as "urban", possibly contributing fewer than 1,000 people during its nadir, so really we're just talking Athens and Rome, whose population, if indeed this was a low point, likely did amount to around 10,000 people.
The most charitable take of the claims does suggest they're not baseless.
We were looking for a 99% decline in city population between 1300 BC and 900 BC, because that's about what Robin's graph shows.
If we use your method and the chart you linked, we find that for 1300 BC, there's about 90k population listed. Compared to 1000 BC where there's something like 20k population listed, and 700 BC where there's around 75k population listed. So at most your method and chart says there might have been an 80% decline, not a 99% decline.
Surely there are other cities not listed in your chart, which would bump up the 1300 BC number from 90k to something closer to 1 million, and similarly bump up the 900 BC number.
If not, then the anomaly in Robin's graph is the 1300 BC number being an order of magnitude too high rather than the 900 BC number being too low.
20k listed, and a 300 year gap for that to fall another 50% before rising.
And surely there are other settlements - whether or not they qualify as "urban", however, is a relevant consideration. A thousand settlements of 500 people each is a huge population - but would, for most conceptualizations of "urban", contribute 0 people to the urban population.
Whatever. The point is, your data does not support a 99% decrease in city population between 1300 and 900.