Historic urban population of China

Feb 2011
6,113
#21
It's also funny how you would say this:
We have good data for Chang'an from the Han Chinese census that gives the city 246,000 people around the year 1
Yet, when it is shown that the same census also shows the population of other cities which flies in the face of a 5% urbanization level, you now argue how inaccurate the census is. Are your opinions modeled after the facts, or are the "facts" modeled after your opinions?
 
Feb 2011
6,113
#22
Don't forget, I also showed how the census figures for city family size are accurate. This you didn't address. Another reason is the food imports which confirms the city size of Chang'an.

Here are more city information for excavated Warring States cities:

Jinan: 4500 meters by 3600 meters. Even if we use your low estimation of 100 people per hectare(which you made up, btw), that still gives 162,000 people.
Eastern Zheng city: 5100 meters by 2900 meters. Using your low estimate on urban density, that gives 147,900.
Yong city: 10 square kilometers, which gives 100,000 people
Shouchun: 26.35 square kilometers, which gives 263,600 people
Yueyang city: 2500 meters to 1600 meters. That gives 40,000 people
Zhouyuan: 15 square kilometers, 150,000 people
Linzi: 16 square kilometers, 160,000 people (though records indicate a family size of 70,000 and that it's impossible to walk the streets without brushing shoulders, so the people per hectare should be much higher)
Xiadu: 32 square kilometers, 320,000 people (largest excavated city of the era)
Yan: 3 square kilometers, 30,000 people

And those were the cities that are excavated, and the area is the wall area. The city size is backed by archeological data. There are also many other cities in the records, including capital cities. Their recorded area is comparable to those of the excavated cities I have listed (excluding the two that have only 3 and 4 squared kilometers)

Assuming that these are the 36 "large" cities Zhao used, then even if the rest of the 36 cities were only of 1000 in size, then the average city size would still be near 40,000. If we include the recorded cities as well, then this number would be much higher. Remember, the city population are all estimated using the low urban density rate that you gave.
 
Last edited:
Feb 2011
6,113
#23
The attitude of the government can be seen through the "Discourse of Salt and Iron". The work is written about an argument between the bureaucracy(represented by various government officials) and the literati. And get this, though the bureaucracy prefers the monopoly over salt and iron, it was also they who prefers mercantile pursuits. It is the laissez faire literati who argues against merchants and artisans in favor of agriculturalists, even though they were against the salt/iron monopoly.
For example:
The Lord Grand Secretary: The ancient founders of the Commonwealth made open the ways for both fundamental and branch industries and facilitated equitable distribution of goods. Markets and courts were provided to harmonize various demands; there people of all classes gathered together and all goods collected, so that farmer, merchant, and worker could each obtain what he desired; the exchange completed, everyone went back to his occupation. Facilitate exchange so that the people will be unflagging in industry says the Book of Changes. Thus without artisans, the farmers will be deprived of the use of implements; without merchants, all prized commodities will be cut off. The former would lead to stoppage of grain production, the latter to exhaustion of wealth. It is clear that the salt and iron monopoly and equable marketing are really intended for the circulation of amassed wealth and the regulation of the consumption according to the urgency of the need. It is inexpedient to abolish them.

The Literati: Lead the people with virtue and the people will return to honest simplicity; entice the people with gain, and they will become vicious. Vicious habits would lead them away from righteousness to follow after gain, with the result that people will swarm on the road and throng at the markets.
A poor country may appear plentiful, not because it possesses abundant wealth, but because wants multiply and people become reckless, said Lao-tzŭ. Hence the true King promotes rural pursuits and discourages branch industries; he checks the people's desires through the principles of propriety and righteousness and provides a market for grain in exchange for other commodities, where there is no place for merchants to circulate useless goods, and for artisans to make useless implements. Thus merchants are for the purpose of draining stagnation. and the artisans for providing tools; they should not become the principal concern of the government.

The Lord Grand Secretary: Kuan-tzŭ is reported to have said:
A country may possess a wealth of fertile land and yet its people may be underfed — the reason lying in lack of an adequate supply of agricultural implements. It may possess rich natural resources in its mountains and seas and yet the people may be deficient in wealth — the reason being in the insufficient number of artisans and merchants. The scarlet lacquer and pennant feathers of Lung and Shu, the leather goods, bone and ivory of Ching and Yang, the cedars, lindera, and bamboo rods of Chiang-nan, the fish, salt, rugs, and furs of Yen and Ch'i, the lustrine yarn, linen, and hemp-cloth of Yen and Yü, are all necessary commodities to maintain our lives and provide for our death. But we depend upon the merchants for their distribution and on the artisans for giving them their finished forms. This is why the Sages availed them of boats and bridges to negotiate rivers and gulleys, and domesticated cattle and horses for travel over mountains and plateaux. Thus by penetrating to distant lands and exploring remote places, they were able to exchange all goods to the benefit of the people. Hence His late Majesty established officers in control of iron to meet the farmer's needs and provided equable marketing to make sufficient the people's wealth. Thus, the salt and iron monopoly and the equable marketing supported by the myriad people and looked to as the source of supply, cannot conveniently be abolished.

And the argument continues..... for quite a length of time.
Note: "Equitable marketing" usually means an attempt for the government to stabilize prices. This is usually done by buying goods when they are cheap and selling them when they are expensive.
 
Last edited:
Mar 2011
5,045
Brazil
#24
For comparison here is the medieval and early modern Western European city size raking:



And here is the total urban population measured in terms of towns over 5,000 inhabitants:



Here is the rate of urbanization:



Now, if this dataset is correct in that it indicates the number of urban families in Han China, using 5 persons per family yields:

--------------------- pop, thousands
1) Chang'an: 80,800 - 404
2) Chengdu: 76,256 -381
3) Maoling: 61,087 - 305
4) Luoyang: 52,839 -264
5) Lu: 52,000 - 260
6) Changling: 50,057 -250
7) Yanling: 49,101 - 246
8) Wan: 47,547 - 238
9) Yangdi: 41,650 - 208
10) Pengcheng: 40,196 - 201

total: 2,757 thousand

Now, the 10 largest cities of Western Europe in 1650, when Western Europe had a rate of urbanization of 11.4% and 80 million inhabitants, were 1,890 thousand and in 1800, when Western Europe had 120 million inhabitants and a rate of urbanization of 13.0% was 3,280 thousand.

Compare now the distribution of the 10 largest cities of Western Europe in 1800 AD:

London: 948
Paris: 550
Naples: 430
Vienna: 247
Amsterdam: 217
Dublin: 200
Lisbom: 195
Berlin: 172
Madrid: 168
Rome: 153

Notice the similarity between this raking and my ranking of Roman city population? That's because I based my Roman ranking on the statistical properties of urban systems as well as in the direct evidence.

1) Rome: 1,250
2) Alexandria: 600
3) Antioch: 500
4) Carthage: 400
5) Ephesus: 350
6) Apamea: 200
7) Pergamon: 150
8) Lugdunum: 100
9) Jerusalem: 100
10) Cesarea: 100

There is also a strong correlation between the number of cities over 100,000 inhabitants and the total urban population in Europe:

date ------- cities ------- urban pop
1500 AD----- 4 ----------- 4.8
1650 AD----- 10 ---------- 8.5
1800 AD----- 18 ---------- 17.1

Zhao's numbers and these census numbers simply do not make sense when one understands the statistical properties of urban systems and the dynamics of pre-modern societies.

Considering the high degree of centralization of Han China as compared with Western Europe: 1 centralized empire in the first case, dozens of states in the second case, it is a consequence that the largest city of Han China probably had a larger proportion of the total urban population of the country and that the largest city was considerably larger than the second larger and third largest cities. The census data is incompatible with that.

Proportion of the urban population of WE in the largest city:

1500 AD - 4.7%
1650 AD - 4.7%
1800 AD - 5.5%

Clearly, the proportion of the urban population of Europe in the largest city usually was around 5% over the 300 year period. In Imperial China it was probably always larger (as in the Roman Empire, the city of Rome probably had 10% or more of the total urban population of the empire).

Some properties of Western Europe in 1800 AD:

1- Iron production of nearly 1 million tons, as compared with 5,000 - 6,000 tons for Han China.
2- Global power projection and colonial empires streching over 40% of the earth's surface.

However, if the census data is correct, then Han China would have an urban population of about 20 million. A rate of urbanization of over 30%.

And Han China did not reach a higher total absolute urban population as reached by Western Europe in 1800 AD, period.

Therefore the census doesn't represent the actual urban family population of these Chinese cities.

So, the difference between the population suggested by the number of families, 404,000 and the actual population measured, 246,000 is probably a product of two different measures. The first may be measuring the total number of families in a district containing Chang'an while the second may be measuring the number of people inside the urban area of the actual city. So 158,000 is perhaps the population of the countryside around the city. If other centers had similar rural populations one can reach a much more realistic distribution of the population among the 10 largest cities.
 
Last edited:
Feb 2011
6,113
#25
1- Iron production of nearly 1 million tons, as compared with 5,000 - 6,000 tons for Han China.
2- Global power projection and colonial empires streching over 40% of the earth's surface.
I have thoroughly disapproved that here: http://www.historum.com/asian-history/28104-han-dynasty-iron-production-2.html#post686990
It was a discussion in which you were a part of. Don't make me keep track of the number of times you would just copy/past things that have already been refuted.

However, if the census data is correct, then Han China would have an urban population of about 20 million. A rate of urbanization of over 30%.
The census data did not say that, not even Zhao's estimate approached near that number. You need to stop making things up. Even from your source, the upper tier cities of Europe in 1330 approaches the tenth largest Han city in the list provided. If the list of European 1330 AD European cities is in any way a reflection of the city distribution of the Han dynasty, then we could add the European/Chinese lists together for the total urban population of the Han dynasty. With a population of ~70 million for both cases, we would add the top ten Han cities which make up 4% of the population(as the upper tier), and the list of 1330AD European cities(assuming they match the size of Han cities of the bottom tier) which would make up ~5-6% of the total population. This results in ~10%, a very believable number. So, if the assumptions are true, then the Han dynasty would have ~10% of its population residing in cities with over 10,000 people. This is more than the urban ratio of Medieval Europe, but less than the urban ratio of Southern Europe.

Considering the high degree of centralization of Han China as compared with Western Europe: 1 centralized empire in the first case, dozens of states in the second case, it is a consequence that the largest city of Han China probably had a larger proportion of the total urban population of the country and that the largest city was considerably larger than the second larger and third largest cities. The census data is incompatible with that.
I also went over how high degrees of centralization does not equate with low urbanization or mercantile activities in post 22. There were plenty of nations whose capital cities were not drastically higher than the next one, and there are plenty of nations in which its capital cities far outweighs the rest of its urban rivals in population. Rome expanded as a city-state, so the empire would naturally have Rome as a huge metropolis. The Han dynasty was not. Centralization only means that the power was concentrated in the central government instead of local ones. This may just as well reflect higher urban density rather than lower ones. For example, local Han administrators do not have the power to raise taxes without central government approval. On the other hand a decentralized government could very well have local administrators upping taxes to their heart's content. The centralization of the Han dynasty also explains why Chinese dynasties tend to last significantly longer than Roman equivalents. It takes about 2 generations before Roman dynasties were overthrown by coups or military revolts.

So, the difference between the population suggested by the number of families, 404,000 and the actual population measured, 246,000 is probably a product of two different measures. The first may be measuring the total number of families in a district containing Chang'an while the second may be measuring the number of people inside the urban area of the actual city. So 158,000 is perhaps the population of the countryside around the city. If other centers had similar rural populations one can reach a much more realistic distribution of the population among the 10 largest cities.
What did I say about making numbers up? Do you not realize how small a Han district is? You can't simply just toss a number into the air as you see fit. So far you have misrepresented both Zhao and Rozman, repeatedly. It's getting ridiculous. I don't know if you're doing this on purpose, or you feel so strongly about one-uping the non-Westerners that you end up reading what you're trying to find, but it's getting old.

Now lets assume you're right: the population of the Han Chang'an was NOT measuring the population within the city walls, but rather the district population. We do not have the specific number of districts in the Han empire, but a district is part of a Han county. A Han country could have a dozen districts, and the Han empire has 1587 counties. With a population of 60 million people, each county on average would have 37,800 people on average. A district would have a dozen times less, so the chance that the true size of the Han cities and the size given by the census would only have a difference of ~3000 people. But lets go one step further. Lets assume the census was telling us about county population, which is several magnitudes higher than a district. A county has on average 37,800 people. So even if we subtract this number from the size of the list of ten Han cities, they would still all be well over 150,000. Instead of being 4.5% of the population, they would instead be 4% of the population. Note that 37,800 people is already subtracting an unreasonable amount, considering how this number includes urban residents.

Plus, if the discrepancy between Changan's recorded population of 240,000 and family population of 80,800(unlikely average family size of 3) is caused by the census measuring both district and population of people within the city walls, then how do you explain Yanling? Yanling has a recorded population of 261,418 with 49,101 families. That's a very reasonable family size of 5.3.

I have already made clear two other reasons why the Han census most likely did NOT county people outside the city. Here is the link: http://www.historum.com/asian-history/37128-historic-urban-population-china.html#post905011

I have given my reasons why their calculation of city family population is accurate, as it's hard to miscount the number of houses in a city by even 10 percent, especially considering that Han cities are planned in a grid. There are 160 residential wards in Chang'an, and it has a census population of 80,800 families. What's the chance that 80,800 is divisible by 160? 1 out of 160, so there's a 99.375% chance that this is no coincidence. If the Han census miscalculated, the chance that they were inaccurate by an exact multiplier of 160 is pretty low.


I also mentioned texts in which 4 million hu of grain/year where recorded to be shipped to Chang'an, and 6 million hu/year a hundred years later. Considering that the average person consumes ~ 360 liters of grain each year, 4 million hu would feed 222,222 people whereas 6 million hu would feed 333,333 people (~400,000 if you include the grain from the surrounding countryside). This matches well with the census data.

Neither of these was addressed. Also,the possibility of a miscount is very small, because Han cities are planned in a grid, separated by wards with specific number of houses per ward (probably no more than just 200 wards in the largest city). As long as they know multiplication, they're fine. The only real chance of a miscount is in counting the family size, as that depends on the honesty of the people involved. They may have also undercounted the people in the countryside, as those were not organized in a grid. But the chance of miscounting the number of families (equivalent to the number of houses) is slight.

Notice the similarity between this raking and my ranking of Roman city population? That's because I based my Roman ranking on the statistical properties of urban systems as well as in the direct evidence.
Comparing two urban ratios set 2000 years apart is not direct evidence, nor does it match any semblance of basic statistical standards. You are also misrepresenting the data. You chose 1800 as a representative for urban ratios 1000 years ago, yet what about the year 1200-1500? The population distribution for these years matches Han city distributions very well, yet you chose 1800. You provided the data, so you should know. Again, you choose to resort to sampling bias.

From Guaporense' own source, here is the list of European cities in 1330(measured by population size in thousands):
Granada---150
Paris------150
Venice----110
Genoa----100
Milan-----100
Florence---95
Seville-----90
Cordova---60
Naples----60
Cologne---54

Here is the number of families of 10 Han cities:
1) Chang'an: 80,800
2) Chengdu: 76,256
3) Maoling: 61,087
4) Luoyang: 52,839
5) Lu: 52,000
6) Changling: 50,057
7) Yanling: 49,101
8) Wan: 47,547
9) Yangdi: 41,650
10) Pengcheng: 40,196

So even though you have data showing cases where city distribution were more equalized, you pick a specific time period where it's not in order to prove the impossibility of such a scenario. This is despite the fact that the same source shows that such a thing is possible. This is selective sampling based on preconceived opinion.

Here is the size of Warring states cities from archeological excavations:
Xiadu: 32 square kilometers, 320,000 people (largest excavated city of the era)
Shouchun: 26.35 square kilometers, which gives 263,600 people
Jinan: 16.2 square kilometers, 162,000 people.
Linzi: 16 square kilometers, 160,000 people (though records indicate a family size of 70,000 and that it's impossible to walk the streets without brushing shoulders, so the people per hectare should be much higher)
Zhouyuan: 15 square kilometers, 150,000 people
Eastern Zheng city: 14.79 square kilometers, 147,900.
Yong city: 10 square kilometers, which gives 100,000 people
Yueyang city: 4 square kilometers. That gives 40,000 people
Yan: 3 square kilometers, 30,000 people

The population was found by measuring the walled area and extrapolating it with an urban density rate of 100 people per hectare(this is the rate that Guaporense insisted on). So as we can see, Warring States cities were large as compared to Greek ones. There are several cities with a 6 digit population, while for classical Greece the only city with a 6 digit number is Athens. It's reasonable to say that Warring States upper tier cities were larger than the classical Greek equivalent. Thus to have an average population of 50,000 for the top 36 cities sounds reasonable, for out of the 9 excavated cities shown above, 7 of them have 100,000 people or over.
 
Last edited:
Mar 2011
5,045
Brazil
#26
Greek Kingdoms & Chinese urbanization

Instead of comparing Greek City States with Chinese Kingdoms why not compare Greek Kingdoms with Chinese Kingdoms?

According to Hansen there were 24 cities located in the Kingdom of Epirus of which we know the walled areas of 18, summing up to 346 hectares, he estimated conservatively an urban population of 26,000, assuming the same average size for the other 6 cities comes to an urban total of 34,500. While it has been estimated that given the size of the territory and the likely population density that Epirus had 425,000 inhabitants. So we had a rate of urbanization for Epirus as 8.1%. Since many of these towns were under 2,000 inhabitants, the actual rate of urbanization may have been as low as 4-5%.

If we discount the walled areas of the cities among the 18 of which were too small to house 2,000 inhabitants confortably (ie. at densities lower than 120 per hectare) we get 262 hectares in 9 cities (a reduction of 25%, reducing the rate of urbanization by 25% yields 6.1%, nearly the same rate of urbanization as estimated by Rozman for China over 1,300 years).

For Macedonia we can apply the same logic and reach a rate of urbanization around 5%. Though we don't know the walled areas for more than 3 cities there, so we cannot do the type of estimate as performed for Epirus.
 
Aug 2013
42
morroco
#28
So Zhao is ridiculous to state 25% urbanization for Warring States period, but we should accept a more ridiculous figure (50-70%) for the Greeks and 40% for the Romans ? LOL.
This is the height of ridiculous bias.

any figure of urbanization that quotes population higher than 25% for a pre-industrialized society needs to be backed by precise data. Or else, it is simply cultural-superiorism masked as 'speculative analysis': a symptom most common amongst 'western' scholarship.
yes you are right, they pretend to have good arguments and they say that the roman rate of urbanization was 40%. they ignore the fact that in china cities houses where smaller so more people can live in a city than the Mediterranean standard, for example japan reached a rate of 17% during 17th century with edo reaching 1 million and osaka 500000 and kyoto 400000, in a time when the ming china was considered more urban.
 
#29
Those population data for Han cities actually represent the population of the entire county as Chinese didn't register urban population differently until song dynasty. According to the Book of Han, The entire Jingzhao commandery has a population of 682,468, in 195,702 households. The Jingzhao commandery has 12 counties. The county of Chang'an, has 80800 households, a population of 246,200. Due to the fact that large parts of Chang'an county are royal hunting fields and the fact this is the capital city of the empire, it's likely that most of the 246,200 people in Chang'an county were urban residents. However, this rule should not be allied to the other counties.
 
#30
While I do believe that Zhao's numbers are too high, Rozman estimate,however, is somewhat too low.
Urbanization rate of china was rising considerably during late Tang dynasty. Starting from Song dynasty, rulers began to realize the importance of urban population, and new government agencies were developed for the administration of urban population.

The urban administration system that can help us the most with urbanization rate estimation is the Lushisi(录事司) system. This system was first developed in Liao(Khitan) dynasty, matured in Jin(Jurchen) dynasty, and continue to be used in Yuan(Mongol) dynasty.
In Jin dynasty, all urban communities with more than 2000 households are to be separated to have one of the three levels of urban administrative agencies instead of ordinary county government. These agencies are Sihousi(司候司, cities with population of 2000+ households), Lushisi(录事司,major cities larger than Sihousi, population standard is unknown), Jingxunyuan(警巡院, the six capitals).

In total, Jin dynasty had 112 Sihousi cities, 66 Lushisi cities, and 6 Jingxunyuan cities. This means that the Jin empire had at least 184 cities with population of 10,000+.

We can estimate the population of permanent urban residents in the 6 capitals from the number of officials put in charge of population registration. The rule is to have 6 officials per 10,000 households as the base ratio, as the population of the city increases, the number of officials per 10,000 households decreases. So Huining(in northern Manchuria) and Dading(in southwestern Manchuria), with 6 population registration officials, both had urban population of 10,000 households. Liaoyang(in southeastern Manchuria), with 8 officials, probably had 14,000 households of urban popation. Datong(in northern China proper) and Kaifeng(in central China proper), with 10 officials, probably had more than 17,000~18,000 households of permanent urban residents. Peking, being the capital of the empire, had 36 population registration officials and probably had nearly 70,000 households of permanent urban residents.
Therefore, the total population of permanent urban residents in the six capitals is probably 140,000 households.

If we assume that the average population of Sihousi cities to be 3000 households, the average population of Lushisi cities to be 7000 households, then the total urban population of Jin empire adds up to 938,000 households. The height population of Jin empire, according to its own census data, is 8,413,164 households. This means an urbanization rate of 11.1 percent.

The rate goes even higher if we consider in the fact that the Jurchen aristocrats had many slaves under their households resulting in huge "families" with hundreds of people, and that the hundreds of thousands of soldiers stationed in the cities were also not included in local population census. The urbanization rate can go lower if you lower the average population estimate for the Sihousi and Lushisi cities, but I don't see much possibility of it going lower than 10%. Although southern Song also registered its urban population, we don't know the exact number. However, with a better and cheaper transportation system consisting large number of rivers and canals, Sothern song's urbanization rate is unlikely to be much lower,if not higher, than that of Jin.

In conclusion, China's urbanization rate in the 12th century is unlikely to be lower than 10%. If the cities with population of less than 10,000 are also counted, the rate is very likely to go into the 15%~20% range.
 
Last edited:

Similar History Discussions