CityReads│City Walls in Late Imperial China

108

City Walls in Late Imperial China

Cities with walls in the past have higher employment and population densities at the present time.

Yannis M. Ioannides, Junfu Zhang, Walled Cities in Late Imperial China, Journal of Urban Economics, Available online 10 November 2016, ISSN 0094-1190, .

Source: 

Picture source: A 17th-century painting showing the city wall of the Old City of Shanghai and the river port outside the wall, 

Archaeological evidence reveals that as early as over 4,000 years ago, human settlements in China were often surrounded by walls. Throughout the recorded history of China, major cities always had defensive walls. In the imperial period, the great majority of urban residents lived in walled cities. It is a surrounding wall that most Chinese people used to essentially distinguish a proper city from towns and villages.

There is no real city in Northern China without a surrounding wall, a condition which, indeed, is expressed by the fact that the Chinese use the same word ch’eng for a city and a city-wall: for there is no such thing as a city without a wall. It is just as inconceivable as a house without a roof (Sirén,1924, pp. 1-2).

City walls represented a most salient feature of Chinese cities throughout history until the mid-twentieth century, when the government sought to demolish city walls all over the country in the name of shaking off “the shackles of the past.” Today, complete city walls have been preserved for only a few Chinese cities, including Jingzhou, Pingyao, Xi’an, and Xingcheng. In most other cities, one can hardly see a trace of a city wall.

City walls in China were built primarily for defensive purposes. Typical city walls were thick enough to allow soldiers, horses, or even chariots to march on the top. They were usually fortified by adding battlements, towers, and barbican gates (see Figure 1). Earlier city walls were generally made of rammed earth only. Starting in the Ming Dynasty (1368-1644), it became a common practice to have city walls faced with bricks. Most Chinese cities had moats surrounding their city walls.

City walls in late imperial China function as an important window to understanding China’s urban development. Defense considerations are closely related to a city’s geographic location. Cities surrounded by a rich hinterland would naturally be preyed upon by bandits or disgruntled nobles. Coastal locations may be more vulnerable to incursions by foreigners, but their wealth may have generated envy by non-foreigners too. Cities in remote locations were critically important in the Chinese state’s ability to fend off invaders, and had to be able to provide for their sustenance during sustained sieges. 

City walls defined city life in many instances. The well-kept records of city walls have allowed researchers to use their physical size as a proxy for their populations in case where historical data on populations are unavailable. City walls were built to protect urban residents and properties from outside attacks, they served as physical and to some extent economic boundaries of cities. Therefore, the land area inside a city wall is a natural proxy for the size of the city.

The walled city model

In an article published in Journal of Urban Economic, Walled Cities in Late Imperial China, Yannis M. Ioannides and Junfu Zhang presented a model to explain the existence of city walls. It has five implications that can be explored empirically:

First, the size of the city, i.e., land area inside the city wall, is positively correlated with the size of (urban and rural) population.

Second, the quality of a city wall is increasing in the probability of being attacked.

Third, the distribution of city size is close to a power law.

Fourth, a fixed component of the cost of maintaining a city wall implies a lower bound in the distribution of city size.

Fifth, presence of a city wall in history implies a higher density of economic activity in the present time.

Data

The first data set was constructed by digitizing hand-collected information from a monumental work, the 130-chapter Important Notes on Reading the Geography Treatises in the Histories (Du Shi Fang Yu Ji Yao), written by Gu Zuyu (1631-1692), an early Qing Dynasty (1644-1911) scholar. In his book, Gu sought to cover the history and geography of all places in China in the late Ming Dynasty. We coded data from Gu Zuyu’s work on the circumference of city wall and population of the associated jurisdiction for 1,182 cities.

Table shows the average circumference of city wall and average jurisdiction population for cities by administrative level. Cities at higher levels tend to have longer city walls. Similarly, cities at higher levels tend to have larger jurisdiction population. The two capital cities are outliers in terms of city-wall circumference, obviously because the emperors could use resources from the whole empire to build them, not only for protective purposes but also to symbolize the grandeur of the empire.

Table Circumference of city walls and population sizes in Ming Dynasty

The second data set has been assembled by a group of researchers led by the anthropologist G. William Skinner (1925-2008). They hand-collected data on city walls for the late Qing Dynasty from more than 900 published gazetteers. Their data contain information on various dimensions of city walls for more than 1,600 cities.

In the following table, we present some descriptive statistics on city walls. The key variable, area inside city wall, is calculated by Skinner’s research team from the length of city wall assuming a typical square-shaped city. It has an average of 0.78 square kilometers. A few other variables are shown to help us envision the physical structure of a typical walled city in the late Qing Dynasty. The average city wall is 7.47 meters high and 7.26 meters thick at its base. It has 4 gates and 8 towers. 96 percent of cities also had moats surrounding their city walls.

Descriptive statistics for city walls in Qing dynasty

Based on the estimated area inside city wall, the ten largest cities in the late Qing Dynasty were Nanjing, Suzhou, Beijing, Xi’an, Hangzhou, Yulin, Quanzhou, Hefei, Dingzhou, and Taiyuan. Four of the ten cities (Beijing, Nanjing, Xi’an, and Hangzhou) were capital cities during different dynasties. Others were well-known in the Chinese history for their economic or military significance.

Model results

First, the size of city wall is always positively correlated with jurisdiction population. That is, when a larger population pays taxes to the local government, the local government tends to be located in a city with a longer city wall and thus it tends to have a larger urban area. We are using jurisdiction population, the sum of urban and rural population.

Second is about the relationship between wall size in Qing Dynasty and geographic fundamentals.

The distance to Beijing is negatively correlated with wall length. Greater distance from Beijing connotes lower labor productivity. The city’s distance to the Silk Road, a major medieval trade route that could proxy for a trade-related advantage, is negatively correlated with wall length but the estimates are not statistically significant.

The south of Yangtze River dummy has negative and highly significant coefficients, which reflects the fact that the southern part of China is more mountainous. With more geographic constraints, cities in the south of the Yangtze River were more difficult to grow, more costly to build, but easier to protect, and thus should need shorter defensive walls. Coastal cities tend to have longer walls, consistent with both higher productivity and greater need for protection.

Third is concerned with the quality of city walls. We use each city’s location relative to different frontiers to proxy the probability of being attacked. We divided cities into five different groups: on inner Asian frontiers: 38 cities; on southwestern frontiers: 26 cities; on maritime frontiers: 125 cities; on internal frontiers: 661 cities; not on any frontiers: 733 cities.

We use three alternative quality measures: height of the wall, number of towers on the wall, and thickness of the wall.

Our model suggests that controlling for city size, cities with a higher probability of being attacked by enemies would build walls of higher quality. Overall, we find that city walls on inner Asian frontiers are higher and thicker and that city walls on maritime and internal frontiers have more towers. We interpret these as evidence that cities facing higher risks built better city walls. City walls on southwestern frontiers are inferior in every respect: They are lower, thinner, and have fewer towers. We think this is because they were in mountainous areas and were unlikely to be attacked by enemies or bandits.

Walled cities in Qing Dynasty and different frontiers

Fourth is about the physical size distribution of walled cities.

Size and log size densities are very similar to those plotted for both Ming and Qing Dynasty. Although the sample size of walled cities increased from 1,182 to 1,623 from the Ming to the Qing Dynasty, the overall city size distribution appears to be stable. The density of log city sizes looks remarkably close to a normal distribution.

Density of city sizes in Qing and Ming dynasties 

For both periods, we find evidence that above a certain size cutoff the physical size of walled cities indeed follows a power law.

The rank-size distribution of walled cities in the Ming and Qing Dynasty is fairly close to Zipf’s law.

Long run effects of city walls on modern Chinese cities

We proceed to examine empirically whether the presence of medieval walls, even if they have been torn down, might affect the density of economic activity in Chinese cities at present. We thus examine whether there is a significant difference in density of employment or population between cities that had walls in the Qing Dynasty and those that did not.

Cities that did have walls have richer histories, where for a long time urban economic activities have found it inescapable to operate within the confines of those walls. Many areas inside the walls had established a tradition for certain types of uses. Urban planners tend to respect these traditions in order to maximize the value of urban land within walls and thus build densely in these city centers even if the walls are no longer standing up. This type of path dependence is demonstrated aptly by the example of Beijing. Within one mile of Tiananmen Square, Beijing’s historical center, there are three highly popular, densely built shopping districts (Qianmen, Wangfujiang, and Xidan).

We report empirical results using contemporary data for 288 prefectural level cities for which presence or not of walls in the Qing Dynasty is known. We show that a higher density of economic activity, measured by population and employment density in 1984 and 2013, is associated with existence of walls in the past.

The city wall dummy has a positive and statistically very significant coefficient. Despite the fact that city walls were torn down many decades ago, cities that did have walls in history tend to have higher densities of economic activities today. This is a powerful demonstration of the lingering effect of a city wall as a man-made amenity that defines the historical boundary of a contemporary city’s historical core and induces higher density. Presence of city walls in medieval times has a lingering effect in population and employment density in Chinese cities at the present time.

Data on population in 208 Chinese prefecture-level cities in Ming Dynasty and in 2010. Raw correlation of log population (all cities): 0.39.

source: 

Related CityReads

（Click the title or enter our WeChat menu and reply number ）

CityReads ∣Notes On Cities

"CityReads", a subscription account on WeChat, 

posts our notes on city reads weekly. 

Please follow us by searching "CityReads"  

Or long press the QR code  above