Wiring China: The Impact of Telegraph Construction on Grain Market Integration in Late Imperial China, 1870-1911

This blog is based on the authors’ article which has been published on Early View:https://onlinelibrary.wiley.com/doi/10.1111/ehr.13119

by Yu Hao (Peking University), Li Yuanzhe (Peking University), John V.C. Nye (George Mason University)

Although scholars have studied the role of modern transportation and the impact of opening the treaty ports on Chinese trade in the nineteenth century, their role in the significant expansion of Chinese commerce in the last quarter of that century was limited.  Railways did not appear until 1895, and only 20 per cent of China’s prefectures had access to railways by 1911.  Steamships were limited to coastal regions and along the Middle and Lower Yangzi River.  The growth in foreign trade was largely concentrated on the coastal areas surrounding the treaty ports.  The network that really mattered – and which has been overlooked – was the telegraph.

Telegraphs were of considerable importance to inland areas, especially those that had limited or no access to waterways.  We study the effects of this new technology on the Chinese economy by measuring how grain market price integration was facilitated by the spread of the telegraph network.  For most areas reliant on traditional courier routes and waterways,  available .communication was slow, cumbersome, and costly.

Because of its importance to the military, especially as a defence against foreign invasion, telegraph construction was heavily promoted by the Qing government from the 1880s.   Its spread was so rapid that 34.4 per cent and 65.5 per cent of prefectures were connected to the network by 1890 and 1911, respectively.  This rapid dissemination of the telegraph system encouraged  long-distance commercial traders to adjust their business decisions which affected  price integration throughout  China.

We exploit the exogenous variation in telegraph construction to identify the telegraph’s impact on prices, adjusting for various geographic features and weather.  Our difference-in-difference (DID) model allows for construction of a prefecture-pair level dataset that has data on rice prices, time of telegraph construction, geographic features, natural disasters, and other characteristics derived from primary and secondary sources.  Our results indicate that the introduction of the telegraph reduced monthly price differences in the rice market by about 20 per cent, on average.  These results are robust to the inherent problems of unobservable characteristics, controlling for treaty ports, and other large technological changes (railways and steamships) during the same period.

Our main findings may be summarised as follows.  First, the effects of the telegraph were much stronger for prefecture-pairs where neither area had access to waterways and for those prefectures featuring rugged terrain (and hence, more costly courier service).  Second, we find that the telegraph reduced price levels when areas were hit by floods and drought, mitigating the effect of grain market fluctuations due to harvest failure.  We also show that the telegraph enhanced the commercial power of dominant traders and their existing networks.  The telegraph’s impact was greater for a prefecture from which a long-distance trader group originated and where a long-distance trader clubhouse was located.  Hence, traditional traders played an important role in promoting market integration.  Finally, we show that by 1910,  prefectures that were earliest  connected to the telegraph had greater population density, higher urbanization rates, and higher levels of commercialization.

Our paper provides the first empirical results on how the spread of modern information technology in late imperial China improved market integration.  Our findings are consistent with research  on the adoption of mobile phones and price integration in Kerala, India between 1997 and 2001 (Jensen, 2007); Aker’s (2010) work on mobile phones in rural Niger between 2001 and 2006; and also Steinwender’s (2018) findings that transatlantic cotton price differences fell after the construction of the transatlantic telegraph line in 1866.

Whereas previous studies have mostly relied on a small number of markets, small regions, or limited time periods, our study uses price data for 114 prefectures and nine provinces in China, over 504 months.  This comprehensive dataset permits the analysis of social welfare and development, especially in comparison to previous work on transportation breakthroughs and forced trade openness.  We have provided new information on the heterogenous benefits enjoyed by traditional traders who leveraged this new technology for their own profit, thereby accelerating market integration.

To contact the authors:

Yu Hao: maxhao1003@pku.edu.cn,

Li Yuanzhe: liyuanzhe@pku.edu.cn,

John V.C. Nye: jnye@gmu.edu