Selected Publications

We have systematically explored the patterns, trends, and main characteristics of China’s ecommerce by relying on data on JingDong’s online sales. One of the most interesting findings is that China’s online consumer spending is positively correlated with regional income. In addition, people’s online consumer spending behavior exhibits regional heterogeneity and age cohort heterogeneity. People in the east region exhibit the strongest online consumer spending capacity. Finally, the most popular products sold online at JD are cell phones, followed by food and beverages, makeup and cosmetics, digital products, and lifestyle and travel goods.
Mandel and Haughwout - Handbook of U.S. Consumer Economics, 2018

This paper explores how power distribution and dynamics among national leaders shape factional competition among lower level bureaucrats under authoritarian regime, using novel data from China. Employing news reports in Chinese national and local newspapers (2000-2014) coupled with elite network data, we find that bureaucrats linked to strong national leaders tend to target the members of weak factions in polit- ically damaging news reports. Our analysis shows that strong factions publish more frequently on corruption investigations ongoing in provinces ruled by smaller factions. These negative reports indeed harm the promotion prospects of reported-on province leaders, weakening the already weak factions and expanding relative power of strong factions. Our findings suggest that elite competition in China has a tendency of power concentration on smaller number of factions. They also imply that the recent trend of power personalization in China may not be a sudden phenomenon but reveals latent patterns of Chinese elites’ behavior.
Working Paper, 2017

How does the media help hold politicians accountable in countries with competitive elections has been well studied. However, whether and how media could discipline non- elected officials remain unclear. In this paper, I examine the effect of negative news coverage on political selection in China. To be specific, I evaluate how news reports on fatal coal mine accidents influence prefectural party secretaries’ likelihood of promotion. I find media coverage, rather than the coal mine accident itself, decreases the possibility of politicians’ promotion. This effect is more substantial when the media is non-local. Our results suggest that even with serious media control, there is publicity-induced accountability in China. The accountability-enhancing effect of media is driven by informing the public rather than informing the principle. I also propose that the M-form hierarchy of bureaucracy and economic decentralization create considerable discretion for media, which although limited, can hold politicians accountable.
Working Paper, 2017



The R Package of Panel Data Approach for Program Evaluation.

Invited & Upcoming Talks

Innovative Uses of Data and Methods in the Study of China
SDSI Data Science Lab
2019 Asian Meeting of the Econometric Society
The Data Analytics Boot Camp
Annual Conference of Chinese Association of Quantitative Economics


  • 858 405 2327
  • Social Sciences Building, Room 351, UCSD, California, 92093, USA
  • Wednesday 9:00 to 10:50 or email for appointment