When rural medical clinics in Pakistan started transmitting real-time performance data to policymakers, doctor attendance and inspections increased.
By Krysten Crawford
Jennifer Pan was a second-year PhD student in political science when she stumbled on a data gold mine. She and a small team were gathering millions of Chinese social media posts for an algorithm they were developing that could analyze Chinese language text. Along the way, they noticed that the links to hundreds of thousands of the posts no longer worked.
"We realized the Chinese government had censored the posts," recalls Pan, now an assistant professor of communication and faculty affiliate with the Stanford King Center on Global Development (King Center). "Since we had collected the content prior to censorship, we saw an opportunity to figure out how the Chinese government was using digital technology to try to suppress dissent.”
For Pan, the moment was serendipitous. Soon, the series of pro-democracy uprisings in the Middle East known as the Arab Spring would highlight the growing role of social media in politics. In the decade since, governments worldwide have been looking to use social media and other digital tools to their advantage — most notably authoritarian regimes like Russia, which has attempted to influence foreign elections, and China, which blocks Google.com and Facebook.
Pan’s research into how non-democratic governments are leveraging online services goes well beyond the headlines. In the last decade, she has focused largely on ways in which China and Russia are attempting to leverage technology to control internal political discourse and prevent popular unrest. Now, with support from the King Center, Pan is developing new, data-driven methods for tracking social media to identify nascent political movements happening offline.
That work contributed to a working paper, in which Pan looks at the impact that physical repression can have on online dissent. Looking at Saudi Arabia’s history of arresting dissidents and Muslim clerics through 2017, Pan finds that the attempted crackdowns silenced those who were detained, but led to more online criticism of the government.
It's just one example of how political scientists like Pan are taking a more interdisciplinary approach to their research by applying traditional theory, advanced statistics, machine learning, economic analyses and other methods to large datasets. In Pan’s case, her approach is raising awareness of how authoritarian regimes in the digital age are pursuing policies that either bolster or inhibit economic development — with potentially far-reaching consequences for their overall stability.
"There is a great deal of opportunity now thanks to large-scale data," says Pan, "for social scientists to understand phenomena that we were previously unable to see."
Pan's preoccupation with authoritarian governments — and China's in particular — stems in part from her upbringing. Born in the southwest province of Sichuan ("known for pandas and spicy food"), Pan's parents came to the United States in the mid-1980s. She lived for two years with her grandmother in China before she was able to join them in Bloomington, Indiana.
After graduating from Princeton in 2004, Pan worked at the Chinese Center for Disease Control and then as a consultant for four years with McKinsey & Company. In 2015, she earned her PhD in government from Harvard. She joined the Stanford faculty shortly after.
Pan committed to a career in academia after discovering a passion for research. "The process of learning something about the world — whether it's trivial or leads to something broader — is incredibly exciting for me," she says.
She felt that thrill when she discovered the censored Chinese social media posts in the summer of 2010. At the same time, she wasn’t buying that the Arab Spring was the “Twitter Revolution,” as pundits were calling it.
"There was no empirical evidence at all that social media leads to democratization," she says.
Her skepticism, combined with a thirst for knowledge grounded in hard evidence, spurred her interest in collective action in authoritarian societies — and the role that social media plays both in galvanizing popular unrest and tamping it down.
"Collective action as a way of engaging in the political process is important in any society," says Pan. "But in non-democratic countries where people can't meaningfully express their preferences by voting, it is one of the most effective ways ordinary people can voice their preferences and participate in politics."
Big data and machine learning, she says, has allowed social scientists like her to design better experiments and observe how people connect, interact, and communicate in unprecedented ways.
Pan has looked at how China and Russia use fabricated social media as subtle manipulation tactics to distract the public instead of confronting skeptics and controversial issues. She has also shown that local politicians in China use the Internet as a vehicle for self-promotion. And she finds that authoritarian governments that adopt China's online censorship strategies are unlikely to succeed.
She has shown how efforts to use online tactics to monitor popular sentiment at the local level can backfire — giving central authorities unreliable information that paints a distorted picture of what's happening on the ground. In a study published earlier this year in American Political Science Review, she shows how lower-ranking officials in China conceal public complaints of corruption posted online even though senior-level leaders want to use these public complaints to keep tabs on local governments.
Offline strategies of social control remain relevant, says Pan, who is writing a book about dysfunction within China’s largest social assistance program.
"The Chinese government's preoccupation with political order and control seeps into many domains. This area has many opportunities for research, and it’s one where large-scale data can play an important role."