Periodic Reporting for period 3 - MEDIACHINA (Social Media and Traditional Media in China: Political and Economic Effects)
Reporting period: 2021-01-01 to 2022-06-30
New media are rapidly increasing the amount of information available to citizens and leaders in many autocracies, including China. The consequences for political accountability and regime stability are unclear. The new media can be used to spread information and organize opposition against authoritarian governments. However, autocrats may censor the information, or even use it for surveillance to solidify regime stability. China is leading this development, both because of the enormous popularity of social media and because of their adoption of state-of-the-art censoring and surveillance technologies. The development in China is of huge importance in its own right. However, social media effects in China today may also point to the direction of social-media effects in other non-democracies approaching this juncture in the near future. For example, within a decade or so, several countries with authoritarian governments in Africa may have reached similar level of social-media penetration and technological sophistication as China has today.
This project will analyse how economic and political outcomes in China are affected by traditional and social media. It will also use content in these media to measure factors that are otherwise difficult to observe, such as political networks and the trade-off between political and economic goals in Chinese firms. A major impetus for this project is that an explosion of social media use in China has produced an information shock to society and its leaders, also supplying a data shock to researchers, which is magnified by the digitization of traditional media content and coupled with new methods for analysing this type of data, originating from the big-data and machine-learning revolutions. As a result, a large set of previously unanswerable questions are now open for research.
The project consists of several modules, listed in Section 4 of the DOA. The four main modules are (i) “strikes and protests”, investigating the effect of social media on these outcomes; (ii) “bad drugs”, the effect of social media on drug quality (social media may affect this market because it provides information that helps consumers and regulators identify the bad drugs), (iii) “Weibo effects on other information providers”, how social media affects reporting by traditional media; (iv) “Firms with dual politico-economic goals”, how the bias in traditional media is affected by competition and other characteristics.
In addition, the DOA lists four tentative modules: (v) “Firm accountability: worker safety and pollution”, how social media affects worker safety and pollution, (vi), “Innovation: political incentives and ICT”, how political incentives and ICT affect innovation; (vii) “Predictive modelling”, predicting social and economic problems, such as bad drugs; (viii) “Political connections”, using newspaper data to uncover political connections.
This project will analyse how economic and political outcomes in China are affected by traditional and social media. It will also use content in these media to measure factors that are otherwise difficult to observe, such as political networks and the trade-off between political and economic goals in Chinese firms. A major impetus for this project is that an explosion of social media use in China has produced an information shock to society and its leaders, also supplying a data shock to researchers, which is magnified by the digitization of traditional media content and coupled with new methods for analysing this type of data, originating from the big-data and machine-learning revolutions. As a result, a large set of previously unanswerable questions are now open for research.
The project consists of several modules, listed in Section 4 of the DOA. The four main modules are (i) “strikes and protests”, investigating the effect of social media on these outcomes; (ii) “bad drugs”, the effect of social media on drug quality (social media may affect this market because it provides information that helps consumers and regulators identify the bad drugs), (iii) “Weibo effects on other information providers”, how social media affects reporting by traditional media; (iv) “Firms with dual politico-economic goals”, how the bias in traditional media is affected by competition and other characteristics.
In addition, the DOA lists four tentative modules: (v) “Firm accountability: worker safety and pollution”, how social media affects worker safety and pollution, (vi), “Innovation: political incentives and ICT”, how political incentives and ICT affect innovation; (vii) “Predictive modelling”, predicting social and economic problems, such as bad drugs; (viii) “Political connections”, using newspaper data to uncover political connections.
Of the four main modules, the project has worked primarily on i, ii and iv.
Module iv: “Firms with dual politico-economic goals”. The work in area iv is published in Qin, Strömberg and Wu (2018). This paper examines whether and how market competition affected the political bias of government-owned newspapers in China from 1981 to 2011. We measure media bias based on coverage of government mouthpiece content (propaganda) relative to commercial content. We first find that a reform that forced newspaper exits (reduced competition) affected media bias by increasing product specialization, with some papers focusing on propaganda and others on commercial content. Second, lower-level governments produce less-biased content and launch commercial newspapers earlier, eroding higher-level governments' political goals. Third, bottom-up competition intensifies the politico-economic trade-off, leading to product proliferation and less audience exposure to propaganda.
Module (i) “Strikes and protests”. The work is almost complete, and results are summarized in a working paper (Qin, Strömberg and Wu, 2019), with planned submission the fall of 2020. This paper studies whether the explosive growth of social media in China affects the spread and incidence of protests. We combine a unique dataset of 13.2 billion microblog posts published during 2009-2013 with detailed information on thousands of protests and strikes during 2006-2017. We use retweets to measure the network of social media information flows across cities, and estimate the effects of this rapidly expanding network. Despite the strict media control in China and the lack of information for explicit coordination, we find that the social media network has a sizeable and significant effect on the spread of both protests and strikes. The spread of events over social media is fast and predominantly local -- between events within the same category (e.g. cause and industry); event spread across categories is still significant, albeit weaker. Furthermore, we find that social media networks increase the incidence of protests and strikes. These findings shed light on the recent debate regarding the political role of social media in autocracies.
Module (ii) “Bad drugs”. The work is well under way. The relevant data has been assembled and most of the analysis is done. We have a first draft of the paper, but it is not quite ready for circulation. Preliminary results suggest that the number of bad drugs is decreasing in Sina Weibo use. Consistent with the prediction of a simple moral-hazard model, we find that the reduction of bad drugs is driven by two mechanisms: Sina Weibo induces more effort from the Drug Administration and it deters the production of bad drugs. Finally, the diffusion of Sina Weibo seem to have a higher marginal effect for disadvantaged groups, consistent with microblogging being a cheap, accessible media.
Module (iii). “Weibo effects on other information providers”. Data has been collected. Analysis has yet to be started.
Of the four tentative modules, the project has worked on modules vi and viii.
Area vi: Innovation: political incentives and ICT. The data collection for this subproject is almost complete and analysis has started. The patenting data has been updated and complemented with additional geographical information. The data on politicians’ careers has been updated and complemented and is almost complete.
Area viii. Political connections. Data on coverage of politicians has been collected. Analysis has started.
Module iv: “Firms with dual politico-economic goals”. The work in area iv is published in Qin, Strömberg and Wu (2018). This paper examines whether and how market competition affected the political bias of government-owned newspapers in China from 1981 to 2011. We measure media bias based on coverage of government mouthpiece content (propaganda) relative to commercial content. We first find that a reform that forced newspaper exits (reduced competition) affected media bias by increasing product specialization, with some papers focusing on propaganda and others on commercial content. Second, lower-level governments produce less-biased content and launch commercial newspapers earlier, eroding higher-level governments' political goals. Third, bottom-up competition intensifies the politico-economic trade-off, leading to product proliferation and less audience exposure to propaganda.
Module (i) “Strikes and protests”. The work is almost complete, and results are summarized in a working paper (Qin, Strömberg and Wu, 2019), with planned submission the fall of 2020. This paper studies whether the explosive growth of social media in China affects the spread and incidence of protests. We combine a unique dataset of 13.2 billion microblog posts published during 2009-2013 with detailed information on thousands of protests and strikes during 2006-2017. We use retweets to measure the network of social media information flows across cities, and estimate the effects of this rapidly expanding network. Despite the strict media control in China and the lack of information for explicit coordination, we find that the social media network has a sizeable and significant effect on the spread of both protests and strikes. The spread of events over social media is fast and predominantly local -- between events within the same category (e.g. cause and industry); event spread across categories is still significant, albeit weaker. Furthermore, we find that social media networks increase the incidence of protests and strikes. These findings shed light on the recent debate regarding the political role of social media in autocracies.
Module (ii) “Bad drugs”. The work is well under way. The relevant data has been assembled and most of the analysis is done. We have a first draft of the paper, but it is not quite ready for circulation. Preliminary results suggest that the number of bad drugs is decreasing in Sina Weibo use. Consistent with the prediction of a simple moral-hazard model, we find that the reduction of bad drugs is driven by two mechanisms: Sina Weibo induces more effort from the Drug Administration and it deters the production of bad drugs. Finally, the diffusion of Sina Weibo seem to have a higher marginal effect for disadvantaged groups, consistent with microblogging being a cheap, accessible media.
Module (iii). “Weibo effects on other information providers”. Data has been collected. Analysis has yet to be started.
Of the four tentative modules, the project has worked on modules vi and viii.
Area vi: Innovation: political incentives and ICT. The data collection for this subproject is almost complete and analysis has started. The patenting data has been updated and complemented with additional geographical information. The data on politicians’ careers has been updated and complemented and is almost complete.
Area viii. Political connections. Data on coverage of politicians has been collected. Analysis has started.
Module iv: “Firms with dual politico-economic goals”. This paper goes beyond the state of the art by demonstrating that, even in a highly controlled environment such as China, media bias is affected by a trade-off between political and economic goals. As long as the system creates enough incentives for media owners to pursue economic benefits, the cost of manufacturing media bias will escalate with market competition. We create a novel measure of media bias in the Chinese context, based on coverage of government mouthpiece content (propaganda) relative to commercial content. We find that a reform that forced newspapers to exit (reduced competition) affected media bias. Moreover, lower-level governments produce less biased content and launch commercial newspapers earlier, eroding higher-level governments’ political goals.
Module i: “Strikes and protests”. Work in this module goes beyond the state of the art in the following ways. It documents the spread of information about protests and strike on social media in China and showing that this impacted the spread of real-world events. It uses retweets to measure the network of social media information flows across cities and to estimate the effects of this rapidly expanding network. It shows that the social media network has a sizeable and significant effect on the spread of both protests and strikes.
For the other subprojects, expected results are as described in the DOA.
Module i: “Strikes and protests”. Work in this module goes beyond the state of the art in the following ways. It documents the spread of information about protests and strike on social media in China and showing that this impacted the spread of real-world events. It uses retweets to measure the network of social media information flows across cities and to estimate the effects of this rapidly expanding network. It shows that the social media network has a sizeable and significant effect on the spread of both protests and strikes.
For the other subprojects, expected results are as described in the DOA.