Journal of Public Economics, Год журнала: 2020, Номер 188, С. 104220 - 104220
Опубликована: Июнь 12, 2020
Язык: Английский
American Economic Review, Год журнала: 2021, Номер 111(3), С. 831 - 870
Опубликована: Фев. 25, 2021
Does the consumption of ideologically congruent news on social media exacerbate polarization? I estimate effects exposure by conducting a large field experiment randomly offering participants subscriptions to conservative or liberal outlets Facebook. collect data causal chain effects: outlets, Facebook, visits online sites, and sharing posts, as well changes in political opinions attitudes. Four main findings emerge. First, random variation substantially affects slant sites that individuals visit. Second, counter-attitudinal decreases negative attitudes toward opposing party. Third, contrast effect attitudes, find no evidence leanings affect opinions. Fourth, Facebook’s algorithm is less likely supply with posts from conditional subscribing them. Together, results suggest algorithms may limit thus increase polarization. (JEL C93, D72, L82)
Язык: Английский
Процитировано
425
Annual Review of Economics, Год журнала: 2020, Номер 12(1), С. 415 - 438
Опубликована: Май 26, 2020
How do the Internet and social media affect political outcomes? We review empirical evidence from recent economy literature, focusing primarily on work that considers traits distinguish traditional off-line media, such as low barriers to entry reliance user-generated content. discuss main results about effects of in general, particular, voting, street protests, attitudes toward government, polarization, xenophobia, politicians’ behavior. also role dissemination fake news, we summarize strategies employed by autocratic regimes censor use for surveillance propaganda. conclude highlighting open questions how shape politics democracies autocracies.
Язык: Английский
Процитировано
400
Journal of Economic Literature, Год журнала: 2023, Номер 61(1), С. 3 - 40
Опубликована: Март 1, 2023
Information provision experiments allow researchers to test economic theories and answer policy-relevant questions by varying the information set available respondents. We survey emerging literature using in economics discuss applications macroeconomics, finance, political economy, public economics, labor health economics. also design considerations provide best-practice recommendations on how (i) measure beliefs; (ii) intervention; (iii) belief updating; (iv) deal with potential confounds, such as experimenter demand effects; (v) recruit respondents online panels. finally typical effect sizes sample size recommendations.(JEL C90, D83, D91)
Язык: Английский
Процитировано
292
Опубликована: Июнь 1, 2020
Media outlets often present diverging, even conflicting, perspectives on reality -not only informing, but potentially misinforming audiences.We study the extent to which misinformation broadcast mass media at early stages of coronavirus pandemic influenced health outcomes.We first document large differences in content between two most popular cable news shows US, both same network, and adoption preventative behaviors among viewers these shows.Through a selection-on-observables strategy an instrumental variable approach, we find that areas with greater exposure show downplaying threat COVID-19 experienced number cases deaths.We assess magnitudes through epidemiological model highlighting role externalities provide evidence contemporaneous information is key underlying mechanism.
Язык: Английский
Процитировано
255
The Journal of Economic Perspectives, Год журнала: 2017, Номер 31(1), С. 117 - 140
Опубликована: Янв. 30, 2017
In this paper, we document basic facts regarding public debates about controversial political issues on Chinese social media. Our documentation is based a dataset of 13.2 billion blog posts published Sina Weibo—the most prominent microblogging platform—during the 2009–2013 period. primary finding that shockingly large number highly sensitive topics were and circulated For instance, find millions discussing protests, these are informative in predicting occurrence specific events. We an even larger with explicit corruption allegations, predict future charges individuals. findings challenge popular view authoritarian regime would relentlessly censor or ban Instead, interaction government media seems more complex.
Язык: Английский
Процитировано
252
Science Advances, Год журнала: 2022, Номер 8(10)
Опубликована: Март 9, 2022
Repression research examines the causes and consequences of actions or policies that are meant to, actually do, raise costs activism, protest, and/or social movement activity. The rise digital media has brought substantial increases in attention to repression activists movements use tools repression, which is spread across many disciplines areas study. We organize review this growing welter under concept by expanding a typology distinguishes based on actor type, whether overt covert, behaviors shaped coercion channeling. This delineation between broadly different forms allows researchers develop expectations about better understand what "new" terms explanatory factors, repression.
Язык: Английский
Процитировано
87
American Economic Review, Год журнала: 2024, Номер 114(3), С. 815 - 850
Опубликована: Фев. 29, 2024
We conducted a nationwide field experiment in China to evaluate the direct and indirect impacts of assigning firms public or private citizen appeals when they violate pollution standards. There are three main findings. First, regulator through social media substantially reduce violations emissions, while cause more modest environmental improvements. Second, appear tilt regulators’ focus away from facilitating economic growth toward avoiding pollution-induced unrest. Third, reductions by treated not offset control firms, based on randomly varying proportion at prefecture level. (JEL D22, L82, P28, P31, Q52, Q53, Q58)
Язык: Английский
Процитировано
64
The Quarterly Journal of Economics, Год журнала: 2023, Номер 138(3), С. 1349 - 1402
Опубликована: Март 13, 2023
Abstract Recent scholarship has suggested that artificial intelligence (AI) technology and autocratic regimes may be mutually reinforcing. We test for a reinforcing relationship in the context of facial-recognition AI China. To do so, we gather comprehensive data on firms government procurement contracts, as well social unrest across China since early 2010s. first show autocrats benefit from AI: local leads to greater new political control, increased indeed suppresses subsequent unrest. innovation benefits autocrats’ suppression unrest: contracted innovate more both commercial markets are likely export their products; noncontracted not experience detectable negative spillovers. Taken together, these results suggest possibility sustained under Chinese regime: entrenches regime, regime’s investment control stimulates further frontier innovation.
Язык: Английский
Процитировано
48
Journal of the European Economic Association, Год журнала: 2023, Номер 22(3), С. 1495 - 1539
Опубликована: Окт. 11, 2023
Abstract We study how social media affects election outcomes in the United States. use variation number of Twitter users across counties induced by early adopters at 2007 South Southwest festival, a key event Twitter’s rise to popularity. show that this is unrelated observable county characteristics and electoral before launch Twitter. Our results indicate lowered Republican vote share 2016 2020 presidential elections, but had limited effects on Congressional elections previous elections. Evidence from survey data, primary text analysis millions tweets suggests relatively liberal content may have persuaded voters with moderate views against Donald Trump.
Язык: Английский
Процитировано
48
American Economic Review, Год журнала: 2024, Номер 114(4), С. 887 - 925
Опубликована: Март 28, 2024
To investigate general patterns in news information the United States, we combine a protocol for identifying major political stories, 11 monthly surveys with 15,000 participants, and model of discernment. When confronted true fake story, 47 percent subjects confidently choose 3 remaining half are uncertain. Socioeconomic differences associated large variations probability selecting story. Partisan congruence between an individual story matters, but its impact is up to order magnitude smaller. (JEL D72, D83, L82)
Язык: Английский
Процитировано
17