We explain this puzzle by focusing on citizens’ lack of knowledge regarding the repressive nature of digital surveillance in dictatorships, which can be attributed to (1) invisible and targeted repression associated with digital surveillance and (2) government propaganda and censorship further concealing its repressive potential.

A field survey experiment on 750 college students in 3 Chinese regions shows that revealing the SCS’s repressive potential substantially reduces support for the system, but emphasizing its social-order-maintenance function does not increase support. Observational evidence from the field survey and a nationwide survey of 2,028 Chinese netizens show that the support is higher if citizens knew about the SCS through state media.

Our findings highlight the role of information and framing in shaping public opinion on digital surveillance.

[Keywords: social credit, surveillance, information, public opinion, repression]

…The results show that reminding respondents of the SCS’s role in maintaining social order does not change their support for the SCS much, but revealing information about the SCS’s role in political control largely reduces respondents’ support for the SCS. Given that the average level of support is 7.5 (scale of 0–10), the repression information treatment substantially reduces individuals’ support by 12%.

…We identify 180 less informed respondents and 557 more informed respondents, and then estimate equation (1) on these two subsamples. Figure 4 shows that the repression information treatment has a larger effect among less informed respondents. The findings suggest that information about repression poses a greater shock to less informed respondents, which provides further evidence for our information argument.

…As predicted, a 1 SD increase in respondents’ reliance on state media for information about the SCS increases support by 0.22 SD, and the effect is statistically-significant even after we control for a number of covariates. This strong positive effect provides evidence consistent with the theoretical argument.

…Tendency to avoid low-score peers: An interesting finding is a positive relationship between individuals’ changing attitude toward friends with bad credits and support for the SCS (Figure 6). Figure 7 shows that, among 2,028 respondents, 62% of them will either look at the friend differently or hesitate to hold a positive attitude. Figure 6 shows that a 1 SD increase in this measure increases support for SCSs by 0.18 SD, and the effect is statistically-significant.

…We obtain 50 CCTV news reports (data 2003^–₁₅2018), 410 articles from People.cn (the online platform of People’s Daily), and 186 articles from the Global Times. We use human-coded sentiment analysis to identify the tone of the articles (Table 2). We find that only 2.8% of articles are negative. The rest of the articles either praise the SCS’s trust-building and social-order-maintenance functions (positive) or simply present facts about the SCS to the general public (neutral). Among the 16 negative articles (excluding 2 identical articles reported by different outlets), 11 articles express concerns over local governments’ overdoing of SCSs’ social-order-maintenance function (eg. punishing jaywalking, unpaid parking fees, and frequent job turnovers), 3 articles raise privacy concerns, 1 article mentions the lack of remedies for people in social credit blacklists, and 1 Global Times article actually defends the SCS against Western criticism. Among the 11 articles concerning local governments’ overdoing of SCSs, only 1 article mentioned a phase “credit deduction for illegal petitioning [闹访、缠访扣分]” that is related to political repression. This phrase is barely noticeable, as the article mainly talks about local governments’ overdoing of SCSs’ social-order-maintenance function. The evidence supports our assumption that Chinese state media discuss SCSs in a very positive way and avoid revealing their political repression function. Even in the 2.8% of articles in which a negative tone can be detected, strictly speaking, only one article has one sentence that can be related to political repression.

See Also: