DATA DRIVEN AUTHORITARIANISM A CASE STUDY OF THE CHINESE SOCIAL CREDIT SYSTEM by Dalton Rupp An undergraduate thesis submitted in partial completion of the Metropolitan State University of Denver Honors Program May 2019 Richard Moeller , Ph.D. Megan Hughes Zarzo, Ph.D. Primary Advisor Honors Program Director
1 Part I. Personal Data The world of information we engage with is now digital. The rapid expansion of information systems and their data gather ing capabilities have created an entirely new domain of human interaction . Financial institutions, government agencies, technology companies which are now some of the largest and most influ ential organizations in history, 1 all of these different bodies have partial access in to the lives of Americans through the collection of personal data. inancial status, criminal history, phone and e mail records, friendships, professional relations, favorite activities, f requently visited places, and current or past location s are all categories of data that are translated and stored by various state and non state bodies . 2 F or the most part, people recognize that giving up their control over their own information is a neces sary cost to participate in the digital world. 3 However, the trust that these institutions will use personal data in a benevolent way is beginning to fade. In recent years a number of s candals have come to light in the U.S. regardin g the collection and use of personal data by the institutions which have access to it; w hether it be government agencies like the NSA intercepting the e mail and phone records of citizens without their knowledge, 4 Facebook making its user data available to third party politica l firms who then utilize it to manipulate public opinion, 5 or Google collecting personal data from homes (e mail records, account passwords, text messages, internet histories) and making it available to advertisers, 6 which led to lawsuits in over 38 1 Comp Fresh Air , National Public Radio Inc. Oct. 26, 2017. 2 Surveillance and Society 12, (2). (2014), p. 198. 3 Journal of Media and Cultural Studies 31, (2). (2017), p. 196. 4 Stuart Sumner, You: For Sale , (San Diego, William Andrew, 2015), 17. 5 The Washington Post , Mar. 19, 2019. 6 Stuart Sumner, You: For Sale , 76.
2 states . 7 Leaks like this have created a sense of unease among Americans . People feel that they have lost control over their personal information . They neither know who has access to their information nor how that information is being us ed . 8 The anxiety surrounding personal data collection and analysis is not unwarranted. Current legislation regarding the prope r collection of personal data is not fully developed. 9 Laws and social norms are lagging behind the advancement of technology, thus the pr oper regulations and ethical standards of personal data policy are not clear. In the realm of social media, both the users and the companies assume that data belongs to them. The ambiguity in U.S. data policy has led to confusion as to how personal data c an and should be utilized by government institutions and private companies . 10 Determining the proper role of the state when it comes to personal data collection will undoubtedly be a major task in the coming years. Might the proper path forward be to increase regul atory and government oversight policies for pla tforms like Facebook and G oogle? 11 If private companies cannot be trusted to self regulate, then how else can it be done? At the same time, government agencies like the NSA have shown that the y too will collect personal data by engaging in operations that are unethical or outright illegal. 12 Handing more regulatory power to the state may not be healthy for social media companies or their users . 13 7 Ibid. 76. 8 PewResea rch , Mar. 7, PewResearch, Jun. 4, 2018. 9 Brookings , Jun. 23, 2014. 10 S tuart Sumner, You: For Sale , 79. 11 The Washington Post , Mar. 19, 2019. 12 Stuart Sumner, You: For Sale , 19. 13 The Conversation , Nov. 28, 2017.
3 Individuals, lawmakers, and social media companies are all hesitant to suggest clear policies regarding personal data collection or use. Even small changes to data policy can create major consequences in the future. Still, the fact remains that people are concerned about their personal data, and they woul d like something to be done about it . 14 In order t o get a better grasp of this predicament over personal data and advance towards some solutions, there are a number of strategies that could be employed . One useful strategy to determine the proper policies moving forwar d, is to examine the respo nse other countries have taken to meet the challenges of the digital age. Investigating another data policy strategy may offer insight and guidance for lawmakers both in the U.S. and el sewhere. To better understand the magnitude of the cons equences that can result from personal data policies , the re is one country in particular we should turn our attention, China. With over 800 million citizens connected to the internet 15 ( more internet users than the U.S. and Europe combined ) and home to some of the largest technology corporations in the world, 16 will ha ve a major global impact going forward . By using China as a case study and investig ating policies regarding the personal data of its citizens , the U.S. can better determine the proper course of its own data policy. While America remains hesitant to implement policies and regulations regarding the use of personal data , 17 China is taking a highly coordinated and multifaceted approach t o in its data policy through the establishment of a N ational Social Credit System (SCS) s s ocial c redit s ystem goes beyond any data policy the world has yet seen . The SCS is an all inclusive data collection 14 PewResearch 15 39 th Statistical Report on Internet Development in China, CNNIC , 2017, p. 49 16 MarketWatch , May 31, 2018. 17 Council on Foreign Relations , Jan. 30, 2018.
4 apparatus which gathers personal information from nearly all facets of Chinese society. The SCS is hailed by the Chinese government to be the guiding force by which it will raise the integrity of the entire nation. 18 This piece will examine how t he Chinese SCS is instead desig ned to be a fully incorporated sta te surveillance infrastructure 19 which uses personal data as a tool to exert coercive social pressure over the majority of the population . The results SCS will be of great interest to burgeoning authoritarian powers eager to find enhanced methods of state coercion . The Chinese SCS should be of great concern to U.S. policymakers and privacy advocates. However, the Chinese SCS also provi des policymakers with an opportunity to clarify the importance of personal data protection and what policies that should be avoided going forward. 18 State Council, Notice of , 1990, No. 19. 19 Policy and Internet 10, (4). (2018): p. 416
5 Part II. The Chinese Social Credit System Th e Chinese Communist Party (CCP) is the centralized p olitical authority in China . It holds control in all of the state ministries and agencies and has done so since the middle of the 20 th century . The CCP crafted an ambitious and bold new data policy in 2014; to establish a National Social Credit System. By integrating the personal data collection and analysis methods from various sources in China, the SCS gives the Chinese government the power to determine is as an actor within Chinese society . The trustworthiness me asure is based on criteria set forth by the CCP itself. 20 The SCS gives the Chinese government an astonishing level of access into the personal lives of individual citizens the country . Implementing coercive social control mechanisms is not a novel strategy for the CCP, nor should it come as a surprise to the international community. However, the growing personal data collection capabilities make the Chinese SCS far more intrusi ve than methods used by the Chinese government in the past. A n understanding of social control policies used in the past will better contextualize the current data policies in China and offer an explanation as to how th e SCS came to be . The CCP has devel oped various social cont rol policies since its rise to power in 1949. Initially the social control methods were applied to group collectives rather than individ ual citizens. This was most notable in the commune system of collectivized agriculture during under the Mao ist regime . 21 Rural Chinese f armers were ordered into working crops from collectiv e agricultural 20 Freie Universit Ã¤ t Berlin , (2019): p. 2. 21 Asia for Educators , Columbia University (2009 ).
6 compounds . G roup ranking systems were established by the regime . Rankings were based on a farming collective work status, com pliance with the authorities, and political standing with the Communist leadership . 22 It was expected that if the group ra nkings were enforced strongly enough , the farming collectives would police their own behavio r and remain subservient to the government . The criteria for rating the farming communes ; i.e. compliance with authority and maintaining good political standing , have remained as benchmarks social policy ever since . , the Chinese government wa s amid st a debt crisis. Many Chinese firms and businesses had defaulted on debt payments to various lenders and the lending enterprises had debt repayments t hey themselves had defaulted on. This resulted in a lending practices. 23 The Chinese State Council , the chief administrative authority in Chi na, and comprised entirely of CCP members , responded by calling for the establishm ent of a corporate credit score to better determine the loan trustworthiness of businesses in 24 Third P arty credit rating firms investigated the fiscal practices of businesses to gauge how trustworthy they were either as lenders, or in repaying loans. By doing so, an air of credibility and integrity returned to the Chinese financial sectors. Clearly, a credit system based on trustwo rthiness has been implemented in various ways through out Originally , it was applied as a group measurement in the collective farming communes crisis morphed into c orporate/ financial credit system for business es and firms. E ventually , the same credit philosophy was applied to citizens at individual and consumer levels. 22 Foreign Policy , April 3, 2018, p.5 23 424. 24 State Council, N , 1990, No. 19.
7 During the mid 2000 , several local regions in China began experimenting with personal data collection methods . L ocal governments would collect what they dubbed public credit information from different regulatory, admin istrative, and public bodies within their communities . In reality, they were collecting personal data , gathering information about activities with in the community . 25 Data such as court records, traffic history , criminal backgrounds, interactions with administrative bod ies, etc. , was collected and examined in order to determine that individuals level trustworthiness as a member of the community . These regional allow the national government to test which methods of data collection and social control they would use to construct the national SCS . Construction of the national SCS began in 2007 . The Chinese State Council established joint ministerial meetings between 18 central government departments , representatives fr om local provinces where the pilot were implemented , and the financial credit investigation firms of the to gether to outline the new social credit s ystem . By 2012 , th e joint ministerial meeting system comprised of 35 central government departments; including the ministries of fin ance, justice, commerce, t he department of Publicity, the Peoples Bank of China , and the National Development and Reform Commission, which would spearhead the SCS program . 26 By integrating the data collection capabilities of these various institutions, the SCS could provide comprehensive personal data on the activities and behavior of hundreds of millions of Chinese citizens . In 2014 , the State Council released the first official outline detailing the construction of the social credit system . It clearly spelled out the goals and methods of the new system, and made no qualms about the objectives of the project . Data collected from citizens and businesses would be analyzed by the government, which then issues each citizen with an overall social cre dit score. The social credit score is 25 The Journal of Comparative Law , 12 No. 2. (2017): p. 364. 26
8 a reflection of how trustworthy or untrustworthy that citizen is as an actor in the eyes of the Chinese government. The precise way this social credit score will be represented once the national SCS is in place is stil l unclear. Local SCS programs in Chinese cities like Beijing and Rongcheng issue citizens with numerical social credit scores on a scale from 500 10 00 27 akin to FICO financial credit scores in resented with a letter A D 28 . Whatever way the score is signified, the social credit score itself is a complete and total representation of that . It is important to note that the SCS will also collect data from businesses and corporations in China , 29 scoring those entities as actors in society as well. This is more or less a radical extension of the corporate credit score policy that was this an alysis will focus primarily on SCS as it pertains to individual citizens and their personal credit score. Virtually all activities can be affected by an social credit score. Higher social credit scores come with enticing perks and benefits f rom the government , while lower scores can result in harsh punishments . The punishment and reward mechanism of the SCS policy will be explored later. The outlin e released by the Chinese State Council in 2014 implies that the primary goals of the SCS are t o reduce corruption and reinforce the populations trust in public institutions in the pursuit of . 30 However, the underlying motives remain the same as the social control policies implemented in the past. One major goal of the SCS, as State Council explicitly 27 Van Vollenhoven Institute, Leiden University , (2018): p. 10. 28 Morning Edition , Oct. 31, 2018. 29 State Council, Notice of the State Council on Printing and Distributing the Outline of the Construction of the Social Credit System (2014 2020) , 2014, No. 21. p. 9. 30 Ibid. p. 1.
9 states, is to ensure that those who respect the rules and norms of the Chinese Communist Party, who er 31 If the SCS functions as planned, it will grant the s tate the ability and the authority to monitor, predict and govern each actor in its political , economic, and soc ial domains. 32 It seems unlikely that any society as expansive and complex as China could successfully implement a policy program as all encompassing a s a national SCS , but that is only because up to this point, it has never been tried. The SCS requires intricate coordination and new partnerships among some financial institutions, major employers, mobile application developers, and national government agencies. As is the case in the U.S., in China e ac h of these various entities already have a level of access to the personal data of individual citizens . The major obstacle for the Chinese government is accumulating the data sets collected by all of these separate institutions into a centralized system . H ere the goal for the Chinese government is to build a comprehensive, nationwide, platform aggregating all related personal data by 2020. 33 Once the data sets are linked , all that is left to do is choose which factors and behaviors determine an individual social credit score, and in turn, the ir trustworthiness . It is vital to note that all the methods of scoring individuals within the SCS lie under the total authority of the CCP government. How much debt is it okay for a citizen to have? How many traffic violations can a citizen have before their score is reduced? The Chinese government will have answers 31 backed authoritarianism: Information technology enhances central authority and Mercator Institute for China Studies , No. 1. (2 016): p. 53. 32 33
10 for all of those questions. The government will establish the criteria. 34 Once the criteria is established , a lgorithms will be created so the SCS can sift through the mass collection of personal data and find the relevant data points which affect social credit scores . 35 s crucial to recognize is that the only complete set of personal data compiled from all the various sources ; tech companies, government agencies, employers, banks and financial firms, websites, etc. , will be under the possession of one body. All of the data related to the social credit system will be either sent to or exchanged with central gover nment databases , 36 ensuring the only entity that has total control over SCS will be the government itself . Beyond data collection and analysis is the reward and punishment mechanism . Depending on the trustworthiness of any individual , (as expressed by their personal social credit score), the SCS will flag those who score below the standard rating, a standard that is determined by the government . Flagged i ndividuals are then blacklist ed and can be identified by various local and nationa l agencies . 37 T his allows all sorts of institutions public or private to know which citizens are blacklisted and enforce restrictions . Th e restrictions affect an , (e.g. travel sanctions and market restrictions ). This is the mechanism by which th e SCS wil l 38 Once the national SCS is fully implemented, expected by 2020, the party and the governme nt will be able to autonomously reinforce and maintain political control over the entire society through nigh omnipotent surveillance capabilities . 34 cn.gov official interview , July 23, 2014. p. 4. 35 Liang et. al. p. 430. 36 37 Genia Kostka, 38 The Economist, Dec 17, 2016.
11 It is cruci al to examine the process of data collection , data aggregation, and enforcement of the SCS in de tail . Several methods being used by the Chinese government are similar to procedures used by state agencies in the U.S a nd social media tech giants operating worldwide . By identifying what kind of personal data are being analyzed by the SCS , this will reveal what kind of personal data is most sought after by the Chinese government. It will also make clear what methods of data protection, whether they be legislative or social, will ensure that SCS policies do not spread beyond Chinese borders, and thus deny the CCP their goal of 39 Data Collection As mentioned above, constructing a comprehensive social credit system at the national level requires a high degre e of cooperation within , a mong, and beyond state agencies. As of 2017 , 47 central state departments were active members of the ministerial joint meeting system that was established in 2007 to begin work on the SCS . 40 These departments are working to integrat e the current communication technologies with on the ground security data that typically used by law enfo rcement (e.g. sophisticated CCTV cameras 41 , highly advanced license plate/ facial recognition tools 42 ). State d epartments like the ministry of justice and its local counterparts have access to individual criminal 39 State Council, Notice of the State Council on Printing and Distributing the Outline of the Constru ction of the Social Credit System (2014 2020) , 2014, No. 21. p. 3. 40 Liang et. al. p. 426. 41 The Economist . 42
12 records and court proceedings . 43 For all of these departments , their challenge is not collecting the data, but in tegrating what they have with the o wn centralized databases . Financial data is al so sought after by the social credit system . The Peoples Bank of China has access to data such as tax history, bank statements, and financial credit ratin gs for businesses and consumers, which can be integrated into the SCS databases . However, the financial data of individual citizens is more expansive than bank statements and tax records. Several tech companies have entered partnerships with the state government to help construct the SCS . Alibaba, the largest e commerce company in China, has been using its own internal credit rating for its users since 2013 called Sesame Credit . 44 A users S esame C redit scores are based on their own lending and spending habits , as well as the habits of their friends on the platform. 45 In fact, Sesame Credit was a major inspiration for the national SCS and thus, Alibaba is one of several tech giants that have received a license from the People s Bank of China to share their user da ta with the state government . to include credit into the operation of its e commerce platforms. We hope Alibaba can be further involved in building the credit system for the entire society 46 Tencent, another Chinese based tech compan y , and largest insurers, have also received licens es to share their user data with the government. 47 The data collection capabilities of the SCS go beyond the financial and legal spheres . It is interested in the behavioral data of citizens as well. Thus, social media sites used by Chinese citizens are a primary tar get for the SCS . WeChat is a Chin ese based social media application developed by 43 44 Wired Magazine , Dec. 14, 2017. 45 46 China Da ily , Dec. 03, 2016 47 Financial Times , Jan. 19, 2016
13 Tencent. WeChat is one of the largest mobile applications touting over 1 billion monthly users, approximately 500 million of which are Chinese. 48 The applicatio t is hard to overstate the pervasive ness of WeChat in Chinese life the app is a phone, messe nger, video conference, e commerce platform and gaming console, not to mention [a] noodle delivery service . 49 By collecting data from applications such as WeChat, the SCS can have information related to personal activities ; opinions, relationships , shoppin g history, and recently visited places of a half billion Chinese citizens. The Chinese government is aware that for data collection to be comprehensive, an online anonymity must be diminished. The Chinese Supreme Court has taken legislative a ction to ensure that Chinese tech companies like Tencent and Alibaba real , 50 which require corporations to link user profi les with personal information, severely limit ing the ability for citizens to remain anonymous online . 51 their applications are among Chinese internet users , comprehensive data collection in this domain would grant the state the ability to oversee the majority of all online activity in Chin a. 52 Data Aggregation and Analysis Once the data collection process of the SCS is in place, the next phase is to aggregate the relevant data within the government. China is currently building at least five centralized data platforms 48 Business Insider , Mar. 06, 2018 49 Ethics and Information Technology (2018): p.5. 50 China Perspectives , No. 4. (2015): p. 10 51 The Economist , Dec. 17, 2016 52 Journal of Contemporar y China , 26 No. 103 (2017): p. 95.
14 to store and evaluate credit related data. 53 A mass aggregation of data such as this presents challenges to the government, as it aims to consolidate all the relevant information collected from individuals into one complete social credit profile . Though difficult, the aggregati on of this data is a top priority for the Chinese government. The local pilot social credit programs in the mid partnerships with private companies like Tencent and Alibaba have served to test relevant systems and technologies in orde r to accomplish such a task. 54 In fact, China has to look no further than the U.S. to see that data aggregation of this scope is possible. The U.S. National Security Agency is capable of collecting and storing up to 42 billion internet records per month, an d five billion mobile geo location records every day. 55 Specific points of data collected from ; facial recognition cameras, financial data, internet use, and mobile location records , enable the Chinese government to integrate these data points and connect t hem to specific individuals. 56 The major challenge for the SCS is the national scale on which it seeks to operate . G iven that the pilot SCS programs in Chinese cities such as Beijing 57 , Rongcheng 58 , and Suining 59 have all been implemented by the local governments with some level of success, it is not a matter of if this data aggregation can be done, but when. Once the data has been successfully aggregated, the SCS immediately begins the next phase, enforcement of rewards and punishments . 53 54 55 The Economist, Dec. 17, 2016 56 57 58 59 Straits Times (Singapore) , Mar. 5, 2018
15 Enforcement The CCP controlled government establishes the criteria upon which the data collected from individuals will be judged. This is the focal point of the SCS policy , where the social con trol and behavioral coercion mechanisms begin to function. Many of the factors which can lower an individual credit score are vague, and intentionally so . We can gleam some of the details by examining punishments and rewards that have already been observed in the cities with local SC S pilots . High scoring individuals are treated to rewards and perks under the credit system. Access to better aca demic institutions, fast tracking for job promotions , 60 expedited travel visas, reduced plane/train ticket pric es , lower heating bill s, 61 car ren tals with no deposit, 62 and lower interest rates on loans , are all benefits that can be accrued by obeying the legal and moral imperatives of the government. Those with exceptional scores may even find their pictures displayed outside public libraries, resi dential areas, and city halls, with glowing testimony of their responsible actions. 63 Most concerning, social credit scores can affect the educational and occupational prospects of family , as well. 64 In most cases, the punishments are far more consequential than the rewards. Causing a disturbance in public carr ies a penalt y in most areas where the credit system operates. T he Civil Aeronautics Administration banned 86 individuals from air travel for one year during July of 2017 with . 65 As of 2019, over 5.5 million Chinese citizens on social credit blacklists have been restricted from purchasing train tickets, and over 17.5 million have been restricted from purchasing airplane tickets. 66 Driving habits are monitored through surveillance 60 61 Stacey Smith interviewed by Steve Inskeep, Morning Edition , p.1. 62 63 Sim 64 The Asian Institute for Policy Studies , p. 11. 65 The Epoch Times , Jun. 5, 2018. 66 The Guardian , Mar. 1, 2019.
16 technology on highway and road systems, and a professional road habits can also influence the credit score of their employer . 67 In Rongcheng, traffic offenses lead to score reductions, as d o more innocuous activities. Spreading rumors online incur s a 50 point social credit score reduction. 68 Late and unprocessed payments on judicial or administrative fines can land one on the social credit blacklist outright, wh ich often carries a ban on high speed train and air travel. 69 In the Jiangsu province, not visiting your parents often enough can carry a penalty , although the meaning of , often enough , is left ambiguous . 70 Activities as trivial as purchasing videogames may flag some one in the system since it reflects lazy and irresponsible behavior . 71 To discourage activity that puts one on the blacklist , public exposure is a tactic the CCP gove rnment employ s regularly. 72 In the D istrict of Hefei, individuals who are on the blacklist , because of debt , have their personal information (photograph, name, id number, and amount owed) posted on giant electronic billboards above heavily trafficked public squares and highways. 73 Individuals who lose in a court disp ute ( e.g. c hild custody dispute) and default on their payments are put on a judgement defaulter list, their names are displayed on an electronic crawl outside court houses. Over 3 million people were on the judgement def aulter list as of 2015. 74 In some localities, if someone places a telephone call to an individual on a blacklist, an automated message will say , 75 So far, public s haming and criticism methods are extensive across China and stark reminders to all citizens that they could be one poor decision away from a social credit blacklist. Unsurprisingly, it is all too common for 67 68 Morning Edi tion , Oct. 31, 2018. 69 70 71 Morning Edition , Oct. 31, 2018. 72 State Council, Outline of the Construction of the Social Credit System, p. 18. 73 South China Morning Post , May 4, 2018. 74 The Economist , p. 3. 75
17 Chinese citizens to be unaware that they have bee n blacklisted . There is no method of notification for citizens until the y realize they have been barred from certain activities in Chinese Society; buying property, taking out a loan, travelling, or applying for a job or school . 76 Remedies for blacklisted individuals are limited . T here is no explicit process to appeal a position on the blacklist with an authority. If one manages to find the requisite authority, it can take weeks to speak with a representative. 77 These punishments all serve the communist part coercion of those deemed 78 The punishment and blacklist mechanism of the SCS allows the Chinese government to determine who gets to parti cipate in virtuall y all activities in Chin ese society. These restrictions are coupled with the threat of intense public and private scrutiny own life and the lives of their families, friends, and co wo rkers, because not only does the SCS evaluate the particular actors in society as individual units, but also evaluates the quality of their social group as well. Thanks to social media, citizens are more than willing to openly share who they associate with in the community, let alone who the ir family is. Thus, if someone receives a low credit score , their own family and friends may be tempted to disassociate with them, fearing their relationship may have negative effects on their own credit score . 79 T herein lies the most sinister weapon of th e SCS , coercing individuals to self police their own actions and relationships with others means that the government occupies an ever present role within their community, their home, and their mind. Without the need to use direct action, the Chinese 76 The Globe and Mail , Jan. 3, 2018. 77 Human Rights Watch , Dec. 12, 2017. 78 Warning and Punishment Mechanisms for Persons Sub ject to Enforcement for Trust 79 Meissner and Wubbeke, p. 54
18 govern ment will have achieved unadulterated access into the lives of its citizens. The goal is clear and is no better elucidated than by the Chinese government itself. According to the deputy director of the National Development and Reform Commission, the agency condemnation of society, we will form a social shock and constrain untrustworthy behavior of members of society . 80 Two questions remain for China and its citizens. First, how have the Chinese people reacted to th e implementation of the SCS ? S econd, what problems will t he SCS face when it is implemented ? regional surveys have been conducted and analyzed. One notable finding reports a lack of awareness among many Chinese that they are a part of local pilot SCS programs. A survey published in the Journal of N e w Media and S ociety reported that although 43% of respondents resided in a locality where a pilot program was un derway, only 11% were aware they were a part of it. 81 citizens with only 19% expressing a neutral value to their attitude, and only 1% expressing disapproval 82 The data show s that overall the public opinion towards a ny kind of SCS, even one operated by the government, is positive. Feelings of positivity towards the SCS are not without their reasons . I n se veral cases , the implementation of a government ran SCS has made l ife better for many Chinese citizens . Personal anecdotes reporting an overall improve ment in community behavior are numerous. As a citizen from Rongcheng, home to a local SCS pilot first, we just worried about losing points, but 80 cn.gov official interview . 81 82
19 now we 83 The SCS has some provisions like the Enterprise Environmental Credit , 84 which rates companies based on their contributions to air and water pollution, which may have a of smog. 85 Finding someone who is willing to voice any concern over the SCS is a difficult task. There are some examples of Chinese media outlets expressing hesitation towards the national SCS and its broad punishment mechanism . As Yeng Gengshen, columnis t for the Beijing Times wrote , I have never opposed the establishment and improvement of a credit 86 This should not com e as a surprise to researchers. In an authoritarian state such as China, the fact there is hardly anyone expressing disapproval of the SCS may reflect the pressure Chinese citizens a lready feel to avoid criticizing the government respondents may have falsified their preferences to a degree due to concerns about expressions of disapproval resulting in reprisals from the 87 Even without a fully implemented SCS, the constraints on individual expression are well in place. It seem s that the only people in China who are weary of the SCS, are those who have already been t argeted by it, while most others remain unaware of the potential dangers. When questioned about privacy issues, there is a common retort among citizens who remain op timistic about the program. Justifications like, to use my data , 88 and , I trust the 89 These a re common responses among citizens involved in SCS pilot programs. 83 84 85 South China Morning Post , Oct. 2, 2018. 86 The Economist p. 5. 87 88 Straits Times , Mar. 5 , 2018. 89
20 The second question remains, if the national SCS does become successfully implemented by 2020, what challenges will it face? For one, ensuring that the vast array of data collected and analyzed by the SCS is accurate and complet e has not been successful . F alse readings, incorrect information, and reporting errors are a major problem for data collection systems across the glob e ; British National Health Service databases, and American marketing databases have been severely corrupted by the collection of inacc urate data . 90 There is little evidence to suggest fare any better. Likewise, Chinese citizens are severely limited in their ability to dispute inaccurate information that can affect their social credit scores . W hether or not an individual has rights to access and rectify their credit information depend s on the local regulations of the province where they reside. 91 Additionally, many of 92 Lack of data accuracy , consistent regulations , and the ability to update information fast enough aside, a centralized hub of personal data lik e a national SCS database would provide massive incentives for cyber criminals. If a hacking syndicate or even a lone actor were to gain access to the central SCS databases, the consequences could be enormous . The social credit score of any individual woul d be susceptible to manipulation. 93 The vulnerability of personal data to hackers in China is exacerbated even whereby the Chinese Supreme Court ordered that networking services and tech companies must register online accounts with the user s real names and information. mobile phone 90 The Economist p. 4. 91 92 Mar. 6, 2018. 93 The Asian Institute for Policy Studies , p. 10.
21 94 Worries of hacking are not a hypothetical. Earlier this year in Beijing, which has its own pilot SCS program, approximately 364 million online records were leaked. These records included information such as; GPS locations, file transfers, and chat logs. 95 There is certainly no lack of weaknesses and There is much work to be done if the CCP wishes to implement the national SCS by 2020. It is unknown if China even has the technological capabilities to have an operational SCS by that time. 96 So me research suggests that a unified national SCS in China should not be a major concern for the U.S at the moment [W] hile there are many things to be worried about in China, a single and all pervasive 97 However, whether or not the SCS will arrive by 2020 , or 2030, or 2040 should not be the primary concern for the U.S, or the international community at large. The details of this oppressive surveillance apparatus have already been sp elled out , so too have its method s of social co ercion been revealed. What remains now is what we can learn from the Chinese approach to better craft our own digital information policy . It may seem strange looking toward China for guidance on U.S personal data policy. However, the borders between nations in the digital world are porous, and domestic policies like the Chinese SCS do not exist in a vacuum. We must prepare for the effects that the Chinese SCS will have on the rest of the world. 94 France MÃ©dias Monde , March 7, 2019. 95 96 97 Brookings , Nov. 19, 2018.
22 Part III. Consequences and Considerations One major concern is how a fully realized Chinese SCS will impact other nations, especially those inclined towar ds authoritarian ism . China fully intends the SCS apparatus to be adopted by other countries, and even aims to be at the center of an international SCS in the future . 98 China has invested heavily in crafting an international presence, focusing on its economic and diplomatic relations with other countries . China s massive an attempt to establish a transnational credit syste m to secure international trade and economic relations with approximately 65 other nations throughout Asia, Africa, and Europe. 99 In effect, China wants to position itself at the heart of a new international trade and development hub by financing infrastructure projects from Pakistan to the Congo . 100 Aside from establishing new economic sectors throughout the hemisphere, China wishes to deep promoting the establishment of an objective , fair, reasonable and balanced international credit rating 101 Imagine how a similar state sponsored SCS may be implemented in countries like Russia, 102 In fact, r internet laws introduced over the past five years require search engines to delete some search results, messaging services to share encryption keys with security services and social networks to 103 Furthermore, what might a state sponsored 98 State Council, Outline of the Construction of the Social Credit System . p. 3. 99 100 China Power , Sep. 11, 2017. 101 State Council, Outline of the Construction of the Social Credit System . p. 3 . 102 Bloomberg Politics , Mar. 18, 2019. 103 NPR World , Mar. 17, 2019.
23 S CS look like in Saudi Arabia, Pakistan , Uganda or a variety of other countries where government repression, journalistic censorship , and the persecution of minority communities are already enforced by the state? 104 For states like this, the Chinese SCS may become golden example of authoritarian data policy . The SCS will be seen by other oppressive regimes as a valuable tool that enhances politic al power and silences dissident voices in society. 105 The SCS in China ma y seem like a threat only i n sofar as it poses dangers for individuals in China and vulnerable populations around the world. However, there are practices of data collection and utilization within the U.S that are str ikingly similar to those used by In the United States, notions of credit and trustworthiness of their financial and consumer data consumer report eligibility for things like housing, insurance, employment, etc. 106 However, credit reporting agencies are increasingly integrating non traditional credit characteristi cs into consumer credit profiles. Personal data location data, their social media usage, shopping history (on and offline) , and even data regarding their family and social connections are now being incorporated as data points that can affect a financial credit score , eligibility for benefits, and employment qualifications . 107 Even dietary information that one re cords in mobile applications are analyzed and used by insurance providers to determine premiums for their members. 108 predictive analytics to determine credit and eligibility is 104 Human Rights Watch Human Rights Watch Human Rights Watch 105 Backed Authoritaria 106 FTC Report , Jan. 2016. p. ii. 107 Washington Law Revie w 89, (1). (2014). p. 4 108 Alice E. Marwick, How Your Data Are Being Deeply Mined , N.Y. Rev. Books, Jan. 9, 2014.
24 becoming more and more common. A process by which agencies mine personal information to assess the future behavior of an individual consumer based on the historical data analytics of consumers who share similar characteristics. 109 Law enforcement is another domain that is beginning to use these data analytics for predictive policing measures . 110 The healthcare field is increasingly using personal data mining and predictive ana lytic techniques to determine patient eligibility for healthcare benefits. Additional data does not always lead to better and more accurate determinations. Usin information to drive predictive data analytics creates a host of problems. Although predictive data analytics provide useful information to credit agencies and compani es, they also create a disparate impact on certain groups or individuals whose actual be havior does not reflect the analytical data. For example, some credit companies have actively penalized individual consumers based not on that person s financial data, but based on the financial data of other individuals because they all shared a similar s hopping history. 111 When used for targeted advertising of financial products, low income consumers are not advertised superior offers despite being eligible. 112 When it comes to employers who rely on data analytics to make hiring decisions, the credit reporti ng agencies themselves have admitted there is no correlation between an individual consumer credit report and their job performance. 113 Data analytics are replete with erroneous and false information regarding individuals. Databases are suffuse with outdated and inaccurate information because humans, not highly precise computer algorithms, provide the data. Moreover , since the algorithms themselves are programed by humans, our biases and values are embedded into the code 109 110 Journal of Communi ty Safety & Well Being 2, (1). (2017). 111 Federal Trade Commission. p. 9. 112 Ibid. 10 113 Committee on Financial Services , 2019, statement of Edmund Mierz winski, Senior Director, Federal Consumer Programs U.S. Public Interest Research Group. February 26, 2019, 10.
25 itself. 114 Unbridle d reliance on personal data analytics does not increase the accuracy of the determi nations made by these systems. Without human oversight, or even the ability to investigate the algorithms and analytical processes that run them , predictive data analytics c an create a self reinforcing cycle of disparate punishments, and unfair discrimination. A Final Glance at China China has the highest population of internet users in the entire world. That alone is incentive for any technology company to desire a positi on in China . This migration to China is beginning to take place. Google has been solicited by China to c onstruct a fully censored state controlled search engine for the Chinese people, to which Google complied. 115 This joint venture between Google and the Chinese government is Project Dragonfly . I t was kept under strict secrecy within the company with software engineers being instructed to avoid referencing it to their co workers and to deflect any questions aimed at them by the Amer ican press. 116 This project sparked the resignation of hundreds of Google employees and demands that the tech giant testify before the U.S. Senate. 117 What happens when an American based software company is found cooperating with a non allied country, to help that country further restrict the access to information and freedom of speech of its own citizens? We will soon find out the answer to that question. The Chinese Social Credit System is not merely an Orwellian curiosity on the other side of the globe; it is an ideological challenge of current international norms and institutions regarding privacy, the 114 115 The Washingto n Post , Oct. 6, 2018. 116 The Intercept , Aug. 8, 2018 117 Committee on Commerce, Science, and Transportation , U.S. Senate. Sep. 26, 2018.
26 role of government, and individual sovereignty. 118 Without reasonable protections and policies regarding the collection and use of personal data, such infor mation is increasingly becoming available to institutions and agencies, and used in ways that disparately and unfairly affect certain individuals and not others. Scoring systems that rely on personal data are alluring to any and every body that is supposed to make determinations about individuals. Scoring systems offer a highly efficient, automated way to find evidence that justifies their own results and determinations. Their simplicity gives the illusion of accuracy , 119 but treating credit systems as oracles of truth in no need of further human tampering would be a disastrous mistake. Policies of reasonable transparency and human oversight are the necessary conditions under which our scoring systems can be just and without corruption. One need s to loo k no further than the Social Credit System in China to witness how close the world is to entering an age o f data driven authoritarianism. To do everything possible to avoid the least desirable future is the first step towards creating a better one. 118 Council on Foreign Relations. Feb. 21, 2019. 119
27 Bibliography CCP Central Committee General Office. Opinions Concerning Accelerating the Construction of Credit Supervision, Warning and Punishment Mechanisms for Persons Subject to Enforcement for Trust Breaking . China: State Council, 2016. Ethics and Information Technology 20, no. 4 (September 2018): 279 289. a Profiling: Privacy and Chinese The Journal of Comparative Law 12, no. 2 (June 2017): 356 378. China Internet Network Information Center. Statistical Report on Internet Development in China. No. 39. China: CNNIC, 2017. Chinese State Council. China Improves Credit Blacklisting Mechanism to Avoid Undue Punishment . Xinhua: State Council, 2018. Chinese State Council. Across the County . No . 19. China: State Council, 1990. Chinese State Council. Notice of the State Council on Printing and Distributing the Outline of the Construction of the Social Credit System (2014 2020). No. 21. China: State Council, 2014. Citron, Danielle K. and Frank Pa Washington Law Review 89, no. 1 (2014): 1 32. Financial Times . January 19, 2016.
28 Asia for Educators . Accessed March 11, 2019. http://afe.easia.columbia.edu/special/china_1950_commune.htm Van Vollenhoven Institute . Leiden University. (May 2018): 1 32. Twenty Journal of Contemporary China 26, no. 103 (September 2016): 1 16. China Perspectives , no. 4 (December 2015): 5 13. and Dataveillance: Big Data between Scientific Paradigm and Surveillance and Society 12, no. 2 (May 2014): 197 208. South China Morning Post . May 4, 2018. The Epoch Times . June 5, 2018. MarketWatch . May 31, 2018. Fulton, Janet and Marj Journal of Media and Cultural Studies 31, no. 2 (2017). The Washington Post . October 6, 2018.
29 Ga The Intercept . August 8, 2018. PewResearch . Jun. 4, 2018. Business Insider . March 06, 2018. Brookings . November 19, 2018. report/2019. Wired Magaz ine . December 14, 2017. China Daily . December 03, 2016. South China Morning Post . October 2, 2018. New Media & Society , (February 2019): 1 29. Bloomberg Politics . March 18, 2019. The Guardian . March 1, 2019.
30 Journal of Community Safety & Well Being 2, no. 1 (March 2017): The Conversation . November 28, 2017. onstructing a Data Driven Policy and Internet 10, no. 4 (August 2018): 415 453. Straits T imes (Singapore) . March 5, 2018. Terry Gross. Fresh Air , NPR. October 26, 2017. Marwick, Alice E. How Your Data are Being Deeply Mined . New York: Rev. Books, 2014. backed authoritarianism: Information technology enhances central Structures and Processes under Xi Jinping , no. 1 . Edited by Sebastian Heilmann and Matthias Stepan, 52 58. Berlin: Mercator Institute for China Studies, June 2016. Brookings . June 23, 2014. Foreign Policy . April 3, 2018. Council on Foreign Relations . January 30, 2018. The Asan Insti tute for Policy Studies . February 28, 2017.
31 Good Job in Protecting Personal Financial Information . No. 17. China: Departmental Regulatory Documen ts. 2011. PewResearch . Mar. 7, 2018. NPR. Oct. 31, 201 8. Sumner, Stuart. You for Sale: Protecting Your Personal Data and Privacy Online . San Diego: William Andrew, 2015. France MÃ©dias Monde . March 7, 2019. China Power . September 11, 2017. U.S. Federal Trade Commission. Big Data: A Tool for Inclusion or Exclusion? Understanding the Issues . by Edith Ramirez, Julie Brill, Maureen K. Ohlhausen and Terrell McSw eeny. Washington D.C. 2016. U.S. House Committee on Financial Services. Repairing a Broken System . statement of Edmund Mierzwinski. Hearings. Washington D.C. 2016. U.S. Senate Committee on Comme rce, Science, and Transportation. Consumers Deserve Clear Answers and Standards on Data Privacy Protection . e mail from Dr. Jack Paulson. September 26, 2018. The Globe and Mail . January 3, 2018.
32 NPR World . March 17, 2019. Human Rights Watch . December 12, 2017. The Washington Post . March 19, 2019. Interview. July 23, 2014.