Monday, October 07, 2019

"Social Credit and Foreign Enterprises Along the Silk Road": Remarks prepared for a Lecture Delivered at the Institute for East Asian Studies Cologne, Germany October 2019

Set out below is the text of remarks prepared for a Lecture Delivered at the Institute for East Asian Studies in Cologne, Germany. The Remarks, entitled, "Social Credit and Foreign Enterprises Along the Silk Road," considers the now controversial issue of the application of Chinese social credit ratings regimes on non-Chinese companies. The application raises important issues about the nature of governance in both China and the West, the convergence of distinct approaches to corporate accountability, and the use of data driven governance as both a means of standard setting and as the means by which such standards are applied.

To those ends the remarks begin from the simplest of starting points: how is it that one can construct a social credit system? From there the remarks considered China’s Social Credit system structures in general terms. That provides the grounding necessary to then focus on that part of the social credit system that targets business integrity. To those ends the remarks turned to a consideration of a July 2019 publication of the State Administration of Markets on business related Social Credit lists, and more importantly, the 16 July 2019 State Council Guiding Opinion. Lastly, the remarks considered its implications for non-Chinese business operating in China and within the BRI sphere of influence.

The text of the Remarks follow below; they may be downloaded HERE.

The PowerPoints of the Remarks may be accessed HERE.

Social Credit and Foreign Enterprises Along the Silk Road
Larry Catá Backer

Remarks prepared for a Lecture Delivered at the Institute for East Asian Studies
Cologne, Germany (
October 10, 2019.

Good afternoon! I am delighted to be here with you now. It is a gloriously beautiful early Autumn day and we embark on the consideration of one of the most interesting transformations of law and politics that has occurred in the last century. For that I thank Björn Ahl, Professor and Chair of Chinese Legal Culture, and the University of Cologne Institute for East Asian Studies for making this possible. I can only hope that I might use this time allotted to me this afternoon to bring to your attention a way of thinking about law in a slightly different way.


As well, I hope to suggest to you both that the character of law is changing to almost the same degree as the locus of law is being shifted. That change in the character of law moves us from law as a system of commands implemented through ever more complex webs of directive regulation to a seamless real time system of data driven governance founded on the principal that behavior can be managed through real time systems of restrictions and rewards. More importantly, this system of data driven governance is as effective a means of managing the behavior of enterprises—and state officials—as it is in managing the behaviors and world views of the most ordinary individual. I refer, of course, to Chinese social credit and big data management systems (usually with a dollop of artificial intelligence (AI) or machine learning and their manifestations as algorithms thrown in to scare the children (and officials)). But I also speak to the economies of data and stakeholder management that has become an increasingly, though undisciplined, element of markets-based behavior management in liberal democratic states. Technology has liberated the state from its ancient dependence on law as a means of commanding behavior even as it has enhanced its constitutive character, constructing social credit systems the way constitutions were once used to construct the state.

The equally significant change in the locus of law moves us more deeply into the state—in the context of both Marxist-Leninist and apex liberal democratic states. At the same time it liberates law from the state in those spaces in which the state is merely one of several actors—most notably within the space of global production. That this is possible is not remarkable in this new era. On the one hand, its borderlessness has prompted the states that harbor the greatest concentration of controlling economic entities to organize their economic model on the basis of the principle of core and collective, or hubs and spokes, or center and peripheries. These states have begun to remodel empire along principles of control and dependency without the clutter of racism and ethno-culturalism: that is the message of Xi Jinping’s Belt and Road Initiative; it is likewise the message first recast for liberal democracies by Barack Obama presiding at what looked like the triumph of the Washington Consensus, and then recast again by Donald Trump more bluntly as America First.

At the same time, the great governance gaps that the state system exposes within Chinese Silk Roads and Western Production Chains has opened governance to entities that occupy organized social and economic space beyond, through and between states. The emergence of “roads” and “chains” across states have also produced the “law” of the emerging fields of transnational or societal law and imbued it with much of its character. That character is founded on the notion that states no longer sit at the center, or must necessarily be invoked, to produce regulation within consenting communities that stretch between and beyond the state. Many of these regulatory communities—we sometimes construct them as multi-national enterprises, or global civil society—are not merely independent sources of law internal to their organization. They are also the vehicles through which states transnationalize their domestic orders by governmentalizing the operation of these economic actors.

The interplay of these trends—the governmentalization of the economic sector, the reconstruction of trade governance, and the rise of data driven governance modalities as a substitute for or supplement to traditional law-regulatory systems—is what brings me here today. I will speak in more detail about the consequences of the union of data driven governance and the governmentalization of the economic sphere across production chains and on silk and maritime roads that lead from a governance core to the collectives that are aligned along paths that are created to satisfy its aims. My aim is to speak not just to its technical characteristics, but also to the underlying concepts and principles that make its construction both compelling and from a certain point of view inevitable.

In other work I have called this the movement toward “Next Generation Law.” For this conversation I will consider only a very small slice of the possibilities of data driven governance in globalization. I will consider its potentially profound effects in one small corner of its operation. I will consider data driven governance regimes as a state centered project—for the purposes of which one must look to China and its Social Credit system. Within that Social Credit system I will look specifically at its application to business, and more narrowly still to its application to non-Chinese business in China and throughout the Chinese Silk and Maritime Roads. Lastly I will briefly consider the implications for non-Chinese and their compatibility to non-Chinese practice.

To those ends I will begin from the simplest of starting points: how is it that one can construct a social credit system? From there I consider China’s Social Credit system structures in general terms.
That provides the grounding necessary to then focus on that part of the social credit system that targets business integrity. To those ends I will look more carefully at a July 2019 publication of the State Administration of Markets on business related Social Credit lists, and more importantly, the 16 July 2019 State Council Guiding Opinion. Lastly, I will consider its implications for non-Chinese business operating in China and within the BRI sphere of influence.

Building Social Credit.

There is little point in talking about social credit systems and data driven governance without a basic understanding of how it works. The best way to do that, perhaps, is to try to build a rudimentary social credit system. So let’s start. To make it interesting, let’s focus on the construction of a system that most Westerners would welcome.

Let’s take the case of modern slavery in business. Assume that there is a general consensus against the practice. Assume further that beyond writing laws and regulations, states find it hard to control the practice because much trafficking occurs between states and it is best discovered within places of employment. The reason for that is that modern slavery is now assumed to be deeply embedded in official and unofficial transnational labor markets. Assume that all businesses are willing to comply with law and are sensitive to stakeholder action, especially of consumers and assume further that a handful of states have enacted so-called Modern Slavery Laws, or supply chain due diligence laws or Anti-Trafficking disclosure laws, all of which require some kind of public disclosure. In addition, companies may be required to disclose additional information about their operations and their supply chain relationships under the securities laws of the states in which their shares are traded or registered. They may also be required to disclose relevant information indirectly pursuant to exchange listing requirements, public and private financing agreements, and voluntary self-disclosure through Environmental, Social and Governance reporting frameworks.

Under traditional approaches, one would focus on compliance—especially legal compliance. That would require pressuring states to enforce and businesses to fulfill their legal obligations with respect to trafficking and modern slavery disclosures. It would be for the state to enforce its laws against slavery and its disclosure laws. That requires the development of state capacity—including the deployment of police, other investigators, prosecutors, and a judicial system with the capacity to deal with these violations. To the extent the actions are not criminal, then the system would depend on individuals to assert the violation of their own rights. There would likely be agitation for more comprehensive legislation. There would be slightly less agitation for the allocation of national resources for compliance and enforcement. There might be scandals about trials that are unsuccessful or judicial decisions that might interpret legislation or regulation in ways that appear to make it harder to find and punish those engaged in modern forms of slavery, or that invalidate either criminal or civil requirements.

Data driven approaches would start from a different point. Let us begin with a protagonist and its laudable objectives. Let us call our protagonist NGO and assume that it is a conventional non-governmental institution with technical capacity and resources. Agency’s principal objective is the eradication of modern manifestations of slavery. NGO also has developed a set of norms and assumptions about the (1) characteristics of modern slavery, (2) its connection to human trafficking, its principal effects on women, children and indigenous populations, (3) its dependency on the willingness of legitimate business entities for its viability, and (4) its intimate connection with the maintenance not just of suppressed markets (in this case for people) but also of creating an alternative structure for financing terrorist and other activities which NGO believes, like the state, ought to be repressed. NGO believes that corporate integrity and honest behavior in markets ought to be founded on programs to prevent, mitigate or remedy all characteristics and expressions of modern slavery. Those actions ought to be judged against the standards and expectations NGO has developed (NGO’s norms). To those ends NGO would like to develop a means of measuring corporate compliance with anti-slavery norms by gauging their performance against our standards and expectations.[1]

How does NGO go about doing this? First, NGO needs data. But that requires NGO to do two things. HGO must first identify relevant data; it must then determine where such data may be sourced. Neither of these tasks can be undertaken in a vacuum. Data does not exist autonomously in social space; data is a function of what the data harvester is looking for, as well as the way that the data harvester sees the world. In other words, norms and objectives shape the way in which NGO (and the rest of us) must approach the issue of data. For most tasks, being able to theorize an ideal data set is only the first obstacle—the second is to either find where such data may be harvested, or to figure out how to construct that data from other sources of information that may be available.

In this case, the task that NGO has set for itself is relatively easy. It will be relatively straightforward to build measurable markers around the modern slavery norms of NGO, and to do that on the basis of a wealth of publicly available information. Some of that information is itself data; some may be converted to useful data. All that is required is the capacity and resources to harvest. Data, of course may also be harvested by request—NGO would have to ask—the state would merely compel. And in this case because states have indeed compelled the production of information that might be useful data, the task is made even easier—as long as NGO is satisfied that measurable information may be extracted from available data. And by measurable, of course NGO would have to mean measurable against the norms and objectives for which they will be used—again to rate companies on their programs to prevent, mitigate and remedy modern forms of slavery as these are understood through the lens of NGO’s own norms and principles. Lastly, NGO may be able to purchase information from others; this purchased data might be useful as data for the task that NGO has set for itself (it objectives). There are robust markets for data within production chains and along Silk roads, the later at least when they pass by states that that require disclosure or reporting. And in this case, because states may require modern slavery disclosures along a production or supply chain (for example Australia and France) it may be possible to find commercially available useful data sets.
NGO now have lots and lots of data. But raw data must be made useful. That requires the application of a three-part process: analytics, assessment, and judgment. Let’s consider each in turn.

Analytics is the process by which data is interpreted. That interpretation starts, again, from the baseline of the core values against which the data is read, and as a function of the objectives of the analytics exercise. In this case let us assume that the object is to obtain ratings on two points: (1) the completeness of disclosures about a company’s modern slavery policy (whether required or not under local law); and (2) an assessment of the completeness of the policy and the effectiveness of its implementation. The first focuses on compliance (either legal or societally mandated); the second focuses on qualitative and effectiveness assessments. NGO sets the parameters for all of this, again in accordance with its norms and objectives. Effective analytics can convert raw data to the factors in the production of a score. A score here represents a relational system in which all companies in the analytical system are compared to others. Numbers provide the coordinates for that comparison. That comparison plan can be two or multidimensional, depending on the computing power and modelling skills of those who put this together.

Assessment is possible only after raw data is converted into numerically relational fields through normatively dependent applied analytics. Assessment can be understood as the process of framing analytics. It gives meaning to relational measures. In effect it is the process of providing the characteristics of the X,Y, Z, etc. axes on the relational field of measures extracted through analytics. Assessment can also be used to reduce spatial comparison to a flat field. That, effectively, is the purpose of rating systems—to flatten complex analytics into a linear system of assessment by assigning an aggregate symbol (usually a number) to a complex analytic systems which preserves a gross relationship quality (e.g., 1 is less than 2 is less than 3, etc.). This flattening is also a product of the application of norms and objectives to the production of relationship fields. Here the process of assessment is fairly simple. On the basis of NGO’s analytics it has reduced its assessment of corporate compliance, policy and implementation to a “score” that ranges between 1 (company does nothing) to 100 (the ideally engaged company).

Most rating systems stop at this point. Such systems produce the scale, it places all companies in its data set within the scale, and it makes the final product (the scale) and sometimes even the factors in its analytics and the sources of its data (though rarely its data sets) publicly available. The expected results are well known, and most people in this room have had a hand in some part of this system: People make decisions about relationships with companies on the basis of the well-publicized rating, and business conforms behavior to the rating standards. These sorts of ratings systems have been quite effective especially in the university sector and in the context of corporate credit ratings. It is now being used by human rights and sustainability actors—and thus my example here—though it is far too early to tell the cumulative effect.

And it is the effect that drives actors such as NGO from law to data driven assessment. If it works well, law provides the framing for the system, and sometimes serves to generate data. That is the result in our hypothetical. Sometimes law provides the normative basis, or at least its principles. This has certainly been the object of soft law in the human rights and sustainability fields. But the most immediate regulatory effect is not a function of principle or structure but in the details of the standards that serve as the basis of analytics. The standards, then become the regulatory norms to which companies must conform, at least if they take the consequences of ratings seriously. In other words, the standards (and their construction) have regulatory effect, norms (laws) have constitutive and structural effect, and finally the conflation of normative and implementation mediated by data.

But one can take one last step to create a behavior driving ratings system. Once an assessment system is imposed as an overlay on analytics, the final step is possible. That requires the development of principles of judgment. That is, it is the step in which NGO can draw lines (conclusions) from its data driven analytics assessments. Here the process is fairly simple. On the basis of NGO’s analytics, NGO has reduced its assessment of corporate compliance, policy and implementation to a “score” that ranges between 1 (company does nothing) to 100 (the ideally engaged company). Again, the process for NGO is straightforward and value laden. NGO will develop a series of position papers and hold conferences of like-minded people (even better if they are of significant intellectual and policy status) to affirm that scores below 30 are unacceptable (on its 100 point scale), and that scores over 75 ought to earn some sort of positive result. With this data driven judgment in place, the rating agency invites others not merely to take the rating into account, but to take the ratings into account in a particular way. In the case of NGO it is likely that NGO will urge consumers and investors to shun companies whose scores are lower than 30 and to patronize companies whose scores are higher than 75. Perhaps NGO might even suggest sets of punitive measures for low scoring companies and privileges for high scoring entities. But discretion still lies with those who might be moved to act on the basis of the ratings.

It is only a small step from rating to social credit systems. The difference between them is the construction of an institutional structure for imposing consequences on judgments derived from data driven analytic assessment. Let’s imagine how one can convert NGO’s rating system into a social credit system. One thing it can do is to enter into agreements (or even memoranda of understanding) with the New York, Frankfurt, London and Shanghai Exchanges through which the Exchanges would agree to penalize within their rule system (up to instituting proceedings for delisting) for companies rating lower than 30 points. NGO might enter into understandings with large sovereign wealth funds that these would remove low scoring companies from the their investment universe. It might also help lending institutions establish a program through which high scoring companies might earn discounts on loans, or even induce company loan rating agencies to include the score in their assessment of creditworthiness. One sees a version of this already where insurance companies now rate insurability on the basis, in part of creditworthiness of the insured. The object, of course, is to hard wire consequences to the rating well beyond the narrow field of modern slavery. If it is possible to impose a system of restrictions and privileges around conforming to NGO’s norms and expectations for “correct” conduct in relation to appropriate measures to prevent, mitigate and remedy modern slavery in corporate supply chains, then it is likely that companies will be far more willing to change their conduct. Even more importantly, NGO can then become the authoritative source of norms and conduct standards in the field.

What I have just described would likely be hailed as a great step forward in the very worthy campaign against global modern slavery and the markets for humans that feed it. NGO would likely be hailed for innovation and states would be pressured to convert the assessment system and its consequences into legally mandatory rules. Indeed, there might well be a push to make disclosure more transparent and to compel intervention for companies with low scores. Lastly, there might well be a call to align the societal regulation of ratings with the legally mandated systems of civil liability by extending the scope of liability and perhaps relaxing a number of legal impediments—such as those relating to causation (from fault to risk), third party beneficiaries, statutes of limitation and standing rules. Perhaps even the rules of veil piercing may be relaxed based on score. Host states may require companies with low scores to pay surcharges for operation or may deny them licenses to operate; states may read their BITs to permit imposition of civil restrictions for investing companies with low scores. Companies with high scores may be granted expedited processing of applications and approvals. And transport companies could impose restrictions on low scoring companies who seek to use their services by way of agreements among themselves and the rating company. If, in return the scores of transport companies are enhanced by such activity then NGO’s objectives are enhanced. When completed, NGO will have established a social credit system structure. The only thing that is missing is the state. And in our hypothetical, the state is around, though at the margins—enhancing the utility of data markets, developing legal structures that align with rating system assessments, and facilitating the translation of judgment from the societal to the legal sphere.

And yet what I have described, in a western context, is the essence of the social credit system being developed aggressively in China since 2014. Indeed, I imagine that had I even once uttered the word “China” in this context, the reaction to the development of something that appears so natural and “western” would likely be different in a quite negative light when appropriated by the Chinese State and party. The very people who hail efforts like modern slavery ratings as an advance on a global human rights project, decry Chinese efforts to establish social credit systems on the same basis.

But it is only after one has engaged in this exercise, and has been confronted with the substantial allure of data driven governance already embedded in Western regulatory systems, that one might more dispassionately consider the issues of Chinese social credit systems, of their application to enterprises, and of the effects of their application to non-Chinese enterprises for example within the Silk and Maritime road. Let me consider each in turn. What should emerge is that beyond its challenges of implementation (even in alignment with its own principles), the core issues around Chinese social credit can be reduced to the political and economic norms which it furthers, and to the threat that its projection outward may pose to the norms and values of states and actors into whose territories or production chains these values-based-data-driven governance systems are projected. It is further suggested that it may well be that the rules of extraterritoriality might eventually apply to data driven systems to the extent it now applies to law.

China’s Social Credit System.

The genesis of China’s Social Credit system is by now well known. I will not burden you with its history but will take a moment to outline the norms that serve as its foundations, and the objectives which it is meant to attain. Its current genesis was announced in a now famous Chinese State Council 2014 Notice.[2] At its core, Chinese Social Credit institutions were meant to be a response, with Chinese characteristics, to what was perceived to be a deep challenge of society and culture around integrity and accountability. Both, in turn, were viewed as essential to the progress of the political and economic model as it assumed a prominent role on the global stage. These were to present a Chinese path toward problems that remained endemic in liberal democratic states. Its success, then, would also leverage Chinese influence in driving the global narrative of governance.

At its most basic China’s Social Credit system was meant to respond to what was perceived as a “two front” challenge. The first front touched on issues of societal improvement. This societal improvement would have to be understood in terms of the vanguard party’s fundamental obligation to guide the nation toward the establishment of a communist society. This, in turn, touched on the development of the norms that were to serve as the foundation of the political and economic order. Those norms increasingly were thought relevant only to the extent that they reflected and advanced contextually Chinese values (understood in relation to their alignment with the values advanced generally through the political-economic model that basic principles of which were enshrined in the Communist Party Basic Line). Even as Chinese elites thought through these normative challenges, they were developing what would emerge in 2012, two years before Chinese Social Credit emerged out of the shadows, as the twelve core socialist values. These were characterized as a set of moral principles summarized by central authorities as prosperity, democracy, civility, harmony, freedom, equality, justice, the rule of law, patriotism, dedication, integrity and friendliness.

The second front was centered on the evolution of the political model at the start of what was seen as the current stage of China’s history. To those ends, it was increasingly thought that China had absorbed all that was of value from foreign sources, and that it might be time to express Chinese objectives in ways more compatible with the Chinese context and with its political and economic model. There was an increasing sense that the liberal democratic machinery of law and regulations overseen by a burgeoning administrative bureaucracy had not produced commensurate forward movement, and that Leninist forms of regulatory organization were necessary to move forward the Marxist political project.

In the face of the wildly successful project of socialist modernization, by the 18th CCP Congress in 2012, it became clear that both societal improvement and political evolution were the most visible elements of China’s transition to a “new era.” That new era centered the search for a Chinese path in society, economics, and politics, that was meant to meet and match that of its greatest global rival, the United States. For the American Dream there was to be the realization of a “Chinese Dream;”[3] and likewise a socialist rule of law, socialist core moral values, and socialist law. It was in the context of the later that social credit systems filled an important political objective. New Era thinking also converged with a change in leadership, to that of Xi Jinping, that also ushered in a new Chinese approach to the political model. That reform now emphasized systems grounded on the relationship between leadership cores and collectives. The core-collective binary, in turn, was to shape all social, political and economic relations, both domestically, and though the Belt and Road Initiative, internationally as well. And there was an intention that within its own part of the global order, China would serve as the core.

Social Credit was not meant to reshape governance completely but was to attack the problem of “integrity” within the political, social, and economic model. It was understood that integrity was tied to social corruption (and also to the corruption campaigns then being ramped up) that required a substantial intervention in the cultural basis for societal interaction. At the center of that intervention was both the vanguard party and its moral project tied to the transformation of the political and economic model in the new era. Thus, as the 2014 State Council Guidance noted, the objective was in part to “persist in correcting unhealthy trends and evil practices of abusing power for personal gain, lying and cheating, forgetting integrity when tempted by gains, benefiting oneself at others’ expense, etc., and establish trends of sectoral sincerity and integrity.”

Taken together, one sees in Chinese Social Credit, beyond its specific politics and techniques, little more than a ramped-up version of the model NGO was able to create around the issue of modern slavery. Both share a string connection between ideology and modeling; both share the mechanics of institutionalization; and both share the approach of data driven analytics to operationalize principles that have been given concrete form through the choices of data and the standards for their analysis. Most importantly, both amalgamate the legislative, executive and judicial authority of governance thorough the seamless connection between data, analytics, judgments and the real time imposition of consequences (restrictions and privileges) tied to behaviors that contribute to rankings that then contribute to assessment and consequences. The one difference that may be critical is the centrality of the state (under the leadership of its vanguard party) in the construction, implementation and projection of Social Credit power.

Most attention has focused on the application of the Social Credit regimes to individuals. This is not the place to consider that engagement in detail other than to suggest its contours and effects on business directed social credit. Social credit for individuals exhibits many of the characteristics NGO teased out in our modern slavery hypothetical. It is based on the premise that individual behavior is subject to a number of core moral principles, that those principles can be reduced to everyday behaviors, that those everyday behaviors may be measured against behavior ideals, and that these measurements may produce accurate assessments of individual compliance with their societal obligations. Most important, once reduced to a score, these summary judgments about a person’s compliance ought to produce effects—privileges and restrictions depending on ranking that they have earned. These privileges and restrictions need have little direct connection to specific behaviors but are developed through an network of agreements among public and private actors to piece together black lists and red lists that produce incentives toward better behavior and higher scores. The system is itself cobbled together from a series of national and provincial programs, from public and private data harvesters and a series of MOUs that are meant to nationalize lists and develop a coordinated system of evaluation (based on aggregate data) and consequences based on ranking demining placement on blacklists and redlists. The consequence were widely circulated stories in the West about people denied express train and air travel, college admission for their children, restrictions on access to credit and visas for travel abroad on the basis of social credit scores. Also widely publicized was the way that the analytics appears to incorporate actions that would have been weighed differently in the West.

The Turn Towards Enterprise Social Credit.

Given these similarities, it should not surprise anyone that Social Credit would turn its attention to economic activity and the behaviors of economic enterprises. Again, it is important to place this in a broader context. The West has been almost obsessed with similar conduct for the last quarter century at least. The focus, however, has been different. As our hypothetical suggests—Western stakeholder interest has focused on human rights and sustainability including climate change now. Chinese interest has focused on aligning business culture to its core socialist values, especially integrity, and to embed economic activities more positively within the vision for integrated economic and political activity under the leadership of the vanguard party.

This application to business was announced in 2014 along with the establishment of social credit for individuals. The latter, of course, has received far more attention in the Western press and among Western civil society organs. In some respects, it has been easier to build because the structures of data harvesting and the ability to develop systems of restrictions and privileges were easier to sketch out. But it is the application to economic activity that ought to merit more attention.

The approach was nearly identical to those applied to individuals, and outlined in the NGO hypothetical with which was considered earlier. One starts with a set of core values and objectives. In this case it is directed toward integrity in business (generally understood in similar ways worldwide). Then a set of specific behaviors must be identified with respect to which the integrity objective is to be directed. These in turn underline the approach to integrity (its political and moral character) that drives the standard setting entity (in this case the State and the Vanguard Party). The key, though, is the identification of the norms to be furthered, and their connection to specific behaviors targeted to that objective. Also central for the analytics is the system for valuing these behaviors against each other for the purpose of coming up with some unified assessment.

Those behaviors cut across a number of regulatory and data collection sectors. So the next stage requires some thought about two things. The first is the way that data is to be identified and harvested, The second is where such data is to be analyzed and among whom is it to be shared. Lastly, such data, processed or not, must be contributed to produce a flat measure—our rating from the hypothetical. That measure, in turn, becomes the core element for the development of judgement measures. That is, the construction of rating lists is made effective by tying the number ranges to a set of coordinated systems of privilege and restraints triggered by specific scores. Those privileges and restrictions are then circulated among those responsible for these consequences through the circulation of lists. Blacklists are lists of persons identified for restrictions. Redlists are meant for those entitled to privileges. The issue of coordinating public and private entities with the obligation to enforce red and black lists becomes a matter of agreements among governmental entities and between the public authorities and private actors. Many of the privileges and restrictions involve the state. They include permissions, procurement, applications, reviews and the like. None of this is perfect. And as one can imagine, until the entire process is somehow centralized, or coordinated, coherence is an ideal likely to be unrealized. That coordination might require substantial analytic power—and here there is a place for AI systems. But that is also beyond the scope of my remarks today.

Much of this was made visible in two central government documents that circulated in July 2019. On July 16, 2019, the General Office of the State Council issued the "Guiding Opinions on Accelerating the Construction of a Social Credit System to Build a New Credit-based Supervision Mechanism" and proposed key policy measures. These called for an accelerated program of constructing the social credit system for enterprises. It focused on several principal policy areas. The first was innovation in the connections between enterprises and the credit commitment systems, which in some measure are meant to connect enterprises to data retrieval systems. This credit commitment system is to be operated by the Chamber of Commerce and Industry, a mass organization under the leadership of the United Front Work Department of the CCP and a constituent organization of the Chinese People’s Consultative Conference (CPCC). The New York Times recently reported that the scope of data acquisition is quite broad: “court decisions, payroll data, environmental records, copyright violations, even how many employees are members of the Communist Party.”[4] The second was a plan for integrity education. This appears to focus on providing information to enterprises about the sort of behaviors expected of them. The third was to expand the application of credit reports. Credit scores will be used to affect “matters of government procurement, bidding, administrative approval, market access, qualification review” and the like according to the Guiding Opinion. The last was to use information gathered to establish a list of privileges and restrictions based on recorded credit behavior. All this credit tracking is to be undertaken by a number of national, regional, public and private entities, and all coordinated in some way by the state. Coordination would include a market supervision complaints report hotline and in formation platform. In addition, companies would be encouraged to voluntarily share additional information with respect to market operations, contract performance, social welfare and other information through a “Credit China” website, or related sites . Such voluntary disclosures would be considered as part of the credit evaluation. This is to occur at least in part under the supervision of the National Development and Reform Commission, an agency under the State Council. A national credit information sharing platform is meant to aid in the coordination of data harvesting, analytics, assessment and judgment producing a credit score that can then be used to impose restrictions and grant privileges.

The State Council Guidance points to the character of restrictions as well as privileges. Among the restrictions contemplated are more frequent inspections. The National Development and Reform Commission has been given the portfolio to complete two tasks in that respect. The first is to oversee “construction of cross-regional, cross-industry, and cross-disciplinary joint disciplinary mechanisms for untrustworthiness, and fundamentally solve the problems of recurring and easy-to-existence of untrustworthy behavior.” The second is to develop the catalogue of restrictions and discipline. These include “restricting the issuance of untrustworthy joint disciplinary object stocks, bidding and tendering, applying for financial funds, enjoying tax incentives and other administrative disciplinary measures, restricting access to credit and flying Market-based disciplinary measures such as high-grade trains and seats, as well as industry-based disciplinary measures such as criticism and public condemnation.”

The resulting system—a more complex and comprehensive system modeled on our initial hypothetical, is to be developed initially through pilot demonstrations. They are likely to be layered together so that eventually a coordinated national system can emerge.

The Expansion of Enterprise Social Credit to Non-Chinese Enterprises.

It is this system that is meant now to be applied to foreign firms. That possibility prompted some concern in the West. These concerns were nicely summarized in the 2019 Report of the European Chamber of Commerce in China.[5] The Report makes for excellent reading. And it is accurate. But it ought not to come as a surprise. The implications were already well developed in the hypothetical with which I started this conversation. Rating systems, especially when combined with assessment and consequences following judgment from a placement on the rating scale are meant precisely to change behavior. The metrics of such systems are meant to serve as the manifestation of regulation; they give effect and enforce the principles and norms written into law. They are indeed the substitute for and drivers of law. Such metrics are meant to transform data and its analytics into the quantitative expression of qualitative principles. Those who rate, those who analyze, those who assess, and those who judge become, in the last analysis, the organs of state most intimately involved in the governance of the things rated. Western companies now have a long history of involvement with such ratings and assessment systems, as well as with the strategic navigation of the consequences of rating. The resulting compliance and risk management has become a way of life for Western companies.

So why the concern? That, too, may be at some level easy to explain.

First, the normative basis for constructing and operating the system are not those of Western companies or the states from which they operate. Indeed, in many respects, they are incompatible with political and economic notions at the heart of Western political and economic organization.

Second, at least as applied to Western companies, the data harvesting may be redolent with ulterior purposes. These may include everything from tech and business secrets mining, to know-how transfers and disguised indirect control of business decisions. Every business decision, in fact, that may annoy a rating producing agency, might contribute to a lowering of credit scores. This is particularly sensitive in the context of government inspections that are part of the social credit mechanisms for business.

Third, to the extent that local agencies might wish to use the rating system strategically, then differentiated data harvesting and application of privileges and restrictions from credit rating could be used as a disguised protection of local enterprises at the expense of foreign ones. That has been a problem in certain areas of China even in the absence of social credit. But the system—a product of fracture at the bottom and coordination at the top, could also be used to enhance protection of local interest the way the judicial system sis sometimes said to do the same.

Fourth, in a sense, the principal worry is that in the absence of a robust social credit system for government, at least government below the level of the central authorities, it is not clear that the integrity principles at the heart of the social credit system, or the robust application of the core socialist values, will actually be embedded in the actions of lower level authorities (public and private) charged with the operation of the business social credit system.

The Challenges These May Pose.

My time with you is almost up. Let me conclude these remarks by suggesting a number of issues that might well be considered as enterprises and states work through responses to the extension of the enterprise social credit regimes t non-Chinese companies.

First, it is only fair that all enterprises operating on Chinese soil be treated the same. The principle of equal treatment has long been a staple of Western requirements in their investment and trade treaties (if not also most favored nation status). That extension, then, should not be viewed as remarkable or odd. But that is also the standard that perhaps ought to be protected, that is equality of treatment in every respect. Thus, social credit is not inherently odd or threatening in theory; it is the practice of it that changes the risk profile of operations.

Second, if, indeed, ratings and assessment will introduce far more intrusive regimes, then non-Chinese companies will have to begin to consider the extent to which they mean their operations to be embedded within China. It has already been the practice of some to minimize risk of exposure by fine tuning the scope of operations in states where national law would reach into trade secrets, know-how, or financial records. Such risks will have to be evaluated. The considerations will be different for European than for US companies, since the latter may already be starting the process of decoupling operations in light of both America First, and the likely result of the conclusion of trade talks.

Third, Western companies are already quite exposed to data mining. Our hypothetical suggests that much information that may be of use to the Chinese social credit system may already be available for harvesting outside of China, or may be accessed or purchased by Chinese institutional actors. To some extent, Western tastes for rating systems itself will fuel the availability of a more robust Chinese social credit environment. And that produces a great irony since it may be that the great human rights and sustainability advocates in the West will become complicit in the construction and maintenance of social credit regimes that can extend their reach beyond the national territory of China.

Fourth, it must be remembered that all such social credit systems, like the domestic legal orders of any states, are primarily deployed as instruments to project the political and normative objectives of those in control of their mechanics. It is too late in the day to complain about China for being China. The Chinese have never hidden their politics or their norms. That other groups thought they could bend Chinese state and party actors to their own advantage makes no difference in that calculus. That they might continue to believe that this is possible only makes matters worse.

Fifth, that insight produces powerful consequences. One is that social credit (like our modern slavery hypothetical) will be used as instruments to challenge the values of competitor states. In this case those are the values on which Anglo-European globalization have been created and are now operating. One ought to think carefully about that as one engages in market competition with increasingly political consequences. Another is the effect of social credit on operations beyond China. Take a small but potent example, last year U.S. airlines were told that their designation of certain places as Chinese territories or not would affect their social credit scores. Their social credit scores of course would affect their lucrative routes into China. But compliance would also affect the way they operated worldwide.

Sixth, social credit is not limited to the territories of China. The Belt and Road Initiative has extended the scope of the operation of Chinese enterprises to virtually all areas of the world. These companies will have to comply with social credit regimes globally. That might also extend those social credit responsibilities indirectly to all of their partners, business associates and the like. In that respect, social credit will go global, especially respecting the harvesting of data and the ability to associate with firms outside of China which would have a lower social credit score.

Seventh, our opening hypothetical also suggests the extent to which Chinese social credit regimes can be hardened along the Silk Road. That can be accomplished in a number of ways. Let me suggest two. The first is the way that the hypothetical suggest—by developing contractually enforceable arrangements among stakeholders in BRI territories. The second is through Memoranda of Understanding among BRI states that may or may not be part of the public Investment and Trade Treaties and other Agreements among BRI states that permit, in effect, extension of Social Credit regimes into “Road” states. Italy may well prove to be a good place to see how this develops.

Eighth, China is not the only state that can develop social credit. I have already suggested that the bones of social credit regimes are also firmly established in the West. While the West may have little taste for such regimes in the hands of the state, they are open to them in the private sphere. And when coupled with national defensive or blocking legislation—which is sure to come when the negative effects of social credit regimes are felt outside of China—these regimes will pose serious challenges for the coherence of global trade. More likely it will accelerate the decoupling of global trade around core and collective groupings of states within which global production will be rerouted.

Ninth, all of this suggests not the weakness but the power of social credit as the next important regulatory mechanism. Compliance, risk management and data driven governance are here to stay. The evolution of their characteristics will be the great task of lawyers, coders and policymakers for the coming decades.

Social Credit and Foreign Enterprises Along the Silk Road

Let me end by suggesting the thrust of the several points I have tried to make here tonight.

First is that Chinese Social Credit is in reality neither unique nor uniquely Chinese in its broad outlines. It emerges as part of a broader and deeper cultural dialogue that appears to be transforming the landscape and language of law. It is not the Chinese but advanced Western liberal democratic states that have begun to embed notions of markets based regulatory governance, and its modalities of compliance, disclosure, and risk management in its economic and political model. It is the West that has pioneered an ideology of prevention, mitigation and remedy as the basis for institutional operation, and that has pioneered regimes of data mining, analytics and markets based consequences in the form of privileges and incentives.

Second, the essence of ratings regimes is the fundamental connection between basic social and political principles and norms, on the one hand, and the construction of systems of data and analytics on the other. Ratings systems are no more than mirrors of a society’s moral and political choices. And to that respect they will be lauded or despised to the same extent as the underlying political-social model is embraced or rejected.

Third, the distinction between Chinese and Western social credit impulses is one of scope and systems of control. Such ratings systems remain tied to markets and the private sphere in the West, with generally direct and indirect regulatory responses by the state. In China, the state and its vanguard Party stand at the center of the system. Where the system is fractured and multi-directional in the West, in China it is to be coordinated and tied directly to state-directed political-economic-and social goals.

Fourth, the direct connection between the state and the ratings-social-credit system that is what triggers anxiety in the West; augmented by the modalities through which that exercise of control is manifested. That anxiety is in part a reflection of the central incompatibilities of the political economic model of liberal democratic states with those of a Marxist Leninist system. That difference expresses itself not just in the context of social credit, but in the organization of global trade and in the relationship between markets, private and public ordering and in the nature of the role and function of the state and its leadership organs.

Fifth, social credit regimes, then, are symptomatic rather than generative. The effects of social credit regimes on enterprises, then, reflects these generative differences. While they are essential for the construction of “new era” governance in China, and are compatible with its development of a Leninist methodology of law and social control, they pose substantial challenges for the West and its own principles of governance, politics and societal relations with the state.

Sixth, the challenges posed by these generative differences are likely to shape the nature and extent of Western investment in China. In the face of the realities of Social Credit, foreign enterprises will have to again exercise caution in the way they develop their footprint in China, and the way in which they shape their relationship with Chinese entities abroad. This is not a criticism but a reality of the problems posed as two distinct global political-economic models again arise.

Seventh, there is a moral for states as well. Liberal democratic states will have to exercise caution in the way that their trade and investment arrangements are now made, not just with China, but with states that are part of the Silk Road. At the same time, China will have to be sensitive to what will be an emerging pattern of blocking legislation for its effects on the integrity of its Social Credit regimes and its ability to project its reach along the Silk and Maritime Road.

Lastly, all of this suggests that contemporary analysis still has a long way to go to understand what people are attempting in the West, what China is attempting to do as it evolves its own economic-political model within its own core-collective sphere, and what will need to be done to bring some sort of possibility to continued robust engagement in the coming decades.
Many thanks.


[1] SustainAbility, “Understanding the Universe of Corporate Sustainability Rankings,” available (“There are a dizzying number and variety of external ratings, rankings, indices and awards that seek to measure corporate sustainability performance. Stakeholders of all kinds – investors, consumers, employees, etc. – are increasingly relying on these ratings to help inform their decisions (to invest, purchase, work, etc.). Companies also rely on such ratings to gauge and validate their own sustainability efforts, with some even linking management performance evaluation and compensation to external ratings.”). For the Report see, Christina Wong, Aiste Brackley, and Erika Petry, Rate the Raters 2019: Expert Views on ESG Ratings (SustainAbility, 2019). Available

[2] State Council Notice concerning Issuance of the Planning Outline for the Construction of a Social Credit System (2014-2020) 社会信用体系建设规划纲要

[3] This was understood as a restatement of the CCP Basic Line of the objective to realize a moderately prosperous society within the framework of the CCP’s Basic Line. One official website captures the spirit well:
In 2017, China made remarkable achievements in various areas, coming one step closer to realizing the Chinese Dream of national rejuvenation. The country's GDP rose to 80 trillion yuan (about 12.3 trillion US dollars), over 13 million jobs were created, more than 10 million rural residents were lifted out of poverty, the first Chinese-built aircraft carrier was launched, and the quality of the environment has improved. China has also moved closer to the global center stage, playing an increasingly prominent role in world affairs. As China celebrates its 40th anniversary of reform and opening up in 2018, President Xi Jinping vowed to press ahead with reform until the ultimate triumph.

CGTN , New Era for China, available

[4] Alexandra Stevenson and Paul Mozur, China Scores Businesses, and Low Grades Could be a Trade WQar Weapon, New York Times 23 September 2019. Available

[5] European Union Chamber of Commerce in China, The Digital Hand: How China’s Corporate Social Credit System Conditions Market Actors (2019. Available[709].pdf.



No comments: