“Visibilizing the Invisible Hand”

Michael W. Dowdle¹

[Chapter 12 from Michael W. Dowdle, ed., Rethinking Law and Public Accountability (forthcoming)]

(Please do not cite without permission.)

Introduction:

This Chapter explores the history and analytic limitations of the traditional, what I will call “regulatory,” model for constitutionalism. It will show that in the United States, this model, which conceptualizes accountability in terms of mass-based electoral democracy wedded to a bureaucratic command-and-control structure, was actually the by-product of processes of late 19^th century industrialization, and presumes a rationalized, transparent “regulatory” environment consistent with what James Scott has termed “modernism.” Not all such constitutional environments are so rationalized, however. For such “non-regulatory” environments, I propose a method for examining for constitutional accountability that proceeds inductively – i.e., through the construction of patterns out of a collection of particularized instances – rather than deductively. I then use this inductive approach to critique a recent analysis of a non-regulatory constitutional environment – that of China, showing how an inductive analysis visiblizes significant accountability processes that the more traditional, regulatory constitutional model misses.

I. The “Regulatory” Model of Public Accountability and its History

Contemporary analyses tend to conceptualize constitutional accountability primarily as a two-part construct. The first part sees public accountability as ultimately flowing from processes of electoral democracy. For those political actors directly subject to periodic election, the processes of being elected by an informed polis via direct, free, fair and universal vote is believed to provide the popular sovereign a necessary and essential opportunity to respond to and, if need be, discipline these public servants for errant political frolics. The second part addresses those political actors who are not elected. It says that such persons can still be made democratically accountable by subjecting them to the bureaucratic command and control of persons who are subject to direct election.

We will refer to this as a “regulatory” conception of constitutional accountability, in order to reflect the idea that embedded in this conception is a presumption that the constitutional environment – the social space that is structured by the constitutional order – is organized in such a way as to be directly comprehensible to a centralized administrative apparatus and directly governable by that apparatus through a rationalized framework of universal rules.² At the core of this regulatory conception of public accountability is a phenomenon that James Scott has recently termed modernism.³ Scott defines modernism is a process by which the central state structures local social activity so as to be comprehensible – “visible” in his terms – to central administrators who lack intimate knowledge of the dynamics of that locality (what he, following Clifford Geertz, sometimes refers to as “local knowledge”).

Scott cites the typical gridded street plan as good example of modernism.1 Pre-modern cities and communities generally sported eccentric, non-rationalized street designs, designs that reflected, indeed were the product of, the particular transportation patterns dictated by local life. However, these street designs frustrated the development of centralized, state administration that began to occur in Europe during the enlightenment. Centralized administration required administrators who did not have direct knowledge of most of the communities under their charge to nevertheless be able to readily locate, distinguish and keep track of a variety of local entities. This proved difficult in environments in which each locale sported a unique and complex road design. Therefore, national governments across Europe began standardizing urban street design into the Cartesian grids that dominate urban environments today. Such grids allow central-level administrators to more readily locate and keep track of relevant local actors without having to learn and keep in mind the unique, complex street layout of every community in their jurisdiction.

Both democratic and bureaucratic components of the regulatory conception of constitutional accountability are founded on modernist presumptions of societal “visibility”. The democratic component, the component that equates accountability with electoral competition, presumes that the major, constitutionally-significant interests to which government must address itself can and will be made comprehensible – i.e., visible – to even those elected governmental officials who lack direct involvement in these interests through nationalized, rationalized electoral procedures. Relatedly, it also presumes that the essential behavior of elected officials will be visible and comprehensible to an electorate that lacks participatory experience in governmental decisionmaking. In a similar vein, the bureaucratic component presumes that the rationales and effectiveness of the lower levels of the regulatory bureaucracy will be comprehensible to elected superiors, and through them to a general electorate, who lack any direct participatory experience in the regulatory environments of these lower-level administrators.

A. History and Development of the Regulatory Model of Public Accountability

The problem is that neither modernism nor its particular, regulatory structurings are inevitable or even essential features of constitutional society. For the first 100 years of American constitutionalism, the dominant model for constitutional accountability was not the regulatory conception, but a more res publica – or “republican” – conception in which constitutional accountability stemmed primarily from exposure to the enlightened, public-minded intelligentsia that the constitution created. Indeed, the framers, and maybe the even country as a whole, would most likely have disagreed with both components of the regulatory model of public accountability. The founders’ belief that electoral politics corrupted rather than engendered accountability to the public good is well recognized. Their suspicion of bureaucratic modes of command and control is less explored, but still fairly evident: the institutional devices that collectively define our particular constitutional structure are overwhelmingly anti-bureaucratic, in that their principal purpose is to disperse and derationalize power whereas bureaucratic command and control demands a unified and rationalized power structure.1 The anti-bureaucratic aspect of the American constitution was so pronounced that the theoretical father of bureaucratic rationalization, Max Weber, found it completely incompatible with the needs of a modern administrative state.

The framers’ rejection of what would later become modern visions of public accountability is sometimes dismissed today as the product of elitism or classism.⁶ But such dismissals invariable overlook the degree to which today’s democratic society depends on a prior modernization. There is good reason to suspect that in the pre-industrial world of the framers, today’s regulatory model of public accountability simply would not work as intended. As first and most famously described by de Tocqueville, citizens of the fledgling country operated within a highly localized political environment. This localization made the centralization and “visibilization” that is the defining hallmark of modern notions of public accountability impractical:

In great centralized nations the legislator is obliged to impart a character of uniformity to the laws, which does not always suit the diversity of customs and of districts; as he takes no cognizance of special cases, he can only proceed upon general principles; and the population is obliged to conform to the exigencies of the legislation, since the legislation cannot adapt itself to the exigencies and the customs of the populations; which is the cause of endless trouble and misery. [1/180]

It was not until the 20^th century that American conceptions of constitutional accountability shifted to what we herein are calling the “regulatory model.” This shift is directly attributable to prior changes in American society that made the modernist assumptions implicit in this model analytically viable. In the forty years following the Civil War, the localized character that had so defined antebellum American political society was effectively erased in an explosion of mass-production and industrialization. At the heart of this revolution was the development of forms of economic organization that fed off of economies of scale. This transformed the corporation from a localized creature of public law into a nation-sized creature of private industrial capitalism. This transformation catalyzed two corresponding transformations in American political society. First, the formation of powerful, nation-sized private economic actors unleashed a corresponding emergence of nation-wide interest groups, such as geographically dispersed workers who worked for a common company, geographically-disperse consumers who purchased in a national market, etc. Second, the formation of powerful, nation-sized private entities and organized interests required the parallel creation of powerful, nation-sized administrative entities to regulate them. Whether out of necessity (see, e.g., Max Weber), path dependency (see, e.g., Theda Skocpol) or simply out of lack of ingenuity (see, e.g., Charles Sabel), these administrative entities took the form of rationalized bureaucracies.

The success of this new style of centralized regulatory governance depended and continues to depend crucially on the corresponding restructuring of society brought about by mass production and industrialization. Modernized bureaucratic regulation succeeded in the US in part because its hierarchical, rationalized, juridified organizational structure meshed with and complemented the hierarchical, rationalized, juridified organizational designs that industrial organizations had earlier began adapting to control their growing organizational sprawl.⁸ In fact, turn-of-the-century industrial leaders often welcomed regulation, including regulation like workmen’s compensation and workplace safety regulation that appeared to work again their “class” interests, precisely because they believed that the rationalizing disciple of regulation would promote rationalizing efficiencies in their own internal operations.

Perhaps nowhere is the synergy between regulatory and industrial ideologies is nowhere clearer than in the ICC’s germinal 1910 decision in the Eastern Rate Case rejecting the railroads’ request for a rate hike. The railroad industry had been under considerable economic strain during the first decade of the new century, and many – including many within the railroad industry – attributed that strain to lack of working capital. For this reason, ICC habitually approved railroad requests for rate hikes. The Eastern Rate Case marked the end of this habitual deference, however, due to arguments developed by Louis Brandeis, whom the shippers had hired to argue against the requested rate increase before the ICC.

Brandeis presented the case against rate increase by compiling statistical data and expert opinion that showed that railroad management practices did not conform to those common to other, more economically successful industries, and in particular to Frederick Winslow Taylor’s newly theorized industrial practice of “scientific management”. Scientific management emerged out of managerial efforts to gain control over increasingly extended industrial firms. It did this by disaggregating corporate responsibilities into sets of discrete tasks, and then cataloguing, rationalizing and costing each of these tasks, theoretically allowing firms to better identify and correct particular sites of waste and inefficiency within their organization. Brandeis showed that as a whole, the railroads had largely failed to make use of principals of scientific management in structuring their operations. He argued that the railroads lack of working capital was therefore due to inefficient management, and that the railroads had therefore failed to demonstrate a real need for a rate increase.

Not only did the ICC embrace Brandeis’s argument insofar as the 1910 request was concerned, it also embraced scientific management as the model by which they would evaluate all future requests for rate hikes.⁹ After a decade of total deference to rate request, the ICC only approved one such request during the next seven years,¹⁰ asserting consistently that the railroads had yet to meet standards of scientific management demonstrated in other industries, and that until they did it would be wrong to make consumers bear the costs, in the form of higher rates, for this mismanagement.

In short, Brandeis used Taylor’s model of scientific management to visiblizes railroad operations to external ICC regulators. But what is even more interesting about this story for our purposes, however, is the effect that Brandeis’s argument had on the railroads themselves. After initial and angry denunciation, the railroads quickly began taking Brandeis’s argument to heart, with numerous railroads undertaking intensive internal studies on how to “Taylorize” operations in order to alleviate inefficiencies and waste. Within a decade, railroad officials were actually praising Brandeis and his Eastern Rate Case argument for impelling them to develop more efficient operations.¹¹

The story of the Eastern Rate Case shows how the bureaucratic efficiency emphasized by modern regulation complemented and synergized the productive and managerial efficiency emphasized by industrialization. Nor was this a unique occurrence. For example, beginning in the late 19^th century, industrial firms and organizations often supported workplace safety regulation – such as workmen’s compensation,¹² ICC regulation of railroad car couplers, and regulation of boilers – in large part because they believed that rationalized regulation in these areas would help rationalize and thus render more efficient their internal operations.¹³ The railroads supported the initial creation of the ICC for similar reasons.¹⁴ And despite the fact that the railroads were the clear losers in actual shape ICC ultimately took in 1887,¹⁵ the force of regulation’s rationalizing appeal was such that the railroads later supported efforts to strengthen the ICC’s regulatory capacities in the 1900s.¹⁶

Modern regulation’s early dependence on the products of prior industrialization is also apparent in the observation that the most effective early regulatory practices and standards were not constructed anew out of whole cloth, but tended to work with and through existing industrial practices and standards. For example, the single most important regulatory tool possessed by early regulatory agencies was the uniform system of accounts, an accounting standard developed by Milo Maltbie in the late 1890s. The uniform system of accounts required all firms within the regulated environment to keep uniformly detailed and integrated records of revenues, expenses and earnings of each part of their operations. The principal responsibility of early regulatory agencies was to prevent private firms operating in markets for public goods – like railroads and public utilities – from using inherently monopolistic tendencies of these markets to impose unreasonable charges on their customers. The uniform system of accounts was crucial to their ability to effectuate this responsibility, because it allowed regulators to visiblize and generate evaluations and comparisons of pricing and profits across and within firms in the regulated industry in a way that could be defended against political attack

The uniform system of account was really just a tweaking of the already established industrial practice of integrated cost accounting. Integrated cost accounting, an accounting system in which firms itemize individual sources of cost, expenditure and revenue, had become a popular practice among the larger, industrialized firms that emerged following the Civil War. By making the corporate activities of local actors more visible to remote, centralized supervisors, integrated cost accounting facilitated centralized control in these larger but more disperse industrial organizations. At the same time, the visibility benefits of integrated cost accounting accrued not only to firm managers, they also accrued to suppliers of firm capital. Capital markets and bank lenders both began demanding access to these accounts as a condition for investing in the firm.

All this resulted in an explosion in the demand for persons trained in integrated cost accounting systems. Such demand was met through a corresponding post-Civil-War explosion in the number of proprietary business colleges and public and private universities in the United States offering training in accounting. This, in turn, created a market for accounting textbooks. As increasing percentages of accountants received their education from these colleges or otherwise from these textbooks, accounting practices became more uniform. By the late 1870s, accountancy practices had become standardized and rationalized enough to support the emergence of accounting theory. One of these theorists was Henry Carter Adams, a professor of economics at the University of Michigan and one of the co-founders of the American Economic Association, who devised an integrated cost accounting system specifically for railroads and public utilities. It was from this system that Maltbie adapted his uniform system of accounts.¹⁸

Modern regulation’s dependence on the prior existence of industrial standards and practices is further evinced in the “case-by-case” strategy that the ICC and other early regulators used to develop regulatory standards.¹⁹ Under this strategy, the regulator focused principally on resolving individual disputes. In resolving such disputes, she would first consult with the effected actors in order to gain a feel for the particular industrial norms and practices that were at issue in that dispute. The regulator would then craft a regulatory (i.e., rule-based) response to that dispute by codifying the particular industrial norms and practices that best resolved the dispute to the regulator’s liking. In short, early regulators found that they were much more effective working with and through the existing practices and expectations of the regulated community, practices that were themselves the product of industrial organization.²⁰

The widespread use of this case-by-case strategy also evinces another aspect of the early-modern regulator’s dependence on prior industrialization. A critical component of that strategy was a continuous dialogue between the regulator and the regulated community, manifested primarily in the form of public and private hearings. Crucial to this dialogue were the private regulatory institutions – trade associations, labor unions, professional organizations – spawned by the growing coordination problems associated with industrialization’s ever more-extended market networks. These institutions “gave voice” to their particular component of the regulatory environment, serving as the principal organizational interfaces through which the regulator and regulated spoke to one another. It was through trade associations that both the railroads and shippers communicated industrial norms to the ICC, for example. .

The other component of the regulatory model of public accountability, electoral democracy, was also the direct product of mass-production industrialization. We noted above that the dominant model for political legitimacy during the 19^th century was “republicanism”: a model that saw political legitimacy as flowing primarily from the enlightened intellectual capacities of the national elite. But the real locus of government was at the local level, and at that local level, political accountability was demonstrated primarily by the local politician’s responsiveness and effectiveness in securing public resources for his community.

As industrialization and regulation pushed the locus of American government upward, they rendered the older, localized mechanisms of political accountability increasingly unworkable. Political attention thus turned to those political devices and practices that allowed political concerns to be transmitted more effectively to an increasingly remote political decisionmaker. One of these devices was the regulated election. Of course, elections had been an important component of the pre-industrial political system. But they had not been definitive of either democratic accountability or legitimacy in the way they are today. Electoral irregularities were not only common in 19^th century America; they were an accepted aspect of localized democratic competition.1 The urban advocates for electoral reform that began appearing in the last quarter of the 19^th century were not concerned so much with promoting “democracy” per se – indeed, early supporters of electoral reforms tended to conceptualize these reforms as necessary counter-measures to democracy. To them, the real purpose of these reforms was to inoculate urban government from more traditional, localized “democratic” politics.

It is not necessary for the purposes of this chapter to explore how the regulated election went from being thought of as a necessary counter-weight to democracy to a core element of democracy. What is important, and often overlooked, is the degree to which the effectiveness of this new ideal depended on the pre-existence of social structures that had been generated by industrialization.

This dependence is perhaps best evinced in the 1896 US presidential campaign. The 1896 presidential election, and in particular the campaign of the Democratic candidate, William Jennings Bryan, was the first national election in American history to feature many of the electoral dynamics that today we normatively associate with democratic accountability. Bryan’s 1896 campaign was the first presidential campaign in American history in which the candidate sought to appeal directly to a broadly constituted national electorate. Before then, presidential candidates functioned more as remote and symbolic figureheads for their respective parties. Such elections were largely divorced from policy issues and their results were driven primarily by regional party loyalties that in turn were driven by spoils and patronage considerations. Bryan and his supporters, by contrast, initiated a campaign in which Bryan appealed directly to a broad electorate by evoking a variety of policy issues – including those involving labor rights, monetary policy, and governmental reform.

While Bryan’s campaign was not successful in winning the Presidency, it was successful in securing for him the Democratic nomination for presidency in the face of opposition from the party elite. Eight years later, the Republican president Theodore Roosevelt would use this same general strategy – direct, issue-based appeal to the voting electorate – to wrest his presidency (which he inherited upon the death of William McKinley) away from the control of Republican party leadership. Hence was inaugurated the modern, democratic presidency – a presidency whose political power flowed directly from the president’s unique institutional capacity to appeal to and gain prestige from the electorate.

The success of Bryan’s (and Roosevelt’s) strategic mobilization of electoral democracy depended crucially on the pre-existence of social organizations and institutions that had formed in the wake of and in response to industrialization. Bryan mobilized the electorate primarily by appealing to three large, nation-spanning interests – silver interests, labor interests, and agrarian interests. Each of these nation-spanning interests had distinctly industrial predicates that allowed them to generate and organize around nation-sized organizational structures. Agrarian interests had earlier formed the People’s (or Populist) Party in order to address nation-spanning social disruptions caused by the early industrialization of American agriculture (particularly the introduction of large-scale cash-crop farming). Labor interests had earlier formed the Knights of Labor in order to address nation-spanning social disruptions cause by the growth of wage-labor. Silver interests were driven to considerable extent by an already organized silver producing industry taking over and uniting remnants of regional Greenback parties that had formed in reaction to the increasing monetarization of the economy that had been catalyzed by industrialization.

Bryan and his supporters used the superior mobilizing capacities of these organizations to simply overwhelm the political opposition the more established, leadership elements of the Democratic Party could mobilize through the Party’s traditional, patronage-based structures. The superior mobilizing capacities of these organizations, in turn, stemmed from the fact that they were national (or at least inter-regional) in scope. And the reason they were national in scope was because they were organized around common experiences and concerns generated by common interaction with the new, nation-sized economic and regulatory institutions generated by industrialization.

Another way in which industrialization enabled the creation of a workable electoral accountability system in the US was by facilitating modern public finance (including governmental finance and party/campaign financing). The biggest obstacle to both electoral and civil service reform was the spoils system – the system popularized by Andrew Jackson in which jobs in public service were the principal currency by which party leadership bought institutional loyalty and coherence. The spoils system impeded electoral accountability both by making the civil service more accountable to the private, corporate interests of the political parties rather than to the public interests of the people as whole, and by making the parties themselves more accountable to the interests of potential spoils-holders rather than to the interests of voters.

The spoils system itself was a creature of necessity. The problem was that political success is invariably a corporate endeavour, and that corporate loyalty invariably has to be bought with something. Access to public office was the only real resource corporatized political parties could use to encourage party loyalty. Even strong ideological opponents of the spoils system, like Theodore Roosevelt and Woodrow Wilson, found themselves dependent upon that system for political effectiveness once they achieved office.

The real death-knells to the spoils system, at least at the national level, came from the invention of modern campaign finance and the implementation of a centralized, direct-tax system – the Federal income tax.1 Both of these developments were only made possible by industrialization. The advent of a centralized, direct-tax system greatly increased the revenue the central government could use to fund civil servants – which in turn caused civil servants to become more responsive to elected government and less responsive to party. Without industrialization, a centralized direct-tax system would have been infeasible. As we saw above, industrialization catalyzed the development of accounting and auditing practices that are crucial for facilitating the central government’s ability to observe, monitor and tax income flows. Industrialization’s concentration of wealth and income into a smaller number of large financial organizations further reduced monitoring and collection costs. Its tendency to replace production-based sources of wealth with wage-based sources of wealth greatly expanded the federal tax base.

Modern campaign finance was the brainchild of Mark Hanna, a 19^th century industrialist who oversaw McKinley’s 1896 campaign against Bryan. During that campaign, industrialists especially were frightened by Bryan’s ability to mobilize disaffected elements of American political society. This made them disposed to contribute money to the Republican campaign, and the Hanna exploited this disposition by routinizing and industrializing the Party’s ability to solicit and make use of industrial campaign contributions. Because of the larger size of these industrial contributors, Hanna’s mobilization of industrial contributions provided the party for the first time with a relatively permanent source of capital. This, in turn, provided the party with an alternative to the spoils system – an alternative in which party loyalty could be purchased through access to capital (in the form of campaign contributions, salary for staff, etc.) rather than through placement in the civil service.

Of course, many will question whether today’s parties’ reliance on corporate donations represents a positive aspect of America’s democratic system. But what is important here is that regardless of how one characterizes it ideologically, the development of modern industrialized campaign finance was crucial to the institutionalization of much of whatever electoral accountability our present system does enjoy.

As with rule of law, the symbiosis between electoral accountability and industrialization is further evinced by looking at those particular environments in which electoral regulation and reform failed to have significant affect on political dynamics. Indeed, electoral reform efforts associated with the creation of our modernist electoral democracy often failed, at least initially, to have the effects intended of them. In the large cities, one of the principal beneficiaries of the pre-modern, machine-style democracy that these reforms were intended to disrupt was the distinctly “pre-modern” class of immigrant poor, a class whose tight community organization, very localized political existence and limited formal education made them much more able to manipulate the localized, relationship-oriented dynamics of machine-style politics than the centralized, rationalized and professionalized dynamics of a more regulated electoral democracy. This class largely resisted the electoral reforms progressive reformers had instituted in their benefit. The machine style urban governance that these reforms were intended to curtail often continued unfazed. What ultimately pushed these machines into obsolescence was the increasing “regulation” of inner-city life through the development of modernized social welfare programs. By offering the urban poor access to public resources that did not depend on municipal patronage, these programs alleviated these citizens’ dependence on the political machine. Shorn of its pre-modern basis of support, these machines quickly succumbed to the same superior mobilizing efficiencies that had allowed rationalized electoral democracy to colonized national politics a generation earlier.

The story is the same for electoral reform and regulation in rural communities. By the opening of the 20^th century, most rural jurisdictions had implemented the core modernizing electoral reforms advocated by Populist agrarian interests during the 1880s and 1890s.[1] But the political interests that dominated both rural politics and policy formation remained largely unaffected. It was only after establishment in 1919 of the American Farm Bureau Federation, a nation-wide, industrial interest organization devoted to promoting agriculture interests,1 that American agricultural interest began to have significant impact on American domestic policy.

B. Limitations of the Regulatory Model

The modern “regulatory” model of public accountability is a vision that rests solidly on the pre-existence of social and economic foundations built out of mass production’s mobilizations of economies of scale, and it is a vision that can easily collapse where this foundation is eroded too severely. Present-day evidence suggests that the effectiveness of modern, regulatory models of public accountability continues to be highly dependent upon prior industrialization.

The positive correlation between industrialization and the sustainability of liberal democracy is well documented.[2] A similar story, albeit one that is far less examined, can be told about the bureaucratic component of that model – what is sometimes equated with “rule of law”. One of the few comparative studies of the relationship between bureaucratic rule of law and economic development, Katharina Pistor and Philip Wellons’ The Role of Law and Legal Institutions in Asian Economic Development 1960-1995, suggests that “market / rule-based law might have a measurable effect on future economic development only after economies have reached a certain threshold of development.”²³ This hypothesis is further strengthened when we examine the experiences of international efforts to promote bureaucratic regulation in non-industrialized economies – i.e., “rule of law” initiatives. To-date, such initiatives have shown little success. The earlier rule-of’law movement of the 1960s is widely considered a failure. The more recent resurrection of that movement in the 1990s may have had somewhat more effective impact in promoting modern regulatory practices in some developing countries (due in large part to increased domestic policy reliance on international trade as a motor for economic growth), but so far the introduction of modern regulation in these countries has not had the kind of socioeconomic or political impact expected of it.

The symbiosis between public accountability and industrial modernization is evinced particularly well in the failure and then success of international efforts to help Thailand provide at least some degree of social welfare protection for vulnerable populations in the wake of East Asian Economic Crisis during the late 1990s. When that crisis hit in June of 1997, many in the international development community attributed it at least in part to a failure on the part of East Asian countries to adopt effective “rule of law” – what we are calling “regulatory” – institutions. Lack of effective rule of law, it was argued, allowed corruption and cronyism to corrode the efficiency of these countries’ economic and political markets. The crisis was simply a return to equilibrium that resulted when these inefficiencies finally became visible to these markets.

Early efforts to assist Thailand in the wake of the 1997 crisis were undertaken primarily by the Asian Development Bank (ADB) and the World Bank, both of which bought into this line of argument (which issued not only from Western analyses, but also from many Thai NGOs). As a result, both these institutions framed their assistance projects in rigorous bureaucratic protections – namely very detailed, rationalized standards for eligibility, project structure, and fiscal monitoring – that were intended to prevent the cronyism and state corruption that was supposedly endemic to this region from diluting the effectiveness of these projects.

The result was a failure. For example, in August of 1998, the World Bank instituted a “Social Investment Fund (SIF)” that was to immediately begin promoting local community welfare, conservation, and skill development projects. By the end of 1998, the fund had approved only 12 out of 836 applications. By June of 1999, the SIF had disbursed only 5% of its available capital. In July of that year, the World Bank itself admitted that as currently designed the SIF had failed and that the Bank would have to “restart” the project.

The failure of the SIF program was due primarily to the fact that rural Thailand – that part of Thai society that was most in need of social-welfare assistance – does not sport an industrialized economy. Thailand’s economy is dominated by small enterprises rather than large enterprises.²⁴ At any particular point in time, only 40% of its labor force are paid employees.²⁵ As of 1996, 49% of its labor force was employed primarily in agriculture, while only 13% worked in manufacturing.²⁶ A significant portion of the industrial workforce is migratory.

The non-industrialized character of Thailand’s labor system made it difficult to centrally identify and target at-risk rural populations. We noted above how the success of early regulation in the United States vitally depended on its cooption of widespread, uniform accounting and managerial practices spawned by industrialization. The SIF’s stringent, rationalized application and monitoring requirements presumed a similar pre-existence of widespread, uniform managerial technologies. Unfortunately, the vast majority of actors operating in Thailand’s largely non-industrialized rural sector had not adopted such technologies because, being embedded in localized, non-industrial environments, there was little need for them to do so. Consequently, very few of the localized, indigenous development and welfare organizations that were best placed to identify and assist these at-risk populations possessed the experiences necessary to allow them to comply with the World Bank’s stringent application and monitoring requirements.

Indeed, many local welfare organizations never even became aware of the SIF program. Again, we noted above that early regulatory success in the US was founded in significant part on that system’s ability to make use of the private coordinating institutions that had previously emerged in order to facilitate the more extended coordination needs for those operating in the larger social environments generated by industrialization. These private regulatory institutions were often the principal means by which the regulator communicated with the regulated. Lacking extensive industrialization of the agricultural sector, however, Thailand’s rural economies remained largely localized. Local welfare organizations thus never had need to develop capacity to readily communicate and coordinate with remote actors. Hence, systemic and comprehensive communication of the existence of the SIF program was functionally impossible, and many potential participants never learned about it.

By contrast, international social welfare assistance programs that abandoned bureaucratic, “rule of law” protections were much more effective in reaching these rural populations. In late 1998, Japan inaugurated its “Miyazawa Scheme” for promoting recovery in East Asia. In stark contract to the SIF, the Miyazawa scheme “abandoned all pretence of careful targeting [and] elaborate bureaucratic procedures.”²⁷ It simply “disbursed funds through local government bodies despite the fact that these bodies were often very new and supposedly prone to corruption by the locally powerful.” Nevertheless, the Miyazawa scheme was significantly more effective in providing social welfare security to vulnerable populations than the SIF program. Both the SIF program and the Miyazawa Fund had budgets of around 10 billion bahts. Both sought primarily to provide immediate and substantial assistance to rural welfare development. But whereas the SIF was only able to disburse some 5% of its funds in nine months of operation, the Miyazawa fund disbursed most of its funds within the same time frame. Moreover, there were no widespread reports of corruption or misuse of funds, and follow-up studies found that localized Miyazawa-funded projects were relatively well managed.²⁸

None of this is to suggest that the regulatory model of accountability is simply a myth used to perpetuate certain power hierarchies brought about by industrialization.¹¹ In fact, quite the opposite: as described above, key aspects of both the bureaucratic component and the democratic component of public accountability actually arose out of efforts to combat and counteract the power hierarchies of industrialization. Indeed, the universal popularity of the regulatory model in the United States, suggests that that model has at least been somewhat successful in crafting a political-constitutional dynamic that most Americans, including those opposed to industrial economic interests, today accept and respect. (See Harrington & Umat.) Nor is it to suggest that rural Thailand might not someday evolve so as to become more receptive to industrialized regulatory practices. What this history does emphasize, however, is that whatever success that model enjoys in effectively fashioning public accountability was and still is vitally dependent on the prior presence of a regulatory society that has been modernized and standardized. And not all constitutional environments are so structured.

The dominance of the regulatory state in Western political society over the last century has caused us to forget just how fortuitous is the development and perhaps even how delicate is the survival of the regulatory state. The success of both mass production and the regulatory state that has been constructed off of it may depend crucially on an “improbably stability” ¹² and improbable standardization of social and economic forces that simply does not exist in much of the world, its history, or even its future. It does not exist in the rapidly evolving societies of many developing and newly-developed countries, in which both stability and standardization find it difficult to gain a toehold in the face of the constant social change provoked by industrial or post-industrial development.¹³ It does not exist in national regulatory environments dominated by flexible production, at least that particular form of flexible production found in upstream production environments – what Frederic Deyo refers to as “quantitative flex” (meaning productive flexibility that derives from flexibility in employment obligations).¹⁴ Moreover, there is even credible suggestion that a destabilizing confluence of technology and communication advances and the explosion of transnational economic and political pressures they have unleashed may be collapsing regulatory stability and its attendant standardizations even in those parts of the world in which it has historically been enjoyed.

A vision of public accountability that pre-emptively excludes these kinds of regulatory environments from its analytic purview is of limited and decreasing utility today, and could even become obsolete tomorrow. Yet this is precisely what the regulatory model of public accountability does. By tacitly presuming the presence of social structures associated with modernism, that vision leaves itself with nothing meaningful to say about political - regulatory environments that have not been colonized by such structures or may otherwise be moving outside their ambit. To the extent that our interest in public accountability and constitutionalism more generally aspires to something that is greater than a conventionalist description of a small, accidental and perhaps even brief convergence of human experiences, we will need to come up with understandings of constitutional accountability that do not demand a modernized regulatory environment.

Some may argue that the quest for such a more expansive vision is in vain. Public accountability could well be a “spontaneous” phenomenon whose parameters exist outside the reach of rational manipulation. (See Scott. See also Mashaw, Rubin, Harrington & Umat.) There is strong appeal to this argument. The point here, however, is not necessarily that we need to construct some new mode of public accountability out of whole cloth, but that we always need to be alert to the new kinds of accountability that non-regulatory environments might spontaneously generate. (See, e.g., Morgan, Courville. See also Freeman.) The structures created by the “invisible hands” of spontaneous social phenomenon are not necessarily themselves invisible, they can simply be undiscovered. In the case of public accountability in post-regulatory and non-regulatory societies, what we need is a method for discovering these structures if, when and where they happen to emerge.

II. Visibilizing Public Accountability in a Non-regulatory State: Inductive Accountability

A. Deductive vs. inductive accountability

We will start our investigation into how we can go about visiblizing “invisible” modes of public accountability that might be operating in non-regulatory environments by first exploring more precisely what it is that makes the regulatory model of public accountability inappropriate for these environments. At its heart, this model revolves around a deductive mode of analysis. Deductive reasoning involves the extrapolation of particularized conclusions about a particular system through application of a set of generalized rules. In the case of the regulatory model of public accountability, this rule is in the form of “if elected or under bureaucratic control of someone who was elected, then constitutionally accountable.”

The limitations inherent in the regulatory model of accountability can be traced ultimately to this deductive character. As described above, what distinguishes a non-regulatory from a regulatory environment is the absence in the former of rational standardization. What this means, in other words, is that non-regulatory environments are environments for which no meaningful set of rationalized performance rules has presently been identified, and without a useful set of rules, useful deductive analyses are of questionable utility.

To say that an institutional environment is “non-regulatory” is not to say the meaningful patterns of accountability to not exist within that environment, however. Indeed, such patterns must almost certainly exist in any functioning institutional environment. At its most fundamental level, an “institution” is simply a stable pattern of coordinated behavior. The fact that the behavior is coordinated means that there must be some device or series of devices – what Nelson and Winter have termed “routines” – that cause actor A to perform act A’ whenever actor B performs act B’. Moreover, assuming that both A and B are volitional actors, then both A and B must themselves be aware of this routine and have confidence in it.

Non-regulatory environments, therefore, are not environments in which meaningful patterns of accountable behavior do not exist, they are environments in which some crucial portion of these patterns may be visible primarily to those to whom they most directly apply – a type of phenomenon that Clifford Geertz has famously captured in the term “local knowledge”. Note in particular that the qualifier “those to whom they most directly apply” excludes not only outsiders, but also insiders who for whatever reason are not involved with that part of the system. In this sense, accountability in non-regulatory environments is like the street patterns of a pre-modern city. Those looking at it from the outside perceive only randomness; others looking at it from the inside perceive a finely grained coherence but only in the limited portions of the environment in which they operate.

The problem of accounting for public accountability in non-regulatory environment thus boils down to a problem of how one goes about identifying systemic patterns in environments whose patterns are presently only visible to direct participants. To explore this conundrum, let’s again consider the pre-modern city. We first enter that city as an outsider. But after walking around a bit, we gradually begin to develop a mental map. We learn how to get to the market and how to get to the bank; we then learn the most efficient route to the tavern during dinnertime. At some point, we become what we might call a “limited” insider – someone whose local knowledge encompasses that portion of the city in which we directly operate.

As our explorations take us farther afield, our local understanding of the city expands. Moreover, if we are so inclined, we can begin communicating at least some of this presently local knowledge to those inside and / or outside the city who lack it. We can draw up street maps; we can write travel guides; we might author disquisitions that point out and explain the Arabesque architectural influences that appear in a particular section of town. Comparing with other cities that we know, or perhaps simply with similar, locally-inspired descriptions of cities we do not know, we might even begin to develop insights into how humans go about organizing their urban communities – insights that, paradoxically enough, could be used deductively to construct understandings of locales to which we have never been.

The cognitive process I have just described is called induction (or inductive inference). Induction extrapolates generalities from particulars, rather than the other way around. Induction is also the antecedent of deduction. The principles we use to deduce conclusions about particulars often emerged out of inductive analyses of particularized phenomenon. Cognitive studies have shown that induction is principal means by which individuals work to gain articulated understanding about phenomenon that defies deductive analysis. This suggests that the answer to the problem of how to account for public accountability in non-regulatory environments should similarly be approached by shifting the cognitive methodology underlying our investigations from one of deduction to one of induction.

B. Inductive Accountability¹⁹

As described above, inductive inference is a two-stage process. During the first stage, the investigator goes about identifying relevant particulars. During the second state, she identifies patterns that conceptually link these particulars. In constructing an inductive process for exploring for public accountability, we will examine each stage in turn.

1. Identifying Incidents of Accountability

Our first task in constructing an inductive understanding of possible, non-regulatory modes of public accountability is to find a way of identifying what might constitute “incidents” of accountability – i.e., those particulars from which we will later generate inductive patterns. In order to get a grasp on how we might do this, we might first consider how someone might seek to gain information about an opaque environment in which she has no direct experience? On way is through direct study – making herself an insider, as it were. But this is time-consuming, kind of like trying to find the post office by simply wandering the streets. More efficiently, she would probably just ask someone from the environment who she feels is likely to know the answer to the particular questions she seeks.

In a non-modern regulatory environment, the problem of environmental opacity is not unique to the environment’s outside observers, it is also a problem faced by the environment’s own members. Assuming the environment is of any significant size, members at least sometimes must coordinate with member from other localized areas that lie outside their direct experience, just like residents of one neighborhood in a city must sometimes go to other neighborhoods that are in the same city but with which they are completely unfamiliar. In such situations, they too would seek direction and local information by communicating with some other member with more local knowledge of that particular sub-environment. When the pre-modern federal government in the United States sought to establish a pension scheme for wounded northern civil war veterans (the first manifestation of the regulatory state), for example, they simply asked local doctors to tell them who in their locality was eligible. When Franklin Roosevelt sought to begin federally regulating American capital markets in the 1930s, he hired consummate market insider and known pool operator Joseph Kennedy Sr., to be the first head of the SEC precisely because of Kennedy’s inside, local understanding of how capital markets operated.

Consistent with this observation, the dynamics of “non-regulatory” constitutional governance are often conceptualizedin terms of discourse. And indeed, in the pre-modern world of antebellum American constitutionalism, the dominant model of constitutional legitimacy was founded on an essentially discursive paradigm: the “town meeting” at the local level; “republicanism” at the national level. We might also note that today’s renewed interest in “republicanism” and deliberative democracy, and the recent identification of new discursive regulatory paradigms such as negotiated rulemaking, and democratic experimentalism, have corresponded to increasing segments of the American socio-economic environment moving into post-regulatory forms of organization and governance.

We thus might approach the question of how to identify incidents of accountability in non-regulatory environments by thinking about how we conceptualize and identify accountability in discursive or deliberative settings. A discursive conceptual of accountability begins by assuming that the involved parties are more-or-less autonomous, at least with regards to each other.²⁰ Being autonomous, we cannot condition their accountability to outside standards. In this way, the autonomy assumption functionally captures the fact that in opaque environments, institutional responsibilities will not be obvious to outsiders. Thus, in a discursive setting, accountability has to be conditioned to standards of the participants’ own, individual choosing. This suggests that in looking to identify particulars from which to construct an inductive accountability, our first step should be to look for obligations and responsibilities that the actors themselves acknowledge in their interactions with each other.

Some may question how credible such acknowledgments really are. Before answering this concern more directly, I would first emphasize at this point that what we are developing here is simply a heuristic for identifying probable incidents of accountability. This heuristic does not claim that every such incident it identifies in fact is a product of functioning accountability mechanisms; nor does it claim to be able to see all such instances. But the point of our exercise is not to provide a precise mapping of such incidents – it is simply to provide an accurate enough identification so as to allow patterns of incidents to become perceivable. (We shall see below, the patterning component of inductive accountability helps overcome the epistemic impact of false positives and false negatives that might slip into our account.)

Returning to our “discursive” for identifying public obligations, there is good reason why we might reasonably trust confessional statements as giving generally true accounts of institutional responsibilities. First, non-regulatory environments tend to organize themselves around network rather than hierarchical institutional forms;²¹ and network disciplining tends to focus on participants’ reputation for keeping their word. If a participant within a networked environment falsely acknowledges or refuses to acknowledge a personal responsibility , her capacity to garner cooperation, support and power within that environment will be compromised. Therefore, such participants would generally be very weary of suggesting the existence of responsibilities that they themselves do not believe exist or have no intention of enforcing.

Of course, such articulation only evinces of the presence of responsibility. It does not evince a functioning accountability per se. Accountability is not simply about defining performance standards. Accountability’s ultimately aim is to compel or encourage actors to conform their actual behavior to these standards. In other words, an incident of accountability must evince, not simply an articulation of the relevant responsibility, but also evidence that actors are altering their behavior in ways that are consistent with these responsibilities. For this reason, our heuristic for identifying incidents of accountability must also feature a descriptive, behavioral component – a component that indicates that the accountability is shaping and conforming actual behavior to formal responsibility.

Notice, however, that this “behavioral” component of our heuristic does not demand complete compliance with the responsibility. It simply requires that the actor change her behavior to be more consistent with that responsibility. Also, the reason why the actor actually changed her behavior is irrelevant – it is the change itself, not the intention behind the change, that concerns us. I will address each of these points in turn.

First, the heuristic only demands some significant alteration of behavior, not actual compliance.²² This is in contrast to many analyses of constitutional systems, particularly comparative analyses of developing constitutional systems, which seem to demand either absolute compliance with a particular norm, or otherwise such compliance at such levels as found in advanced, developed constitutional democracies such as the US. The problem with this more rigorous standard is that actual compliance with a responsibility can be frustrated by factors outside the relevant actors’ or the institution’s control. This is particularly likely to be a problem insofar as non-regulatory environments are concerned, because as we have seen, such environments are subject to many more unforeseeable, non-visiblized forces than regulatory environments.

Indeed, often the institution itself does not actually intend its formal responsibility standards be met completely. This is true even for advanced regulatory environments. Kal Raustiala cites speed limits on freeways as an example. He notes “[s]peed limits are rarely complied with in a strict sense – most traffic exceeds speed limits by a comfortable margin – but speed limits appear to dampen traffic speeds nonetheless. In cabining some driving behavior the speed limits are effective, even if compliance with the legal standard is very low or even non-existent.”²³ Again, given the higher degree of non-visiblizable forces that buffer non-regulatory environments, one might well expect that a significantly greater portion of responsibility standards in such environments would of this kind.²⁴

Where an actor fails to meet a responsibility for reasons outside her and her institution’s control, we do not commonly attribute such failings to problems of accountability. Nor would we say that speed limits are ineffective disciplining devices simply because they only cabin and do not dictate vehicular speed. Similarly, in the context of American constitutionalism, Bruce Ackerman has recently noted:

Any robust philosophically-developed conception of ideal democracy . . . would lead a clear-minded observer to the conclusion that there has never been a truly legitimate democracy in the world. It has only been since 1965 that black people in the United States could vote. In our democracy right now, probably thirty-five percent of the population can't read well enough to make informed political judgments. Is this a legitimate democracy? This Whiggish idea – we are almost a well-ordered democracy, we are in collaboration with the People, we are almost there – it seems to me so banal, such an obvious form of wish-fulfillment. The truth is that this government does not fulfill the minimal conditions presupposed by Habermas or Dworkin - or, for that matter, the conditions I lay down in Social Justice in the Liberal State.

Does that mean that we should not continue the constitutional project? Well, I think that's a serious question. My tentative answer is that the constitutional tradition, defective though it is, nevertheless provides the best chance of pursuing justice when compared to the dismal alternatives. But don't suppose that the American project has already fulfilled some minimal conditions of genuine democratic legitimacy.²⁵

A behavioral standard that demands absolute compliance could thus well render crucial accountability aspects of a constitutional system invisible to us, both in non-regulatory and even regulatory environments.

Secondly, we noted that the reason why the actor chooses to act in conformity with the standard is irrelevant. This is also in contrast to many analysesthat tend to dismiss possibly progressive institutional implications of liberal political behavior where that behavior is seen to promote the personal utilitarian interests of the actor. Even under the best of conditions, efforts to identify motive are extremely suspect – in fact, one popular form of political analysis, that of “public choice,” argues that we should simply posit that all significant political behavior in every political environment is strategically self-serving. In more opaque regulatory environments, efforts to identify motive frequently devolve into exercises in tautology: postulating strategic motives from implications in behavior, and then turning around and attributing that behavior to the postulated strategic motive.

Moreover, many accountability systems actually work by aligning institutional responsibility with the strategic, personal interests of the relevant actors. Financial auditing does not work by making corporate executives less greedy and more moral. It works by making the costs of greedy and immoral behavior exceed the rewards. As his famous comparison of men and angels made clear, Madison constructed our own constitutional accountability system in express recognition that most politicians could never be made moral in thought, only moral in action.²⁶ Thus, consistency between conforming behavior and utilitarian self-interest can and frequently is completely compatible with accountability effectiveness.

We might also flag two other important concerns with the behavioral component of our heuristic for later discussion. The first is its post hoc, ergo propter hoc character. The simple fact that an alteration in behavior is consistent with institutional demands does not mean that it was actually caused by those demands. Note in particular that our heuristic does not require that the actor herself expressly acknowledge or even recognize a linkage between her behavior and a corresponding responsibility in order for that linkage to constitute what we are calling an incident of public responsiblity. This is because actors’ behavior can sometimes be motivated by “socialized information that the possessor knows but is not consciously aware that she knows.”²⁷ This would be particularly likely in non-regulatory environments. Not having been rationalized, the ordering of such environments is much more likely to be structured significantly by “spontaneous” processes – processes that operate outside of individual design. Operating outside of human design, we might well expect that some of the norms and routines generated by such processes would exist in the form of this kind of pre-articulate, “socialized” information. Support for this hypothesis comes from the observation that in the United States, many new constitutional norms have often been identified ex post facto from pre-existing patterns of experience and performance.²⁸

The other concern with conceptual efforts to link responsibilities with behaviors involves the phenomenon that Edward Said famously identified in the context of European analyses of Arabic and Middle Eastern cultures as “orientalism.” Orientalism refers to the propensity of analysts of other culture to tacitly and perhaps even subconsciously import their own cultural biases into their analyses. A famous example of this might be A. V. Dicey’s slander of French administrative law because of its failure to conform to supposedly universal forms of “natural justice.” Obviously, natural justice represents only one of a number of ways by which administrative behavior can be cabined to law. As it turns out, not only was the French system effective insofar as the French system was concerned, it may well have proved to have been more effective globally than Dicey’s own English system. Dicey’s belief that the doctrinal demands of natural justice were universal was simply a cultural prejudice. But a heuristic that allows us to draw our own connections between responsibility and behavior, even where not supported by intra-cultural articulations, seems to invite such prejudices into our analytic process.

What both of these concerns share in common is a deeper concern with the false positives and false negatives that might arise under the heuristic. We will therefore address these concerns in the process of explorating the second stage of our inductive model, that of patterning.

2. Patterns

As describe above, induction is a two-stage cognitive process. The first stage consists of identifying particulars – “instances” we have called them. The second stage consists of extrapolating generalities from those particulars. Under inductive inference, generalities emerge as patterns that unify particularized observations. Ultimately, however, there is no coherent formula for this stage of the inductive process. (Indeed, if there were such a formula, it would ultimately render the cognitive process deductive – since the formula would serve as the “universal principle” from which our conclusions about the particulars would be derived.) This does not mean that we cannot evaluate the quality of these patterns, or of the generalizations they produce. We can always evaluate whether a particular pattern-induced generalization proves useful in helping us understand other particulars of that system. Where it does, then the pattern and its resultant generation will survive as a successful part of our analytic repertoire. Where it does not, it will quickly fade from collective reference. But the worth of a pattern is can only be demonstrated by its subsequent use. At its heart, induction – and by extension inductive accountability – is very much a pragmatic endeavor.

It is precisely this pragmatic character that allows it to insulate its epistemic structure from the false positives and false negatives that might be captured by our heuristic for identifying incidents of accountability. Perhaps the most common of these would be false negatives, since as our above example of Dicey’s orientalist rejection of French administrative law illustrates, orientalism operates primarily by preventing us from seeing the rationalities inherent in other cultural practices. But false negatives will be generally irrelevant insofar as knowledge generation is concerned. Such negatives can make the identification of patterns more difficult, but they rarely prevent such identification, particularly when we are aware of their possible existence. If I observe a group of numbers “2, 6, 8, 10, 14, 16, 18, 20, 22, 26, 30 . . .”, I can easily identify the pattern of even integers that underlies this sequence, particularly if I am aware that my heuristic is inclined to producing false negatives, .

False positives are somewhat more epistemically destructive, but not fatally so. To return to our example above, false positives could cause our observed pattern to read: “1, 2, 4, 5, 8, 10, 11, 12, 14, 18 . . .” Now the actual pattern underlying our observed pattern is less obvious. But even in such cases, the pragmatic nature of inductive reasoning helps shield the rest of the epistemic system from such flaws. Due to the ultimately non-formalizable nature of pattern recognition, we have no means for affirmatively determining that no pattern actually exists for any particular collection of unsolved incidents. For this reason, failure to identify a meaningful pattern only produces what we might call “false ignorance,” i.e., a belief that a truth does not exist where in fact it does.

False ignorance does not impact our larger epistemic system the way that false knowledge does. Knowledge is interdependent: what we know about A affects what we can know about B, because our knowledge of A and B must be consistent. A false idea can thus breed many other false ideas, and become imbedded into our knowledge system, making it increasingly difficult to correct. By contrast, what I do not know about A has much less impact on what I know about B. There is no epistemic demand that knowledge be somehow consistent with ignorance (indeed, it is difficult to conceptualize how such a command would work). Of course, ignorance about A may breed some degree of ignorance about B, but it does not at all prevent us from recognizing the veracity of B when that truth is revealed through other means. We might say that false ignorance merely makes us blind, whereas false knowledge makes us deluded.

In fact, the successful generalities created by inductive reasoning can actually work to repair false positives and false negatives. To return once more to our number example, once we recognize that we are indeed dealing with a series of even integers, we can explore why our heuristic failed to identify or falsely identified particular integers, and work to improve our use of the heuristic to correct such failures. Once British jurists understood enough about the actual operations of the French administrative law system, they were also able to recognize that Dicey’s “false negatives” were the product of cultural biases that were deeply embedded in British jurisprudence. Scholars and even judges then set out to correct these biases, and explore for aspects of the French system that might even be used to improve the same British system that so blinded Dicey’s comparative analyses. In this way, inductive accountability not only buffers the threat of orientalism, it actually can work to reduce its incidences.

III. A Demonstration: Finding Public Accountability in a non-Elected Parliament – China’s National People’s Congress

The utility of inductive accountability can be demonstrated by comparing its analytic product with that of more conventional constitutional analysis in the context of non-regulatory constitutional systems. William Alford and Benjamin Liebman’s recent study of the legislative and constitutional dynamics underlying statutory reforms to China’s Air Pollution Prevention and Control Law, entitled “Clean Air, Clear Processes? The Struggle over Air Pollution Law in the People’s Republic of China”²⁹ [hereinafter Alford & Liebman], provides an excellent subject in this regard. This is for two reasons. First, Alford & Liebman have clearly relied on what we have termed the regulatory model of constitutional accountability in drawing conclusions from their data. At the same time, however, their study includes such a rich and detailed description of the details of the reforms they study that it allows for exploration of alternative analytic approaches.

As per Alford & Liebman’s account, China’s first clean air law, the Air Pollution Prevention and Control Law [APPCL] was passed in 1987. The regulatory regime set up by that law assumed a highly controlled economy and society. By the early 1990s, it was clear that that assumption did not hold; there was widespread recognition among China’s political elites about the need for more effective law. This awareness arose at the same time that the National People’s Congress [“NPC”], China’s national parliament, had began wresting control of China’s statutory development from the State Council (China’s administrative branch), and so in 1993, it was the NPC’s own, newly formed Environment and Natural Resources Protection Committee [ENRPC], rather than to the State Council’s more established National Environmental Protection Agency [NEPA], that assumed principal responsibility for drafting amendments to the APPCL

In drafting these amendments, the ENRPC nevertheless received strong cooperation from NEPA. Both NEPA and ENRPC were very much in favor of stringent statutory protections for clean air, and their draft reflected this. However, opposition to this position began to coalesce once the draft was passed on the NPC’s Committee for Legislative Affairs [CLA], a body that functions somewhat like a central clearinghouse of all bills presented to the NPC. In particular, the CLA complained that ENRPC / NEPA had not given these opposition concerns adequate opportunity to participate in the development of their draft. Alford & Liebman find that this complaint was indeed valid, but ENRPC, with the support of NEPA, refused to compromise on the contents of its tendered draft. As a result, the CLA consulted with these other interests on its own, and despite lacking expertise in environmental issues, on its own significantly amended on ENRPC draft to reflect some of these interests (as is their right under the NPC’s procedural rules). This resulted in a much watered-down version of the amendments, and it was this watered-down version that the NPC ultimately passed in 1995.

The passage of the 1995 amendments did not end the story, however. Many administrative and local entities resisted complying with even the CLA’s watered-down provisions. NEPA and some local urban jurisdictions, for their part, sought to strengthen the effect 1995 amendments by trying to pass some of the central provisions removed by the CLA as administrative regulations. Such attempts were often rejected, however, by constitutional overseers (which in at least one case included ENRPC itself) on the grounds that allowing executive and local agencies to enact as national and local regulations provisions than had been expressly considered and rejected by the NPC itself would be contrary to the spirit of “rule of law”.

At the same time, China’s air quality continued to worsen at an alarming rate, thus attracting the attention of more and more political elite. During the course of the formulation of the 1995 amendments, the central apparatus of the Communist Party of China [the “CPC”] and its political elite generally did not concern themselves with the content of these amendments, leaving the development of this content to the constitutional system. After the passage of the 1995 amendments, however, prominent party leaders, including Party Secretary Jiang Zemin, began to show much more interest in the issue, to the point of expressly endorsing in public speeches some of the central provisions that had been removed in the CLA’s version of the 1995 draft. The CPC also instituted a national awareness campaign intended to promote popular awareness of environmental problems and popular support for environmental protection. By 1997, there was widespread agreement that the 1995 amendments were inadequate, and the NPC began organizing the drafting of a succeeding set of amendments. ENRPC was again given principal responsibility for the development of this draft.

This time, however, ENRPC officials had clearly learned their lessons.³⁰ ENRPC and NAPA started the drafting process by developing strong support among a wide range of governmental, academic and GONGO institutions. They conducted a much more open drafting process that solicited and seriously considered the opinions of entities that opposed stringent clean air legislation. Their draft, which included many of the core provisions that had removed from their 1995 draft amendments, encountered little opposition from the CLA. In 2000, it was enacted as law in virtually the same shape as it was when it left ENRPC.

Despite the relatively happy ending from the perspective of environmental protection, this story leaves Alford & Liebman fairly pessimistic about the ability of the NPC to shape the direction of political-legal reforms in China. They conclude that while it does show some increased capacity on the part of the NPC to shape statutory design, it more importantly shows that this capacity is still very much dependent on the support – or at least on the apathy – of political elite in other institutions, most importantly the CPC and the State Council:

[The] irony and [] larger lesson about the evolving role of the NPC [is that] [b]ecause the Communist Party did not treat the environment as involving core political issues through the mid-1990s, [environmental agencies] were able to achieve some victories through parliamentary maneuvers and a careful tapping of public opinion, but these triumphs were modest. It was only when the central leadership began to appreciate the extent to which environmental problems might be destabilizing and to mount vigorous national campaigns to foster environmental awareness that [NPC] was able to push through strong air pollution legislation.³¹

Ultimately, they suggest that the experiences detailed in that case study “provide scant consolation to those who hope to bring law to bear in addressing the PRC’s many and serious environmental problems”³² and, a reader might extrapolate, perhaps on China’s larger social and political problems more generally.

Alford & Liebman’s interpretation is very much driven by a regulatory model of constitutional and public accountability. They focus on ambiguities in the institutional role of the CLA (i.e., part drafter, part supervisor). They argue that this ambiguity contributes to a “opacity of governmental processes that remains a deterrent to the creation of ‘pluralistic political dynamics’ . . . involving more than political elites.”³³ They also point to how the NPC “lacks capacity to secure agency [and sub-national] compliance” with their laws.³⁴ This lack of capacity they attribute in part to the “attenuated public mandate” of the NPC, caused by the fact that NPC delegates “are not elected by the public.”³⁵ Their emphases on transparency, bureaucratic command-and-control, and mass democracy parallel the emphases of what we are calling the regulatory model of public accountability. Indeed, their complaints are strikingly similar to those that American constitutional reformers in the late 19^th and early 20^th centuries levelled against the pre-industrial American political system.

We noted above, however, that the effectiveness or accuracy of the regulatory model of constitutional accountability relies to considerable extent on the presence of a modernized, regulatory social environment. There is real doubt whether China’s constitutional environment conforms to these prerequisites. China has undergone great political, legal and socials transformations over the past 20 years. Such transformations would have greatly discouraged the systemic rationalization and standardization that is the hallmark of regulatory modernism. This being the case, then China’s constitutional environment might be much better suited to analyses that approach the issue of accountability inductively, rather than from the deductive perspective of regulatory modernism.

As described above, an inductive analysis would begin by looking to identify some of the key institutional norms and responsibilities operating in China’s constitutional / political environment. We would do this by looking primarily at what actors within this environment are actually saying about their political responsibilities. Alford & Liebman do not look at this issue directly (it is not really necessary given their deductive approach to the problem), but their narrative suggests general agreement with other scholars who find that since the 1980s, the NPC has persistently appealed to norms of participatory democracy and legality in its internal discussions with other constitutional and political actors in China. Indeed, appeals to these norms have been core to the NPC’s efforts to justify its growth in political power and prestige.³⁶

Of course, many scholars are quite sceptical as to the degree to which the NPC actually reifies these norms. The gap between such articulated norms and political reality in Chinacauses them to dismiss these appeals as meaningless, strategic speech. But we explored above how an inductive approach to constitutional accountability would not be so demanding. Recognizing that there are many reasons why an institutional action might not be able to conform to even strong institutional norms, inductive accountability focuses on finding behavior that simply evolves somewhat in response to these norms. Alford & Liebman’s account reveals a number of such incidents of accountability.

In particular, with regards to the norm of participatory democracy, they note how the CLA not only appealed to norms of participatory democracy in successfully justifying its decision to take control of the drafting of the 1995 amendments away from the ENRPC, but also sought to rectify what they identified as democratic deficiencies in the ENRPC’s drafting process by giving these excluded interests voice in their own drafting procedures. They also note that when the ENRPC changed its drafting processes to be more inclusive in the drafting of the 2000 amendments, the CLA did not intervene – a fact they attribute both to the ENRPC’s greater sophistication and the CLA greater willingness to respect the special competencies of parliamentary committees.

Alford & Liebman’s account also exposes apparent incidents of accountability attaching to the CPC. Alford & Liebman attribute the relatively successful passage of the 2000 amendments in part to “the fact that the Party leadership, which did not view the environment as politically freighted in 1995, had come to embrace at least the rhetoric of environmentalism by 2000.” But they also note that in the lead-up to the 2000 amendments, the ENRPC seems to have devoted considerable energy towards developing and articulating the strong public support behind its position, while at the same time giving significant opportunity and attention to opposing views, and that “impetus for [the Party’s embrace of stronger clean air regulation] owns something to rising public opinion that the ENRPC has helped articulate.”³⁷

Reformatted into the framework of inductive accountability, what this actually suggests is a possible incident of CPC accountability to the NPC. But even after acknowledging this possibility, Alford & Liebman dismiss its significance. They note:

[T]he NPC’s role appears less that of a pluralistic, democratic body providing routinized outlets for the expression of citizens viewpoints generally and more that of another elite bureaucratic entity whose role remains incompletely defined in a world in which political power, at least in central government and party circles, continues to be tightly held.³⁸

But this dismissal could be the product of an underlying “modernist” desire for of environmental visibility. Once we recognize that in non-regulatory environments, such visibility is much more limited, Alford & Liebman’s conclusion that the NPC’s role “appears . . . more of another elite bureaucratic entity whose role remains incompletely defined” [emphasis added] might simply describe limitations in environmental transparency rather than limitations in the relevance of the NPC within that environment. This hypothesis seems even more plausible when one notes their highly modernist suggestion that the NPC could only really distinguish itself from the mass of China’s other “elite bureaucratic entities” by being a “pluralist, democratic body.” There is no question but that becoming a “pluralist, democratic body” would indeed distinguish the NPC in constitutionally important ways, and would have tremendous implications for our understanding of constitutionalism in China. But recall also that the concept of a “pluralist, democratic” congress is a very modernist discovery. As our exploration of pre-industrial American constitutionalism suggests, there are other important ways through which a parliament could distinguish itself in a non-regulatory constitutional environment.

To see how this is so, let’s continue our inductive exploration for potential incidents of constitutional accountability. One that stands out is the State Council’s rejection of a proposed regulation tendered by NEPA on the grounds that such regulation, while not contradicted by the text of those amendments per se, did not accord with the spirit of the 1995 amendments in that it included provisions that had been expressly removed from the 1995 law. Another would be how the law and its surrounding debates seems to have shaped the efforts of sub-national entities to develop their own clean-air standards.³⁹ Particularly interesting in this regard was the decision by the Beijing municipality to seek approval from ENRPC before enacting its own clean air regulations.⁴⁰ The impact of the 1995 amendments on the content of local and national regulation seems particularly noteworthy given Alford & Liebman own observation that that local jurisdictions are generally able to ignore NPC legislation with impunity.

From all this, a pattern does begin to emerge. Assuming that Alford & Liebman are being comprehensive about the ongoing evolution of China’s national clean-air regulation, and not simply limiting their focus to the ENRPC’s particular role in that evolution, than we begin to notice how most of these “incidents” of accountability seem to flow through the NPC and ENRPC. There is no account of NEPA directly shaping local policy, or vice versa; there is no account of NEPA directly shaping Party policy; there is no account localities directly shaping central Party policy. In fact, surprisingly enough given general claims about the omnipotence of CPC power in China, there is no accounted incident of the the CPC (as a corporate entity) directly shaping the content of NEPA or local policy.⁴¹

Of course, there is little question but that NEPA, localities, and the Party did indeed influence one another. But the pattern suggested by our exploration for incidents of accountability suggests that such influence tends to flow through the NPC. Localities and NEPA influenced Party policy through the NPC’s mobilization of these interests. The Party’s interests become reflected primarily in the form of NPC law, and it is this law – rather than the interests directly – that seemed to be the principal referent in the shaping of administrative and local behavior.

In 1992, Ronald Burt famously articulated the idea of the “structural hole.”⁴² A structural hole is a gap in the communications network of an institution that prevents two or more significant segments of institutional actors from communicating effectively with each another.⁴³ Burt and others have shown that institutional actors that are able to bridge structural holes – i.e., are able to provide conduits that allow relatively direct communication between otherwise separated entities – enjoy a competitive advantage in shaping the opinions of the larger institutional environment.⁴⁴

By showing that incidents of constitutional accountability invariably seem to flow through the NPC, inductive accounting of Alford & Liebman’s story suggests strongly that the NPC is bridging a structural hole. And if that is the case, than the NPC is not just “another elite bureaucratic entity,” it is a bureaucratic entity whose particular position within the larger political environment gives it a significant competitive advantage in propagating its norms through that environment. Such an advantage is particularly relevant in a rapidly evolving non-regulatory environment such as China’s political environment. Many scholars have argued that that environment is in desperate need of guiding political norms. If this is the case, then an inductive analysis of Alford & Liebman’s story exposes the possibility that the NPC actually may be in a very special position to contribute to the normative development of that environment. The accountability patterns we have just seen – CPC responding to concerns generated primarily by ENRPC rather than other identified political actors; Beijing and other localities responding to norms generated primarily by ENRPC and the NPC, rather than by other identified political actors; and even the State Council supporting norms generated by the NPC over competing norms offers by their own NEPA – appear to confirm this possibility.

So where does this leave us with regards to Alford & Liebman’s ultimate conclusion that their account “provide scant consolation to those who hope to bring law to bear in addressing the PRC’s many and serious environmental problems?”⁴⁵ If we are looking at this problem primarily through a regulatory lens, then we have no reason to doubt this claim. Alford & Liebman clearly show that regulatory law is not a particularly effective device in China. But if we look as law in a non-regulatory light – i.e., as a framing device within which (rather than through which) autonomous political actors develop and shape future social development⁴⁶ – then the picture changes. As I and others have shown, the NPC is somewhat distinctive in the degree that its core institutional identity encompasses express norms of democratic participation and legality. The NPC’s conceptions of these norms also appear largely compatible with our own, Western ideas about how such norms should work in society. If this is the case, then our inductive analysis of Alford & Liebman’s account actually gives significant “consolidation” to those of us interested in a more constitutionalized China. The combination of the NPC’s progressive constitutionalism and its competitive advantage at disseminating these norms fuels a hope that China will ultimately evolve into what Alford & Liebman themselves would probably regard as a more democratically and legally effective constitutional state.

---

1 Fellow in Law, Australian National University Regulatory Institutions Network (RegNet). I would like to thank and acknowledge the following for their important contributions to this paper: William Alford, Alfred Aman, Mark Barenberg, Roy Schotland, Sasha Courville, Michael Dorf, Kent Greenwalt, Peter Vincent-Jones, Colin Scott, and Adrienne Stone. This is not to imply that any of the above agree with its content, however.

2 This use of the term “regulatory” was suggested to me by Colin Scott. See Chapter 10.

3 See James C. Scott, Seeing like a State: How Certain Schemes to Improve the Human Condition Have Failed (1998).

1 Pages 53-63

1 See, e.g., Gordon Wood, The Creation of the American Republic 1776-1787, at 46-48, 53-65, 75-90 (1969). Cf. Herbert J. Storing, What the Anti-federalists Were For 84 n. 12 (1981) (arguing that framers’ debates over the dangers of a standing army was motivated in significant part by concern over the dangers of what later would be called “bureaucracy”).

6 See, e.g., Howard Zinn, A People's History of the United States (1980).

⁸ In the case of the 19^th century railroad industry, see John Fabian Witt, “The Transformation of Work and the Law of Workplace Accidents,” 1842-1910,” 107 Yale L.J. 1467 (1998). Cf. Dennis W. Carlton & J. Mark Klamer, “The Need for Coordination Among Firms, with Special Reference to Network Industries,” 50 U. Chi. L. Rev. 446 (1983). Networked industries were the principal focus of early modern regulation.

⁹ Skowronek, p. 269-270.

¹⁰ Between 1910 and 1917, the ICC only approved one rate request. Skowronek, p. 270.

¹¹ See Oscar Kraines, “Brandeis’ Philosophy of Scientific Management,” 13 (1) West. Pol. Q. 191, 195 (1960).

¹² Witt, Colum. L. Rev. 37

¹³ See Witt, Yale; John Fabian Witt, “Speedy Fred Taylor and the Ironies of Enterprise Liability,” 103 Colum. L. Rev. 1 (2003) [check also William Graebner, Coal-Mining Safety in the Progressive Period: The Political Economy of Reform 128 (1976), boiler safety article.]

¹⁴ See especially the writings of Charles Francis Adams, Jr. See Skowronek,133-138

¹⁵ See Swokronek, p. 148.

¹⁶ See Skowronek, p. 251

By the late 1890s, Supreme Court caselaw had effectively eviscerated the ICC’s regulatory authority. See, e.g., Cincinnati Ry., Co. Et al. V. ICC, 162 US 184 (1896) (courts can ignore ICC findings of fact); ICC v. Alabama, 164 US 144 (1897) (same); The Import Rate Case, 162 US. 197 (1896) (rejecting ICC’s capacity to interpret ambiguous statutory text); ICC v. Alabama Midland Ry. Co. et al., 168 US 144 (1897) (same); Maximum freight Rate Case, 167 US 479 (1897) (denying ICC legislative competence). See generally Skowronek, pp. 154-160. In 1910, the Court reversed its opposition to the ICC’s legislative and administrative autonomy. see ICC v. Illinois Central Railroad Co., 215 US 452 (1910). This appears due in significant part to the fact that opposition to regulation by judiciary had reached such heights that political threats to judicial independence were now credible. See, e.g., Charles A. Prouty, “Court Review of the Orders of the Interstate Commerce commission,” 18 Yale. L.J. 297, 310 (1909); see also Skowronek, supra note __, at 347 n. 29.

¹⁸ Uniform systems of accounting are just one example of the way that early effective regulatory rules tended to rely on existing socioeconomic practices occasioned by industrialization. Another prominent example is workmen’s compensation, the principal mean by which American governments regulated workplace safety prior to the 1960s,¹⁷ which emerged out of increasingly standardized industrial practices of firms providing private, firm-specific “accident relief funds” that compensated worker for on-the-job injuries without regard to fault. See Witt, Columbia 21-28.

¹⁹ With regards to the ICC, see Skowronek, p. 153, see e.g., Strauss, op cit. With regards to the role that the case-by-case strategy played in the development of state regulatory regimes, see especially Mark A. Covaleski, Mark W. Dirsmith & Sajay Samuel, “The Use of Accounting Information in Governmental Regulation and Public Administration: The Impact of John R. Commons and Early Institutional Economists,” 22(1) The Accounting Historians Journal 1 (1995).

²⁰ Morton J. Horowitz, The Transformation of American Law 1870-1960: The Crisis in Legal Orthodoxy 145-167 (1992). R. Hoksbergen, “Postmodernism and institutionalism: Towards a Resolution of the Debate on Relativism,” Journal of Economic Issues 679 (September 1994) 1994 PP. 705-708 See also Mark A. Covaleski, Mark W. Dirsmith & Sajay Samuel, “The Use of Accounting Information in Governmental Regulation and Public Administration: The Impact of John R. Commons and Early Institutional Economists,” 22(1) The Accounting Historians Journal 1 (1995). See also Peter L. Strauss, “When the Judge is Not the Primary Official with Responsibility to Read: Agency Interpretation and the Problem of Legislative History,” 66 Chi.-Kent. L. Rev. 321 (1990) (describing how this strategy worked in the development of railroad safety regulations).

1 See Peter H. Argersinger, “New Perspectives on Election Fraud in the Gilded Age,” 100 # 4 Political Science Quarterly 669-687 (Winter 1985-1986).

1 See Clifton K. Yearling, The Money Machines: the Breakdown and Reform of Governmental and Party Finance in the North, 1860-1920, at 167 (1970). See also Carolyn Webber & Aaron Wildavsky, A History Of Taxation And Expenditure In The Western World 416-27 (1986); David N. Hyman, Public Finance: A Contemporary Application of Theory to Policy 448-49 (4th ed. 1983); Edwin R.A. Seligman, The Income Tax: A Study of the History, Theory, and Practice of Income Taxation at Home and Abroad 388-419 (2d ed. 1921).

[1] Richard Hofstadter, The Age of Reform: From Bryan to F.D.R. 108 (1955).

1 Robert H. Wieber, The Search for Order 1877-1920 (1967).

[2] See, e.g., Christopher Clague, Suzanne Gleason, and Stephen Knack, “Determinants of Lasting Democracy in Poor Countries: Culture, Development, and Institutions,” 573 Annals of The American Academy of Political and Social Science 16 (2001); Adam Prezeworski & Fernando Limongi, “Political Regimes and Economic Growth,” 7(3) J. of Econ. Perspectives 51, 62, 63-64 n. 16 (1993) (surveying results of various studies); Ronald Inglehart, Modernization and Postmodernization: Cultural, Economic, and Political Change in 43 Societies 184 (1997); Diane Ethier, “Democratic Consolidation: Institutional, Economic, and External Dimenstions,” in Abdo I. Baaklin & Helen Desfosses, eds., Designs for Democratic Stability: Studies in Viable Constitutionalism 259, 270-271 (1997); Ross E. Burkhart & Michael S. Lewis-Beck, “Comparative Democracy: The Economic Development Thesis,” 88 Am. Pol. Sci. Rev. 903 (1994).

²³ Katharina Pistor & Philip A. Wellons, The Role of Law and Legal Institutions in Asian Economic Development 1960-1995 at 111 (1998).

²⁴ Frederic C. Deyo, “Reform, Globalization and Crisis: Reconstruction Thai Labor,” 42 J. Ind. Rel. (2000).

²⁵ Pasuk Phongpaicht & Chris Baker, Thailand’s Economic Crisis 82 (2001).

²⁶ Id. At 84.

²⁷ See Pasak & Baker, supra note __, at 81.

²⁸ See Pasak & Baker, supra note __, at 81; United Nations Economic and Social Commission for Asia and the Pacific [UNSCAP], Country Study, “Social Safety Nets in Thailand” at 45 (http://www.unescap.org/safetynet/csthailand.doc) (accessed 6 May 2003). In this, too, there are interesting parallels between the Miyazawa Fund and the pre-industrial Civil War Pensions, which also appear to have been more effective than rule-of-law-based suppositions predicted. See note __ supra.

11 Cf. E. P. Thompson, “Time Work Discipline,” 39 Past & Present 56 (Dec. 1967).

12 I borrow this phrase from Bob Jessop. See Bob Jessop, “Regulationist and Autopoieticist Reflections on Polanyi’s Account of Market Economies and the Market Society,” 6 New Political Economy 213, 213 (2001).

As I have explained elsewhere, this presumption of stability is very much built into the foundation of our constitutional thought.

Mature constitutional environments are concerned primarily with preserving the overall stability of and promoting the efficacy of their current constitutional / political arrangements. This static emphasis is readily evident in the constitutional theories found in the United States. Separation of powers and federalism work to prevent particular collections of actors at the central and local levels respectively from accumulating quantities of power that would allow them to disrupt the democratic character inherent in the present constitutional order. The difficult amendment process seeks to ensure that the constitutional / democratic structure is able to withstand destabilizing majoritarian passions. Judicial review seeks to insure that the existing constitutional arrangements are uniformly preserved in the actions of individual political actors. Our constitutional focus on stability and permanence is further demonstrated by the pride that many Americans take in the fact that the United States Constitution has undergone relatively little formal amendment over the two centuries of its existence.

Michael William Dowdle, “Of Parliaments, Pragmatism and the Dynamics of Constitutional Development: The Curious Case of China,” 34 NYU. J. Int'l L. & Pol __ (2002).

13 See, e.g., id.

14 See Frederic C. Deyo, “Introduction,” in Economic Governance and the Challenge of Flexibility in East Asia 1 (Frederic C. Deyo, Richard F. Doner& Eric Hershberg, eds., 2001). See also Frederic C. Deyo,

19 Purists might complain that my notion of “inductive accountability” actually conflates two distinct cognitive processes – that of induction and that of analogy. Cf. Dan Hunter, “No Wilderness of Single Instances: Inductive Inference in Law,” 48 J. Legal Educ. 365 (1998) (complaining that legal scholars often wrongly conflate these distinct cognitive phenomenon). To this charge I have to plead guilty. I do so because I feel that it is probably unnecessary to distinguishing the inductive from the analogical aspects of this model insofar as my particular analysis is concerned, and that to do so would more likely confound than clarify.

20 Note that “autonomy” is not the same as independence. Participants in dialogue are not really independent – a dialogue requires a certain degree of coordination and cooperation, which in turn constrains the actions of its participants. Nevertheless, a dialogue also assumes these two participants must have some mutually independent capacity to manipulate their own agenda and perspectives during the course of that dialogue. If the participant’s respective agendi and perspectives are completely fixed, there is no benefit to discussion. This means that under conditions of dialogue, actors are presumed to have some significant control over the scope of their own responsibilities vis-a-vis the other party.

21 See, e.g., David Starks & Laszlo Bruszt, Postsocialist Pathways: Transforming Politics and Property in East Central Europe (1998).

22 See also Terry M. Moe, Politics and the Theory of Organization, 7 J.L. Econ. & Org. 106, 120-26 (1991).

23 Kal Raustiala, “Compliance and Effectiveness in International Regulatory Cooperation,” 32 Case W. Res. J. Int’l L. 387, 395-396 (2000).

24 See also Cass R. Sunstein, “Incompletely Theorized Agreements,” 108 Harv. L. Rev. 1733 (1995).

25 Comment by Bruce Ackerman, “Symposium on Fidelity in Constitutional Theory: Fidelity as Synthesis: Colloquy,” 65 Fordham L. Rev. 1581, 1584 (1997).

26 The Federalist No. 51, at 322 (J. Madison) (C. Rossiter ed., 1961):

If men were angels, no government would be necessary. If angels were to govern men, neither external nor internal controls on government would be necessary. In framing a government which is to be administered by men over men, the great difficulty lies in this: you must first enable the government to control the governed; and in the next place oblige it to control itself.

(Emphasis added.)

27 See, e.g., Philippe Baumard, Tacit Knowledge in Organizations (Samantha Wauchope, trans., 1999). See also Peter L. Berger & Thomas Luckman, The Social Construction of Reality: A Treatise in the Sociology of Knowledge (1980).

28 See Dowdle, supra note [2003]. See also Berger & Luckman, supra note [1980].

29 52 Hastings L.J. 703 (2001).

30 Id. at 747.

31 Id. at 747-748.

32 Id. at 748.

33 Id. at 739.

34 Id.

35 Id. at 741.

36 Skeptics may question whether these appeals were honest – i.e., whether the NPC leadership making these appeals actually believed in the values (and concurrent responsibilities these values imply) they professed. But as we saw above, inductive accountability is concerned only with behavior per se (in this case, the behavior being the appeal). The motive behind the behavior is analytically irrelevant.

37 Id. at 478.

38 Id. at 747-748.

39 See id. at 732-733.

40 Id. at 734.

41 Alford & Liebman suggest that the central Party shaped local policy indirectly through campaigns. But they provide no concrete suggestion that the Party directly directed the formation of local environmental protection regulations. In fact, Alford & Liebman’s account actually implies that this did not happen. A local entity whose environmental policies were approved by an all-powerful party would have no reason to then turn-around and also seek approval from the NPC, as did Beijing and other municipalities. And it would certainly have no reason to respect an NPC veto of that local policy, given the fact that any NPC veto would clearly be trumped by an approval from the omnipotent party.

Skeptics may argue that we can presume direct Party influence, given the party’s all powerful control of China’s political environment. As we have seen, however, “all-powerfulness” is itself very much a phenomenon of regulatory modernity. Given the non-regulatory character of China’s constitutional environment, the idea of an all-powerful central Party seems highly unlikely. It may be the case, but that case would have to be demonstrated rather than presumed. Nothing in Alford & Liebman’s account provides any demonstration of that particular claim.

This highlights one of the advantages of inductive accountability that we discussed above, its tendency to treat orientalist biases as false ignorance rather than false knowledge. The common claim of CPC political omnipotence is a claim that has little basis in actual fact. It is also a claim that is completely inconsistent with what we know about how complex organizations and societies operate. Instead, it is a claim that exists primarily as a tautology: everyone knows the CPC is omnipotent; therefore all political decisions are ultimately controlled by the CPC; and because all political decisions are ultimately controlled by the CPC, the CPC must be omnipotent. This tautology works because it is easily incorporated into deductive analyses as a presumptively universal principle. Our inductive analysis of Alford & Liebman’s account gives us no purchase for the initial assumption, since we do not observe any pattern of accountability that points towards party domination (in fact, just the opposite), and thus allows us to escape the bias.

42 Ronald S. Burt, Structural Holes: The Social Structure of Competition (1992).

43 See also Ronald S. Burt, “The Social Capital of Opinion Leaders,” 566 Annals 37, 48 (1999):

The structural hole between two groups does not mean that people in the groups are unaware of one another. It means simply that the people are focused on their own activities such that they have little time to attend to the activities of people in the other group. Holes are buffers, like an insulator in an electric circuit. People on either side of a structural hole circulate in different flows of information. Structural holes are thus an opportunity to broker the flow of information between people and control the form of projects that bring together people from opposite sides of the hole.

44 See generally id.

45 Id. at 748.

46 See also Gunther Teubner, “Juridification: Concepts, Aspects, Limits, Solutions,” in Gunther Teubner, ed., Juridification of Social Spheres 3, 18 (1987) (discussing pre-industrial, “classical formal law”).