A Model of Political Economy: Embedded Coordination, Cooperation, and Conflict

OVERVIEW

Synthesizing these diverse efforts to understand political economy allows us to outline a model of the economy as embedded coordination, cooperation, and conflict. Actors and organizations are always already embedded in a ma­terial (nature + technology) and social (institutions + ideology) context, in social relations inherited from prior rounds of political, social, and economic struggle.²⁰ There is no Archimedean point on which to perch microfounda- tional agents from whose universal, time-invariant nature (e.g., an imagined propensity to truck, barter, and exchange) social structures emerge.

It is critical to emphasize, though, that once market society emerges in the transition to capitalism, the dynamic of improved productivity through technological and institutional innovation is too powerful to permit incum­bents to retain their position purely through power-seeking investments.

AT&T could not prevent the emergence of voice over Internet protocol any more than canal or turnpike companies could prevent railroads. But within a broad range, institutional power matters: craft producers in Lyon or North­ern Italy did succeed in resisting Fordism and maintaining more flexible, less exploitative labor relations at the same technological and productivity fron- tier,²¹ and Danish unions today urge adoption of robots, secure in their power to obtain a fair share of the productivity gains in a global marketplace. Under these realistic market conditions, firms and individuals mix strategies. They trade off investments in improving productivity to stay ahead of the compe­tition for investments in obtaining power. They seek market power horizon­tally, against competitors and disruptive innovators to create larger rents, and bargaining power vertically, against workers, consumers, suppliers, and distributors to obtain a larger share of these rents from other claimants. And they trade off investments in short-term exploitation of the existing mate­rial and social context for investments in shifting the long-term context toward arrangements that increase their future power.

Definitions

Power is a property of a relationship between A and B, describing A's capacity to shape B's behavior, outcomes, or context so that the respective behaviors, outcomes, or context of A's and B's relations is closer to A's preferred rela­tions than to B's, short term (within context) or long term (about context).

Institutions are explicit or implicit instructions for who should do what in which social relation, serving as constraints and affordances on behavior for persons in the social relation to which they apply. Law is a system for producing such instructions susceptible to enforcement by legitimate vio­lence. Social norms are systems of such instructions enforced through social coercion—gossip, shaming, ostracism—or internalized social conformism. Organizational or professional norms are explicit or implicit instructions produced by a given set of social actors to govern behavior (i.e., create affor­dances and constraints) in the social context for which they are developed so

Power and Productivity 37 as to constitute the social relations they constitute—for example, the work­place or the profession.

Ideology is that subset of institutions that shape how we understand the world, what causes what, what goes with what, what is valued and what is loathed.²²

Technology is congealed practical knowledge embedded in material culture.²³ “Practical knowledge,” which is knowledge applied functionally to achieve desired outcomes, is a universally adopted element included in the defini­tion of technology. I add the notion of “embedded in material culture” to distinguish technology from institutions and ideology, each of which is of­ten treated as a form of practical knowledge (i.e., how to behave, how to in­terpret). While some definitions of technology treat any practical knowledge that allows us to do new things or old things more efficiently as “technology,” and others emphasize “sociotechnical systems” to underscore that all tech­nologies take their meaning from specific social relations, those definitions are less useful for distinguishing between institutions and technology.

In sum, power in social relations, its magnitude and distribution, is a function of institutions, technology, and ideology. Institutions are the “rules of the game”—that is, the instructions about who can do what in which con­text that define the relation and distribute power within it. The fact that you can touch the ball with your hands, cannot hold it for more than 3 seconds, and must dribble to advance makes basketball a distinct social practice from soccer. Technology describes the material conditions under which a practice so constituted is carried out. The fact that the hoop is of a certain size and located 10 feet off the ground means that taller players are more talented in

basketball than they would have been had the same hoop been set at 2 feet. Ideology is the conception people have of what they are doing, the frame through which they understand the practice and define their preferences and beliefs and understand their constraints in the situation.

DYNAMICS

How the preponderance of agents and organizations act, what outcomes they obtain, and what practices form their competitive environment is shaped by the institutions, ideology, and technology that make up the context of the re­lation. Firms know this and act strategically in interactions both within mar­kets and about the institutional, ideological, and technological determinants of power in market relations to increase their ability to extract quasi-rents in all their interactions—horizontally, against competitors or innovative disruptors, and vertically, against workers, consumers, suppliers, and dis­tributors. Workers and consumers know this, so they organize to resist and reshape the power relationships, sometimes through market organizations like unions or cooperatives and often through nonmarket organizations and social movements.

Cyrus McCormick's reaper was a quintessential productivity-increasing technology, transforming American agriculture in the second half of the nineteenth century. But when Cyrus McCormick Jr. inherited his father's company, he harnessed technology in a very different way. Soon after inher­iting the company, McCormick declared wage cuts, despite the firm's banner year in 1883. He had not counted on the iron molders union, whose unique craft skills made them irreplaceable and who had been the driving force of labor organizing in Chicago for 20 years. In 1884, McCormick lost a violent three-week strike against his arbitrary wage cuts. Almost immediately, the company invested in technology and politics to gear up for the next strike. It purchased pneumatic molding machines and replaced the entire workforce of craft molders with unskilled laborers working the new machines.

Chicago's mayor in exchange for the appointment of the new, anti-labor police inspector who would lead the 1886 Haymarket Massacre of protesters fighting for an 8-hour workday, still commemorated as May Day.²⁴

More generally, firms deploy technologies that increase their bargaining power over labor in three primary ways—homogenization, monitoring, and fissuring.²⁵ Automation that standardizes and simplifies labor inputs weak­ens labor bargaining power, whereas automation that increases productiv­ity more but requires experienced operators strengthens labor. Monitoring similarly alters bargaining power by removing workers' credible threat to slow down production if treated unfairly. When firms cannot perfectly ob­serve effort or quality, they pay workers a premium to induce them to make firm-specific investments and work beyond what is observable. If technolo­gies make effort more observable, workers' bargaining power declines, and with it wages. Technology that gives employers finely honed sticks reduces their dependence on carrots. Levy and Barocas, for example, show how retail firms repurposed data collection systems designed to monitor customers to homogenize experienced salespeople, making them more readily replaceable; monitor employees more finely to impose starker discipline; and externalize the risk of the ebb and flow of business onto workers by forcing algorithmi­cally set “flexible” work schedules.²⁶ Finally, employers can deploy technolo­gies that fissure the workplace and undermine worker collective action, as Rogers emphasized,²⁷ and which Gray and Suri demonstrated in their study of how contemporary “artificial intelligence” systems incorporate human “last mile” operations, harnessing individuals in the United States and India working behind a one-way mirror so that everything they do is observed and measured, while disabling worker coordination.²⁸

Similar actions of individual firms abound throughout the history of mod­ern market societies, not only with technology but with institutions and ide­ology as well. In the 1960s, Kelly Girls and other temporary personnel firms launched a campaign to circumvent then-still-powerful labor union's resis­tance to temporary workers. They framed their new employment model in patriarchal terms, recruiting married women “to make some pocket money” to fill the ranks ofjobs in the new services industry, precursors of the fissured workplace.²⁹ Dubal's study of a century of labor struggle in the San Francisco taxi industry situates Uber's strategy of spending investor capital to mo­nopolize the market in rider apps: Uber designed its app to evade regulators and control drivers and lobbied to undermine the municipalities' traditional jurisdiction over livery services and ensure the designation of drivers as inde­pendent contractors.³⁰ Schor examines how these strategies produced highly differential outcomes for occasional gig workers, mostly white and working for extra income, who benefit from opportunities to add income more than they lose from the power imbalance with the company, and gig workers who depend on platforms for a livelihood who are mostly minority and who see the power asymmetry cutting into their basic income.³¹

This microfoundational story (what individual firms or agents do) must be complemented by meso-level analysis of collective action, and both only happen in inherited macro-level institutions, ideology, and technology. The 1970s saw a dramatic increase of business lobbying efforts, which in turn supported institutions that weakened the power of government to con­strain business and redirected government power toward weakening labor.³² Unions had played a critical role for three decades during the postwar period, not only on wage setting and enforcement of compensation norms, includ­ing managerial compensation, but also as the central countervailing political power over broad questions of economic policy and redistribution.³³ A core focus of Organized Business since the 1970s was to undermine Organized La­bor and harvest the rents from that transformation in bargaining power. Its decisive victory was President Ronald Reagan's breaking the PATCO (Profes­sional Air Traffic Controllers Organization) strike in 1981. Declining union membership since then has been a major cause of median income stagnation in the United States. Broad, cross-industry collaboration among businesses on institutions like antitrust or labor law was complemented by industry­level collective action as a major determinant of market structure and both horizontal and vertical power in product and labor markets throughout the economy. The structures of the telecommunications industries, both wired and wireless, across the Organization for Economic Cooperation and De­velopment (OECD) are the fossil record of the two-decades-long battles be­tween firms and regulators.³⁴ So too the structure of the pharmaceutical, au­tomobile, energy, and any other large-scale sectors. But collective action does not necessarily result in exploitation. Where power is symmetric and power­seeking offers few gains, coordination and cooperation between labor, man­agement, and a well-functioning state actually help make high-productivity egalitarian arrangements work.³⁵

These discrete political battles occurred on the background of a macro­level historical shift in institutions and ideology that responded to the po­litical and economic shocks of the 1960s and 1970s. The ideological work had been in the works for decades, ready to coalesce when the postwar settlement was shocked out of equilibrium by the Great Inflation of the 1970s. Hayek and Friedman were both on the margins of academic and policy circles through­out the 1940s until the 1970s, but they built a network of academics and or­ganizations that would be ready when the winds shifted. Initially funded by ideologically committed individuals, the neoliberals built organizational ca­pacity through think tanks and special-purpose programs within academia. In some cases, as with Henry Manne's successes in fundraising for the law and economics movement, there were direct appeals to the self-interest of companies like ITT or US Steel, which wanted to loosen antitrust law to fund a movement that would nudge law in that direction.³⁶ Subsequent study con­firmed that these efforts were successful and that judges who participated in Manne's Pareto in the Pines program rendered systematically more pro­business verdicts and tended to rule against regulatory and tax agencies more often for decades thereafter.³⁷ These appeals fit well the changed politi­cal program of business organizations in the 1970s and were foundational to the victory of neoliberalism as the dominant economic theory of the 1980s through the Great Recession. Some of these organizational beachheads were located in traditional academic departments with a critical mass of members who then influence future appointments to build a “school.” The Chicago economics department was one such place, as were Buchanan and Tullock at the Virginia Polytechnic Institute and later George Mason University. So too with the Olin Foundation's support of the establishment and expansion of law and economics programs at law schools. To this “inside” strategy the movement added think tanks that housed scholars focused on translational work or academics translating their academic work for consumption by poli­cymakers and elite opinion makers. The American Enterprise Institute and Foundation for Economic Freedom were soon joined by the Heritage Foun­dation, the Cato Institute, and others that attracted funding to support a steady flow of papers and events criticizing regulation and redistribution at a detailed level of analysis of both policy and politics. To these think tanks, the movement added public-facing programming to educate elites—such as Manne's “Pareto in the Pines”—and mass audiences, as Milton Friedman's Newsweek column and television show did so remarkably. By moving from big ideas to technically well-worked-out details, from academia to think tank to popular culture, and from idea development to education and training, the movement was able to create a large cadre of elite actors who, some more consciously than others, had come to adopt a worldview, a way of interpret­ing the world, that saw markets as efficient and liberating and government planning as doomed to fail, corrupt, and tending to tyranny.

These institutional and ideological shifts of the 1970s were complemented by broad technological deployments of information and communications technologies that enabled offshoring and outsourcing of production. Then, by the 1980s, computers and spreadsheets enabled ever-more complex finan­cial products that led to financialization. In combination, these moves put management and finance in a position to disinvest from labor, adopt short- termism, embrace the earnings game, and use the newly found legal freedom to suppress competition and extract a larger share of the resulting rents. Free trade agreements that emphasized investor protection and financial flows, but not labor or environmental standards, vastly expanded competition in the labor market between domestic and offshore labor and severely limited the power of labor domestically. Weakening antitrust enforcement since the Reagan administration, looser financial regulation, weaker labor regula­tion, and monetary policy aimed at keeping inflation in check by sticking to a relatively high “natural rate of unemployment” that kept labor markets relatively slack—all contributed to increased horizontal and vertical power in product and labor markets. These micro, meso, and macro dynamics com­bine to explain the observed patterns of the American economy over the past 40 years—declining business dynamism, increasing concentration and markups, slower productivity growth, and the particular pattern of American inequality—a top 1 percent and 0.1 percent takeoff coupled with broad-based income stagnation and economic insecurity.

The emergence of the Internet as the basic infrastructure of the twenty- first century, where I focused most of my research, is rife with examples of this dynamic. Microsoft's suite of antitrust cases in the 1990s revolved around technical choices (intentional incompatibilities with DR-DOS in the early 1990s; intentional incompatibilities with Java or HTML in the mid- 1990s) and institutional strategies (licensing terms that made it impossible for PC manufacturers to replace Microsoft's products with competing op­erating systems without incurring huge costs) designed to build the firm's market power horizontally—against potential disruptive innovators or competitors—and vertically, against consumers, suppliers, and distributors. Apple's App Store is no more technically necessary for loading software on a mobile device than it had been for the PC. But it creates a bottleneck that allows Apple to extract rents from complementary app developers and delay or degrade apps that threatened to decrease its rents or bargaining power.³⁸ Cisco developed “policy routers” in 1999 to enable newly emerging cable broadband providers to extract rents from suppliers of complementary prod­ucts. The result was two decades of political struggle and litigation over net neutrality. The dynamic is replicated in the advertising platforms of Google and Facebook and Amazon's relations with sellers in the Amazon market- place.³⁹ As Julie Cohen masterfully showed, a series of strategic actions, in litigation and legislation, shaped intellectual property, on the one hand, and privacy and data protection law, on the other hand, to make data about in­dividuals a “public domain” free for unconstrained harvesting by the major firms of our era, while making the data aggregations collected by these firms “private property” protected from both competitors and regulators.⁴⁰ And as Amy Kapczynski showed, transformations of trade secret and constitutional law since the 1980s have strengthened many companies' bargaining power in labor markets and increasingly hampered state regulation of firms across a broad range of industries.⁴¹ In all these cases, firms engaged not only in technical changes but in extensive litigation and lobbying to create an insti­tutional setting conducive to their continued power, and in extensive ideo­logical work to reinforce the inevitability and benevolence of unconstrained technological change.

Throughout this period, individuals—acting alone or in networks, or col­lectively in civil society organizations—tried to push back in both institu­tional battles and by building technologies. Efforts by organizations like the Electronic Frontier Foundation (EFF), EPIC, Public Knowledge, Free Press, or Fight for the Future were conjoined with periodic mobilization efforts in the “copyright wars,” encryption and privacy, access to knowledge, and more recently, fairness in algorithms. These institutional and ideological battles were complemented by technological interventions, such as strong encryption tools to protect privacy and decryption tools to circumvent dig­ital rights management, each designed to defeat the efforts of companies and governments to impose arrangements that activists deemed oppres­sive. Nowhere was this dimension more clearly embodied in practice than in the free software movement.⁴² Here, as in the free culture movement that followed it, we saw direct conflict between firms seeking to bring more of the economy into market relations and a social movement of people seek­ing to construct a context that allowed more of the economy to function on nonmarket models. Advocates of the commons (myself included) pointed to successful commons-based practices like free software and Wikipedia as ex­istence proof that the economy need not be purely cleared by prices in mar­kets supported by ever-more perfect deployment of property and contract. Indeed, we argued at the time, user innovation, socially motivated hacking, and norms-driven knowledge production offered important checks on purely market-based information, knowledge, and cultural production.⁴³

Battles over privacy or algorithmic use of data are now the dominant front of consumer-oriented struggle in technology politics. Several of the world's most valuable companies are focused primarily on developing technologies whose core task is to extract information from and about, and run behav­ioral experiments on, consumers. One requires Panglossian optimism to imagine that pervasive surveillance and personalized, experimentally vali­dated behavioral advertising was designed to inform rather than manipu­late consumers. Successfully manipulating demand increases the value of quasi-rents by manipulating users' willingness to pay. It also increases the half-life of the quasi-rents by delaying competitive entry: manipulating in­formation about substitutable products and magnifying perceived differen­tiation between the manipulator's product and substitutes. Applied to poli­tics, the translation of this power into shaping the institutional dimension of power is obvious. While there is little quantitative evidence showing that these technologies work,⁴⁴ it is clear that their purpose is to develop such power over consumers and that, even without evidence, advertisers are buy­ing enough of the promise to make these technology companies the most valuable in the world.

Most of the battles of the 1990s and 2000s focused on individual free­dom and the market reach relative to nonmarket production, rather than on distributive justice. Nonmarket, nonproprietary production was celebrated largely as a degree of freedom from the power of market actors to invade our privacy and set the terms of information, knowledge, and cultural pro­duction. The Access to Knowledge movement was a first reorientation to­ward distribution-sensitive politics of technology, influenced by the Access to Medicines movement.⁴⁵ Since the Occupy moment we have seen more ef­forts to include concerns with economic power, insecurity, and inequality. Platform cooperativism,⁴⁶ the purpose-driven OuiShare festivals, the Sharing Cities Alliance, and the National Domestic Workers Alliance are all seeking to reorient technologically mediated economic practice toward egalitarian relations of production, using different mechanisms to embed production in solidaristic social relations (reflecting the focus of various chapters in this volume: cooperativism in Schor and Eddy; purpose-driven organization in Henderson; municipalism in Stears; and social mission-driven nonprofit in Shelby). In academia, leading examples of this reorientation are Cohen's work on the legal construction of informational capitalism;⁴⁷ Zuboff's on surveillance capitalism;⁴⁸ Schor's work on the sharing economy;⁴⁹ Kapczyn- ski's analysis of the cost the price system to innovation and culture;⁵⁰ Levy on monitoring of truckers,⁵¹ or her work with Barocas on the impact of con­sumer data collection on workers;⁵² Ajunwa, Crawford, and Schultz's work on workplace surveillance;⁵³ Dubal's on the precarity of the gig economy driv­ers;⁵⁴ Pasquale's on algorithmic black boxes;⁵⁵ Barocas and Selbst's focus on big data's disparate impact;⁵⁶ and Rogers's work on the major dimensions of technological power employers seek as leverage over employees.⁵⁷ A distinc­tive feature of this newer work is a call for reviving state power as a counter­balance to market power—nowhere more forcefully than in Khan's work on

Power and Productivity 4 5 antitrust⁵⁸ or Rahman's call for leveraging democratic governance to contain the domination of market actors.⁵⁹

In contrast to this power-based story, the most influential neoclassical explanations of rising economic inequality centered on technology as an ex­ogenous and politically neutral force: skills-biased technical change (SBTC)⁶⁰ and the economics of superstars in winner-take-all markets.⁶¹ These expla­nations form the intellectual origin of current arguments that robots will create structurally high levels of unemployment and platforms will casualize work. Technology in these explanations develops exogenously, is roughly de­terministic (some things are easier to automate, others harder), and interacts with efficient labor markets to change the relative value of different kinds of labor (skilled/unskilled workers; routine/nonroutine tasks). This interaction makes highly skilled workers valuable, the super-skilled few superstars, and relegates low- or mid-skilled workers to stagnant or declining wages. These technological explanations naturalize inequality as an inevitable function of the most distinctive dynamic in market society—productivity growth through technological change. Efforts to address inequality must there­fore focus on fitting the poorly trained workers to inevitable technological change. The primary weakness of SBTC and winner-take-all theories is that they fail to explain how countries at the same technological frontier embrace these technologies with widely differing social consequences. The Nordic so­cial democracies, Germany, France, and Japan all operate at the same tech­nological frontier as the United States, yet they exhibit different patterns of inequality, and none exhibits the escape of the 1 percent that characterizes American inequality. Political economy offers a more plausible explanation of these divergent patterns than any generalized, apolitical, and ahistorical model.

The diversity among market societies at the production frontier is a prod­uct of the history of struggle in each society and its resulting social rela­tions of production. These battles shaped how much happened in a market sphere as opposed to through nonmarket relations; how embedded those aspects that do occur in markets are in social relations of mutual obligation and solidarism; and how institutions that diverged between these clusters of democratic market societies shaped power in the economy, both within markets and between market and nonmarket spheres.⁶² The differences in productivity and inequality in Demark and Mississippi are not usefully un­derstood in terms of universals but in terms of specific historical struggles, over class, race, gender, and immigration, and how the social and material context inherited from these struggles determines current distributions of

power within markets and dependence on and alternatives to markets for people satisfying their basic needs and developing basic capabilities. One cannot understand the relative weakness of labor in the United States, by comparison to Europe, or the relative stinginess of American social insurance without understanding the central role that enslaved Black workers played as the core workforce of America's leading export industry in its first seven decades and the role racial ideology has played as the central strategy for undermining labor mobilization and social insurance in America ever since.⁶³ Nor can one understand American capitalism without understanding how waves of immigration and anti-immigrant sentiment were used by capital to undermine labor organizing throughout the nation's history. It is to these dynamics that I turn next.