2 Power and Institutions
Why do some people’s preferences become policy and other people’s don’t?
- Explain what “politics” actually is and what the three faces of power are.
- Explain why concentrated interests usually defeat diffuse ones.
Think of a recent product decision at your company, or a recent change in the taxes you pay. Who actually decided on the change? Who set the agenda so that this was the decision on the table, and whose interests won? Those three questions are the core of this chapter. Power is not just about who wins a visible fight: it is about who decides what is worth fighting about in the first place, and who shapes what people think they want.
Similarly, in order to understand how regulation works in the real world, we first have to understand why and how some interests prevail over others, which means we have to understand power and politics.
2.1 The Three Faces of Power
The most intuitive account of power is the one Steven Lukes calls the first face: who wins when there is a visible decision (Lukes 2021). When a legislature passes a bill or a manager picks one proposal over another, the winner is exercising the most obvious kind of power, and it is easy to see because it happens in public.
Power’s second face is harder to see because it happens before the vote. It is the power to decide what gets onto the agenda in the first place. A product roadmap that doesn’t include work on privacy or a meeting whose agenda is fixed before the people most affected walk into the room are both examples. The people who control what is even considered never have to win the argument, because the argument never happens. This is why “neutral” defaults or “technical” changes are actually agenda-setting dressed up as inevitability.
The third face of power is the power to shape what people even want, or equivalently to ensure that some options never even occur to them. The third face is at work when an arrangement feels natural, obvious, or inevitable rather than chosen, and its catchphrase is, “We don’t do politics here.” What that sentence actually means is that the winners already have what they want, and can afford to describe it as just the way things are.
2.2 Institutions as Rules
Douglass North defined an institution as “the rules of the game”: the formal rules such as laws and contracts, and the informal constraints such as norms, conventions, and taboos that structure human interaction (North 1990). The definition is deliberately broad because institutions do more than constrain: they allocate advantage. Every rule makes some outcomes easier and others harder; in other words, they give some people’s agendas an advantage and others a disadvantage, which means that rules are never actually neutral. The question is not whether a rule is neutral but who decided it was a rule, when, why, and who was dispossessed in the process.
The English enclosure movement shows how this plays out in practice. Between 1750 and 1850, Parliament passed hundreds of Enclosure Acts, each one extinguishing common rights over a specific area and converting that land into private property. The process was not neutral arbitration between competing claims: the landowners who stood to benefit were the same class that controlled Parliament, while the commoners whose rights were extinguished had no equivalent political representation (Bollier 2025). Before enclosure, common land was governed by overlapping, community-enforced use rights: the right to graze a specific number of animals, to cut timber for fuel, or to fish particular stretches of a river. Elinor Ostrom’s research documented how these systems were managed sustainably for centuries without either private ownership or state control (Ostrom 2015). The commons worked; the argument that they were inherently prone to overuse was made by enclosure’s beneficiaries and repeated long after it had been refuted. The result was a large-scale transformation of semi-independent smallholders into wage laborers.
Colonialism carried the enclosure model outward from England as one of the primary instruments of dispossession. In Ireland, India, Africa, and the Americas, colonial law systematically refused to recognize collective or customary land tenure. Land that was not individually titled under a system legible to European courts was declared waste, Crown land, or legally available for settlement. This did to colonized peoples what the Enclosure Acts had done to English commoners: extinguish practices that had sustained communities for generations, convert resources into commodities that could be extracted or sold, and make the people dependent on their new overlords (Linebaugh 2014).
This process continues today. Brazil’s grilagem (the fraud of converting indigenous and public Amazon land into private titled property using falsified documents) has been digitized in recent decades, with forged records fed into government registries to launder claims at a scale that replicates the Enclosure Acts across half a continent. The same playbook now operates on the web. Email, RSS, and the early web were a commons that anyone could build on. Social platforms converted that commons into walled gardens, moving the audience inside and charging rent for access to it.
2.3 Design as Governance
A piece of software is an institution. Default settings, permission models, and rate limits are rules that allocate advantage just like zoning ordinances or tax codes, which means a design decision is a governance decision, whether or not anyone in the room is willing to say so.
The 1886 American railroad gauge change makes the point concrete. Southern railroads ran on a five-foot gauge while northern lines ran on four feet eight and a half inches, which meant that freight crossing the boundary had to be unloaded and reloaded. Southern carriers profited from that: the incompatibility was a market weapon, and the “technical” choice was political all the way down. The eventual conversion to a single gauge was a decision about who would capture the value of moving goods. It was also an instance of path dependence: the gauge had been chosen for reasons that no longer applied, but once the tracks, the rolling stock, and the jobs were built around it, structural lock-in meant tearing up an entire landscape.
Today’s software platforms reproduce the same logic. The medieval Church provided civil infrastructure, such as record keeping, courts, hospitals, and schools, and collected a tithe of roughly one-tenth of productive output in return, whether you were a believer or not. The App Store’s thirty percent commission is every bit as much a tithe. Excommunication—removal from the Church’s services—meant you could no longer conduct legal business; deplatforming does the same to app creators. They lose access to customers and markets, and like peasants in the Middle Ages, have no higher authority to appeal to (Southern 1990).
Institutions also produce behavior. When Wells Fargo fired 5,300 employees over fake accounts (the “eight is great” scandal), wrongdoing did not flow from five thousand bad individuals. It flowed from impossible quotas, aggressive measurement, and a reward structure that made bad behavior the rational response. The institution shaped the incentives that produced the wrongdoing; the people who followed those incentives were punished, not the ones who had created them.
The Olympus fraud makes the same point at the top of the hierarchy. In October 2011, the Japanese optics company Olympus fired Michael Woodford, its newly appointed chief executive, a few months after he uncovered $1.7 billion in concealed losses that had been running for nearly twenty years (Woodford 2012). The notable thing was not that individuals behaved dishonestly, but that successive layers of management had found it easier to maintain the deception than to stop it. It was as if the organization had developed a mind of its own. Wirecard showed the same pattern in a European setting: when the German payments company collapsed in 2020, €1.9 billion that had been reported as revenue turned out never to have existed. In this case, Germany’s financial regulator attacked the journalists who exposed the fraud instead of the fraud itself.
Institutions can also be self-fulfilling. Terms of service written for adversaries, and moderation systems that treat everyone as a suspect, produce the distrust they assume. This contrasts with Elinor Ostrom’s work on the commons: communities have governed shared resources sustainably for centuries but through locally developed rules, monitoring, and graduated sanctions (Ostrom 2015). Design a system as if people are untrustworthy, and you get untrustworthy behavior; design it as if they can govern themselves, and you (sometimes) get self-governance.
The same point applies to the corporate form itself. The usual excuse for why tech companies are not worker cooperatives or other equitable structures is that cooperative governance is too slow and cooperative financing too limited. That explanation is conveniently incomplete: it omits that the founders, lawyers, and venture capitalists making the choice benefit most from the conventional structure (Whyte and Whyte 1991; Kelly 2012). The “neutral” form is a choice with distributional consequences.
The legal scholar Joel Bakan made this point more sharply. He asked the psychologist Robert Hare to score a publicly traded corporation against the clinical checklist used to diagnose psychopathy; Hare concluded that it fit the profile (Bakan 2005). In most jurisdictions, corporate executives have a fiduciary duty to shareholders: they are legally required to pursue shareholder interest, and a board that sacrificed profit to benefit workers or communities could be held liable. The resulting entity is therefore prohibited from having a conscience in the way an individual person might.
2.4 Platforms as Quasi-Governments
When a seller on eBay disputes a transaction, eBay adjudicates the claim. When a developer’s app is removed from the App Store, Apple’s internal review process is the only available appeal. When Facebook removes content in a country with regulated speech, it is making regulatory decisions in a jurisdiction where it has been granted no regulatory authority. Platforms are quasi-governments, and the word “quasi” is doing a lot of work.
The clearest test of what governance actually is came from an unlikely source. In the days after the March 2011 Tōhoku earthquake and tsunami, the yakuza—Japan’s organized crime syndicates— supplied food, water, and emergency goods to affected communities faster than official relief channels could mobilize (Kaplan and Dubro 2012). The lesson is not that organized crime is admirable; it is that governance is whoever can actually get things done. When formal institutions are slow, whoever controls logistics, relationships, and trust governs in their place.
That is the position platforms are building for themselves. The company town is the industrial version of a pattern that goes back further. E.P. Thompson’s account of the transition from putting-out to factory production made a point that factories were initially not more efficient at producing cloth: they were more effective at eliminating worker autonomy. Before industrialization, workers controlled their own time and pace, could share work across household members, and could resist disadvantageous terms through slow work, variable quality, and informal coordination. The factory closed off these responses. Workers arrived at fixed times, worked at a supervised pace, and until labor unions emerged, had no practical means of collective withdrawal. App-based platform work is doing the same thing today. The app is not merely a more convenient interface. It is supervisory infrastructure that monitors pace and completion rates and eliminates the degrees of freedom that less tightly managed arrangements permitted (Thompson 1966).
George Pullman built a company town in Illinois in the 1880s, where the company owned the workers’ housing, stores, church, library, and bank. When Pullman cut wages during the depression of 1893 while holding rents fixed, workers could not cushion the blow by cutting other expenses, because the company controlled those too. They had no recourse and nothing left to lose, and the 1894 Pullman Strike paralyzed rail traffic across the country and required federal troops to suppress— which tells you something about what total control eventually produces (Lindsey 1942). Pullman justified all of this in the language of paternalistic improvement. The company town was presented as a planned community designed to give workers superior housing, sanitation, and amenities. This was partly true: Pullman, Illinois was better built than most industrial housing. But the justification also framed the absence of collective bargaining as a feature. Workers who were being improved did not need unions. They needed to trust the judgment of the person improving them. Conveniently, this converted a power arrangement into a benevolent one.
Pullman was not an isolated case. Coal patches in Appalachia and the Canadian Maritimes ran on the same model: company housing, company stores where scrip could be spent, and company control of access roads and community infrastructure. In the post-Civil War South, plantation stores extended dependency across the formal end of slavery, tying sharecroppers to specific land through debt that was nearly impossible to escape. Mining camps across Latin America reproduced the same structure under different flags. The company town recurs because it solves a real problem for whoever controls it: it makes exit nearly impossible while appearing to be voluntary.
The distinction between short-term dependency and structural lock-in is what keeps workers and developers in arrangements that are getting worse. Short-term dependency is a rational calculation: you stay because the current terms are good enough and leaving has costs. Structural lock-in is different. The costs of leaving have been deliberately raised so that the calculation does not change even as the terms worsen. A platform can worsen its terms incrementally; each worsening is individually insufficient to justify the exit cost, even if the cumulative total is large.
A developer whose business depends entirely on the App Store, a seller whose inventory and customer relationships live on Amazon, or a creator whose audience exists only within a single platform is in an equivalent position: the terms can be changed at any time, the cost of departure is prohibitive, and the entity that provides the housing also adjudicates disputes about it.
The difference between an employer and an infrastructure provider is a matter of degree, not kind. An employer controls your income; an infrastructure provider controls the conditions under which you can earn, spend, and participate in social life. Mao-era China’s danwei work units controlled employment, housing, healthcare, food rations, and education through a single institution, so that to be removed from a danwei was to lose all of them at once. A super-app like WeChat integrates messaging, payments, healthcare bookings, government services, and social identity into a single system; a user whose account is suspended loses not just a communication tool but the infrastructure through which they conduct their civic and commercial life (Stoller 2019; Walder 2017). The Pullman Company controlled the economic conditions of workers’ lives. The super-app, at its most developed, aims to be a danwei. Foxconn’s Shenzhen campuses, which house hundreds of thousands of workers in company dormitories with company stores and company transportation, are the Pullman model at a scale Pullman could not have imagined.
2.5 Legibility and Institutional Money
James C. Scott argued that states standardize people so they can be taxed, drafted, and policed. Surnames, house numbers, passports, and identity registers exist not because the state wants to know you, but because it wants to be able to act upon you (Scott 1998). Scott called this legibility: the degree to which an institution can observe, measure, and administer a population through standardized categories. The technical history of identification has a concrete origin. In 1879, a Paris police clerk named Alphonse Bertillon proposed measuring the skull and the length of the forearm so that a repeat offender could be pinned to a record even when he gave a false name. His anthropometry spread across Europe, the United States, and colonies in Asia and Africa, and was the first large-scale attempt to use the body as a database. Bertillonage had a fundamental weakness: measurements had to be taken correctly by trained operators. By 1900 fingerprinting was replacing it almost everywhere because fingerprints were more reliable and required less skill to record. But the desire of states and institutions to pin individuals permanently to a record did not die with the calipers. Fingerprinting replaced it within a few decades because it needed less skill to record, but the desire to pin individuals permanently to a record did not die with the calipers. The technical history of identification is also a political history of control, and South Africa’s apartheid-era pass system— which became the world’s first large-scale biometric register by 1986— is the sharpest reminder of it (Breckenridge 2016). The pass book recorded a person’s employer, their designated “homeland,” and their permission to be in urban areas. Police could stop anyone and demand the book; failure to produce it meant arrest. The pass system was one of the most sophisticated identity management projects in history, and it was designed entirely to restrict freedom of movement and force labor into mines and factories at wages set by the government. Colonial governments introduced population registers, caste certificates, and tribal designations across India, Africa, and Southeast Asia that served double purposes: administration and taxation on one hand, and identifying potential troublemakers and restricting movement on the other. The state’s desire for legibility does not have a built-in limit. The state that registers births so it can provide schooling can use that register to conscript soldiers. The government that issues identity documents to allow people to vote can use those documents to deport people.
Legibility also cuts in a contradictory direction. Democratic participation depends on being able to identify voters, so electoral systems require voter registration, and registration requires identity documentation. In the United States, that requirement has a long history of being weaponized: poll taxes, literacy tests, grandfather clauses, and, since the 1990s, photo ID requirements that are nominally neutral but fall hardest on the poor, the elderly, and people of color. Communities that have reason to distrust state identity systems— people who remember population registers being used to intern Japanese-Americans, or to deport undocumented immigrants in mass raids— are the same communities such systems are most likely to exclude. The registration built to enable participation keeps being turned against it, and the Voting Rights Act of 1965 exists because marchers and organizers made the status quo more costly than change (Cole 2002; Torpey 2000).
India’s Aadhaar system is the sharpest present-day version of the dilemma. Launched in 2009, it has enrolled over 1.3 billion people in a biometric database of fingerprints and iris scans linked to a twelve-digit number. Its designers argued that a universal biometric identifier would eliminate fraud and exclusion in government benefit programs, and it has reduced certain kinds of fraud while bringing some previously excluded people into the formal economy. It has also created new exclusions: people who cannot authenticate are cut off from food rations and pension payments, and a database this concentrated cannot be “reset” if it is breached— you can change a password, but you cannot get new fingerprints (Khera 2019).
The same logic has a commercial endpoint. China’s Social Credit System bars individuals with low scores from purchasing plane and train tickets, enrolling their children in private schools, and accessing certain financial services: a score derived from past behavior now determines present civic participation.
2.6 Concentrated Interests, Diffuse Costs
In 1936, Harold Lasswell defined politics as the study of “who gets what, when, how.” More recently, it has been defined as what people do when they don’t agree on goals or values, but have to work together anyway. As Isaiah Berlin argued, these disagreements aren’t a temporary condition awaiting the right institutional fix (Berlin 1991). The goods people pursue are genuinely plural and irreconcilable. liberty and equality cannot both be maximized at the same; neither can security and freedom or efficiency and fairness. Liberal democracy is not a system for finding the “right” answer: it is a way to manage the permanent tension between right answers that contradict each other.
Mancur Olson explained why some groups reliably win and others reliably lose. It is easier to organize a small group with a large stake than a large group with a small stake (Olson 1965). Ad-tech firms—few, organized, and rich— can coordinate, hire lobbyists, and file comments on proposed regulations. Individual users—many, unorganized, and each with a small stake— mostly cannot. This single asymmetry predicts much of tech policy: producers beat consumers and platforms beat users, not because anyone is smarter or more virtuous, but because the costs of organizing are distributed unevenly.
Selectorate theory sharpens the point. Leaders of countries, companies, and volunteer organizations need enough support to stay in power, and they get it by distributing benefits to a minimum necessary winning coalition rather than to the whole population (Mesquita and Smith 2011). Everyone else is interchangeable, whatever the language about family. In November 2022, after laying off about a third of its workforce, Elon Musk emailed the remaining Twitter employees asking them to click a button confirming that they were committed to working “hardcore,” with those who did not click treated as having resigned; a few months earlier, Twitter’s leadership had described the company as a family. The winning coalition is smaller than the family, and it is the coalition that gets served first. The winning coalition is not the whole organization: it is the subset of people whose support the leader actually requires. In an autocracy, this might be the military brass, a security service, and a handful of oligarchs. In a publicly traded company, it is the board, major institutional shareholders, and a small number of indispensable senior executives. Everyone else—the people who are told they are family—is interchangeable. They are what the theory calls the selectorate: large enough to give the winning coalition options if any member defects, but not powerful enough to claim a significant share of private benefits. The perks are cheap ways to signal belonging to people who are not actually in the winning coalition. The rhetoric costs nothing but extracts real commitment. The “family” language comes with real perks attached. Amazon preaches a “Day 1 culture,” and Google has offered free meals, nap pods, and climbing walls, inviting employees to treat the campus as home. The metaphor does real work: it extracts commitment, discourages outside offers, and makes employees—particularly younger ones— feel that the relationship is something other than a transaction. What it does not change is what the relationship actually is. In most of the United States, and in varying degrees elsewhere, employment is at-will: either party can end it, at any time, for any reason not specifically prohibited by law. The “family” exists at the employer’s pleasure.
None of this is a secret to the people who succeed at it. Jeffrey Pfeffer found that political skill predicts advancement better than technical competence, and that managers who ignore politics systematically lose to those who do not (Pfeffer 1992). That is an uncomfortable finding for an industry that likes to imagine decisions are made by the best argument. They are made by the people who can assemble a coalition, set an agenda, and make their own interests look natural. Interests and beliefs are not opposites. People pursue what they think is right, and what they think is right is shaped by their position in the organization. A sales leader who believes the product team should prioritize enterprise features is not lying. She is telling the truth as experienced from where she sits. She also reaps the benefits if her plan is adopted.
The phrase “we don’t do politics here” appears regularly in tech companies. It usually means one of two things. The first is a genuine belief that technical decisions should be made on technical merit, that coalition-building is a form of corruption. That belief is reasonable, but almost entirely wrong about how decisions get made. The second meaning is that the people in the winning coalition have already gotten what they want from the current structure, so there is no need for them to engage in visible political activity. People who are well-served by existing arrangements can afford to describe those arrangements as natural and political contestation as illegitimate.
In 2021, the founders of Basecamp banned “societal and political discussions” on internal company channels. They framed this as keeping the workplace professional and focused. Roughly a third of the company’s employees resigned within days, recognizing that the decision to ban discussion of politics was itself a political decision, made unilaterally by the winning coalition.
The family metaphor is not unique to American tech companies. Japan’s postwar lifetime employment system produced what selectorate theory would describe as a very large nominal selectorate of permanent employees with a winning coalition drawn from senior management and major shareholders. When the asset bubble collapsed in 1990, the permanent employment guarantee at the heart of the “family” bargain was broken through expansion of temporary and contract employment that covered the same work under worse conditions. The family language remained; the security it implied did not.
The question is never whether power is present; it is whose interests it is serving.