The New Corporate Enclaves

Executive Summary

Dominant corporations like SpaceX are devising new ways to escape democratic accountability through jurisdictional secession. Their recent efforts parallel efforts from previous industrial barons, like George Pullman and Walt Disney, to secure governing independence through different institutional organizations of the historical company town. Rather than constraining these efforts, local government law has facilitated the emergence of the contemporary corporate enclave—while also blocking the remedy that central cities like Chicago had previously wielded against the corporate capture of local politics: municipal annexation. These trends present a serious challenge for efforts by economic progressives to revive local democracy in response to the ongoing backsliding of federal institutions into entrenched oligarchy.

This report documents this trend through four case studies:

• Disney World’s self-governed district near Orlando
• The nationwide bidding war that Amazon instigated during the search for its second headquarters
• Wisconsin’s Foxconn deal
• Nevada’s proposed innovation zones

Synthesizing the academic literature, it finds several economic and democratic harms from these new corporate enclaves, including distortions to competition, higher taxes and worse services, harms to workers, and harms to democracy. It then looks at some of the challenges that local governments have faced and will face if they attempt to better regulate these enclaves.

Introduction

Earlier this year, 218 residents of Starbase, Texas—almost all employees of Elon Musk’s aerospace transportation company SpaceX or their family members—voted overwhelmingly to incorporate the company’s landholdings into an independent and self-governed jurisdiction.¹ The incorporation represented the culmination of a yearslong effort by Musk, first confirmed in a 2021 tweet on the social media platform that he now owns: “Creating the city of Starbase, Texas.”²

Almost immediately after the incorporation vote, the company (operating through its new shell government) consolidated its governing authority over a wide range of local policymaking that affected its interests. The Starbase city commissioners—all of them current or former SpaceX employees³—approved a request from SpaceX to cut off public access to the community by installing gates, accessible only by access code, at points of entry.⁴ Although the new city has legal authority to issue public debt, its initial operations were funded by a direct $1.5 million loan from SpaceX,⁵ forgoing the legal processes and transparency mechanisms that typically accompany the issuance of municipal bonds. And in October 2025, Starbase announced its intention to contract with SpaceX itself for fire response and other emergency services.⁶

At every step, the Texas legislature has sought to keep Musk happy—and retain his companies’ investments in the state—by shielding SpaceX from taxation, regulation, and democratic accountability. Over three years before Musk selected the launch site near Brownsville that is now Starbase, he held a competition between states (including Florida and Georgia) for tax breaks and regulatory relief. In 2011, Texas officials approved Musk’s demands, awarding about $20 million in financial incentives, changing state law to create a process for closing public beaches during launches, and legally shielding the company from noise complaints and other liability statutes.⁷ The county agreed to withhold corporate information that would otherwise be subject to open records laws. More recently, the new incorporation has facilitated this pattern: In June 2025, SpaceX secured state legislation transferring to its new government the sovereign authority to shut down state highways and public beaches during launch events, without input from the county.⁸ After a recent approval from the Federal Aviation Administration, those events may increase to 25 times per year.⁹

Those incorporation efforts are consistent with similar pushes by gigantic, dominant corporate conglomerates around the country. Recent headlines have covered conflict between Florida Governor Ron DeSantis and Disney over the governance of the company’s territory outside of Orlando, the still-mysterious effort by a group of tech billionaires to start a private city outside of San Francisco, and the for-profit city of Próspera, located in Honduras but privately operated by a Delaware-based company.¹⁰ As the New York Times recently reported: “Enticed by the potential profits and eager to have more control over their footprint, investors and businesses are backing new town and city concepts that . . . are guided by a private hand.”¹¹

Starbase represents the culmination of two concerning trends: (1) dominant firms’ use of structural power in the interjurisdictional “market” for mobile investment to extract firm-specific tax relief and, increasingly, deregulation through the corporate location incentive megadeal; and (2) the reemergence of the historical “company town” as incorporated corporate enclave for these dominant firms. In particular, the episode reveals how structural power can be used to secure legal autonomy from local democratic accountability—including, in a growing number of cases, the attainment of sovereign power. Both strategies harm workers, distort market structure, and increase costs for taxpayers and consumers.

This report describes how dominant corporations are successfully seceding from public control through the combination of structural power and various forms of jurisdictional fragmentation. Changes in local government law—in particular, increased hostility to efforts by central cities to eliminate corporate enclaves by annexation—have facilitated these efforts by dominant firms. Particularly in a moment of federal retrenchment from the enforcement of labor rights and environmental protections, many progressives are rediscovering the promise of local democracy. Inspiring grassroots campaigns—like Zohran Mamdani’s mayoral election in New York City—offer generalizable lessons for how to exercise greater democratic control over dominant corporations through local organizing. But the new corporate enclave reveals the limits of those strategies, in the absence of legal reforms. When Elon Musk can escape from these forms of accountability by simply incorporating his own city—populated only by loyalist management—the tool of local democracy has limited utility.

Structural Power and the Corporate Enclave

Economists often conceptualize corporate location incentives as reflecting the market-clearing “price” that governments “pay” to influence the spatial allocation of private capital, established through a process of competitive bidding from local jurisdictions. But recent empirical evidence suggests that this interjurisdictional bidding market for mobile economic activity is severely distorted by firms’ ability to leverage (through capital flight) their dominance in those negotiations. This form of market power helps explain the robust relationship researchers have established between firm size and the receipt of generous—and economically unjustified—location incentives.

During the Gilded Age and Progressive Era, reformers concluded that the dominant firms had attained a destructive and undemocratic power over local communities. As legal theorist K. Sabeel Rahman has recounted, Progressive Era reformers observed that large corporations “had acquired a size and a degree of economic and political power that could affect a wide range of other actors in society—not only [their] own workers, but also small businesses, and governments threatened by competitive pressure and corporate influence.”¹² Over subsequent decades, comparative political economists have closely studied this form of power by private capital over public governance—conceptualizing it as structural in nature.¹³ As scholars Jacob Hacker and Paul Pierson observe, firms’ structural power is particularly heightened where exit threats are both credible and consequential—in particular, where (1) investment capital is highly mobile, (2) political institutions with decision-making authority are spatially fragmented, and (3) fiscal authority is decentralized.¹⁴

Particularly following the Great Recession, a growing share of this public spending has been tied up in a handful of “megadeals”—available only to, and increasingly demanded by, the largest and most dominant corporations. These large firms have become adept at both leveraging cities’ economic vulnerability to extract public subsidies and devising new schemes by which jurisdictions can reduce their business costs and other liabilities in a targeted way.

Describing the distinctive features of the American political economy, one recent account highlighted how these dynamics “open possibilities for mobile actors to venue shop, maneuvering across states and localities to find the most favorable structures of political authority, often putting downward pressure on regulations and taxes.”¹⁵ Indeed, by one measure, state and local governments now spend nearly $50 billion annually bidding against each other to entice corporations to locate mobile economic activity within their jurisdiction.¹⁶ These deals have grown in recent decades, becoming both more frequently negotiated and larger in size when awarded. Particularly following the Great Recession, a growing share of this public spending has been tied up in a handful of “megadeals”—available only to, and increasingly demanded by, the largest and most dominant corporations. These large firms have become adept at both leveraging cities’ economic vulnerability to extract public subsidies and devising new schemes by which jurisdictions can reduce their business costs and other liabilities in a targeted way.

Economists Cailin Slattery and Owen Zidar, who have studied these deals extensively, thus describe state and local business tax incentives as the “primary place-based policy in the United States.”¹⁷

Headlines about these megadeals tend to focus on the public dollars transferred through state and local programs of tax relief. But many of the more recent megadeals include various provisions that exempt specific corporations from legal liability and regulatory control. Corporate giants like Amazon are becoming more sophisticated at “devising new schemes by which state and local governments can act to reduce their business costs and other liabilities in a targeted manner” through the location incentive megadeal.¹⁸ Many such strategies—which seek broadly to reduce business costs by eliminating taxation, offloading private costs onto the public, and avoiding local regulation—share an underlying logic with the contemporary corporate enclave.

When local governments compete to entice corporate investment, their efforts often require policy choices that undermine economic fairness and local democracy. Classic theories of fiscal federalism predict that competition among local governments for mobile capital—the wealth held by residents as well as firms—restricts their practical ability to implement redistributive policy, regardless of local preferences.¹⁹ The same forms of jurisdictional fragmentation that have facilitated the creation of corporate enclaves like Starbase also open new possibilities for mobile actors to venue shop, maneuvering across states and localities to find the most favorable structures of political authority, and putting downward pressure on regulations and taxes.

Starbase represents the culmination of this trend, but it is not the only Texas community that Elon Musk is trying to incorporate to secure state power for his companies’ operations. He has acquired thousands of acres of farmland in Bastrop County, outside of Austin, with the stated goal of securing a similar institutional arrangement that could support an entire private city.²⁰ His opposition to locally enforced workplace safety regulations (in addition to state taxes) motivated his initial relocation of Tesla’s corporate headquarters from California to Texas.²¹ At the time, Alameda County had imposed a pandemic public health closure that required the temporary halt of Tesla’s manufacturing operations; Musk sued the county, also threatening to shift his operations to other states “immediately” if his demands were not met.²² The state of California took the side of Tesla, and Alameda ultimately relented²³—but in the meantime, lawmakers in Texas (as well as in Nevada, Georgia, Utah, and Oklahoma) had already put together generous incentive packages to entice relocation.²⁴

A Continuation of the Historical Company Town

During the Gilded Age and Progressive Era, proprietors of company-owned towns exercised absolute control over workers through the private law of property and contract.²⁵ Mining bosses and industrial barons like George Pullman intentionally rejected the municipal form, using dismissal and eviction to enforce company policy as the operative governing authority within their dominions. This strategy became less effective after the New Deal, leading observers to pronounce the demise of the company town.²⁶ But parallel developments in local government law have allowed the company town to proliferate through new institutional forms, enabling corporate titans to wield public powers without accountability to any broad public. Rather than facilitating domination over captive worker-residents, these institutional forms are now used primarily to externalize costs and escape democratic obligations like taxation—goals that also deeply motivated the governance of historical company towns.

Indeed, as a matter of public law, historical company towns were not “towns” at all: Almost without exception, they operated without a municipal charter. We call them towns because they featured the hallmarks of a residential community: housing, schools, and places of worship, in addition to the workplace. As the Supreme Court summarized in its famous Marsh decision, the company town of Chickasaw, Alabama, had “all the characteristics of any other American town” except that it is “owned by the Gulf Shipbuilding Corporation.”²⁷ But this was not a legal designation. The historical company town had no mayor, and no municipal elections. These communities were characterized by the absence of any sort of representative policymaking bodies.

This common characteristic was no coincidence. It represented a conscious decision of town bosses like Pullman, who went to great lengths to avoid the municipal form. Democracy was the main deterrent. When asked by a reporter whether he planned to form a city government, Pullman states his intentions clearly: “As long as the town of Pullman is owned by one association, there’s little necessity of agitating the subject of its control.”²⁸ But although company towns were not governments, they were governed through highly organized processes where company policy operated as the governing authority within the community. Residents were often bound by strict morality codes banning alcohol and sex work, speech was highly restricted, and any efforts by workers to assert their collective rights, and to form a union, were severely restricted.²⁹

Even though company policy was not public law, it could be enforced by state power through the rights of contract (the employment relationship) and, in particular, the property right to exclude. Outsiders were banned, and if you violated company policy or opposed its interests, you were fired and then evicted. One historian described the result: “The company alone was the town legislative and executive body. It handed down all the rules and regulations and was responsible for putting them into effect.”³⁰ Rather than seeking to commandeer local democratic institutions, industrial barons during this period used spatially consolidated property ownership and nonnegotiable adhesion contracts to displace public law as the governing authority within their dominions. Through this process, town bosses eliminated any democratic friction between the company’s governing interests and the authority that governs public life within a territory (private law, in these cases). Although the legal mechanism that facilitates Musk’s domination today is different, the outcome is the same.

Case Studies

The following case studies demonstrate how large, dominant firms like SpaceX and Amazon are using their structural power over local communities to evade democratic accountability. Their governing arrangements go beyond the forms of tax relief that are typically associated with corporate location incentive deals: In important ways, as described below, they represent forms of secession from the jurisdictional authority of local governing institutions. A key lesson is that the corporate enclave is a vehicle for new forms of targeted deregulation, an effective blocking maneuver for gigantic corporations to prevent regulation and taxation by (more democratically inclusive) local governments. And the outer limits of this strategy have not yet been realized. Indeed, the available evidence suggests that dominant firms will continue to innovate new forms of jurisdictional exit at the expense of economically vulnerable jurisdictions competing for their capital investment.

The Disney Playbook

The Starbase incorporation represents the complete realization of a strategy that corporations seeking to evade democratic control have adopted over the years. Walt Disney was an early innovator of this strategy, through Disney’s self-governed district near Orlando: the Reedy Creek Improvement District.

During the 1960s, Walt Disney oversaw a national competition between states to land his new “Disneyland of the East.” The company was very clear in its negotiations with local officials: If you want our investment, you need to give us our own government.³¹ In a special act in 1967, Florida delegated to Disney—through a special purpose district—the ability to provide its own public services, while shifting various costs of its amusement empire to the public fisc. Florida gave the company full control over land use and safety regulation within its district and then made it responsible for monitoring its own compliance. More broadly, the arrangement provided legal protection from taxation demands of local neighbors (outside the special district boundaries) and various forms of municipal and county regulation.

Disney pitched a plan to create a residential community to the state legislature, but, once the state approved the Reedy Creek district, Disney maintained its political control by limiting the residential population to hand-selected loyalists.³² For example, eligibility to serve on Reedy Creek’s governing board required ownership of land within the district. Disney has managed this access by temporarily deeding five-acre plots of land to approved candidates, to be returned to Disney at the conclusion of the owners’ “public” service. As a state audit described: Although empowered with the authority of a municipality, Reedy Creek is “nothing like a real city, with only a dozen residents in each, all current or former Disney employees or tenants.”³³

Amazon’s HQ2 Search

In September 2017, the tech conglomerate Amazon publicly launched its search for a second North American headquarters. The company’s press release proclaimed that this “Amazon HQ2” would be a “full equal” to its current base in Seattle, bringing “billions of dollars” in new investments and “as many as 50,000 high-paying jobs” to whichever metropolitan area put together the winning proposal.³⁴ Amazon invited local and state government officials to prepare formal bids, expressly requiring that the winning bid include a lucrative package of location incentives.

The company identified specific tax exemptions and other economic incentives that it expected jurisdictions to provide, describing its costs of doing business as “critical decision drivers.” Amazon demanded strict secrecy from competing jurisdictions about their promised incentives, which the Request for Proposals claimed were “confidential, proprietary, and constitute trade secrets.” Over the next five weeks, government officials behind the 238 submitted proposals—coming from 43 states, Washington, DC, and Puerto Rico—more than met this challenge.³⁵

Chicago reportedly promised Amazon at least $2.3 billion in combined state and local tax breaks and other financial incentives, including payroll credits, sales tax exemptions, property tax breaks, and direct investments.³⁶ Maryland’s Montgomery County offered the largest known subsidy package, totaling $8.5 billion.³⁷ Over half of this amount took the form of an unusual arrangement whereby every dollar that normally would be paid by Amazon employees in state income taxes (and used for collective public needs) would be diverted from the public fisc and transferred directly from workers to the private company that employs them. As part of Newark’s bid, New Jersey Governor Chris Christie offered Amazon a $10,000 tax credit per employee.³⁸ The state’s accompanying incentive package ultimately amounted to around $7 billion in tax breaks for the company—$2 billion more than Amazon had expected to invest in the new headquarters. And in perhaps the most dramatic preview of recent incorporation trends, one Georgia city council voted to form a new city of “Amazon” by de-annexing 345 acres from the jurisdictional boundaries of an existing suburb just outside Atlanta.

After years of wrangling, Amazon announced in 2023 that it was indefinitely pausing the second phase of its HQ2 project in Arlington, Virginia—“a decision that coincides with the company’s deepest ever job cuts and a reassessment of office needs to account for remote work.”³⁹ As recently reported in the Washington Post—the newspaper owned  by Amazon’s former CEO and major investor, Jeff Bezos—the company remains “behind on jobs promised for funding to build Virginia headquarters.”⁴⁰

Foxconn’s Regulatory Relief

Shortly before Amazon announced its HQ2 search, Wisconsin Governor Scott Walker announced an agreement to provide Taiwanese manufacturer Foxconn between $3 billion and $4.8 billion in subsidies in exchange for building a large flatscreen panel plant near Racine. The subsidy package was estimated to be the largest, by a significant margin, ever awarded to a foreign firm in US history—the equivalent to paying Foxconn a 30 percent wage subsidy for 20 years, over 10 times as generous as the current average incentive offer.⁴¹

By 2018, the expected value of the subsidy package grew to almost $5 billion—over half of which Foxconn would receive directly in cash as a discretionary tax abatement assessed against already-eliminated state tax liability. The state also agreed to make road improvements worth over $252 million in addition to providing tax breaks for construction worth $150 million. Local governments also added their own incentives to the state’s offer, including $1.5 billion through a new tax increment financing district. To fund those promises, Mount Pleasant and Racine County together issued more than $310 million in debt;⁴² in August 2018, Moody’s downgraded Mount Pleasant’s credit rating.

Wisconsin’s enticement package was not limited to direct financial incentives. As part of the recruitment deal, Wisconsin exempted Foxconn from various state environmental regulations.⁴³ Governor Walker also offered a statutory provision creating a legal right, exclusive to Foxconn, to appeal any unfavorable judicial or regulatory decision immediately and directly to the state’s business-friendly Supreme Court. This provision—later dropped amid public outcry—represents an extreme and unusual form of targeted regulatory relief intended to deliver only one company broad exemption from otherwise binding state policies and legal constraints.

Nearly a decade later, the deal stands as an example of the precarious nature of the jobs promises that motivate these deals. Although highly secretive about its plans, Foxconn already by 2020 appeared to be making just a fraction of its initial promised investment—building a factory with less than 5 percent of the square footage of the initially proposed manufacturing plant, with robots doing much of the assembly work.⁴⁴ Wisconsin Governor Tony Evers (Walker’s successor) later succeeded, over legislative opposition, in renegotiating the terms of the state incentives. But, as reported by the Financial Times, “the sprawling complex and 13,000 jobs never came, and the project has become a case study in the gap between dealmaking hype and on-the-ground reality.”⁴⁵

Nevada’s Proposed “Innovation Zones”

In his 2021 State of the State address, Nevada’s Democratic Governor Steve Sisolak outlined a series of policy proposals to attract economic investment to the state.⁴⁶ In addition to initiatives supporting job training, infrastructure improvements, and community education, Sisolak proposed authorizing a new form of political subdivision that would be available to “new companies creating groundbreaking technologies.” If enacted, qualifying tech companies would have the legal right to govern, as sovereign, through autonomous “innovation zones.” Upon promising to make significant capital investments, such companies could secede from the jurisdiction of the existing county and govern through an unelected local board of supervisors—a majority of whom (two of three) would be nominated directly by the company. Provided that the total number of registered voters residing within the zone remained under 100, no general elections would be required.

Subsequent reporting revealed that the proposal had been championed by the cryptocurrency company Blockchains LLC, which had provided $60,000 to Governor Sisolak’s campaign and affiliated political action committee.⁴⁷ In 2018, Blockchains purchased 67,000 acres of land at an industrial center located outside Reno; it later announced plans to develop a futuristic “smart city” on the site.⁴⁸ According to a promotional memorandum prepared by Blockchains, this new authority would “provide the legal, fiscal and governmental structure for [the company’s proposed] Smart City to develop, operate and grow.” Blockchains’ proposal would allow the company, and other qualifying tech firms, to control a wide variety of local services within their self-governed zones. The proposed statutory language granted the governing body the sovereign power to “perform the duties and provide the services” of an extensive list of county offices, which could be consolidated into a single office as determined by the (self-governed) board. The language further clarified that the board’s authority “supersedes the exercise of that power or duty by the county in which the Innovation Zone is situated.”

Among the state powers expressly delegated to qualifying firms was the right to establish an independent local judicial system with “jurisdiction over cases filed [within the Zone] on or after the date that the creation of the justice court takes effect.” Amid opposition, Governor Sisolak later backed away from the Innovation Zone legislation.⁴⁹ Citing this lack of support, Blockchains announced that it would no longer pursue the construction of its smart city.⁵⁰ But the proposed legal framework previews the likely destination of these trends.

The Harms of the Location Incentive Megadeal and the New Corporate Enclave

But the contemporary corporate enclave—and location incentive megadeals more broadly—does not only harm local communities in the abstract democratic sense. These case studies illustrate some of the distinctive harms presented by the self-governed corporate enclave and, more broadly, the exercise of structural power over local communities seeking to attract capital investment: distorting market competition, harming local workers and other (nonresident) neighbors, and undermining democratic accountability.

Recent empirical research confirms that larger firms are more likely to receive generous subsidy packages than their smaller rivals. Slattery and Zidar’s 2020 study assesses how firm characteristics affect the ability to secure these subsidies from competing jurisdictions. In a review of 543 firm-specific subsidies between 2002 and 2017, they find that “large, profitable firms are more likely to receive firm-specific subsidies”—describing the effect as “striking.”⁵¹ Between 80 and 96 percent of incentives go to firms with over 100 employees, even though such firms are responsible for only two-thirds of private-sector employment.

Not only are the largest firms most likely to get the largest deals (even on a per-jobs basis), but recent research also suggests those megadeals may be least likely to prove economically beneficial to the offering jurisdiction. The authors of one such study suggest a connection between firms’ dominance (relative to competing state and local jurisdictions) and their ability to extract incentives—observing that “large establishments have greater bargaining power and thus are able to extract incentives, even though they are not producing or promising higher-than-average employment gains.”⁵²

Distortions to Competition

In 2017, just 20 firms accounted for $6 billion in promised tax incentives and subsidies—one-third of states’ economic development budgets. By one count, the four major tech companies—Apple, Amazon, Google, and Facebook—have collectively received over $5 billionin recent support from state and local governments.⁵⁵ According to researcher Pat Garofalo, Amazon alone received nearly $3 billion in location incentives between 2000 and 2020. Although little empirical work has directly addressed how these megadeals affect the dynamics of market competition,⁵⁶ researchers have speculated about how the dominant companies benefiting from these firm-specific incentives convert this advantage into market power, negatively affecting competitive conditions. Slattery and Zidar hypothesize that “the fact that only the largest firms receive large discretionary subsidies may facilitate increasing industry concentration.”⁵⁷

Extensive suggestive evidence demonstrates that superior access to state-provided capital—secured through structural power in the market for corporate location—can be leveraged to undermine competition among market rivals. Location incentives are legal commitments, demanded from competing local jurisdictions, that transfer public resources to specific favored firms—but not to their rivals. When structurally powerful firms receive these targeted benefits from states and localities, they secure an artificial market advantage (in the form of access to capital) that is unavailable to their competitors that lack the same bargaining power vis-à-vis the state. By conferring to the receiving firm superior access to capital, these preferential forms of state aid may ultimately affect market structure.

The contemporary corporate enclave supercharges these market distortions—as well as the race-to-the-bottom dynamic that already characterizes the interjurisdictional market for corporate investments. The result of this arms race is a distorted playing field that shifts the determinants of market success away from fair competition and toward efforts to secure targeted state favors—including, as highest aspiration, the sovereign power of the state itself.

Consider the advantages, from a business perspective, that Disney has derived from its special district. When its competitor Universal Studios, located just a few miles away, wants to expand its amusement infrastructure, it must raise capital through private borrowing and then navigate local land-use processes. Once that new park is in operation, Universal pays local taxes on its assessed value and complies with various local regulations. By contrast, Disney can fund capital projects by issuing a tax-preferred municipal bond; with financing secured, the company gives itself permission to build. And Disney’s special district allows it to avoid redistributive local taxes; the company writes its own building codes and then enforces them. A recent state audit concluded that “being exempt from the requirements that bound other large developments clearly gave Disney a leg up on the competition.”⁵⁸

After decades of competing on these uneven terms, Universal has more recently demanded that its own local authorities spend public money to reduce its capital cost disadvantage. And in 2023, Universal Studios petitioned to create a special purpose district encompassing its new Epic Universe theme park, which opened in the summer of 2025. Orange County’s board of governors approved the request by unanimous consent, granting sovereign authorities—including the ability to issue tax-exempt bonds—to the territory’s two landowners, Universal and Hilton.⁵⁹

Higher Taxes and Worse Public Services

In order to offset the budget costs of these massive deals, jurisdictions may increase costs on other rival firms, including through higher taxes. As Slattery and Zidar observe, “if the tax revenue generated from additional employment and economic activity falls short of paying for the incentive, it will need to be financed through higher taxes or lower public service provision, reducing the welfare of local workers.”⁶⁰ Timothy Bartik observes that “the net financial costs of incentives, after netting out fiscal benefits, must come from somewhere in state and local budgets.”⁶¹ Jurisdictions may seek to recoup those losses through new taxes and user fees for critical services.

This basic dynamic can take different forms. Recent reporting established that Google data centers use 29 percent of all water in one Oregon city.⁶² In his book about Amazon’s fulfillment centers, journalist Alec MacGillis reports on one Virginia jurisdiction’s effort to construct a new 230-kilovolt power line to a substation—built at a public cost of $65 million to serve Amazon specifically.⁶³ Under state law, an extension intended for a single user must be paid for by that user. But the Virginia legislature approved a proposal to pay for the power line with a monthly fee on all ratepayers. Amazon then filed a 78-page application to state regulators “seeking a special discounted rate for the power it would be using at its data centers,” information about which the company insisted be made off limits to public request.

For example, in 2022, Oklahoma state legislators scrapped plans to eliminate the state’s regressive tax on groceries after a majority agreed to allocate $698 million for an incentive package to entice an unknown company, rumored to be Panasonic.

More broadly, writing in 1984, Clarence Stone observed that “regressive taxes, like sales taxes, have become more prevalent” as local governments have subsidized and assumed some of the risks of economic transformation.⁶⁴ For example, in 2022, Oklahoma state legislators scrapped plans to eliminate the state’s regressive tax on groceries after a majority agreed to allocate $698 million for an incentive package to entice an unknown company, rumored to be Panasonic. One state representative who opposed the deal—Rep. Collin Walke (D-Oklahoma City)—articulated his frustration that residents would pay higher (tax-inclusive) prices as a result of the incentive package: “If the grocery tax [relief] doesn’t pass as a result of having to give three quarters of a billion dollars to an unnamed company, I think that’s pathetic. . . . Oklahomans need help. Panasonic does not.”⁶⁵

One national study of corporate location incentives estimated that property tax abatements—selective reductions in the taxation of corporate property—constitute nearly 30 percent of total corporate tax incentives in the average state.⁶⁶ Those incentives directly affect funding for local public schools, which receive, on average, around half of their funding from local tax revenues. The cumulative effect is significant. One recent study estimates that “economic development tax abatements given to corporations cost public school districts at least $2.37 billion in foregone revenue in FY 2019.”⁶⁷ In one Missouri school district, for example, this loss totaled around $2,000 per student.⁶⁸ Similar consequences can be seen at the state level. Kansas, for example, dramatically scaled back funding for public education—with some schools forced to close early or eliminate instruction days—after historic business income tax cuts reduced state revenues by $700 million.⁶⁹

Harms to Workers

Extensive evidence shows that the conditions of labor monopsony—scenarios in which a (usually rural) locality is dominated by a single employer—suppress wages. One recent study, for example, concluded that the arrival of new Amazon fulfillment centers is associated with a reduction of retail wages within 100 miles by 2.5 percent, equivalent to an $825 decrease in annual income.⁷⁰ In addition to their direct power as employers in labor markets where alternative options are scarce, these firms’ dominance can harm workers by undermining state and local governments’ willingness to enforce legal protections. In one notable example, a 2020 investigation by the Occupational Safety and Health Administration found merit in allegations that Indiana’s workplace safety agency had—during an elaborate state-led effort to bid for Amazon’s second headquarters—taken inappropriate steps to clear Amazon in the 2017 workplace death of one of its Plainfield warehouse employees.⁷¹

In addition to regulatory forbearance, this form of corporate power can also be exercised through affirmative state action. In an effort to forestall a recent union drive in its Bessemer, Alabama, warehouse, for example, Amazon asked local officials to alter traffic light patterns outside its warehouse during shift changes to make it more difficult for organizers to talk to commuting workers. After city officials initially denied the report, Jefferson County later confirmed they had “increased the green light duration in an effort to clear workers off the worksite faster.”⁷² Here again, the corporate enclave supercharges these dynamics: The institutional forms facilitating corporate secession harm workers by undermining the various labor protections that are authorized by state and local authority, from worker safety regulation to minimum wage ordinances. Stated differently: The ability of marginalized workers to overcome mobile capital’s structural power and successfully shape local policymaking processes requires that those workers be included in the relevant democratic process in the first instance.

One recent study, for example, concluded that the arrival of new Amazon fulfillment centers is associated with a reduction of retail wages within 100 miles by 2.5 percent, equivalent to an $825 decrease in annual income.

The stakes are most clearly illustrated in the different organizing conditions for workers at Disney’s Florida and California parks. In California, Disney’s original theme park (Disneyland) is located within the municipal jurisdiction of Anaheim, a city of around 350,000 residents. In early 2018, after years of unsuccessful demands on Disney for higher pay, the local union representing Disney workers succeeded in getting a local ballot question mandating that qualifying “area resort workers” be paid $18 an hour by 2022 (higher than the state minimum of $11). After the initiative qualified for the ballot, Disney gave more than $1.2 million to a political action committee supporting a slate of “Disney-friendly” candidates who would be responsible for implementing the ordinance. (These local decision-makers mattered: For example, the city attorney had acted to protect Disney by issuing a legal interpretation that “Measure L would not apply to Disney . . .”⁷³) But the living wage referendum passed, finally taking effect in 2023, after years of legal challenges. Indeed, a settlement providing retrospective pay under the ordinance was the largest workers’ compensation settlement in California history.

By contrast, workers at Florida’s Disney World—which operates through its own special district—are not able to access this avenue to a living wage; they commute from their own localities to labor in a corporate enclave, where the local government that regulates their employer is their employer. The institutional structure of the corporate enclave removes local ballot initiatives as a site of political organizing from workers’ tool kits entirely. This structure was no accident: Walt Disney’s expansion to Florida, made contingent on the associated demand for his own jurisdiction there, was specifically motivated by his frustration with the local democratic processes to which he was subjected in Anaheim.

Democratic Harms

Concerns about the democratic harms of location incentives helped motivate the United States’ early antimonopoly protections. In 1886, the Senate Select Committee on Interstate Commerce issued a report about the power of the dominant railroad trusts that concluded: “Upon no public question are the people so nearly unanimous as upon the proposition that Congress should undertake in some way the regulation of interstate commerce.”⁷⁴ That report paid special attention to reports of widespread corruption in the process of extracting special concessions from state legislatures and local governments—and to the democratic harms that collectively undermined the ability of local communities to govern themselves.

This concern informed legislators’ conclusion that federal regulation was needed. As legal scholars Joseph Fishkin and William E. Forbath noted in their 2022 book, “radical, reform-minded, mainstream, and conservative analysts [all] agreed that ‘the modern corporation and association of incorporations called trusts’ put both republican self-rule and equal opportunity in peril.”⁷⁵ Rahman explains that “the goal of antitrust reform was to prevent the concentration of economic power by trusts and to enhance individual freedom by protecting against economic—and potentially political—domination.”⁷⁶ Sen. John Sherman described his 1890 legislation as “a bill of rights, a charter of liberty”⁷⁷; Sen. George Hoar warned that monopolies were “a menace to republican institutions themselves.”⁷⁸ These fears are realized in both the contemporary location incentive megadeal and corporate enclave.

Empirical evidence shows that dominant firms’ structural power over state and local decision-making distorts location incentive awards. Researchers have also found that higher corporate subsidy spending in a state is associated with increased campaign contributions for incumbent politicians. Specifically, when a state starts awarding “large” subsidies—defined as a big jump from the historical baseline—annual campaign contributions increase by “20.5 percent (or $158,600) from lobbyists and lawyers who represent large firms in the political process, and 106.8 percent (or $122,000) from large business advocacy and trade organizations.”⁷⁹ Corporations that make state-level campaign contributions are almost four times more likely to receive subsidies—and when they do, “the award is 63 percent larger.”⁸⁰ One 2016 study of state economies finds “that a greater number of lobbyists and campaign contributions from businesses leads to more subsidy spending, all else equal.”⁸¹

Corporations that make state-level campaign contributions are almost four times more likely to receive subsidies—and when they do, “the award is 63 percent larger.”

These ulterior motives can take the form of outright corruption, facilitated by the lack of transparency and magnitude of the dollars at stake. A task force investigating New Jersey’s state economic development agency, for example, found that extensive lobbying and insider dealing had significantly inflated the cost of awarded subsidy deals.⁸² Indeed, according to a study by the Kansas City Federal Reserve, locations with more “troubled political cultures” spend more on corporate subsidies. Specifically, the study finds that “increasing the rate at which government officials are convicted of federal corruption crimes by 1 per 100,000 residents over a 13 year period is associated with a 2.9 percent greater chance that a community will offer business incentives.”⁸³

In recent work, I have argued that the “company town” is best understood as a democratic phenomenon: the absence of friction between (1) the governing interests of spatially concentrated private capital and (2) the authority that governs public life within a territory.⁸⁴ Although the corporate enclaves of today are distinct from the historical company town in many respects, both share this democratic feature. We should understand both Starbase and Pullman as facilitating, through different legal mechanisms, a particular form of localized corporate domination within the industrialized American political economy.

Local Government Law and the Proliferation of the Corporate Enclave

Across our history, the relationship between corporate power and local democracy has been mediated by state laws governing local jurisdictional boundaries. History demonstrates that political decentralization creates opportunities for private domination by powerful firms not only indirectly (through the structural power of mobile capital and capture of public institutions) but also directly: through the facilitation of corporate fiefdoms. This section illustrates how the democratic valence of city power—and “local” control, more broadly—is contingent on the spatial concentration of taxable wealth and the institutional design of metropolitan governance.

Where political control can be maintained through strategic manipulation of the electorate, the municipal form allows firms to escape democratic obligations to the wider public. This institutional arrangement facilitates alternative mechanisms for pursuing the consistent goal of profit maximization—namely, by protecting corporate wealth from taxation, externalizing costs onto a broader public, and escaping local regulation. But rather than prevent this outcome, local government law has actively facilitated its proliferation over the past century.

For much of American history, private fiefdoms had been maintained through the playbook articulated by Pullman that required workers to reside within communities owned (and therefore controlled) by their employers. But as this strategy became more costly and less effective, parallel developments in local government law provided an attractive alternative. Rather than exercising domination over worker-residents through private law, employers have increasingly adopted the public form—and then schemed to exclude from the voting population any residents who might oppose their prerogatives. The result from decision-making processes would remain guaranteed—not by quashing dissenters’ voices, but by designing governing institutions that prevented their preferences from having legal consequence.

Contrast: Pullman and Vernon

When George Pullman built his model town in 1880, American cities like Chicago grew primarily through jurisdictional expansion.⁸⁵ Throughout the first decade of the model town’s brief existence, Chicago was trying to annex the suburban village of Hyde Park, which at the time included the town of Pullman and not much else. George Pullman, fearing increased taxes and the loss of governing authority, led the effort to block it: underwriting a massive political campaign and firing any workers who demonstrated support for annexation. Local political debate centered on these efforts. Decrying the democratic stakes of Pullman’s governing authority, for example, the Chicago Herald argued, “Annexation is the remedy.”⁸⁶

Ultimately, Pullman failed: Chicago annexed his town in 1889. But in the decades after this annexation fight, state laws governing municipal formation would provide town bosses with new options for maintaining political control that were not available to Pullman. By manipulating the jurisdictional boundaries of local governance to match the territorial contours of their influence—and then excluding any residents who might ever oppose their interests—companies could protect their proprietary domains from outside interference. They could not only guarantee the failure of any attempted annexations but also entirely escape local efforts to tax their wealth or regulate their affairs—and attain the formal powers of the state.

The consequences can be observed in the industrial city of Vernon, located right in the middle of Los Angeles. The urban historian Mike Davis described it as a “phantom city” that was created by “powerful industrial landowners . . . to exploit land-use control and hoard their lucrative tax base.”⁸⁷ In a 1991 profile, the Los Angeles Times described Vernon as the county’s “most industrial city with a daytime population of 51,000 and a nighttime population of 152”⁸⁸ —whose management and ownership constituted a majority of the city’s voting residents and so could set municipal policy without democratic constraint.

Vernon was founded in 1905 as an industrial town; its motto remains “Exclusively Industrial.”⁸⁹ When manufacturing declined, the jurisdiction sought to attract new industry through promises of regulatory forbearance and minimal taxation. Vernon provided so few actual municipal functions that revenues went almost entirely to compensation. Vernon is operated as an industrial city precisely through the same mechanism that guaranteed Disney’s control. Almost nobody lives in Vernon: Everybody who does live there, until recently, either worked for the city or was related to someone who did.⁹⁰

Here you can identify the influence of local government law: Los Angeles has repeatedly tried to annex Vernon, just like Chicago tried to annex Pullman. But due to changes in local government law—now deferring, nearly as a rule, to the preferences of the local community resisting the change—all those efforts have failed. This is a strikingly different outcome from the Pullman experience, back when local government law was more accommodating of annexations. As a result, Vernon remains independent. And neighboring communities—including those nearby residents affected by, for example, the notorious pollution that emits from these industrial cities and the nonresident workers affected by labor conditions within them—have no way of exercising democratic control over its industries through local-level political organizing.

Identifying Corporate Enclaves

How widespread is this phenomenon? It is difficult to study these trends at scale, but the available evidence suggests that dominant firms—wielding structural power over fragmented (and economically vulnerable) jurisdictions competing for mobile capital—will continue to innovate new forms of jurisdictional exit.

In a recent publication with coauthors Robert Manduca and Jacob Waggoner, I developed a methodology for measuring tax base fragmentation: the unequal distribution of taxable wealth across political jurisdictions within a defined geographic territory.⁹¹ We used a dataset of 138 million geocoded property tax records to directly quantify the extent to which the territorial boundaries of local governments overlap onto spatial patterns of economic segregation to create interjurisdictional fiscal inequalities within metropolitan areas. We used the results to make an interactive web map that allows anyone to look up how the property wealth in their own metro area (or any other) is fragmented across different local municipalities. This tool visualizes tax base fragmentation across the United States: www.taxbasefragmentation.net.

This analysis allowed us to identify a complete list of what we term “municipal tax havens”: independent general-purpose local governments whose boundaries contain enormous levels of property wealth relative to neighboring jurisdictions. (Specifically, our analysis identified jurisdictions whose taxable property wealth per capita is at least three times the average for their metro area.) Using this criterion, we identify 502 distinct municipal tax havens, with a total combined population of 1.8 million residents.

Zoom in on the map to view the locations of these tax havens in your state, county, or city.

Among these jurisdictions are some of the most famously wealthy municipalities in the country—Malibu, California; East Hampton, New York; Palm Beach, Florida—as well as locally prestigious suburbs of many major cities. But our dataset (from CoreLogic) included all reported categories of assessed property wealth—including corporate property. This allowed us to identify, empirically, the phenomenon of “corporate enclave” municipalities that is represented by Vernon: incorporated local jurisdictions that contain very few actual residents but tens of millions of dollars in corporate property wealth that is effectively shielded from taxation

Sorting by the “fiscal capacity ratio”—the per-capita property wealth of a jurisdiction compared to its surrounding metro area—we found that 5 of the top 10 “municipal tax havens” that we identified have a tax base that is more than 80 percent nonresidential properties. This includes Vernon as well as the Disney-owned “cities” of Bay Lake and Lake Buena Vista, Florida, located within the company’s Disney World resort. Other examples that we identified include: a city in Colorado (Black Hawk) that has nearly 20 large casinos and just 111 residents; a borough in New Jersey (Teterboro) whose boundaries include a private airport and corporate jet storage facility, but only 61 residents; and a municipality in St. Louis county (Champ) that is home to a $20 million, privately owned solid waste disposal facility, occupying the vast majority of the town’s 512 acres, but only 10 residents. These jurisdictions shield their contained corporate wealth from redistributive taxation (and associated services) within and across their respective metropolitan areas. Under our federalism design, the wealth of these jurisdictions is effectively shielded from the fiscal needs of poorer residents elsewhere in the metropolitan community.

In these corporate enclaves, tiny numbers of handpicked (and unrepresentative) residents are able to shield their immense resources from democratic control by the (more numerous and representative) residents just beyond their local borders. If residents of Los Angeles wish to regulate the toxic emissions of Vernon’s industrial pollution, or the residents of Orlando wish to fund their local schools by taxing the immense wealth of these Disney properties—there is no local election through which they can secure the desired policy change. It doesn’t matter how much they organize: Our institutional design—facilitated by these forms of jurisdictional fragmentation—effectively blocks these pathways to policy change. This dynamic is well understood in the context of residential segregation, but our extreme jurisdictional fragmentation has consequences also for corporate power.

Policy Response

These outcomes are, at least in significant part, a downstream consequence of the economic capture that results from the decision to block redistribution across geographic space. The United States’ comparatively anomalous fiscal federalism design—which requires local communities to generate large portions of their budgets from taxes on economic activity within their borders—puts local governments at a distinct disadvantage in negotiations with holders of mobile capital, including structurally dominant firms. Partly as a result of this design, corporations are able to leverage the economic vulnerability of local communities to override the public will and secure unpopular policy concessions—such that, even where formal structures for democracy exist, private actors exercise an effective veto over public policy.

Under the United States’ fiscal federalism design, local public goods are funded predominantly from wealth contained within constructed and contested municipal boundaries. This design is not inevitable: Other systems characterized by political and programmatic decentralization feature considerably higher interjurisdictional redistribution, typically through centralized fiscal transfers. One group of researchers recently noted “how distinctive the United States is in the way it combines institutional arrangements that facilitate metropolitan fragmentation (through jurisdictional proliferation) and those that reward such fragmentation (through opportunities for resource hoarding).”⁹²

But in the absence of redistributive transfers or other policies that could rebalance the relationship between dominant firms and vulnerable jurisdictions, local governments will find it difficult—or even impossible—to resolve these policy failures on their own. In particular, the fragmentation of metropolitan tax bases alters the dynamics of corporate relocations. Extensive research on the urban policy implications of the United States’ knowledge economy has emphasized “agglomeration effects” accruing to “superstar cities.”⁹³ But such spatial efficiencies are regional in nature, whereas local tax bases are jurisdictionally bounded. Even if network effects pull economic activity toward particular metropolitan areas, corporations retain their leverage over fiscally vulnerable municipalities within those metropolitan areas.

By contrast, the federal government—under a different system of fiscal federalism—could allow cities to make democratic investments with the immense wealth produced by the US economy, no matter how it is distributed over geographic space. The forms of interjurisdictional tax competition that have, now and across history, harmed antimonopoly efforts ultimately derive from the absence of centralized fiscal transfers of the sort that are provided in every other wealthy democracy. And during a moment when federal funding is being withdrawn from critical local funding priorities, states will need to increase their own efforts to reduce the disparities between the wealthy and poor jurisdictions within their (larger) boundaries. Instead of asking how to convince Amazon to locate in excluded communities, policymakers should focus on ensuring that all communities are able to access the determinants of economic opportunity—whether or not some dominant firm decides to locate within the same jurisdiction.

In addition to reducing the consequences of jurisdictional fragmentation, states also can directly reduce its existence through reforms to the laws that govern municipal incorporation. Current state law makes it possible for a concentration of corporate interests, without much difficulty, to access public powers by incorporating into the municipal form. Indeed, the past century of developments in local government law has facilitated the preservation of the historical corporate enclave across new forms. Localized corporate domination is contingent on our institutional design—and this history helps to reveal a path not taken. Rather than attending rarely, indirectly, and only belatedly to the downstream consequences of corporate fiefdoms, public law should directly prevent their emergence.

Conclusion

Starbase exemplifies the institutional form of the contemporary corporate enclave, as well as its consequences. Jurisdictional fragmentation, and deference to forms of private governance, has facilitated the continuation of the “company town” for gigantic corporate conglomerates through new institutional forms. There is little reason to believe that the outer limits of this strategy have yet been realized. Indeed, these various forms of effective secession will remain attractive to large firms as long as they remain effective vehicles for escaping democratic regulation, evading shared obligations like taxation, and accessing the powers of public authority.

Particularly in a historical moment when long-settled labor rights and other forms of corporate regulation are under resurgent legal and political threat at the federal level, we should expect to see increased efforts by dominant corporations to escape various forms of local regulation. Under these dynamics, the task of ensuring that workers and other vulnerable groups retain the ability to pursue basic protections—through policymaking processes in which their interests are represented—will require new efforts to prevent (and reverse) enclaving through the municipal form. Indeed, if not prevented by law, why would dominant corporations seeking competition advantage (or other benefits of sovereign power) not demand their own state? We may eventually come to see Reedy Creek and Starbase not as curious outliers but as early templates: the next iteration of an arms race that has already begun to transform the political economy of American local governance.