The Institutional Demand Lock-In Theorem: Why Captive Markets Resist Reform

Reform fails when the decisive coalition earns income from the harmful activity

The paper defines institutional demand lock-in as a structural dependency: employment, revenue, or institutional survival depends on continuation of welfare-destroying activity X while system welfare C is outside the payoff space.

• The flawed game G is bilateral: A and B can win while C loses
• The equilibrium is Hollow Win (0,1,1): private actors and incumbent institutions gain while the system degrades
• The dependency ratio δ measures the share of the decisive political coalition whose income depends on X
• When δ exceeds δ* ∈ [0.15, 0.35], no democratic reform coalition can form regardless of the welfare gain

The threshold depends on mobilization and electoral bias, not on welfare losses

The theorem states δ* = 1 - b/(2p), where p is the mobilization rate of non-dependent reform supporters and b is electoral bias toward concentrated interests.

• The welfare term β_W does not enter the coalition condition
• Even infinite welfare gains do not form a coalition once δ ≥ δ*
• Electoral bias b raises the weight of concentrated dependent groups
• Low mobilization p among non-dependent voters lowers the reform threshold

Lock-in is not the same as regulatory capture or fiscal capture

The paper separates three mechanisms. Regulatory capture is industry influence over regulators. Fiscal capture is government reliance on harmful revenue. Institutional demand lock-in is broader: workers, towns, agencies, creditors, and export systems can all become dependent on X.

• Regulatory capture requires industry action; demand lock-in requires measurable survival dependency
• Fiscal capture explains budget dependence; demand lock-in also explains employment and community dependence
• The locked-in constituency resists reform because reform removes the institution around which income, location, and community are organized

Arizona’s Red Rock contract made empty beds a budget liability

In 2010, Arizona guaranteed 100% occupancy at CoreCivic’s Red Rock Correctional Center in Eloy. The state owed 45,000 per empty bed per year, so a 10% vacancy across 2,000 beds created 9 million in annual penalty exposure.

• Red Rock generated $45 million in annual CoreCivic revenue
• CoreCivic’s 2014 total revenue was $1.7 billion
• The private-prison welfare calibration is β_W = 12.08
• Eloy had 6,200 workers and 2,500 prison-related jobs, giving δ = 0.40

Sentencing reform died because decarceration threatened Eloy’s jobs and revenue

The paper’s Arizona case ties the contract to a local veto constituency. When Arizona considered a 2013 sentencing reform that would cut prison population by 15%, Eloy’s mayor testified against it and the bill died in committee.

• A 15% population reduction would trigger $13.5 million in penalty payments
• The same reduction would eliminate 300 jobs in the paper’s account
• Pinal County prosecutors charged more aggressively, judges sentenced longer, and parole boards denied release
• The mechanism is demand for filled beds, not only lobbying by prison firms

Pará’s clearing economy ties municipal survival to land conversion

In Pará, agriculture and extractive industries account for 42% of GDP and 55% of formal employment, giving δ = 0.42. The paper estimates $20.5 billion in annual soy, cattle, and timber revenue against β_W = 7.21 for commodity-driven deforestation.

• Agricultural GDP in Pará is $42 billion annually
• Soy exports were 8.5 billion in 2022 and cattle ranching revenue was 12 billion
• Pará’s estimated share of Amazon deforestation welfare cost is $260 billion
• In Novo Progresso, 62% of the workforce depends directly on land-clearing for cattle ranching or soy farming

The soy-cattle-timber cascade collateralizes the town against future clearing

The paper describes a staged dependency: clearing creates de facto property claims, cattle ranching adds jobs and taxes, then soy intensification brings processing plants, credit, and equipment dealers.

• IBAMA’s 2021 enforcement in Novo Progresso seized 15,000 cattle and fined 200 producers
• The municipality declared an economic emergency after enforcement
• Bank loan portfolios, municipal budgets, and equipment inventories become tied to future expansion
• EU deforestation-free requirements reduce clearing incentives but do not create substitute employment

Nevada combines job dependency with fiscal dependency

Nevada’s gaming sector creates both employment lock-in and budget lock-in. The paper reports 15.5 billion in 2023 gaming revenue, 1.5 billion in gaming taxes, and 452,000 gaming-related jobs.

• Gaming taxes provide 25% of Nevada’s general fund
• Gaming jobs represent 29% of the state workforce, so δ = 0.29
• Clark County has 73% of the population and generates 89% of gaming tax revenue
• Commercial gambling has β_W = 7.30 through bankruptcy, family dissolution, suicide, crime, and productivity loss

Sports betting becomes hard to unwind once states build budgets around it

After Murphy v. NCAA in 2018, state sports betting revenue started modestly and then grew into a budget line. The paper argues that staff, bonds, education transfers, pension funds, and infrastructure projects become dependent within 3 to 5 years of legalization.

• The American Gaming Association spent $14 million on lobbying in 2023
• A Nevada 2023 bill limiting casino advertising during problem gambling awareness month died without a floor vote
• The Nevada Resort Association spent $2.3 million lobbying against that bill
• Estimated electoral bias b = 1.4-1.6 lowers the effective δ* to approximately 0.18

Bangladesh’s garment export model locks workers, factories, and the state into throughput

Bangladesh garment exports were 42 billion in 2023, representing 84% of exports and 13% of GDP. The sector employs 4.2 million workers, 75% women, at an average monthly wage of 95.

• The paper uses δ = 0.84 based on export dependency
• EU textile imports from Bangladesh were $18 billion in 2023
• Fast fashion has β_W = 7.01 in the paper’s calibration
• Bangladesh’s estimated share of global fast-fashion welfare cost is $35 billion

Rana Plaza changed factory safety rules but not the throughput model

The Rana Plaza collapse killed 1,134 workers and injured 2,500. The Bangladesh Accord improved inspections, fire protection, and exits, but the paper argues it did not replace income tied to high volume, low durability, and rapid turnover.

• Factory owners absorbed safety costs by reducing wages, extending hours, or subcontracting to non-compliant facilities
• Capital channel: loans assume throughput volume, so durability mandates can trigger default
• Labor channel: piece rates make reduced volume a direct income cut
• Trade channel: higher costs push buyers toward lower-cost competitors

Substitute income before restricting the harmful activity

The paper’s rule change R converts G into G1 only when R1 is legally, fiscally, and operationally guaranteed before R2. R1 supplies substitute income; R2 restricts the welfare-destroying activity.

• Private prisons: rural employment, municipal revenue replacement, retraining, economic development grants, and fiscal equalization transfers
• Deforestation: standing-forest income, carbon-credit forestry, agricultural intensification without clearing, and municipal fiscal equalization
• Gambling: replacement revenue for schools and workers before advertising and product restrictions
• Fast fashion: an industrial upgrading compact that pays for durability-oriented production without collapsing worker income