Applying the System Asset Pricing Model to Data Brokerage: Measuring the System Welfare Cost of Behavioral Surplus Extraction

Data brokerage generates 6.13 in welfare costs for every 1 of revenue

The System Asset Pricing Model (SAPM) calibrates a median system beta of 6.83 (90% CI: 4.7–8.0) across 100,000 Monte Carlo draws. No draw produces a beta below 1.0.

• Private payoff Π = 290B (narrow) to 667B (broad behavioral ad industry)
• Total welfare cost W = $2.0T across six channels
• Cooperative counterfactual Π_C = $89B preserves 27.6% of revenue while eliminating >95% of welfare costs

CAPM for investors becomes SAPM for society

The same formal grammar—beta, risk premium, Sharpe ratio—but the accounting unit shifts from individual risk to system welfare. Data brokerage lacks transactional legibility: 78% of Americans have little to no idea what companies do with their data.

• System beta β_W = W/Π = 6.83 (MC median)
• System welfare gap S_W = W − Π_C ≈ $1.89T
• Break-even institutional efficiency μ* = 0.088 — close 8.8% of the gap for Pareto improvement

Standard externality frameworks break in data brokerage

Pigouvian taxes fail because welfare costs are emergent from aggregation, not attributable per data point. Coasean bargaining fails because property rights are undefined, transaction costs are astronomical, and parties are unidentifiable.

• A single data point has negligible cost; 10,000 attributes across 2.5B profiles enable identity theft and voter suppression
• 4,000+ firms operate in a 'shadow economy' — production is systematically unobservable
• Solove (2013): reading all relevant privacy policies would require hundreds of hours annually

Welfare costs flow through six transmission channels

Each channel satisfies causal linkage to data brokerage, measurability via empirical studies, and materiality (>100B). The largest channel is autonomy degradation at 480B.

• C1 Security externalities: 186B — breach costs, identity theft (20.9B from four major broker breaches alone)
• C2 Autonomy & privacy degradation: $480B — chilling effects, self-censorship, cognitive depletion
• C3 Democratic manipulation: $320B — voter suppression, micro-targeted disinformation, trust erosion
• C4 Algorithmic discrimination: $260B — bias in housing, lending, employment, criminal justice
• C5 Consumer manipulation: $390B — dark patterns, addiction-by-design, price discrimination
• C6 Governance failure: $310B — regulatory capture, lobbying paralysis, consent theater

Centralized aggregation creates catastrophic single points of failure

The Equifax breach (147M records) cost 1.5B; T-Mobile (76.6M records) cost 500M. The global scam economy, reliant on breached broker data, stole an estimated $1 trillion in 2024 — exceeding the global drug trade.

• Average breach cost for AI-processed data: 5.8M (IBM), rising to 9.2M for regulated industries
• Enterprise defensive spending: ~$80–100B globally
• Identity theft losses from just four major broker breaches: $20.9B (JEC Minority Report 2026)

Privacy degradation costs $480B annually in cognitive and democratic harms

This channel prices intangible welfare costs that standard cost-benefit analyses exclude: autonomy loss, chilling effects, and trust erosion. It is the largest single channel.

• Acquisti et al. (2016): privacy's value includes 'economic dark matter' — autonomy, dignity, fairness
• Tim Wu's estimate of attentional externalities: $120B/year from cognitive depletion alone
• GDPR fines (€4.5B through 2024) are <1% of behavioral ad market growth ($440B)

Data brokerage enables voter suppression and democratic manipulation

Cambridge Analytica used 5,000 data points per voter across 220 million Americans. Haenschen and Hersh (2022) found 3.5 million Black Americans categorized under 'Deterrence' in the 2016 Trump campaign's Project Alamo database.

• Non-white users in battleground counties received suppression ads nearly 10× more frequently than white users in non-battleground states (PNAS)
• ODNI (2024): commercially available information poses counterintelligence threats to the US
• Total democratic manipulation cost: $320B/year

No opt-out strategy can reduce system beta below 1.0 under current architecture

Three axioms — Data Generation Inevitability, Re-identification Persistence, and Consent Theater — prove a formal impossibility. This is the first impossibility result in the SAPM domain catalog for an information-economy sector.

• Axiom 1: Humans cannot cease producing behavioral data (inevitability)
• Axiom 2: De-identification fails against modern linkage attacks (persistence)
• Axiom 3: Notice-and-choice cannot produce informed consent at scale (theater)
• Result: β_W > 1.0 for any individually rational strategy within current institutions

A privacy-respecting data economy retains $89B while cutting welfare costs by 95%

Π_C = $89B is grounded in observable market data: contextual advertising CPM differentials, privacy-premium WTP estimates, and data cooperative revenue projections. The cooperative alternative is feasible, not only desirable.

• Contextual advertising: ~$33B
• Privacy-preserving DaaS: ~$19B
• Data cooperative / privacy-premium revenues: ~$37B
• μ* = 0.088 means closing <9% of the welfare gap suffices for Pareto improvement

The FTC's 430M budget faces a 667B industry — a 1,500:1 asymmetry

When the EU imposed GDPR, fines represented less than 1% of market growth. The regulatory capture ratio of 1,500:1 calls for institutional rather than marginal budgetary remedies.

• FTC budget: 430M/year; behavioral ad industry: 667B/year
• GDPR fines (€4.5B total through 2024) vs. market growth ($440B)
• Data brokers use dark patterns and no-index tags to hide opt-out pages from search engines

Data governance must shift from consumer protection to structural externality

The impossibility theorem proves marginal reform cannot work. The break-even threshold μ* = 0.088 shows that perfect regulation is unnecessary — but continued inaction is welfare-indefensible.

• Policy question shifts from 'Is reform affordable?' to 'Is inaction defensible?'
• Architectural redesign required: data minimization mandates, fiduciary duties, cooperative governance
• SAPM positions data brokerage at rank 43 of 71 in welfare intensity, but absolute cost ($2.0T) is among the largest due to universal exposure