The
Problem.

One Transaction, Fully Traced

Where Every Dollar Goes When You Buy Something

The most effective way to understand the extraction economy is to follow a single dollar through a single transaction and watch what happens to it. Not in the abstract. Not as a percentage. Dollar by dollar, fee by fee, hand by hand.

A product costs $3.00 to manufacture. It retails for $10.00. Here is every person and institution that touches the difference, what they take, and what they provide in return.

The Manufacturer

Raw materials cost $1.50. Labor cost $0.80. Overhead $0.70. Total manufacturing cost: $3.00. The manufacturer sells to a distributor at a wholesale price of $5.00. Gross manufacturing margin: $2.00.

From that $2.00 the manufacturer must: service debt taken on to build the factory, pay corporate taxes, cover insurance, and fund compliance costs. Average net manufacturing margin in consumer packaged goods: 8-15%. On this transaction: $0.24-$0.75 to the manufacturer.

The Distributor

The distributor buys at $5.00 and sells to the retailer at $6.25 — a 25% markup. But the stated markup is not the real cost to the brand. Research

by SupplyPike and industry analyses of UNFI and KeHE financial disclosures document the full fee stack:

— Stated distributor markup (25%): $1.25

— Spoils and shrink fee (4% of invoice, whether or not actual spoilage occurs): $0.20

— Freight padding (15% above actual freight cost): $0.08

— Data access fee (KeHE Connect 1.5% ongoing allowance): $0.08

— Compliance fine (averaged across all orders): $0.05

— Mandatory promotional buy-in: $0.10

— Real distributor take: $1.76

The distributor stores the product, moves it to a truck, and delivers it to the retailer. That is the service. $1.76 for warehousing and last-mile delivery on a $5.00 wholesale transaction.

The Retailer

The retailer buys at $6.25 and sells at $10.00. But the stated retail margin is also not the real cost to the brand. The fee stack continues:

— Slotting fee (amortized per unit, to earn shelf placement in the first place): $0.15

— Chargeback deductions (3% of invoice; 75% of chargebacks contain errors the brand pays rather than dispute): $0.19

— Mandatory promotional scan allowance: $0.20

— Payment float (brand finances retailer working capital for 72 days at credit card interest): $0.06

— Retailer margin plus brand-funded costs: $3.35 + $0.60 in deductions = $3.95 total retailer extraction

The Technology Infrastructure

To process this transaction, the brand needed: EDI compliance software to communicate with the retailer ($0.08 per unit amortized), trade show appearance to get discovered ($0.06 per unit), ERP system to manage orders ($0.05 per unit), and sales rep commission at 10% of wholesale ($0.50).

— Technology and sales infrastructure: $0.69

The Final Distribution

— Manufacturer: $0.24-$0.75 (8-15% net margin)

— Distributor (stated + hidden fees): $1.76

— Retailer (stated + deductions): $3.95

— Technology infrastructure: $0.69

— Brand keeps: $0.16 on a $10 product they manufactured for $3.00

The brand that conceived, manufactured, marketed, and took the financial risk keeps $0.16. The intermediaries who moved and shelved the product keep $6.40. The rest covers infrastructure and taxes.

This is not a worst-case scenario. This is the median. This is what Book 1 of this series documented in forensic detail for the wholesale industry. Every chapter in Part Two of this book documents the identical architecture in eleven other industries.

The mechanism is always the same. The industry language changes. The outcome does not.

The Seven Faces of Extraction

What the Research Shows, Made Human

Labor economics research, platform earnings studies, marketplace seller analyses, restaurant industry surveys, short-term rental host data, warehouse worker testimonies, and artisan seller research across six industries and seven regions document a pattern so consistent it functions as a law: people who build their livelihoods on infrastructure they do not own follow a predictable trajectory. Here is that trajectory given seven human faces, each drawn from documented research rather than invention.

The Platform Driver (North America)

Rideshare driver earnings research (Economic Policy Institute, UC Berkeley Labor Center, Ridester Annual Driver Survey, MIT Center for Energy and Environmental Policy Research) documents a consistent arc: net hourly earnings after platform fees, fuel, maintenance, insurance, and self- employment taxes decline over time as platforms raise take rates. Year-one average: approximately $17-22/hour net. Year-six average: approximately $9-14/hour net in most markets.

At the bottom of that range, a driver working 60 hours per week earns less after expenses than a full-time minimum wage employee with employer- paid payroll taxes and potential benefits. The difference: the minimum wage

employee has legal protections, workers' compensation eligibility, and unemployment insurance. The platform driver has none of these, bears full vehicle depreciation, and carries no path to equity or advancement.

The platform's market capitalization during the same period the driver's earnings declined: up tens of billions. The platform's argument that its drivers are "independent entrepreneurs" is the legal architecture that makes this math possible. Uber and Lyft spent $200 million in California to defend that architecture when legislators tried to reclassify drivers as employees (Proposition 22, 2020). They won. The architecture holds.

The Marketplace Seller (North America)

Marketplace seller research (FTC marketplace investigations 2020-2023, academic analyses of Amazon seller economics, EU antitrust proceedings) documents a consistent pattern: successful sellers on major e-commerce platforms face algorithmic demotion when the platform identifies their product category as commercially attractive for private label.

Amazon Basics was launched after Amazon identified high-margin, high- volume categories from aggregated seller data (EU antitrust charge, 2020; Amazon settlement, 2022). The EU found Amazon employees accessed seller data for specific product categories — car trunk organizers, seat covers, phone holders — before launching Amazon Basics competitors in those exact categories. Amazon settled by committing not to use non-public seller data for private label decisions. The products remain on shelves. The practice continues under modified structures.

Account suspension without explanation is documented across thousands of seller testimonies. The FTC's 2023 complaint against Amazon includes documentation of suspension practices. No seller has a right to appeal to a human reviewer within a defined timeframe. The terms of service — which

sellers must accept to participate — prohibit discussion of account suspension publicly. Legal recovery costs ($10,000-50,000 in attorney fees for account reinstatement) exceed the financial benefit of fighting for most sellers with affected accounts under $500,000.

The Restaurant Owner (West Africa)

Delivery platform analyses across West African cities (World Bank Digital Economy Reports, Statista delivery platform market analyses, GSMA Mobile Economy reports for Sub-Saharan Africa) document a specific dynamic in markets with thin margins: restaurant commission rates of 25-30% on orders where restaurant net margins run 8-12% produce negative operating margins on delivery-channel sales.

The platform's competitive response to this math is documented: cloud kitchen operations using aggregated order data to identify high-demand menu items, then competing directly in delivery-only formats. The restaurant owner who trained the customer loyalty, developed the recipes, and absorbed the commission cost provides the market research the platform uses to build its competing product. This dynamic is documented not only in West Africa but across delivery markets in South and Southeast Asia and Latin America.

The Delivery Worker (South Asia)

Quick-commerce platform worker research (ILO Gig Economy Reports, International Labour Organization South Asia studies, Worker Rights Consortium field research) documents per-delivery earnings of 20-35 rupees ($0.25-$0.42 USD) while platforms charge 49+ rupees per delivery in service fees. Workers bear vehicle costs, fuel, insurance, and injury risk while classified as "delivery partners" rather than employees.

The algorithm's treatment of work-related injury is documented: missed days due to accident reduce reliability scores, which reduce order allocation, which reduce earnings, which extend financial recovery. The platform does not distinguish voluntary absence from injury. Workers injured on the job face both physical recovery and algorithmic penalty simultaneously, with no compensation mechanism and no appeal process.

The Short-Term Rental Host (Europe)

Short-term rental platform research (host survey data across European cities, platform analytics studies, academic housing economics research in Barcelona, Amsterdam, and Lisbon) documents a pattern of algorithmic demotion when platforms shift preference to professional property management companies with multiple listings.

Individual hosts who invested in property improvements to achieve high ratings face rating collapse from single negative reviews the platform will not remove, algorithmic burial below commercial operators, and no path to recover accumulated standing. The 30,000 euros invested in renovations to meet the platform's quality standards remains locked in the property while the platform's ranking system routes buyers to commercial operators instead. The investment was made for the platform's benefit and is not recoverable.

The Warehouse Worker (South America)

Distribution center worker research across Latin American logistics facilities (labor ministry data, ILO working conditions studies, worker testimony documentation from Brazil, Mexico, and Colombia) documents consistent conditions: 10-hour shifts, productivity surveillance through algorithmic management systems identical to systems deployed from New

Jersey to Nairobi, timed bathroom access, and wage growth below inflation producing effective pay declines.

The promotion mechanism is documented: warehouse floor expertise does not qualify for supervisory roles that require formal credentials. Workers who train colleagues, who know every process, who demonstrate operational leadership are passed over for people with university degrees and no floor experience. The algorithm that surveils productivity produces data that goes upward to shareholders. The worker who generates that productivity data has no access to it and no claim on the value it represents.

The Artisan Seller (West Africa)

African e-commerce seller research (World Bank Digital Economy Reports, cross-border digital trade analyses, GSMA Mobile Economy Africa) documents specific friction that domestic sellers in wealthy countries do not face: payment processor holds of weeks under "fraud prevention for high-risk regions" (a geographic classification, not a fraud history), currency conversion fees stacked on transaction fees, and search ranking disadvantages against sellers with advertising budgets proportionally larger than the artisan's annual income.

International patent filing costs ($5,000-15,000 minimum) exceed annual income for most artisan sellers in developing markets, making intellectual property protection effectively unavailable. Fast-fashion manufacturers copy designs documented on the platform and list machine-made versions above the original in search results funded by advertising at scale. The platform's terms of service provide an IP complaint mechanism. Resolution timelines run months. By resolution, the copied product has captured the market.

The Common Architecture

Seven people. Seven industries. Seven continents. The platform lifecycle is identical in every case: hope, hustle, hook, harvest, discard, repeat. The specific fee numbers differ. The legal structures differ. The currency differs. The outcome does not.

This is not a collection of individual bad luck stories. This is documented research across millions of participants showing that the architecture produces this outcome consistently. The extraction is the feature, not the bug.

The Loneliness Extraction

What Commerce Without Community Costs

In 2023, the United States Surgeon General declared loneliness a public health epidemic, issuing an advisory comparable in urgency to previous advisories on smoking and opioids. The research behind that declaration: loneliness has the same measurable health impact as smoking 15 cigarettes per day (Holt-Lunstad et al., Perspectives on Psychological Science). Sixty percent of American adults report feeling lonely (Cigna Loneliness Index, 2023). Gig workers report depression at twice the rate of traditionally employed workers (American Journal of Public Health).

This is not incidental to the extraction economy. It is produced by it.

Commerce used to mean community. The marketplace was where people gathered. The shop owner knew your name, your family, your preferences. The sales representative was a trusted advisor who visited regularly, remembered your kids' birthdays, and told you honestly when a product was

not right for you. These relationships were not sentimental extras. They were the infrastructure of social life for hundreds of millions of people.

Platform commerce dismantled this infrastructure systematically and deliberately. Not as a side effect. As a design choice. Every element of platform architecture that removes human interaction is an efficiency gain that gets celebrated in quarterly earnings calls. Automated customer service. Algorithmic matching. Contactless delivery. No-touch checkout. Each one eliminates a moment of human contact that used to exist.

The local bookstore that closed because Amazon undercut it on every title was not just a retail casualty. It was the place where a particular kind of community event happened: the author reading, the children's story hour, the book club, the place where a person who loved mysteries found other people who loved mysteries. Amazon does not host these events. Amazon does not know your neighborhood exists.

When you spend $100 at a local business, research consistently shows that approximately $68 stays in the community: the owner pays local employees, who eat at local restaurants, which buy from local suppliers. When you spend $100 on Amazon, $0 stays in your community. The full $100 flows to Seattle, to shareholders, to a system that does not know or care whether your town's main street is alive or dead.

The economic loss is documented. The social loss is harder to quantify but equally real. GRAJ's seven physical infrastructure pillars — garages, markets, studios, parks, fests, gear, games — are not secondary features of a digital platform. They are the deliberate restoration of commerce-as- community. The weekly market is not just a distribution channel. It is the place where the network becomes real.

Commerce

The Foundation That Runs Everything Else

Book 1 of this series examined the wholesale commerce industry in exhaustive forensic detail: 24 chapters, named companies, documented fees, verified mechanisms. It opened with the real P&L of a real brand — DREAMS AREN'T THIS GOOD, built in Brooklyn — that put $778,800 in product on retail shelves every year and lost $49,000 to $60,000 doing it. Not because the product was bad. Because the system was designed to extract everything before the brand could profit.

That P&L documents the $4.8-$6.1 trillion in documented annual extraction specific to the wholesale commerce industry — the identifiable gatekeepers, fee stacks, and compliance systems whose extraction is quantified with full methodology in Book 1's Appendix B. The $25+ trillion figure used throughout this book is the broader global extraction economy across all twelve industries. Both figures are real. They measure different scopes. This chapter is where they connect: wholesale is the infrastructure that every other industry's supply chain runs on.

Global wholesale commerce: $62 trillion per year. More than retail. More than e-commerce. More than both combined. It is the largest single sector of global economic activity, and it runs on technology from 1997. While direct-to-consumer commerce was transformed by Amazon, Shopify, and Stripe into a one-click, real-time, transparent system, the backbone of

global commerce is still processing orders by fax, tracking inventory in Excel, and manually re-entering data into eight disconnected systems that do not talk to each other.

The Documented Extraction in Wholesale Commerce

— Distributor markups plus hidden fees (spoils, freight padding, data access, compliance fines, promotional buy-ins): 25-45% of invoice value. A brand invoiced $47,000 to UNFI and received $31,200. A brand invoiced $32,000 to KeHE and received $21,000. Both documented.

— Slotting fees: $9+ billion extracted from brands annually before a single unit sells. $1.5-$2 million per SKU for a national retail launch. Some retailers earn more from slotting fees than from product sales.

— Retailer chargebacks: $18-20 billion annually in the U.S. alone. 75% of chargebacks contain errors. Brands pay them because disputing costs more.

— Payment terms: 55-56% of all B2B invoices overdue at any given time. The top 10 U.S. retailers hold $80-$100 billion in money owed to suppliers, earning float income while brands drain personal savings.

— EDI compliance: $50,000-$200,000 per retailer to implement a protocol invented in the 1970s. Mandatory for business with every major retailer in 2026.

— ERP systems: 50-75% implementation failure rate. Average cost overrun: 189%. A $500,000 to $2 million investment with a coin-flip chance of success.

— Data: Circana (formerly Nielsen/IRI) charges $50,000-$500,000 per year for sell-through data that brands generated with their own products. Walmart charges suppliers $10,000-$100,000 per year to see their own sales data in Walmart's stores.

Why commerce is chapter 4 rather than chapter 16: the wholesale supply chain is the infrastructure through which every other industry's products move. The extraction mechanisms documented in commerce — middleman markup stacks, opaque pricing, payment delays, data hoarding, compliance costs — are not unique to food and beverage. They are the operating system of every supply chain. The chapters that follow prove it, industry by industry.

Healthcare

How a $3 Bandaid Becomes $10, and Who Takes the $7

The United States spends $4.5 trillion per year on healthcare — more per capita than any other developed nation. It ranks 46th globally in life expectancy. The gap between what Americans spend and what Americans receive is not a quality problem. It is a commerce problem. The supply chain between the person who manufactures medical goods and the patient who needs them has been built to extract, not to deliver.

The Bandaid Supply Chain

A standard adhesive bandage costs approximately $0.03 to manufacture at scale (medical supply industry cost data; U.S. Government Accountability Office pharmaceutical pricing reports). In an American hospital, it bills at $10-15. Between the manufacturer and the patient:

— Medical device manufacturer sells to a Group Purchasing Organization (GPO). GPOs negotiate contracts on behalf of hospital systems in exchange for administrative fees from the manufacturer — a fee structure that creates incentives to favor higher-priced products.

— GPO sells to a hospital distributor (the dominant players: McKesson Medical-Surgical, Cardinal Health, Medline Industries). Distributor markup: 15-25%.

— Hospital distributor delivers to hospital system. Hospital system marks up for overhead, compliance, insurance, and "chargemaster" pricing — the list price from which negotiated discounts are calculated.

— Insurance company negotiates a "contracted rate" with the hospital. The contracted rate is a fraction of the chargemaster price but still multiple times the actual cost.

— The patient pays their co-pay or deductible portion of the contracted rate.

The same bandaid costs approximately $3 in Europe. Same manufacturer. Same supply chain at the manufacturing level. Different commerce infrastructure between manufacturer and patient.

The Pharmacy Benefit Manager Extraction

Three companies control approximately 80% of U.S. pharmacy benefit management: CVS Caremark, Express Scripts (Cigna), and OptumRx (UnitedHealth Group). PBMs negotiate rebates from drug manufacturers, then keep a portion of those rebates rather than passing them to patients or payers. The U.S. Government Accountability Office has published multiple reports stating that it cannot trace pharmaceutical pricing through the PBM supply chain due to opacity in rebate structures.

The documented mechanism: a drug manufacturer sets a list price of $10. The PBM negotiates a 40% rebate ($4 returned to the PBM from the manufacturer). The PBM keeps approximately $2 of the $4 rebate. The patient's co-pay is calculated based on the $10 list price, not the $6 net price. The insurer's premium reflects the $10 cost basis, not the $6 net. The

PBM earns $2 per prescription for processing a transaction between a drug company and a pharmacy. At scale, across hundreds of millions of prescriptions, PBMs extract $100+ billion annually (Drug Channels Institute estimates, corroborated by congressional testimony).

The Three Distributors Who Shipped 500,000 Americans to Their Deaths

McKesson ($276 billion annual revenue), Cencora ($238 billion), and Cardinal Health ($205 billion) distribute 90%+ of all pharmaceuticals in the United States. These three companies — which manufacture nothing, prescribe nothing, and provide no clinical care — earn $719 billion in combined annual revenue for moving boxes from manufacturers to pharmacies.

All three settled with the U.S. Department of Justice for a combined $21 billion for their documented role in the opioid crisis. McKesson shipped 5 million doses of hydrocodone to a single pharmacy in Kermit, West Virginia — population 400. Cardinal Health shipped 780 million oxycodone and hydrocodone pills to West Virginia between 2007 and 2012 — a state with 1.8 million people. Cencora identified suspicious orders and shipped them anyway. All three knew. All three shipped. 500,000+ Americans died.

The $21 billion in settlements represents approximately 3% of one year's combined revenue. No executive was imprisoned. The companies' revenues increased after the settlements. The commerce infrastructure that enabled the opioid crisis — the three-company near-monopoly on pharmaceutical distribution, the opacity that prevented accountability, the disconnect between volume shipped and clinical need — remains intact.

This is the healthcare commerce problem in its most extreme expression. A supply chain built without transparency, without accountability, and without

any mechanism connecting distribution volume to human need produced a preventable catastrophe at industrial scale.

Food

The 14-Cent Farmer and the $3.49 Markup

The United States Department of Agriculture publishes an annual "Food Dollar" series tracking how every dollar spent on food in America is distributed across the supply chain. The 2023 data shows that farmers receive approximately 14 cents of every food dollar. The other 86 cents go to processors, distributors, retailers, platforms, and various intermediaries.

In 2024, 673 million people globally faced hunger (Food and Agriculture Organization). In the same year, 1.05 billion tonnes of food were wasted globally (United Nations Environment Programme). These two facts are not unrelated. Both are commerce problems.

The Coffee Supply Chain: $0.04 to the Farmer, $5.00 at the Cafe

Coffee is the most traded agricultural commodity after oil. Its supply chain is one of the most extensively documented examples of extraction in global agriculture.

— Ethiopian or Colombian coffee farmer receives: $0.04-$0.08 per cup equivalent at commodity prices. This represents 1-2% of the $5.00 the consumer pays at a specialty cafe.

— Local collector or buying agent (called "pisteurs" in West Africa): adds the first markup layer, typically 10-15% above farm gate price.

— Commodity trader (Neumann Kaffee, Volcafe, ED&F Man collectively handle significant portions of global green coffee trade): 10-15% of value.

— Export processor and shipping: 5-10%.

— Importer/green coffee merchant: 5-10%.

— Roaster: 20-30% of retail value.

— Retailer/cafe: 40-60% of retail value.

"Fair trade" certification adds $0.20 per pound premium to the farm gate price. The consumer pays $2-3 more per cup for fair trade coffee. The difference between the $0.20 that reaches the farmer and the $2-3 that the consumer pays: certification overhead, auditing, and margin expansion by every intermediary above the farmer.

The ABCD Cartel: Four Companies, 70-90% of Global Grain Trade

ADM (Archer-Daniels-Midland), Bunge, Cargill, and Louis Dreyfus — collectively known as the "ABCD" firms — control an estimated 70-90% of global grain trade. Their combined annual revenues exceed $350 billion. They do not grow grain. They trade it. They control the logistics, storage, financing, and distribution infrastructure through which the world's food supply passes.

Cargill, the largest, reports $177 billion in annual revenue. It is the largest private company in the United States. The Cargill family — 14 billionaires with a combined net worth exceeding $50 billion — has operated this private company for over 155 years with minimal public disclosure requirements. ADM paid $100 million in fines for price-fixing lysine and citric acid in the 1990s; three executives were imprisoned. The company continues to operate as one of the four controllers of global grain.

500 million smallholder farms worldwide produce 80% of the food consumed in developing countries (Food and Agriculture Organization). These farmers receive 5-10% of the final retail price of the food they grow. The ABCD firms and their trading networks capture the difference.

Post-Harvest Food Loss: The Commerce Failure That Causes Hunger

Post-harvest food losses in developing countries run 30-40% of total agricultural production (Food and Agriculture Organization, USDA Economic Research Service). This is not a farming problem. Farmers grow the food. It is a commerce and distribution problem: the infrastructure to move food from where it is grown to where it is needed does not exist, or exists only in the hands of intermediaries who extract too much for smallholders to afford.

In sub-Saharan Africa, an estimated 30-50% of fruits and vegetables spoil before reaching markets due to lack of cold chain infrastructure, inadequate transport, and fragmented distribution networks. People go hungry 50 miles from full warehouses. That distance is not a geography problem. It is a commerce infrastructure problem.

Education

The Textbook That Costs $20 to Print and $300 to Buy

U.S. student debt totals $1.75 trillion across 45 million borrowers (Federal Reserve Bank of New York, 2024). The average monthly payment: $503 for 20 years — $120,720 in total payments for principal and interest. Average student loan balance at graduation: $37,000.

In 1970, a minimum wage worker could pay for a year of public university with 306 hours of work. In 2024, that same year of public university requires 1,568 hours of minimum wage work — five times as many hours for the same credential. This is not because universities became more expensive to operate. It is because the commerce of education was redesigned to extract.

The Textbook Supply Chain

A standard 400-page university textbook costs approximately $18-25 to print and bind. It sells for $150-$350. The author typically receives 10-15% royalty on net sales — not the cover price, the net price after returns and distributor discounts. On a $250 textbook, the author may receive $15-25.

The publisher — the dominant players are Pearson, McGraw-Hill, Cengage, and Elsevier, all controlled by private equity or institutional investors — captures the remainder. Their business model depends on frequent new editions: a new edition renders used copies from the previous year unsellable, eliminates the secondary market, and resets the price at full retail for every student in every course. The educational value of the 12th edition versus the 11th edition is, in most cases, marginal. The commercial value of eliminating the used book market is enormous.

The Online Course Platform Extraction

Udemy, Coursera, and Skillshare charge instructors 37-63% of revenue on sales generated through platform marketing, and 3-97% on sales generated by the instructor's own promotion (the instructor brings the student, the platform charges a 3% processing fee that grows to 50% when the platform "assists" in conversion). The person who created the knowledge, built the curriculum, recorded the content, and established the reputation

keeps a fraction. The platform that hosts the video and processes the payment keeps the rest.

The 2 million+ instructors on Udemy collectively teach over 900 million students. The platform's take rate is not disclosed publicly in aggregate, but independent instructor analyses consistently report effective rates of 25-63% depending on sale channel. Course builders report that spending years creating curriculum earns them less than a customer service employee at the company that hosts their course.

Housing

How a $400,000 Home Becomes an $800,000 Cost

Median U.S. home price in 1970: 2.5 times median annual income. In 2024: 7.5 times. The house did not become three times harder to build. The commerce of housing — the system of intermediaries, financing, and transaction costs between builder and buyer — was redesigned to extract more at every step.

The Transaction Fee Stack on One Home Purchase

On a $400,000 home purchase financed with a 7% 30-year mortgage:

— Real estate agent commission (seller's agent, typically 2.5-3%): $10,000-$12,000

— Buyer's agent commission (typically 2.5-3%): $10,000-$12,000

— Title insurance (lender's + owner's): $2,000-$4,000

— Loan origination fee (1-2 points): $4,000-$8,000

— Appraisal, inspection, survey: $1,500-$3,000

— Transfer taxes and recording fees (varies by state): $1,000-$8,000

— 30-year total interest at 7%: approximately $480,000

— Total cost of the $400,000 home: approximately $908,000-$927,000

The home cost $400,000. Buying and financing it cost an additional $508,000-$527,000 — more than the home itself. The vast majority of that additional cost flows to lenders, agents, title companies, and various intermediaries who did not build the home.

The Institutional Investor Capture

Institutional investors — private equity firms, REITs, and investment funds including Blackrock, Invitation Homes, and American Homes 4 Rent — owned approximately 2% of single-family rentals in the United States in 2010. They own approximately 25% in 2024 (ATTOM Data Solutions; National Association of Realtors research). The mechanism: leverage low- cost institutional debt to purchase single-family homes at scale, convert to rentals, raise rents to market, and capture the appreciation.

The homes being purchased and converted to rentals are the starter homes that were historically the entry point into homeownership for first-time buyers. Their removal from the for-sale market reduces supply, raises prices, and forces the first-time buyers they displaced into the rental market — often renting from the same institutional investors who outbid them. The wealth that would have accumulated in those households instead accumulates in institutional investor portfolios.

Energy

The $0.03 Kilowatt and the $0.16 Bill

The cost of generating solar electricity has fallen approximately 90% since 2010 (International Renewable Energy Agency, Lazard Levelized Cost of Energy analysis). Utility-scale solar now generates electricity for approximately $0.02-$0.05 per kilowatt-hour. The average American household pays $0.12-$0.23 per kilowatt-hour (U.S. Energy Information Administration, 2024).

The gap between generation cost and consumer cost is not transmission infrastructure. Transmission and distribution add approximately $0.03- $0.05 per kilowatt-hour in most markets. The remainder is utility profit margin, regulatory compliance overhead, billing and administrative systems, and the carrying costs of legacy infrastructure that utilities are not incentivized to replace.

The Utility Monopoly Architecture

In most U.S. states, electricity distribution is a regulated monopoly. One company controls the wires that deliver electricity to homes and businesses. Competitors cannot build competing wire networks. The regulator sets the rates the monopoly can charge, typically guaranteeing a 9-12% return on invested capital.

This structure creates a perverse incentive: the utility earns more money by investing more capital in infrastructure, because its profits are calculated as a percentage of its asset base. A utility that replaces an aging coal plant with a cheaper renewable source reduces its capital base and therefore its guaranteed profit. A utility that builds expensive new natural gas plants

increases its capital base and its guaranteed profit. The rate of return regulation was designed to ensure utilities had incentive to invest. It also ensures utilities have incentive to invest in expensive infrastructure rather than efficient infrastructure.

The Named Extractors in Energy

The five largest U.S. investor-owned utility holding companies — NextEra Energy ($24 billion revenue), Duke Energy ($28 billion), Southern Company ($24 billion), Dominion Energy ($15 billion), and American Electric Power ($19 billion) — collectively control electricity distribution across large portions of the country. Their CEOs earned $14-23 million in annual compensation (proxy statements, 2023). Their customers pay rates that have increased at roughly twice the rate of inflation over the past decade, not because electricity became harder to generate, but because the regulatory architecture guarantees returns regardless of efficiency.

The three largest oil and gas companies operating in the United States — ExxonMobil ($398 billion revenue), Chevron ($200 billion), and ConocoPhillips ($58 billion) — reported combined profits of approximately $80 billion in 2022, a record year driven by post-COVID price spikes. In that same year, the U.S. government paid approximately $15 billion in direct fossil fuel subsidies and the IMF estimated implicit subsidies (unpriced environmental damage) at $750+ billion annually for the U.S. alone. The companies received public subsidies while reporting record profits and paying executive compensation in the range of $20-36 million.

The Fossil Fuel Subsidy

Global fossil fuel subsidies totaled approximately $7 trillion in 2022 (International Monetary Fund estimate), including both explicit subsidies (direct payments and tax breaks) and implicit subsidies (the unpriced cost

of environmental damage). This represents approximately $1,300 per person on earth, every year, to subsidize an energy system that is producing catastrophic climate consequences.

Energy Poverty: The Regressive Extraction

Low-income households spend a disproportionate share of their income on energy. The U.S. Department of Energy documents that households below 200% of the federal poverty line spend approximately 8.6% of income on energy costs, compared to 3% for higher-income households. This energy burden is highest in states with the oldest utility infrastructure and the least competitive energy markets. The monopoly structure that produces guaranteed returns for utility shareholders produces guaranteed financial burden for the lowest-income customers who have no alternative supplier and no ability to invest in efficiency improvements.

Manufacturing

The $0.50 Worker and the $150 Shoe

A Nike running shoe retails for $100-$180. The factory that makes it receives approximately $20-25 from Nike. The worker who stitches it earns approximately $0.50-2.00 of that factory price, depending on the country of manufacture. The consumer pays $150. The worker receives $1.

The Garment Worker Wage Data

Current documented garment worker wages (Clean Clothes Campaign, Worker Rights Consortium, 2024 field research):

— Bangladesh: $95/month minimum wage. The world's second-largest garment exporter. 4 million+ workers. A garment worker producing 500+ pieces per month — with wholesale value of $2,500-$10,000 and retail value of $10,000-$50,000 — receives $95.

— Cambodia: $190/month minimum wage after years of labor organizing.

— Myanmar: $78/month — among the lowest garment wages globally.

— Ethiopia: $26/month. The newest garment manufacturing hub, chosen specifically because wages are lower than Bangladesh. H&M, PVH (Calvin Klein, Tommy Hilfiger), and other major brands have manufacturing there.

Rana Plaza: The Price of the Lowest Bid

April 24, 2013. The Rana Plaza building in Savar, Dhaka. Five garment factories. Approximately 3,000 workers. The day before collapse, structural cracks appeared throughout the building. Store owners on the lower floors closed immediately. Factory managers ordered garment workers to return. The building collapsed. 1,134 workers died. 2,500+ were injured. It remains the deadliest structural failure in modern industrial history.

The factories produced garments for Primark, Benetton, Walmart, and other Western brands. The brands had demanded the lowest possible price. The factories had cut costs on everything that could be cut: structural maintenance, worker safety equipment, building inspection compliance. The margin between the price brands would pay and the price at which the factory could survive left no room for the concrete reinforcement that would have held the building up.

The brands issued statements of shock and condolence. The Accord on Fire and Building Safety in Bangladesh — a binding legal agreement between brands and unions — was signed by 220+ companies in the aftermath. The price demands that created the conditions for the collapse continued. By

2020, many brands were demanding further price cuts, citing COVID-19 business pressures, while refusing to pay for completed orders already in production.

The Audit Fraud

The Worker Rights Consortium and Clean Clothes Campaign have documented across hundreds of factories: factories audited for labor compliance maintain two sets of books. One presented to auditors: compliant hours (below legal maximum), above-minimum wages, proper safety equipment. One actual: 70-80 hour weeks, below-minimum effective wages when production quotas are applied, locked fire exits, no ventilation, no safety equipment.

Brands rely on third-party auditors they pay to certify compliance. The auditors' business depends on brands continuing to hire them. The incentive to report non-compliance is structurally undermined by the incentive to retain clients. The audit certifies. The conditions continue. This is not a failure of the audit system. It is the audit system working as designed: providing legal cover for brands to claim they do not know what the people making their products live through.

Money

The System That Extracts From the People Who Have the Least

The American financial system charges the highest fees to the people who can least afford them. This is not an accident or an oversight. It is the

logical output of a fee structure built to extract maximum value from captive customers with no alternatives.

The Unbanked and the Underbanked Tax

Approximately 5.9 million U.S. households are unbanked (FDIC National Survey of Unbanked and Underbanked Households, 2022). An additional 18.7 million are underbanked — they have bank accounts but still use alternative financial services. These households pay:

— Check-cashing fees: 1-5% of check value, or $2-8 flat fee per check. On a $500 paycheck: $5-25 per paycheck, $130-$650 per year, just to access money already earned.

— Prepaid debit card fees: monthly fees of $5-10, plus transaction fees, ATM fees, and reload fees. $120-$240 per year to hold and spend your own money.

— Money order fees: $1-5 per money order. For households that pay bills by money order (because they lack checking accounts), this is a recurring tax on every payment.

— Payday loan interest: 300-400% APR documented across the industry (Consumer Financial Protection Bureau). A $300 loan repaid in two weeks costs $345-$395. Annualized: the cost of borrowing is higher than virtually any other legal financial product.

A low-income household without a bank account paying $1,300/year in alternative financial services fees is paying approximately 5-10% of their annual income for access to financial services that a middle-income household with a checking account receives essentially free.

The Overdraft Fee Machine

U.S. banks collected approximately $15 billion in overdraft fees annually in the years before regulatory pressure began reducing them (Consumer Financial Protection Bureau). Three banks — JP Morgan Chase, Wells Fargo, and Bank of America — collected the majority. The customers who paid the most overdraft fees were those with the least money: people who were overdrawn because they did not have enough to cover their expenses, not because of reckless financial behavior.

The CFPB documented a practice called "high-to-low transaction ordering": banks would process transactions in order from highest to lowest dollar amount, rather than chronologically, to maximize the number of overdraft events. A customer with $100 in their account who made a $95 gas purchase and five $3 transactions would incur six overdraft fees (one per transaction after the account balance hit zero) rather than one or two under chronological ordering. This practice — later restricted — was an engineered mechanism to extract maximum fees from the customers with the least financial cushion.

Transportation

The Platform That Owns the Road

Uber and Lyft collectively control approximately 98% of the U.S. rideshare market (Bloomberg Second Measure, 2024). They did not build this market by making transportation better. They built it by losing $31 billion (Uber) over years of below-cost pricing to crush competitors, then raising prices once competition was eliminated. This is the platform lifecycle documented in Chapter 20 of this book, executed at the largest possible scale.

The Driver Economics Documentation

Net driver earnings after accounting for platform fees, fuel, vehicle depreciation, insurance, and self-employment taxes: the MIT Center for Energy and Environmental Policy Research found median net earnings of $3.37 per hour in 2018, revised upward in subsequent analysis but remaining below minimum wage in many markets when all costs are properly accounted. The Economic Policy Institute's 2023 analysis found median net earnings of $9.21/hour after expenses.

The platforms dispute these figures using gross earnings rather than net earnings — before subtracting the vehicle costs that drivers bear. A driver earning $25/hour gross who spends $12/hour on fuel, depreciation, insurance, and self-employment tax earns $13/hour net. The platform reports $25/hour. The driver takes home $13. The gap between how platforms report driver earnings and how drivers experience driver earnings is the difference between a story about good wages and a story about below-minimum-wage work.

The Freight Broker Extraction

Freight brokers match shippers (companies with goods to move) with carriers (trucks with capacity to move them). The freight brokerage market generated approximately $90 billion in revenue in 2023 (Transport Topics, FMSCA data). Freight brokers take 15-35% of the shipping cost as their fee.

The carrier does the driving. The shipper provides the goods. The broker provides a phone call or, increasingly, a digital matching algorithm. For this matching service, the broker takes $15-35 of every $100 in shipping value. C.H. Robinson, the largest freight broker, reports $18 billion in revenue and

a net income margin of approximately 3-6% — on revenue they collected by sitting between people who were trying to find each other.

Governance

The $4.4 Billion Purchase of Political Access

In 2024, corporations and interest groups spent a record $4.4 billion on federal lobbying in the United States (OpenSecrets.org, Center for Responsive Politics). This is not political speech in any meaningful sense of the term. It is the purchase of access, attention, and favorable treatment from legislators and regulators who control the rules of the extraction economy.

The Largest Lobbying Spenders and What They Bought

The five largest lobbying sectors in 2024 (OpenSecrets.org):

— Healthcare/Pharmaceutical: $744 million. The return on investment: drug pricing legislation that does not allow Medicare to negotiate prescription prices, preserving the $500+ billion PBM extraction documented in Chapter 5. The pharmaceutical industry's lobbying investment prevents Americans from accessing drug prices available in every other developed country.

— Finance/Insurance/Real Estate: $726 million. This lobbying maintains the overdraft fee structures, resists fiduciary standards for financial advisors, and has blocked meaningful regulation of private equity's practices in healthcare, housing, and distribution. 39

— Communications/Technology/Electronics: $586 million. Platform industry lobbying prevents the antitrust enforcement that would address the monopolies documented throughout this book. Amazon, Meta, Google, and Apple collectively employ more lobbyists than the FTC and DOJ antitrust divisions employ staff.

— Energy/Natural Resources: $419 million. Maintains the $7 trillion in annual global fossil fuel subsidies. Delayed U.S. climate legislation for two decades.

— Transportation: $317 million. Prevents driver reclassification that would require gig platforms to provide workers' benefits.

The Cost Per Vote

The average cost of winning a U.S. Senate seat in 2024: approximately $26 million (Campaign Finance Institute). In 1990, adjusted for inflation: approximately $5 million. Quintupled in 34 years. The cost of a House seat in 2024: approximately $3.4 million, up from $600,000 in 1990.

These costs are not driven by the expense of reaching voters. Digital advertising has made voter communication cheaper per impression than at any previous point in history. The costs are driven by the volume of spending required to compete against opponents who are also receiving corporate money. The equilibrium is set by what corporations find it rational to spend to protect their regulatory environments.

The Government Procurement Extraction

The U.S. federal government spent approximately $700 billion on contracts in fiscal year 2023 (USASpending.gov). Approximately 65% of that value went to the largest 100 contractors. The top five defense contractors — Lockheed Martin ($65 billion), Raytheon Technologies ($67 billion), Boeing

($77 billion), General Dynamics ($42 billion), and Northrop Grumman ($36 billion) — collectively received over $150 billion in federal contracts in 2023.

The procurement system's complexity — thousands of pages of acquisition regulations, specialized compliance requirements, and established contractor relationships — creates barriers to entry that prevent smaller, potentially more innovative or cost-effective companies from competing for government work. The result is a closed market where incumbent contractors capture year-after-year contract renewals at prices that would not survive competitive bidding in a genuinely open market. The government pays. Shareholders receive.

The Revolving Door: Documented

The revolving door between government and the industries it regulates is not an anecdote. It is a documented, quantified phenomenon. According to the Project on Government Oversight, more than 380 former Pentagon officials, senior executives, and military officers took jobs with defense contractors between 2008 and 2018 — many within a year of leaving government positions with direct oversight of those contractors. The SEC, FTC, FDA, and other regulatory agencies show similar patterns.

The effect: regulators who expect to seek employment in the industries they regulate have structural incentives to maintain relationships with those industries throughout their regulatory careers. Enforcement that would damage those relationships is avoided. Rules that would harm future employers are weakened in implementation. The regulatory system is not captured through corruption. It is captured through career logic.

Media

The $0.003 Stream and the Creator Who Gets Nothing

Spotify pays approximately $0.003-$0.005 per stream to rights holders (Spotify for Artists data, Music Industry Research Association analysis). The rights holder is typically the record label, not the artist. After the label takes its share (typically 80-90% of streaming revenue under standard major label contracts), the artist receives approximately $0.0003-$0.001 per stream. One million streams — a milestone that marks a genuinely successful song — generates $300-$1,000 for the artist.

The Label Extraction Architecture

The major labels — Universal Music Group, Sony Music Entertainment, and Warner Music Group — collectively control approximately 68% of the recorded music market (IFPI Global Music Report, 2024). Their standard contract structure: the label advances recording costs as a "recoupable" loan against future royalties. The artist does not earn royalties until they have repaid the advance at the royalty rate. At $0.0003-$0.001 per stream per artist, recouping a $500,000 recording advance requires 500 million to 1.67 billion streams.

The label, meanwhile, collects streaming revenue from the moment the music goes live. It applies the income toward the advance at the royalty rate, earning the difference between what it collects and what it applies toward recoupment. An artist on Spotify generating $3,000 per million streams for the label generates $300 toward their advance recoupment. The label earns $2,700 on that million streams before the artist is "recouped." The advance was a loan. The label is both the lender and the entity that determines the repayment rate.

The Live Events Extraction

Live Nation Entertainment controls approximately 70% of major concert venues in the United States and approximately 40% of major festival and concert promotion globally (Department of Justice antitrust documentation, 2024). Its subsidiary Ticketmaster processes approximately 500 million tickets annually. The combined entity — which the DOJ sued in 2024 for antitrust violations — charges service fees of 20-35% on top of the face value of every ticket sold through its system.

Artists who want to tour major venues have limited alternatives to Live Nation's vertically integrated system: venue, promoter, and ticketing are often controlled by the same entity. The artist is offered package deals that bundle these services in ways that prevent price comparison or competitive negotiation. Fans pay the service fees. Artists cannot avoid venues that charge them. The DOJ's 2024 lawsuit alleged that Live Nation's market dominance allows it to engage in anticompetitive behavior that harms both artists and consumers while generating extraction at scale.

The Platform Advertising Extraction

Google and Meta together control approximately 50% of global digital advertising revenue (eMarketer, 2024). Publishers — newspapers, magazines, local news organizations, independent journalism — produce the content that audiences read. Audiences arrive to read the content. Google and Meta's advertising algorithms place ads next to the content, charging advertisers for the audience's attention. The publishers who created the content that attracted the audience receive a fraction of the advertising revenue that their content generated.

The result: newspaper newsroom employment in the U.S. fell approximately 57% between 2008 and 2020 (Pew Research Center). More than 2,200

local newspapers closed between 2005 and 2020. The information infrastructure of local democracy — the journalists who covered city councils, school boards, and local courts — was defunded by a platform advertising model that redirected the revenue their journalism generated away from journalism and toward platform shareholders.

The Creator Economy Extraction

YouTube retains 45% of advertising revenue from videos on its platform (YouTube terms of service). Meta takes 30% of all purchases made through its virtual goods and subscription systems. TikTok's creator fund pays approximately $0.02-$0.04 per 1,000 views — a rate so low that a video with 10 million views generates $200-$400 for the creator while generating significantly more advertising revenue for the platform. Twitch takes 50% of subscription revenue from most streamers (reduced to 30% only for its highest-tier "Partner Plus" program introduced under competitive pressure in 2023).

The platforms frame this as "democratizing" content creation. The economics are extraction: the creator provides the content, the audience, and the labor. The platform provides the infrastructure and takes 30-50% of the value in perpetuity. The creator owns nothing — not the audience relationship, not the viewer data, not the algorithmic ranking that determines whether their content is seen.

Identity

Your Data, Their Billions

Google and Meta earned a combined $400+ billion in revenue in 2024. Their products — search, email, social networking, photo sharing, maps — are provided at no charge to users. The business model: collect user behavioral data (search queries, locations visited, content consumed, relationships maintained, purchases considered), use that data to build detailed psychological profiles, and sell advertisers access to those profiles.

The user provides the raw material (their behavior, their data, their attention). The platform processes that material into a commercial product (the advertising profile). The advertiser buys the product. The user receives nothing except access to the product they are powering.

The Data Broker Market

Data brokers — companies that collect, aggregate, and sell personal information — operate largely without consumer awareness or consent. The major players include Acxiom, Experian, Equifax, LexisNexis, and Oracle Data Cloud. The data broker industry generates approximately $200+ billion annually (International Association of Privacy Professionals estimates).

Your personal record — name, address, age, income range, purchase history, political affiliation, health conditions, family composition, social media activity — is sold for approximately $0.005 to $0.50, depending on richness and recency. You receive $0. The data broker earns recurring revenue from selling your information to insurers, employers, marketers, political campaigns, and law enforcement. In most U.S. states, this trade is legal and unrestricted. The General Data Protection Regulation in Europe

provides significantly stronger protections; European residents can request deletion of their data from brokers. American residents largely cannot.

The Credit Score Extraction

Equifax, TransUnion, and Experian — the three major credit bureaus — collect financial behavior data on approximately 220 million Americans without those Americans' active consent. The bureaus sell this data to lenders, insurers, landlords, and employers. They charge the subject of the data — you — to access their own record. They charge you again to dispute errors. The error rate in credit reports is documented at approximately 1 in 5 Americans having at least one error in their credit file (Federal Trade Commission, 2013; updated analyses suggest the rate persists). An error can cost thousands in higher interest rates over the life of a loan. The bureau that created the erroneous record is not liable for the financial damage.

The bureaus earned a combined $12+ billion in revenue in 2023 from selling data that their subjects generated, charging subjects to see their own data, and charging subjects to correct errors in their own data. The economic model of the credit bureau industry is extraction from the people whose economic lives it documents.

The AI Training Data Extraction

Every major artificial intelligence system — the large language models powering ChatGPT, Google Gemini, Anthropic Claude, and dozens of others — was trained on text, images, and other content created by humans who were not compensated for that training contribution. Writers, artists, programmers, journalists, scientists, and billions of internet users collectively produced the corpus from which AI capabilities were derived.

They received nothing. The companies that trained AI systems on their work earned valuations in the billions.

The legal status of this training data use is actively contested in courts in the United States and Europe as of 2026. Class action lawsuits filed by authors, artists, and news publishers allege copyright infringement. The AI companies argue fair use. The core economic question — whether the people whose creative and intellectual work trained a multi-billion-dollar technology should receive any share of its value — has not been resolved. What is clear: the extraction occurred, the value was created and captured, and the people whose work enabled it received nothing.

The Platform Surveillance Architecture

The European Commission charged Amazon in 2020 with using non-public marketplace seller data to inform Amazon's private label product decisions — a practice that gave Amazon the benefit of millions of sellers' market research without their consent. Amazon settled in 2022. Meta has been fined multiple times by European data protection authorities for GDPR violations totaling over $2 billion. Google has been fined over $9 billion by the European Commission for various antitrust and privacy violations since 2017.

These fines represent a fraction of the revenue generated by the practices they penalize. For platforms earning hundreds of billions in revenue, billion- dollar fines are a cost of doing business, not a deterrent.

The Platform Extraction Leaderboard

Every Major Platform, Every Number, Every Source

What follows is a documented record of extraction rates, annual extraction volumes, and executive compensation for the major platforms of the extraction economy. All figures are from public SEC filings, company earnings reports, and independent third-party market research. All are verifiable.

Amazon

— Total seller fees: Referral fee (8-15% of sale price) + Fulfillment by Amazon fee (20-35% of sale price) + advertising (10-15% of sale price, now effectively required for visibility) + storage fees = 45-55% total effective fee rate (Marketplace Pulse annual analysis, Institute for Local Self-Reliance "Amazon's Toll Road" 2021)

— Annual extraction from sellers: $150B+ (estimated; Amazon does not disclose seller fees separately from third-party revenue)

— Sellers: 2 million+ active sellers on the U.S. marketplace

— Average seller profit margin after fees: 10-15% (Jungle Scout State of the Amazon Seller, 2024)

— Warehouse worker compensation: $19/hour average. Injury rate: approximately 2x industry average (OSHA data; Strategic Organizing Center analysis)

— CEO compensation: $29.1 million (2023 proxy statement)

Uber

— Take rate: 25-40% of gross fare (SEC filings; Economic Policy Institute analysis). Up from approximately 20% at platform launch.

— Annual driver payment: approximately $7.5 billion to drivers on $15+ billion in gross bookings (Uber Technologies 10-K, 2023)

— Driver net hourly earnings after expenses: $9-14/hour in most U.S. markets (EPI, UC Berkeley Labor Center)

— CEO compensation: $24.3 million (2023 proxy statement)

— Prop 22 lobbying spend (to maintain driver contractor classification): $200 million (California Secretary of State campaign finance records)

DoorDash

— Commission to restaurants: 15-30% per order (tiered based on subscription level)

— Additional marketing fees (required for visibility): additional 5-15%

— Total effective rate: 30-40% of order value (National Restaurant Association analysis)

— Driver base pay: $2-4 per delivery before tips and expenses

— Annual revenue: $8.6 billion (2023 10-K)

— CEO compensation: $413 million (2020, IPO year, mostly stock grants per proxy statement)

Instacart

— Service fees: 5% service fee + variable delivery fee + optional 15%+ tip + markup on grocery items (typically 15-25% above in-store prices)

— Shopper effective hourly rate after expenses: varies widely; surveys consistently document rates below $15/hour net

— Total effective consumer extraction rate: 35-45% above in-store cost (Consumer Reports analysis)

Airbnb

— Host service fee: 3% of booking subtotal

— Guest service fee: 14-16% of booking subtotal

— Combined take rate: 14-20% of total booking value

— Annual revenue: $9.9 billion (2023 10-K)

— CEO compensation: $8.4 million (2023 proxy statement)

Etsy

— Transaction fee: 6.5% of sale price including shipping

— Payment processing fee: 3% + $0.25

— Listing fee: $0.20 per item

— Offsite ads fee (mandatory for sellers over $10,000 annual sales): 12-15% of sales from offsite traffic

— Effective total fee: 15-20% for active sellers

— Annual revenue: $2.7 billion (2023 10-K)

Faire

— First order commission: 15% + $10 flat fee + 3%+ payment processing = 19-28% total

— Reorder commission: 15% + 3%+ payment processing = 18%+

— Next-day payout option: additional 3.5% + $0.30

— Peak valuation: $12.6 billion (2022)

— Current valuation: $5.2 billion (2025) — 59% decline

— Workforce reduction: 250 employees (20% of staff) in 2023

The Aggregate

These platforms collectively extract an estimated $200+ billion annually from the workers, creators, sellers, and small businesses that provide the labor, inventory, and content that make them valuable. This is not the platforms' total revenue. It is the portion of revenue that represents value created by participants and captured by the platform rather than returned to participants.

The people who make things, move things, cook things, drive things, make things for people to sleep in, sell things, and create things — they created the value. The platforms sit between them and their customers and take a third or more of every transaction.

The Five Layers of Extraction

The Complete Anatomy

The extraction documented across the twelve industries in Part Two is not twelve separate problems. It is one architecture expressed in twelve industries. That architecture has five compounding layers. Understanding how they compound is essential to understanding why reform fails and why only a different architecture can succeed.

Layer 1: Wage Extraction

Worker productivity in the U.S. has increased approximately 400% since 1970 (Economic Policy Institute, "The Productivity-Pay Gap," updated annually). Median worker compensation has increased approximately 26% over the same period, adjusted for inflation. The difference between 400% productivity growth and 26% wage growth represents the value workers created and did not receive.

CEO-to-worker pay ratio in 1965: 21:1. In 2024: 285:1 (Economic Policy Institute, "CEO Pay Has Skyrocketed," 2024). CEO compensation grew 1,094% since 1978. Typical worker compensation grew 26%. The companies employed the same workers. The profits came from the same work. The distribution changed.

Layer 2: Rent Extraction

Median home price as a multiple of median annual income: 2.5x in 1970, 7.5x in 2024 (National Association of Realtors; U.S. Census Bureau median

income data). Institutional investors own 25% of single-family rentals, up from 2% in 2010 (ATTOM Data Solutions). Every dollar of rent builds the landlord's equity. The renter's $1,500/month payment, sustained over 30 years, funds the landlord's $540,000 asset accumulation while the renter ends with nothing.

Layer 3: Debt Extraction

Global debt: $315 trillion and growing (Institute of International Finance). U.S. student debt: $1.75 trillion. Credit card debt: $1.14 trillion at 20%+ average interest. Medical debt: the leading cause of personal bankruptcy in the U.S. (66.5% of bankruptcies involve medical debt; American Journal of Public Health). Every interest payment transfers value from borrower to lender. The monetary system creates money as debt, requiring permanent interest payments that flow upward.

Layer 4: Platform Extraction

Documented in Chapter 16. $200+ billion annually from workers, creators, and small businesses to platform shareholders. Take rates increasing 50-100% over the past decade across all major platforms. Network effects creating monopolies that cannot be escaped. The worker who cannot leave because their ratings, their customers, and their income are all trapped in the platform's system.

Layer 5: Financial Extraction

Financial services: 4% of GDP and 10% of corporate profits in 1980; 8% of GDP and 30% of corporate profits in 2024. Private equity buys companies, loads them with debt, extracts management fees, and sells the remnant. Hedge funds charge 2% of assets plus 20% of gains for underperforming index funds most years. A 2% annual management fee

compounds to eliminate 50%+ of investment returns over 40 years. The retirement you thought you were building: half of it went to fund managers who added no demonstrable value.

How the Layers Compound

These layers do not operate independently. Each one makes the others more effective. The worker whose wages are suppressed (Layer 1) cannot afford to own (Layer 2) and must borrow (Layer 3). The borrower's debt payment reduces the margin available to leave a platform (Layer 4). The worker who cannot leave the platform has no savings to invest except in fee-laden financial products (Layer 5). At every step, the extraction leaves the participant more dependent on the next layer.

The system is not five separate extraction mechanisms. It is one interlocking architecture that captures value at the point of earning, the point of living, the point of borrowing, the point of transacting, and the point of saving. There is no economic activity within the system that does not have an extraction mechanism attached to it.

The 50-Year Construction Project

How We Got Here

The extraction economy at the scale documented in this book is not the natural outcome of capitalism. It is the result of specific choices made by

specific people at specific moments over 50 years. Understanding those choices is essential to understanding that they can be unmade.

1970: The Friedman Doctrine

On September 13, 1970, economist Milton Friedman published "The Social Responsibility of Business Is to Increase Its Profits" in the New York Times Magazine. The argument: corporations have no obligation to workers, communities, or society. Their sole duty is to shareholders. Everything else is distraction that reduces profits and therefore harms the only legitimate corporate purpose.

Before Friedman: the dominant theory of corporate governance balanced multiple stakeholders — workers, communities, shareholders, customers. The Business Roundtable's 1981 statement on corporate governance explicitly named workers and communities as constituencies corporations should serve. Corporate executives throughout the postwar period understood their role as including social obligations.

After Friedman: stock price became the only metric. Shareholder value became the only objective. Workforce reduction became "efficiency." Environmental cost-shifting became "competitive practice." Community abandonment became "market rationalization." The Friedman doctrine was not economic science. It was ideology dressed in academic language. But it became the operating system of every American public company within two decades.

1971: The Powell Memo

On August 23, 1971, Lewis F. Powell Jr. — then a corporate lawyer, two months before his nomination to the Supreme Court — submitted a confidential memorandum to Eugene Sydnor of the U.S. Chamber of Commerce titled "Attack on American Free Enterprise System."

The memo documented a perceived threat to corporate interests from consumer advocates, labor unions, and academic critics, and proposed a comprehensive strategy for corporate America to reclaim political and cultural power: infiltrate universities with pro-business faculty and curriculum, fund think tanks to generate favorable research, cultivate media relationships, build legal infrastructure to defend corporate interests, and lobby aggressively at every level of government.

Within ten years of the Powell Memo: the Heritage Foundation (1973), the Cato Institute (1977), and dozens of other corporate-funded think tanks were operating. Business school curricula shifted toward shareholder primacy. The infrastructure for regulatory capture was built. This is documented history, not speculation — the Powell Memo itself is a public document, available at the Washington and Lee University library where Powell's papers are archived.

1980s: The Policy Turn

The Reagan administration translated the Friedman doctrine and Powell Memo strategy into policy: top marginal tax rate reduced from 70% to 28%, antitrust enforcement effectively suspended, union protections undermined (beginning with the firing of the PATCO air traffic controllers in 1981), financial deregulation accelerating under Treasury Secretary James Baker. The wealth that had circulated through progressive taxation, union wages, and regulated financial markets began concentrating at the top. Top 1% wealth share: 22% in 1980, 43% in 2024.

1990s-2000s: Financialization and the Platform Emergence

Financial deregulation produced the financialization of the economy: the financial sector growing from 10% to 30% of corporate profits while adding diminishing productive value. The Glass-Steagall Act (separating

commercial and investment banking) was repealed in 1999. Derivatives markets expanded without oversight. Private equity developed the leveraged buyout model.

The platform economy emerged from this infrastructure: venture capital (itself a product of financial deregulation) funded the "subsidize, dominate, extract" playbook. The platforms that now control digital commerce, transportation, and media could only pursue predatory pricing at billion- dollar scale because venture capital was willing to absorb years of losses betting on future monopoly extraction. Uber's $31 billion in losses before profitability was not an accident. It was the playbook working as designed.

The Enablers

Congress, Wall Street, Courts, Media, Universities

The extraction machine does not run alone. It has institutional infrastructure. Every institution listed below enables extraction not through conspiracy but through rational responses to the incentives their structure creates.

Congress

Corporate lobbying: $1.4 billion in 1998, $4.4 billion in 2024. Tripled in 25 years (OpenSecrets.org). Amazon alone spent $21 million on federal lobbying in 2024. The pharmaceutical industry spent $744 million. The financial industry spent $726 million.

The lobbying investment is rational: if $744 million in pharmaceutical lobbying prevents a Medicare drug price negotiation bill that would reduce pharmaceutical revenue by $100 billion annually, the return on investment

is 134:1. Corporations lobbying for regulatory environments favorable to extraction are not being irrational. They are responding to incentives. The problem is not corrupt corporations. It is a campaign finance and lobbying system that makes corporate political investment enormously profitable.

Wall Street

Venture capital funded the platform extraction playbook. The logic: invest $31 billion in losses to capture a market, then extract for decades. Uber's investors at IPO received enormous returns on the $31 billion in losses that predecessors funded. The economics worked because the losses bought monopoly. The monopoly produced extraction. The extraction funded the returns.

Private equity developed the complement: buy established businesses, load them with debt, extract management fees, cut costs (primarily labor), and sell within 5-7 years. The businesses frequently decline or fail after the private equity exit. The workers who were cut are not recoverable. The communities that depended on the employment are not recoverable. The private equity firm has already distributed its returns.

The Courts

Citizens United v. FEC (2010): the Supreme Court held that political expenditure is a form of speech protected by the First Amendment, removing limits on corporate political spending. The practical effect: corporations can spend unlimited amounts on political advertising and lobbying. Money became speech. Corporations became speakers with effectively unlimited budgets.

The courts have also consistently ruled in favor of "independent contractor" classification for gig workers, against antitrust challenges to platform monopolies, and for the enforceability of mandatory arbitration clauses that

prevent workers and consumers from suing platforms in court. The legal architecture that protects the extraction economy was built over decades by judges appointed by politicians funded by the industries the judges regulated.

The Universities

Business schools have taught shareholder primacy as gospel since the 1980s. The MBA curriculum at the top schools — Harvard, Wharton, Chicago, Stanford — centers on financial modeling, shareholder value maximization, and strategic frameworks that treat worker welfare and community impact as externalities rather than objectives. The executives running the platforms documented in this book were trained in these frameworks. They are doing exactly what they were taught.

The Business Roundtable's 2019 "Statement on the Purpose of a Corporation" — signed by 181 CEOs including Amazon's Jeff Bezos — repudiated shareholder primacy and committed to serving all stakeholders. In 2020, academic research (Lucian Bebchuk and Roberto Tallarita, Harvard Law School) found that signatories' actual corporate actions showed no meaningful change from before the statement. The statement was rebranding. The curriculum remained.

The Extraction Playbook

The Five Phases Every Platform Follows

Every major extraction platform in the digital economy follows the same five-phase strategy. This is not coincidental. It is taught in business schools, documented in venture capital frameworks, and rewarded by public market

investors. Understanding the playbook makes every future platform predictable.

Phase 1: Subsidize

Platforms enter markets offering below-cost services, funded by venture capital losses. Uber lost $31 billion. Amazon lost money for 20 years. DoorDash lost money from inception through IPO. These losses are not failures. They are the price of eliminating competition. No competing service can operate below cost for years without venture backing. The VC- backed platform starves out competition by running a longer loss-funded war. When competitors exit, the monopoly is established.

Phase 2: Dominate

Network effects create lock-in. The more riders use Uber, the more drivers join Uber, which makes Uber more useful for riders, which attracts more riders. The more sellers use Amazon, the more buyers shop Amazon, which attracts more sellers. Once network effects reach critical mass, switching to an alternative means losing the network. Customers are trained to use the app. The platform becomes essential. Competitors cannot replicate the network even if they can replicate the technology.

Phase 3: Extract

Once dominance is established and switching costs are high, the platform raises take rates. Amazon seller fees: 19% in 2014, 45-55% in 2024. Uber driver share: 80% at launch, 50-60% today. DoorDash restaurant commission: 20% at launch, 30-40% today. The extraction happens gradually — small increases over years — so that no single change triggers mass exit. By the time the cumulative increase is severe, exit is impossible.

Phase 4: Optimize

Algorithms replace human judgment to squeeze harder. Dynamic pricing identifies the exact moment of peak demand and extracts maximum consumer surplus. Dynamic wages identify the minimum driver payment that still produces enough supply. Algorithmic management eliminates the human relationship that might produce leniency. The algorithm does not negotiate. It does not explain. It optimizes for the platform's revenue and is not accountable to any individual participant.

Phase 5: Replace

When participants burn out, new ones are recruited. The platform keeps everything: customer relationships, purchase history, ratings, geographic optimization data. The departing participant takes nothing. The platform's documentation of the participant's skills and customer base makes replacement straightforward. The participant is disposable. The data they generated is not.

The Generational Theft

The Broken Contract by the Numbers

The American social contract promised intergenerational progress: each generation would do better than the one before. The data now shows this contract is broken for the first time in recorded American economic history.

The Wealth Gap

— Millennials (born 1981-1996) own 4.6% of U.S. wealth (Federal Reserve Distributional Financial Accounts, 2024). At the same age, Baby Boomers owned 21%.

— Gen Z is the first American generation projected to be poorer than their parents (Pew Research Center, 2024).

— Between 2020 and 2024, $50 trillion in wealth transferred from younger generations to the asset-owning class through real estate appreciation, stock market gains, and inflation dynamics (Federal Reserve flow of funds analysis).

The Housing Contract Broken

— Down payment savings time (20% of median home price vs. median annual income): 4 years in 1970, 14 years in 2024 (National Association of Realtors; BLS median income data).

— 30-year mortgage total cost: a $400,000 home at 7% costs approximately $958,000 in principal and interest over 30 years.

— The Boomer who bought a $45,000 house in 1975 (2.5x median income) now holds an asset worth $400,000+, appreciated largely through inflation and demand dynamics they did not create. The Millennial who buys a $400,000 house in 2024 (7.5x median income) carries $480,000 in interest costs for the same shelter.

The Education Contract Broken

— 1970 minimum wage worker: 306 hours of work to pay one year of public university tuition.

— 2024 minimum wage worker: 1,568 hours — five times as many (College Board, "Trends in College Pricing"; BLS minimum wage data).

— Average student loan balance at graduation: $37,000 (Federal Reserve Bank of New York).

— Average monthly payment: $503 for 20 years = $120,720 in total payments on principal that was likely less than $37,000.

An entire generation did everything right. Went to college as instructed. Graduated as expected. Entered a labor market that had changed the rules. The ladder had been pulled up, not through individual failure, but through policy choices made by the generation ahead of them.

The Birth Lottery

What the Mobility Data Actually Proves

The following is an economic data analysis, not a political argument. The research cited comes from Harvard's Opportunity Insights project (led by economist Raj Chetty), the Federal Reserve, the World Bank, the Pew Research Center, and peer-reviewed social mobility research. None of these findings are contested empirically. They are documented.

The Intergenerational Mobility Data

— A child born to parents in the top 20% of income has a 40% probability of remaining in the top 20% as an adult (Chetty et al., Opportunity Insights, Harvard).

— A child born to parents in the bottom 20% has a 4% probability of reaching the top 20% — a 10-to-1 advantage for the already-wealthy, determined at birth.

— Moving a child from a low-income neighborhood to a higher-income neighborhood increases lifetime earnings by approximately 25% on average, even when the families are identical in every other measurable characteristic (Chetty et al., "The Effects of Exposure to Better Neighborhoods on Children," American Economic Review).

— Life expectancy in the United States varies by 10-20 years between neighborhoods in the same city (JAMA Internal Medicine, 2016). The same human body, born in different zip codes, has different life expectancy.

The Global Lottery

— A child born in Denmark: expected adult income above the global median, universal healthcare, free university education, strong social safety net.

— A child born in the Democratic Republic of Congo: 25% probability of dying before age 5 (WHO), average adult income under $1,000/year, minimal social infrastructure.

— Same human potential. Radically different probability of that potential producing a viable life. The difference is not character, intelligence, or effort. It is geography — an unchosen circumstance.

The Meritocracy Data Problem

The meritocracy narrative holds that success comes from hard work and talent. The mobility data contradicts this. If success were primarily a function of effort and ability, we would expect to see:

— Wealth distribution roughly matching effort distribution (most people try hard; most people would succeed)

— High social mobility (talent is distributed equally across income levels; it should rise regardless of birth)

— Birth circumstances weakly predicting outcomes (since outcomes should be determined by choices, not starting points)

What the data shows instead: wealth concentrates regardless of effort, social mobility is declining, and birth circumstances predict outcomes with consistency that individual variation does not substantially disturb. The meritocracy narrative functions not as description of reality but as ideology: it tells the successful that they deserve what they have, and tells the unsuccessful that their situation reflects their choices rather than their circumstances. The economic consequence is to reduce pressure for structural change.

Why Reform Has Never Worked

The Structural Impossibility

Fifty years of reform attempts. The extraction gets worse. This is not evidence of insufficient reform effort. It is evidence that reform within the system cannot address an architecture that the system is designed to protect.

The Reform Cycle

Every significant reform attempt follows the same pattern: crisis occurs, reform is demanded, legislation passes, industry lobbies the regulatory implementation, rules are weakened through the administrative process, enforcement is captured by the regulated industry, the reform is effectively reversed, and the crisis eventually recurs in a different form.

The New Deal regulatory framework was eroded over 40 years of lobbying. Glass-Steagall survived 66 years before repeal in 1999. Union protections declined from 35% membership in 1954 to 10% today — not through repeal but through administrative weakening, enforcement gaps, and the structural shift to gig work that places workers outside union protections by design. Environmental regulations are enforced in proportion to the political fortunes of the administration in power.

The Minimum Wage Problem

Platform workers are classified as independent contractors. Minimum wage does not apply. California's AB5 in 2020 reclassified gig workers as employees eligible for minimum wage and benefits. Uber and Lyft spent $200 million on Proposition 22 to create a carve-out exempting themselves from AB5. They won. The legal architecture that makes wage extraction possible survived democratic challenge at a cost of $200 million. That $200 million bought the legal right to continue paying below-minimum wages to millions of drivers. The return on investment was substantial.

The Antitrust Enforcement Gap

The last successful major structural antitrust breakup was AT&T in 1982. Since then: Microsoft won its antitrust case. Google has been in EU and DOJ antitrust proceedings for over a decade with limited structural remedies.

Amazon has faced antitrust investigation but no structural intervention. Facebook acquired Instagram (2012) and WhatsApp (2014) with FTC approval; the FTC attempted to reverse those acquisitions in 2021 and failed in court.

The platforms have outgrown the regulatory apparatus designed to check them. Their legal teams are larger than the agencies regulating them. Their lobbying budgets dwarf the agencies' total operating budgets. The revolving door between regulatory agencies and the industries they regulate means that the most experienced regulators often move to industry, and industry veterans move into regulatory roles.

The Fundamental Problem

All reform attempts try to constrain extraction while leaving the fundamental architecture intact: platforms with fiduciary duty to shareholders, shareholders demanding growth, growth requiring extraction. You cannot regulate away the incentive to extract when the legal structure of the corporation requires it. The solution is a different legal structure — one that does not have extraction as its legally mandated purpose. That is what the LLC → C-Corp → Public Benefit Corporation pathway and protocol governance are designed to provide.

The Graveyard

Every Alternative That Failed, and the Exact Reason

If the extraction economy can be reformed neither from within nor through regulatory intervention, the only remaining option is to build a replacement.

Many people have tried. Every attempt failed. Understanding exactly why each one failed is the precondition for building something that doesn't.

Fairphone (2013) — Couldn't Compete on Price

Fairphone built a smartphone with fair wages for factory workers, conflict- free minerals, and a repairable design. The product was genuine. The mission was real. The economics were fatal. Ethical production costs multiples of exploitative production. At the price point required to cover fair wages, consumers chose cheaper alternatives in the overwhelming majority of cases. Fairphone exists with a small and loyal market share. It did not reform the smartphone industry.

The lesson: you cannot ask consumers to pay a premium for fairness in a competitive market. Fairness must be built into infrastructure that is competitive on its own merits. GRAJ does not ask participants to pay more for fairness. The 5% model is economically competitive because it does not carry the overhead of shareholder return requirements.

Stocksy United (2012) — Couldn't Scale Governance

Stocksy United created a photographer-owned cooperative for stock photography: artists own the platform, share revenue equitably, govern democratically. It works. It serves a small and loyal membership. It did not challenge Shutterstock or Adobe Stock for market leadership.

The reason: cooperative governance requires consensus. Consensus requires deliberation. Deliberation takes time. Markets do not wait for deliberation. While Stocksy debated feature decisions by member vote, platform competitors moved faster. The governance model that makes Stocksy fair also makes it slow.

The lesson: build fair governance that does not sacrifice speed. GRAJ starts as a company with founders who can make fast decisions, then transitions governance to participants progressively as the network matures. Speed enables growth. Decentralization enables permanence. The sequence matters.

Early eBay (1990s-2000s) — Couldn't Resist Shareholder Pressure

Early eBay was the closest thing the internet had to a genuinely democratic marketplace: person-to-person, low fees, accessible to anyone. The eBay of 1996 is the closest the internet produced to what GRAJ is building. Then eBay went public. The IPO created fiduciary obligations to shareholders. Shareholders demanded growth. Growth required higher take rates, more advertising requirements, preferred treatment for high-volume professional sellers. The garage-sale democracy became a commercial marketplace for retailers.

This is the most important case study for GRAJ. Early eBay was right. The mission was right. The technology was right. The business structure destroyed the mission. The lesson: good intentions do not survive a fiduciary structure that legally requires shareholder return maximization. GRAJ's 5% is written into the protocol, not the intention. The protocol does not have good intentions. It has rules that cannot be overridden by shareholder pressure.

Juno (2016) — Couldn't Prevent Acquisition

Juno launched in New York with an explicit promise to drivers: equity in the company. Actual shares, not gestures. A meaningful stake in the enterprise they were building. Drivers responded. Juno grew. Within two years, Gett acquired Juno for $200 million. The driver equity program was effectively

discontinued. Drivers who had been promised ownership received nothing meaningful. The founders received the acquisition price.

The lesson: a promise to give equity is not equity. A verbal commitment to fairness is not a legal obligation. GRAJ's founder equity is locked for 20 years in a shareholder agreement. The $15 million annual earnings cap for all GRAJ team members is in the operating agreement. The protocol transition that distributes governance to participants is the plan, not the aspiration. These are contracts, not commitments.

Ben & Jerry's, Burt's Bees, Plum Organics — Couldn't Prevent Acquisition

Each was a mission-driven company acquired by a large corporation: Ben & Jerry's by Unilever, Burt's Bees by Clorox, Plum Organics by Campbell's. Each maintained mission language in marketing after acquisition. Each saw extraction resume in supply chains, labor practices, and financial structures. Ben & Jerry's has been the most vocal exception, using its contractual independence to take public positions its parent company opposes. It remains an outlier.

The lesson: mission survives acquisition only when it is legally protected, structurally separated from the acquiring entity, and has contractual independence with teeth. Most acquisitions produce mission-washing: the language of social purpose wrapped around the practice of extraction. GRAJ's anti-acquisition provisions are not idealism. They are the structural response to watching every mission-driven company that could be bought get bought, its mission kept as marketing.

The Four Failure Modes and GRAJ's Architectural Response

— Couldn't compete on price — GRAJ counter: 5% is competitive from day one. No premium required.

— Couldn't scale governance — GRAJ counter: company first (fast), protocol second (democratic), decentralized by year 10 (permanent).

— Couldn't resist shareholder pressure — GRAJ counter: 5% written into protocol code and unchangeable without 67% supermajority participant vote, LLC → C-Corp → PBC structure, governance transition eliminates shareholder override.

— Couldn't prevent acquisition — GRAJ counter: 20-year founder lockup, no acquisition provisions, protocol decentralization makes acquisition structurally meaningless by year 10.

AI and the Great Displacement

The Middle Class Being Erased Now

Artificial intelligence is not a future threat to middle-class employment. It is a present one. The jobs being automated in 2024 and 2025 are not factory jobs (most of those left decades ago) and not executive jobs (those remain protected by organizational structure). They are the cognitive

middle-class jobs that grew to replace the manufacturing jobs that left: customer service, bookkeeping, data analysis, content creation, paralegal research, translation, graphic design, entry-level software development.

McKinsey Global Institute estimates that 12 million American workers will need to change occupations by 2030 due to automation (McKinsey, "The Future of Work in America," 2019, with updated projections in 2023). The World Economic Forum projects 85 million jobs displaced by 2025 against 97 million new roles created — but the displaced jobs are concentrated in the occupational middle, and the new roles require skills that displaced workers do not currently have and cannot acquire quickly.

Who Captures the Value

When a company replaces a $60,000/year customer service employee with an AI chatbot, the $60,000 in annual savings does not flow to customers as lower prices. It flows to shareholders as higher profit margins. This is documented in every quarterly earnings call where AI-driven cost reductions are announced: the announcement causes the stock price to rise, which creates wealth for shareholders, which concentrates wealth further at the top of the existing distribution.

The productivity gains from AI will be the largest in human economic history. Who captures those gains is an architecture question, not an inevitability. If AI operates through platforms with the current extraction architecture — platforms that own the AI, the customer relationships, and the infrastructure — the gains will concentrate in the fewest hands of any technological transition in history. If AI operates through infrastructure with a 5% fee structure and participant governance — infrastructure like the protocol GRAJ is building — the gains can distribute to the people whose work is being augmented rather than replaced.

The Wealth Gap Compounds

Exponential Inequality

Wealth concentration is exponential, not linear. The mechanism is simple and documented: money that is invested earns returns. Larger amounts earn larger absolute returns at the same rate. The compounding effect means that wealth gaps, once established, grow automatically even without any additional extraction.

The Federal Reserve's Distributional Financial Accounts data shows: the top 1% of American households earned approximately 7% annual returns on their wealth over the past decade. The bottom 50% earned effectively 0% — because they hold virtually no financial assets to appreciate. Their "wealth" consists of checking and savings account balances that earn fractional interest rates while inflation erodes purchasing power.

The Trajectory

— Top 1% wealth share: 22% in 1980, 30% in 2000, 43% in 2024 (Federal Reserve).

— At current growth rates: top 1% projected to own 50%+ by 2035.

— The bottom 50% collectively own less than 1% of total wealth. Their share is not declining because there is essentially nothing left to decline.

Every economic system in history that reached these concentration levels experienced a corrective event: the French Revolution, the Russian Revolution, the New Deal reforms responding to 1930s conditions in America. The corrections were not designed. They were forced by the social

instability that extreme concentration produces. The question this book poses is whether the correction happens through design — through building different infrastructure before the instability becomes ungovernable — or through the kind of disruptive social event that nobody controls.

Political Capture Is Structural

The Numbers That Make It Permanent

Corporate lobbying spending: $1.4 billion in 1998, $4.4 billion in 2024. Tripled in 25 years. The average cost of winning a U.S. Senate seat: $5 million in 1990, $26 million in 2024. Quintupled. These are not separate trends. They are the same phenomenon: as the economic stakes of regulation increased with platform-scale revenues, the rational corporate investment in political access increased proportionally.

This is not a left or right issue. Both parties receive corporate campaign contributions. Both parties' legislators move between government and corporate positions through the revolving door. The structural reality is independent of partisan preference: when elections cost $26 million and corporations have both the money and the incentive to supply it, the outcome is predetermined regardless of the individual integrity of specific politicians.

The only political path to reducing extraction would require campaign finance reform sufficient to sever the connection between corporate money and political access. This reform would need to be passed by legislators

whose campaigns depend on corporate money, and upheld by a Supreme Court that has twice (Buckley v. Valeo and Citizens United) ruled that political spending is protected speech. The legal architecture makes meaningful campaign finance reform structurally unavailable through normal political processes.

The exit: build infrastructure that does not need political permission to be fair. GRAJ does not lobby. GRAJ does not make campaign contributions. GRAJ does not ask regulatory permission to take 5% instead of 45%. It builds the infrastructure and demonstrates through operation that the alternative is viable.

The Convergence

When All Four Accelerants Hit at Once

Each of the four accelerants described in this part is serious in isolation. What makes this moment historically distinct is that all four are operating simultaneously and feeding each other.

AI displaces middle-class workers. Displaced workers have less income, less savings, and less political power. Less political power means less resistance to corporate lobbying. More concentrated lobbying produces less regulation of AI and platforms. Platforms and AI tools accelerate wealth concentration. Concentrated wealth buys more AI development. The cycle repeats at higher intensity.

The flywheel runs in both directions. The extraction economy's flywheel: AI efficiency → higher corporate profits → more concentrated wealth → more political capture → less regulatory resistance → more AI efficiency. The GRAJ flywheel: lower fees → participants earn more → more

participants join → more transactions → more revenue at 5% → better infrastructure → lower costs per transaction → more participants.

The question is which flywheel gets the initial push. The extraction economy's flywheel is already running and accelerating. Building an alternative flywheel requires enough initial momentum to escape the gravitational pull of the incumbent architecture. That is the purpose of the Detroit and NYC launch: establish enough transaction volume to demonstrate that the alternative flywheel is real, then let compounding do the rest.

The window during which the alternative flywheel can be started is finite. As AI-powered extraction becomes more efficient, as wealth concentration deepens, as political capture strengthens, the cost of entry for alternatives rises and the network effects of incumbents grow. Build now or accept the acceleration.

Germany

What Capitalism Looks Like Without Maximum Extraction

Germany has the fourth-largest economy in the world by GDP. It maintains a consistent trade surplus. It has a manufacturing base that the United States dismantled in pursuit of shareholder returns. Its companies are among the

most competitive in the world in industrial machinery, automotive, chemicals, and precision manufacturing.

Germany also has a CEO-to-worker pay ratio of approximately 6:1, compared to 285:1 in the United States. German workers have legally mandated representation on corporate boards (Mitbestimmung — codetermination). German companies are overwhelmingly long-term oriented rather than quarterly-earnings oriented, because the Mittelstand — the small and medium family-owned businesses that represent 99% of German companies and employ 60% of German workers — is not subject to public market shareholder pressure.

The Mittelstand Architecture

The Mittelstand companies are not publicly traded. They are family-owned, regionally rooted, and multigenerational in orientation. Because they are not optimizing for quarterly earnings or share price, they invest in apprenticeship programs, maintain employment through economic downturns by cutting hours rather than jobs, develop long-term supplier relationships rather than squeezing suppliers annually, and retain skilled workforces rather than outsourcing to the cheapest available labor.

The German apprenticeship system: approximately 325 recognized training occupations with standardized dual education — classroom instruction and on-the-job training — produces skilled workers without the debt burden of four-year university. Approximately 60% of German young people enter apprenticeships rather than university. The system is funded cooperatively by government and employers. The result: low youth unemployment and a manufacturing workforce that maintains world-class precision capabilities.

This is not a Nordic welfare state. It is not socialism. It is capitalism with different rules — rules that preserve long-term value creation over short- term extraction.

Mondragon

The 70-Year Worker-Owned Enterprise That Works

The Mondragon Corporation in the Basque region of Spain is the world's largest worker cooperative. Founded in 1956 by José María Arizmendiarieta, a Catholic priest, with five engineering graduates and a small machine shop, it has grown to 80,000+ worker-owners across more than 100 cooperative businesses: banks, supermarkets, engineering firms, appliance manufacturers, educational institutions, and a university.

The Numbers

— Annual revenue: approximately $12 billion (Mondragon Corporation annual report, 2023)

— CEO-to-worker pay ratio: approximately 6:1 (Mondragon has a policy-set maximum ratio, currently at 9:1 for the most senior positions, with most at 6:1 or less)

— Response to the 2008 financial crisis: instead of layoffs, the cooperative voted to reduce hours and share the economic pain across the membership. Employment was preserved. The decision was made by workers voting on the company's response to a crisis that threatened their livelihoods.

— Failure rate of Mondragon cooperatives vs. conventional Spanish businesses: significantly lower over 70 years of operation (research by Mondragon University and independent academic analyses)

Mondragon proves that worker ownership at significant scale is viable for generations. It does not prove that every business should be structured this way from the beginning. It proves that the destination — participant ownership and governance — is achievable and sustainable.

GRAJ's protocol phase (years 5-10) and peace phase (years 10-30) are building toward the kind of distributed ownership and governance that Mondragon has maintained for 70 years. The path is different: company first, protocol second, decentralization third. The destination is similar: infrastructure owned and governed by the people who use it.

The Nordic Model

The Countries That Chose Different Rules

Sweden, Norway, Denmark, and Finland consistently rank at the top of global indices for happiness (World Happiness Report), social mobility (OECD), quality of life (UN Human Development Index), and worker satisfaction (Eurostat). They are also capitalist economies with competitive private sectors, innovation ecosystems, and global corporations.

The difference is not the absence of markets. It is the rules governing markets.

The Structural Differences

— Union membership: 67-74% in Nordic countries versus 10% in the United States. Sector-wide collective bargaining means wages are set across industries, not just at individual companies. A non-union employer in Denmark pays approximately the same wages as a union employer because the collective agreements set the floor for the entire sector.

— Social safety net: universal healthcare, heavily subsidized childcare, generous unemployment insurance, active retraining programs. The safety net is not charity. It is the precondition for labor market flexibility: workers can accept job transitions without existential risk because losing a job does not mean losing healthcare or the ability to pay rent.

— Tax structure: top marginal income tax rates of 50-60%, funding the social infrastructure. The tax system redistributes a portion of the returns from capital and high income, reducing the compounding of wealth concentration documented in Chapter 26.

— Corporate governance: strong worker representation on boards. Shorter- term executive compensation tied to long-term metrics rather than quarterly stock price. Less pressure to extract for short-term shareholder returns.

The Outcome Data

— Social mobility: a child born into the bottom income quintile in Denmark has a 12% chance of reaching the top quintile. In the United States: 7.5% (Chetty et al., Opportunity Insights).

— Life expectancy: Denmark 81 years, Finland 82, Sweden 83, Norway 83 (WHO, 2024). United States: 76 years. 80