Milton Friedman's 1970 case for shareholder-profit-maximization had a condition built into the same sentence everyone quotes half of: a business should pursue profit "so long as it stays within the rules of the game, which is to say, engages in open and free competition, without deception or fraud."^[1] The popular version of "the Friedman doctrine" keeps the conclusion and drops the condition. But the condition is the whole argument — profit-maximization was only supposed to serve the public because open competition disciplines it. Over the following decades, three separate sets of rules that define "the game" changed: how executives get paid, whether a company can buy back its own stock, and how much market concentration antitrust law will tolerate. None of the three required rewriting Friedman. They just quietly removed the condition his conclusion depends on.

Start with pay. Until November 17, 1982, a public company repurchasing its own stock on the open market risked prosecution for price manipulation under the 1934 Securities Exchange Act. That day the SEC adopted Rule 10b-18, creating a safe harbor for buybacks that meet specific volume and timing conditions — not full legalization, but enough certainty that repurchase volume roughly tripled within a year.^[2] Eleven years later, Congress tried to rein in runaway executive pay with a new tax rule: Section 162(m) capped the deductibility of cash compensation above $1 million, but exempted "performance-based" pay — including stock options — from the cap entirely. The result was the opposite of the intent. Total executive pay rose faster after 1993, not slower, because boards simply shifted compensation into the exempted category, tied directly to the stock price a buyback program can move.^[3] Neither rule told a CEO to prioritize the stock price over the business. Both made prioritizing the stock price the highest-paid available option.

The same year Friedman's doctrine turned eight, a second and arguably larger rule change was underway on a completely different track. In 1978, Robert Bork published The Antitrust Paradox, arguing that antitrust enforcement had drifted into protecting small competitors and pursuing political goals instead of one measurable standard: consumer welfare, meaning price and output effects. Bork's framework won — not by statute, but by persuading courts and regulators — and became the de facto standard for roughly the next four decades.^[4] A consumer-welfare standard sounds like exactly what Friedman would have wanted: competition judged by outcomes, not politics. In practice it set a much higher bar for blocking a merger — a deal only looked dangerous if it was likely to raise prices soon, not if it simply removed a competitor from the field. Concentration followed. Since the late 1990s the Herfindahl-Hirschman concentration index has risen in more than 75% of U.S. industries, by an average of roughly 90%, and the number of publicly traded U.S. companies has fallen by nearly half — even as the economy has more than tripled in real size since the early 1970s.^[5]

Airlines make the mechanism visible, because the industry ran the experiment twice, in opposite directions, inside a single lifetime. The Airline Deregulation Act of 1978 — the same year as Bork's book — genuinely increased competition: it eliminated federal control over routes and fares and let new carriers enter freely.^[6] That lasted until the weakened antitrust standard caught up with it. Between 2008 and 2014 the Justice Department approved a wave of mergers — Delta-Northwest, United-Continental, American-US Airways, Southwest-AirTran — that took the industry from eight major carriers to four. Those four now control roughly 80% of U.S. domestic passenger traffic.^[7]

Consolidation alone doesn't fully explain why a traveler can't just pick a cheaper airline at their home airport. Gate access does. Since at least 1989 the GAO has documented long-term exclusive-use gate leases as a structural barrier: an incumbent airline that controls most of an airport's gates can keep a would-be competitor out even where no rule technically bans them, and GAO found fares ran proportionately higher at airports with more gates under exclusive lease.^[8] Washington Dulles (IAD) is the live, current version of this. United and United Express account for roughly 70% of Dulles departures, and United is the sole carrier on Concourses C and D — even though the airport authority's own stated policy is "no exclusive-use gates."^[9] Southwest, the one carrier that might have pressured that share, gave up after two decades: it announced on March 12, 2026 that it would end all Dulles service by June 4, 2026, having never grown past two gates in Concourse B.^[10]

That isn't an abstraction if you live near IAD. Flying anything other than United out of Dulles routinely means driving to a different airport — Reagan National (DCA), across the Potomac — for the same trip. When Troy's kids flew to Fort Lauderdale, that's exactly what happened: a drive to DCA to get a real choice of carrier. The reason that option exists at all is that the DMV happens to have three major airports within reach. Most of the country doesn't. A traveler in a region with one commercial airport and one dominant carrier has no equivalent drive available — the "free and open competition" Friedman's whole argument depends on simply isn't there to discipline the price.

The competition point compounds with a bigger one: geography doesn't just set your price, it sets your reach. International nonstop service is heavily concentrated at a handful of U.S. gateway hubs — JFK alone accounts for roughly one in ten U.S. international departures, and is the country's top gateway for Europe, Africa, and the Middle East.^[11] Dulles is one of those gateways, and so are Chicago O'Hare and LAX — from any of those metros, "anywhere in the world" is a realistic nonstop or one-stop trip. That reach isn't universal, but exactly how far it doesn't extend is a check, not an assumption: Des Moines currently has zero scheduled international routes, while smaller airports elsewhere carry a handful. The DCA drive proves a real domestic choice exists nearby; the fact that IAD is in the same international-gateway tier as O'Hare and LAX is a second, separate form of the same geographic luck — one Chicago and LA share, and most of the country doesn't.

That reach carries a price tag, not just a convenience one. For any job where the work is showing up in person — client-facing consulting, dealmaking, sales, executive roles that run on face time — hub-airport proximity isn't neutral background, it's part of what the job pays. Economists studying hub airports have found real, measurable spillover effects on local employment and productivity, layered on top of the long-documented "urban wage premium" that larger, better-connected metros carry generally.^[13] For those industries, "where I live" isn't a lifestyle choice sitting next to the paycheck — it's proportional to it.

None of this requires abandoning Friedman's argument — it requires taking his own condition as seriously as his conclusion. Profit-maximization "within the rules of the game" was never a blank check; it was contingent on the rules actually preserving open, free competition. When the rules governing pay incentives, buyback legality, and merger review all moved in the same direction over the same two decades, the condition eroded quietly enough that the conclusion kept getting quoted without it. The fix Friedman himself would likely recognize isn't a new theory of corporate purpose — it's rules that actually do what his sentence assumed they already did.