Vault

Arthur Laffer Interview — “All I Think About Is Incentives”

A 102-minute interview. Laffer’s whole frame is one principle: tax something and you get less of it, subsidize it and you get more. A skeptical British host presses him on market failures: monopoly, 2008, China, and the pre-1913 tariff state.

Date: watched 2026-09-24 (publish date: TBD) Source: YouTube — v=3L0W9BK_nao (The Winston Marshall Show, confirmed by Chris) — Transcript · clean Participants: Dr Arthur Laffer (supply-side; adviser to Nixon, Reagan, Thatcher and Trump) vs Winston Marshall (self-described layman, market-failure skeptic of neoliberalism) Vault relevance: Failure as Mandate, What Is Money?, Business Cycles, Inflation, Plato’s Ship of State, History Repeats — Small-Pain-Tolerance

Argument Structures (what was actually said)

Laffer

Incentives are the whole of economics. “If you tax people who work and you pay people who don’t work… you’re going to get less people working.” Taxes are negative incentives that tell you what not to do (report income), but not how to avoid it. [0–4m, 72m]
The Laffer curve, stated correctly. Below the peak, raising rates raises revenue, but every tax shrinks the base “from the very first percent.” “I don’t have any idea why people worry so much about government revenues and not about people’s welfare.” [4–6m]
The “transfer theorem.” Redistribution lowers total income on both ends (the taxed rich produce less; the subsidized poor have an alternative to working), citing the Slutsky equation. Lemma: the more you redistribute, the bigger the loss. Limit case: perfect equality means taxing 100% of everything above average and subsidizing up to average, so “we’ll all be equal at zero.” [6–9m]
The ideal tax is a low-rate, broad-based flat tax. It’s fair by an “objective standard”: earn 40× and pay 40×, not 40,000× and not zero. Poverty is the problem, not inequality (Harrison Bergeron). [2m, 24–29m]
Market failure is not sufficient for government action. The intervention has to do less damage than the failure. Monopoly is a matter of degree; patents and copyright are a deliberate balance. [10–13m]
Healthcare: transparency as the fix. He pushed an executive order on medical price transparency. The same operation costs “$5 to $5,000” depending on the insurer; HIPAA guarantees nobody knows; hospitals collude. Allow insurance across state lines. [14–19m]
Pay politicians on commission. Link congressional pay to GDP growth (3% = base, 4% = double, 1% = pay it back). Merit pay for teachers and bonuses for civil servants. The host objects that this creates false incentives. [18–21m]
Money: 1776–1913 was private, and it worked. Government defined the dollar in gold and silver and audited banks, but private mints and banks did the rest: “no inflation for 139 years… no great depressions.” The Fed “nationalized money” in 1913, and prices are up 55–60×. He opposed leaving gold in 1971. No Fed, but yes to bank regulation. Crypto is “wonderful.” [32–37m]
2008: let them fail. “Sit down and shut up… I’m not listening until it’s over.” The assets don’t vanish; they get new owners. “It’s failure that creates carefulness.” His own specimen: he turned down a lifetime chair at USC because “if I don’t have any fear of loss, I’ll atrophy.” [37–41m]
The Rust Belt was hollowed out by state taxes, not China. Firms moved to zero-income-tax Tennessee and Texas. Detroit has property taxes near 20%. [42–44m, 98–101m]
China: trade is good; security is a case-by-case exception. Cheap Chinese AI and EVs are good for consumers. “Critical industry” claims must be specific (costs vs benefits), not arm-waving. Foreign ownership stays under your political control. Trade integration makes war costly. Kennedy/Reagan: peace through strength, “the best defense spending is always wasted.” [44–68m, 94–98m]
Trump’s tariffs: his first-term trade deals lowered barriers; tariffs as a negotiating tool are “to be seen.” Smoot–Hawley caused the Depression (Mundell). [74–84m]
The pre-1913 tariff state isn’t a model. Spending was tiny and tariffs were easy to collect at the docks; today’s spending levels can’t be funded that way. [86–88m]
Spending rule: spend until the marginal benefit equals the marginal cost of the tax. The level is for the electoral process to set; his view is “a lot lower.” Enterprise zones (no income or payroll tax, a building-code review, no minimum wage) for inner cities. Right-to-try. Raise the pension age. “The most immoral act a government can perpetrate is to destroy the production base from whence all benefits flow.” [88–94m]

The host

Monopoly is the market failure even the right accepts (Pareto, Teddy Roosevelt’s antitrust, Smith). [8–11m]
Transparency isn’t enough without an alternative to choose; state insurance markets are duopolies (citing The Myth of Capitalism). [14–19m]
Feelings matter: if people feel the system is unfair, they stop playing along. [24–25m]
QE made the rich richer, so the rising tide failed; bailouts in 2008 were “the worst of both worlds.” [32m, 38m]
China uses state subsidy to undercut critical industries (AI, EVs, steel). Trade didn’t prevent WWI or WWII, and didn’t liberalize China; instead the West self-censors. [44–64m]
Pre-1913 America, Laffer’s golden age, was funded by tariffs. [86m]

Discussion seeds

Laffer is sitting on your fence. No Fed and no government money, but yes to bank regulation and fractional-reserve private banks: close to the free-banking package on What Is Money?. His history is the fixed-currency test you proposed an hour earlier. Grounding flags: “no great depressions” skips the panics of 1819, 1837, 1857, 1873 (the Long Depression), 1893 and 1907. The Civil War greenbacks roughly doubled prices. The National Banking Acts (1863–64) federalized note issue fifty years before the Fed. Roughly flat prices over the whole span, but not stable. Does the record help the free-banking case or hurt it?
Commission pay is an attempt to fix the Ship-of-State ratchet. It targets exactly the delegate-grading failure from the morning: pay the delegate on outcomes. But your lag point bites: policies show up 20 to 100 years later, and GDP growth in a 2-year term is mostly inherited. Does commission pay close the inverted loop in Failure as Mandate, or just give politicians a new metric to game?
“Let them fail” is the anti-mandate. The 2008 answer and the USC tenure story are both the principle that losses must land on the one who erred. That’s the positive image of failure-as-mandate and of the small-pain note’s prevented liquidation. It’s a specimen for both.
Transparency targets the credence-good check. His healthcare fix (publish prices, allow insurance across state lines) goes after both checks on the failure-as-mandate page: the customer can grade the service, and a rival can collect. The host’s objection (transparency without an alternative is useless) is the claim that the competition check is missing. Who’s right?
The transfer theorem’s own citation cuts against half of it. Laffer cites Slutsky, but a tax on the rich has an income effect that pushes them to work more (to restore lost income), against the substitution effect. “The taxed produce less” requires substitution to dominate, which is an empirical question, not a theorem. The limit case (100% equalization → zero) is sound; the monotonic lemma needs conditions.
Symmetric check on the Rust Belt. Laffer: state taxes, not China. The “China shock” literature (Autor, Dorn and Hanson) finds large, localized manufacturing job losses from Chinese imports. Both can be true: national gains from trade and concentrated regional losses. Laffer’s “no American is ever made better off by pulling another down” doesn’t address who bears the adjustment.

Other factual slips, noted but not seeds: “government spending ~3% of GDP in 1910… about the same as today” is right for 1910 federal spending but badly wrong for today (federal ~20%+). To verify.

Discussion

Round 1 (2026-09-24)

Commission pay is Goodhart’s law. Chris on seed 2: “a new number to game :) Goodhart’s law.” When a measure becomes a target, it stops being a good measure.

The transfer theorem: right in principle, Slutsky is overreach. Chris:

“I think he is right in principle on his transfer theorem, but it might be overreach using Slutsky. You can tell from the beginning his main driving idea is ‘incentives’, and I think this is a big part. But he is an academic, so his solutions won’t always match his results; he would have to adjust.”

Chris’s restatement of the theorem: shaving off the top.

“The transfer theorem is the point that you can always reach equality by chopping off the top. His point is that transfers always subtract out of the economy because it takes capital away from highly productive people and gives it to low production people. You would have to take from everyone above the median so they get the median, to those below the median, and this is a loss in efficiency; the median will necessarily lower, rinse/repeat until nothing. Will some people work more? No, not if it gets them nothing, as they are above the median.”

Two moves here: (1) a capital-allocation reading (resources move from high-return to low-return hands), which doesn’t depend on work effort; (2) an iterated version, where each round lowers the median and the next round equalizes to a lower median, until nothing is left. And the answer to the income-effect objection: above the median, extra work yields nothing, so there’s no income to restore.

Market failures: claimed, not proven. Chris:

“The debate about ‘market failures’ was good. These are often claimed, but not proven. Did the market really fail, or was it that you just don’t like the results? Though he does admit that there can be information failure.”

Crypto and profit-and-loss signals. Chris: Laffer ties government monetary policy to crypto, and “this goes back to his incentives: profit and loss must be clear and uninterfered with so the signals are clear.”