George Selgin on the New Deal, Regime Uncertainty, and What Really Ended the Great Depression

George Selgin, the monetary economist behind False Dawn: The New Deal and the Promise of Recovery, 1933–1947, walks Tyler Cowen through a policy-by-policy audit of the New Deal, recorded live at the Cato Institute. His argument cuts against both left and right memory of the period: the New Deal was never really a fiscal or monetary stimulus programme, the 1933 gold revaluation did the real work of reflating the economy, and the recovery it started was then throttled by price-fixing schemes and the corrosive effect of not knowing what Washington would do next.

Key ideas

1. Gold revaluation, not New Deal stimulus, drove the early recovery. The 7–8% manufacturing growth rates of 1933–34 came from firms front-running the incoming NRA price controls, ordinary bounce-back after the banking crisis, and gold flowing in from Europe — first from devaluation, later from fear of Hitler and war. Roosevelt’s own gold-purchase scheme, based on George Warren’s now-discredited theory, achieved almost nothing; Keynes said so publicly at the time and was, in Selgin’s account, largely right.
2. Regime uncertainty, not any single bad policy, strangled investment. Selgin credits economic historian Robert Higgs with the term: businesses cannot price the risk of an investment if they cannot even guess what rules will govern it. Unlike an anticipated bad policy (Selgin rates Smoot-Hawley a small factor, deferring to Douglas Irwin’s research), this uncertainty specifically hit investment spending, which had all but collapsed for the whole decade.
3. A mixed scorecard, not a verdict on ‘the New Deal’. Selgin rates individual measures independently: Glass-Steagall’s bank/investment-bank split irrelevant to recovery; the Agricultural Adjustment Act net harmful (paying farmers to destroy output); the RFC’s early lending programme harmful but its later bank-recapitalisation programme genuinely stabilising; deposit insurance modestly positive but overrated relative to other stabilising factors; the 1935 Federal Reserve reorganisation a wasted opportunity under a monetarily passive Marriner Eccles.
4. The 1937–38 downturn was a coordination failure, not one mistake. The Fed doubled reserve requirements over three steps to head off feared inflation from the same gold inflows that had been driving recovery, while the Treasury simultaneously sterilised those inflows and the administration pursued fiscal retrenchment to balance the budget — three independent tightenings landing at once.
5. Keynes’s direct advice to Roosevelt was sound, even where his general theory wasn’t. Keynes urged Roosevelt to stop attacking business, prioritise recovery over reform, and abandon both the gold-purchase scheme and NRA price controls — advice Selgin judges correct, distinct from more dubious Keynesian positions elsewhere (protectionism, nationalising investment). Selgin’s institutional reform preference is a rules-based, NGDP-targeting ‘night watchman Fed’, paired with free banking under a fiat standard rather than any return to gold.

Content

What the New Deal did and didn’t accomplish

Selgin opens by rejecting the popular one-line summary of the New Deal — that it was a stimulus programme that got America spending its way out of depression. In his account, the New Deal ‘made very little use of fiscal or monetary stimulus’. The genuine early-recovery burst in manufacturing output (7–8% growth rates through 1933–34) had three unglamorous causes: manufacturers rushing to buy inputs and build inventory ahead of the incoming National Recovery Administration’s price controls, ordinary bounce-back once the banking crisis stabilised, and gold rushing in from Europe — initially chasing the dollar’s devaluation, later fleeing fears of Hitler and approaching war. None of it reflects the New Deal’s own programmes working as designed.

Roosevelt’s actual monetary intervention — a months-long gold-purchase programme built on farm economist George Warren’s theory that raising the price of gold would drag other prices up with it — achieved almost nothing; general prices barely moved. Selgin credits Keynes with publicly ridiculing the scheme at the time, and judges history vindicated him. What Selgin argues should have happened instead was a swift, decisive devaluation of the dollar the moment gold payments were suspended — not months of price-tinkering before finally settling on a rate.

A policy-by-policy scorecard

Cowen runs Selgin through a rapid-fire audit of named New Deal measures, each rated for its effect on recovery:

• Glass-Steagall (1932), in its famous bank/investment-bank separation sense, gets ‘totally irrelevant’ — mixing commercial and investment banking was never a cause of the Depression, so ending it could not have helped end it.
• The Reconstruction Finance Corporation mostly hurt during its early lending phase — Congress forced it to disclose which banks it lent to, which made banks reluctant to apply — but its later recapitalisation programme, buying bank shares outright rather than lending to them, genuinely stabilised the banking system and gets credited as one of the RFC’s few successes within an otherwise oversized, mixed-record agency.
• The Agricultural Adjustment Act is rated straightforwardly harmful: it paid farmers to destroy crops and livestock to raise prices through supply reduction, a policy whose aggregate-demand rationale (higher farm incomes spent by high-marginal-propensity-to-consume farmers) was ‘rather doubtful’ even on its own terms, and whose measured effect on recovery is, per the literature Selgin cites, small to none.
• The Banking Act of 1933’s deposit-insurance provision is rated net positive but overrated. Selgin notes the political irony that Roosevelt actually opposed deposit insurance until the eleventh hour, signed only because a veto would likely have been overridden, and then took public credit for it — while its real contribution to reopening the banking system was real but smaller than commonly assumed, alongside other stabilising factors.
• The Banking Act of 1935, which concentrated Federal Reserve authority in Washington and created the modern role of Fed chair, is judged a wasted opportunity: new chair Marriner Eccles was personally a fiscal-stimulus advocate but ran a flatline monetary policy, so centralising control changed nothing about the actual stance taken.

Regime uncertainty as the recovery’s real brake

Asked whether Smoot-Hawley mattered much as a cause of the Depression, Selgin defers to Douglas Irwin’s finding that it was a small factor — and uses that to sharpen a distinction. An anticipated bad policy, however severe, is still a known quantity that can be priced. Regime uncertainty is categorically different: investors cannot judge whether they will keep the returns on an investment if they cannot even guess what the rules governing it will be. Because the Depression’s core problem was a near-total collapse in net investment spending — not consumption, which Selgin calls comparatively easy to revive — anything that specifically discourages investment, regime uncertainty above all, becomes a first-order obstacle to recovery. Selgin regards this as the mechanism behind why the New Deal’s genuine early gains stalled rather than compounded.

The 1937–38 ‘second Depression’ gets a similarly precise diagnosis: not one policy mistake but an unplanned convergence of several. Fearing inflation from the very gold inflows that had powered the 1933–36 recovery, the Fed doubled bank reserve requirements over three steps; simultaneously, the Treasury began ‘sterilising’ new gold inflows so they no longer expanded bank reserves; and, believing recovery was nearly complete, the administration pursued fiscal retrenchment to balance the budget. Three independent tightenings — monetary from the Fed, monetary from the Treasury, and fiscal — landed together, and the downturn that followed was severe precisely because none of the actors involved was coordinating with the others.

Keynes’s sound advice and the Austrians’ inconsistency

Selgin deliberately picks Keynes as the economist whose specific advice to Roosevelt holds up best from the period — while distinguishing that from Keynes’s more dubious general-theoretical positions (his writing on protectionism, his call to nationalise investment). What Keynes told Roosevelt directly, in letters and in their one meeting, Selgin judges essentially correct: stop attacking business, since regime uncertainty about Roosevelt’s own hostility toward businessmen was itself discouraging investment; prioritise recovery over structural reform, since reform unsettled the same confidence recovery depended on; and abandon both the gold-purchase scheme and the NRA’s price controls, both of which Keynes criticised sharply at the time.

By contrast, Selgin judges the Austrian economists — Hayek specifically — inconsistent between theory and practice. Hayek’s own model implied that a central bank must stabilise aggregate spending (or restore it) once it has collapsed; yet Hayek recommended gold-standard austerity and, in Britain, appeared willing to let the downturn discipline trade unions rather than press to end it. The early Chicago School (Simons, Viner) gets a mixed grade: they backed public-works stimulus well ahead of Keynes, but Henry Simons’s proposal to abolish fractional-reserve banking altogether is judged a misdiagnosis — Canada kept fractional-reserve banking throughout the period without a banking collapse, which Selgin attributes to better regulation rather than to the absence of fractional reserves.

On the charge, from John Flynn’s As We Go Marching, that the New Deal brought semi-fascist planning to America, Selgin answers comparatively rather than absolutely: Roosevelt looked moderate set against genuine home-grown authoritarian populists of the period — Huey Long, Francis Townsend — whose rising popularity he was actively trying to defuse by co-opting some of their appeal. Judged against an idealised classical-liberal baseline, Selgin concedes, Roosevelt looks illiberal; judged against the realistic political alternatives on offer, he looks like a compromise.

Banking structure, the quantity theory, and reforming the Fed

On why America’s historically fragmented, small-bank system nonetheless produced the world’s deepest capital markets, Selgin distinguishes two different things that can undermine banking stability: restricting what banks are free to do (the pre-1933 problem — no branching, limited diversification) versus introducing deposit guarantees that create moral hazard once banks are freed to take more risk (the post-1933 problem). He doubts a genuinely free banking system — no bailouts, no entry restrictions — would consolidate into a Canadian-style handful of giant banks; Canada’s concentration, in his account, came from strict entry restrictions rather than from branch-banking freedom itself.

On the quantity theory of money, Selgin calls it ‘badly misunderstood’: the claim that inflation tracks the growth rate of a monetary aggregate like M2 held up loosely across countries when he taught it in the 1980s and 1990s, but depends on a stable velocity of money that financial and regulatory innovation keeps disturbing — and ‘all bets are off’ since 2008, once the monetary base decoupled from the overall ease of credit.

His own reform programme follows from these premises. A return to the gold standard is impossible, in his view, because a government that has already broken a gold-convertibility promise once — every government that has ever left a gold standard — cannot credibly recommit not to break it again. Instead he favours free banking under a fiat standard, with the monetary base managed by a fixed rule: a ‘night watchman Fed’, literally a computer targeting stable NGDP growth rather than a discretionary committee. Asked about proposals to fold the Federal Reserve under direct executive control, Selgin warns this would tend toward ‘fiscal dominance’ — looser money to accommodate government spending without raising taxes — and would rather keep pressing for NGDP targeting within an independent Fed than risk that outcome.

See also

• George Selgin — the speaker
• Tyler Cowen — the host
• Monetarism — Friedman and Schwartz’s account of the Depression’s monetary cause; Selgin’s account of what did and didn’t end the recovery is complementary, not contradictory
• Regime Uncertainty — the episode’s organising concept, drawn out in full
• Knightian Uncertainty — a related but distinct risk/uncertainty distinction