Notes — Scott Sumner on Monetary Rules, Blooming Late, and the Death of Cinema

Notes on Scott Sumner in conversation with Tyler Cowen — Conversations with Tyler (https://conversationswithtyler.com/episodes/scott-sumner/), 8 January 2025.

Four questions [Adler frame]

Q1 — What is it about as a whole? Monetary economist Scott Sumner, originator of Market Monetarism and the world’s leading advocate of nominal-GDP-level targeting, ranges across the intellectual history of fiat currency, the mechanics and limits of monetary rules, why the 2008 financial crisis is best read as a nominal-GDP collapse rather than a credit-market collapse, China’s and Japan’s tolerance of deflation, his own maxim ‘never reason from a price change’, and a long closing stretch on cinema — before turning to his own biography as a deliberate late bloomer who did almost all his research after being denied tenure.

Q2 — How is it argued? Almost entirely through historical case comparison rather than formal modelling. Sumner repeatedly sets contrasting national episodes side by side — the US and Sweden in the Great Depression, the US and Europe in 2008, Argentina’s currency board versus post-2001 devaluation, Japan under Abe versus pre-Abe Japan, Hong Kong’s dollar peg — and reads the pattern of nominal shocks preceding real ones as evidence for a causal story running from monetary policy to the business cycle. Cowen’s method is Socratic pressure: he repeatedly proposes an alternative causal story (Congress constraining the Fed; credit-market shocks as autonomous; the tautology risk in Sumner’s own reasoning) and asks Sumner to defend against it in real time.

Q3 — Is it true, in whole or part? Sumner’s narrower historical claims — the timing of the US banking crisis relative to the onset of the Depression, the European-versus-US monetary tightness comparison in 2008 and 2011, the Hong Kong dollar-peg correlation — are testable and, on the evidence he cites, hold up. [?] His larger causal claim, that nominal GDP shocks are the primary driver of real output shocks and that credit-market distress is mostly a downstream symptom, is contested territory: Cowen presses exactly this point [§ On the causes and response to the 2008 financial crisis], suggesting Sumner’s account risks near-tautology given how tightly nominal and real GDP move together in economies with sticky wages, and Cowen states outright he thinks credit-market shocks carry independent causal force Sumner discounts too readily. Sumner’s own answer — that the US/Zimbabwe contrast shows nominal and real variables can and do move independently, so the US correlation is informative rather than definitional — is a reasonable rebuttal but does not close the question. [?]

Q4 — What of it? The episode is the wiki’s first sustained, source-level treatment of Market Monetarism from its originator, and gives the clearest available statement of the ‘never reason from a price change’ heuristic and its lineage back to Sumner’s 1989 JPE paper on real-wage cyclicality — decades before he coined the phrase. It also documents an unusual academic career (denied tenure, did nearly all his research afterward, found his largest audience blogging in his 50s) that complicates any simple story about when intellectual contribution peaks.

Glossary

Market Monetarism — the school of monetary thought, originated by Sumner from his blog The Money Illusion after 2008, holding that a central bank should target the level (not merely the growth rate) of nominal GDP rather than an inflation rate or a money-supply growth rule, and that markets — not the central bank’s own internal models — are the best available forecaster of whether policy is on track. [§ On NGDP targeting]

NGDP (nominal GDP) targeting — the specific policy rule Market Monetarism recommends: stabilise the path of nominal GDP (real output times the price level) rather than inflation alone, on the view that most business-cycle pain comes from nominal GDP deviating from its expected growth path, not from the price level moving on its own. [§ On NGDP targeting]

‘Never reason from a price change’ — Sumner’s maxim that the economic implication of a price movement (an interest rate, an exchange rate, or an inflation reading) cannot be inferred from the movement alone; the same price change means opposite things depending on whether it is driven by a demand shock or a supply shock. [§ On never reasoning from price changes]

Rule-like policy (versus a strict rule) — Sumner’s preferred way of framing the rules-versus-discretion debate: not a binary choice, but a spectrum along which a central bank’s behaviour can be made more predictable and consistent over time even without ever becoming a perfectly automatic formula. [§ On NGDP targeting]

Symmetallism — a historical monetary-standard variant in which the unit of account is defined as a weighted basket of two commodities (e.g. gold and silver) rather than one; Sumner frames it as a midpoint on a continuum running from the gold standard through bimetallism to full inflation targeting, which stabilises the average of all prices in the economy. [§ On the causes and response to the 2008 financial crisis]

Key claims by section

Reading the New York Times through the Depression [§ On when in history fiat currency became viable]

Sumner spent years reading contemporaneous 1920s–30s New York Times coverage rather than retrospective history, and found real-time observers had no sense of how transformative the rise of the Nazi Party would prove — experts at the time expected Hitler to moderate as he neared power.
Fiat currency became politically viable only once the memory of post-WWI hyperinflation faded enough for governments to risk abandoning gold; Argentina’s four painful years under its currency board in the early 2000s, before eventually devaluing, is Sumner’s modern analogy for this reluctance.
Sumner speculates fiat money might have been technically workable even in the late nineteenth century, but was not yet conceived of as a responsible option — an inertia he compares to Milton Friedman’s proposal for floating exchange rates, viewed as fringe when first made and standard within a couple of decades.

Monetary policy as more-or-less rule-like, not rule-versus-discretion [§ On NGDP targeting]

Sumner rejects the binary rules-versus-discretion framing (raised via Carl Schmitt’s dictum that ‘the sovereign is he who decides the exception’): US inflation held near 2 percent for roughly 30 years is evidence of more rule-like Fed behaviour than the prior 30 years, without ever becoming a perfectly automatic rule.
Against Cowen’s suggestion that the Fed uses discretion partly to keep an interventionist Congress at bay, Sumner argues the Fed’s 2020–21 overshoot reflected a genuine belief — informed by a decade of below-target inflation after 2008 — that more aggressive ‘makeup’ policy was needed, not pressure from Congress.
Sumner distinguishes traditional stabilisation policy, where he judges the Fed mostly apolitical, from areas like bank regulation, where political constraints visibly shape decisions.

The 2008 crisis as a nominal-GDP story [§ On the causes and response to the 2008 financial crisis]

Sumner disputes Brad DeLong’s view that NGDP-level targeting implies bailing out firms like General Motors; he argues stabilising nominal GDP resolves most structural problems on its own, leaving individual failures (a poorly managed bank, a single firm) for other policymakers to address case by case.
His account of the causal order: in 1929 and in Argentina in the early 2000s, the monetary-policy mistake preceded the financial crisis, which then worsened as falling nominal GDP crushed the real value of fixed nominal debt contracts — an analogy Sumner draws to a cold turning into pneumonia. In 2008, financial distress hit first, but he still attributes the depth of the ensuing recession to the Fed then mismanaging the equilibrium interest rate and letting nominal GDP fall.
Europe’s 2008–09 recession, and its 2011 double dip after two rate hikes into weak nominal growth, were worse than the US’s despite the US having the original subprime shock — evidence, in Sumner’s reading, that monetary tightness (not the initial credit shock) set the depth of each recession.
Cowen challenges this as close to tautological given how tightly real and nominal GDP move together under sticky wages, and argues credit-market shocks carry independent force; Sumner’s rebuttal cites the US/Zimbabwe contrast (simultaneous recession and hyperinflation) as evidence nominal and real variables are not definitionally linked. [?]

China’s and Japan’s tolerated deflation [§ On China’s monetary policy]

China’s persistent deflation, like its similar deflation in the late 1990s/early 2000s, follows from prioritising a strong exchange rate over domestic macroeconomic stability — the same mechanism, in miniature, as Hong Kong’s 40-year dollar peg, whose business cycle tracks the strength of the US dollar almost exactly.
Sumner recalls US Treasury officials pressuring Japan in the early 2000s not to devalue the yen, under an implicit threat of protectionism — even as American economists urged Japan to devalue to escape its liquidity trap — and speculates similar external pressure may partly explain China’s reluctance today.
Abe’s 2012 campaign on a platform of higher inflation, followed by genuine improvement in Japan’s economy, is Sumner’s key evidence that apparently ‘structural’ deflationary stagnation is often actually a correctable monetary-policy failure — a ‘mistake theory’ of policy he argues applies unusually well to monetary economics specifically, versus areas like tariffs that are better explained by special-interest politics.

‘Never reason from a price change’ [§ On never reasoning from price changes]

Sumner traces the idea’s origin to a 1989 Journal of Political Economy paper he wrote with Steve Silver on real-wage cyclicality, criticising prior literature for reasoning directly from price-level changes rather than asking what was driving them — long before he coined the phrase itself, roughly ten to fifteen years ago.
The generalised rule: never infer the economic meaning of an interest-rate move, an exchange-rate move, or an inflation reading from the move alone, because the same directional change implies opposite things depending on the underlying driver (e.g. the liquidity effect versus the Fisher effect for interest rates).
Sumner names the confusion between low interest rates and ‘easy money’ as a recurring, still-unresolved case of the fallacy — highlighted by the ongoing Keynesian/Neo-Fisherian dispute over whether low rates signal loose or tight policy.

Cinema, taste, and the limits of receptivity [§ On movies]

Sumner reads film primarily as a visual medium — cinematography and directorial visual signature over screenplay — which he offers as the reason cinephiles tend to prefer film over television, where the writing dominates.
He and Cowen agree Hitchcock’s Vertigo is his most personal and ‘deep’ film, versus lighter entertainments like North by Northwest; Sumner names the late-1990s Taiwanese New Wave as the cinematic equivalent of what the French New Wave was for an earlier generation, and doubts he will recapture that intensity of connection with newer films by directors he judges equally talented.
Citing Susan Sontag, Sumner frames the apparent decline of masterpiece-producing eras (Dutch Golden Age painting 1600–75; Coppola’s four-film run in the 1970s) as an open question between two explanations: fewer masterpieces are being made, or ageing audiences become less receptive to them — and credits deliberately cultivated patience with letting him appreciate ‘quiet’ directors like Ozu that would have bored him when young.

A deliberate late bloomer [§ On the Scott Sumner origin story]

Sumner chose the University of Chicago PhD programme for its free-market economics and its emphasis on economic intuition over the more technical, mathematical training then associated with MIT; he was almost flunked out under Chicago’s contentious seminar culture and did his dissertation, under Robert Lucas, on currency hoarding — an unusual topic for a Lucas student.
He was denied tenure at Bentley for insufficient publications, reapplied after quickly producing a few papers (including the 1989 JPE piece), and did the great majority of his research — including the Great Depression book he considers his one substantial academic contribution — after receiving tenure, not before.
He attributes his unconventional, late-blooming career to personality rather than external obstruction, and located his largest audience only in his 50s and 60s through blogging, roughly fifteen years before this conversation.