Larry Summers on Macroeconomics, Mentorship, and Avoiding Complacency
Larry Summers — Harvard professor and former US Treasury Secretary — joins Tyler Cowen in Ep. 28 to range across secular stagnation, the Federal Reserve’s inflation misjudgements, labour unions, monopoly, China’s debt, and what makes a person keep improving well into their sixties.
Key ideas
- Secular stagnation is structural, not cyclical. Summers argues that persistently low real interest rates reflect a deep mismatch: savings propensity is high and investment propensity is low, so the interest rate that would produce full employment — the neutral rate — has fallen dramatically. Historical policy settings that once looked expansionary are now close to neutral, and the Fed has been slow to grasp this.
- The Fed’s 2 per cent inflation target is not genuinely symmetric. If the Fed truly treated 2 per cent as a symmetric midpoint, it would have been above target roughly as often as below; instead, not a single Federal Open Market Committee dot has ever been projected above 2 per cent since the financial crisis. Summers reads this as a disconnect between stated policy and actual practice that explains the persistent undershoot.
- Labour unions do more than raise wages. Pressed by Cowen on whether a union wage premium of 7–8 per cent justifies advocacy, Summers points to non-wage benefits: lower staff turnover, more respectful grievance procedures, better workplace safety, and broader political support for social insurance. The playing field for unionisation is also not level — the cost-benefit for employers of firing an organiser is currently too attractive.
- Capital income tax rates should be closer to ordinary rates, not zero. The theoretical case for zero capital taxation rested on an infinitely elastic savings supply; that premise has since been empirically refuted and partly shown to rest on mathematical errors. A large share of capital income reflects rents, and wealth concentration means the revenue is not neutral in distributional terms.
- Avoiding complacency drives continuous improvement. Summers attributes his ongoing intellectual development to two practices: surrounding himself with exceptionally able, younger collaborators (from whom he learns at least as much as they do from him), and cultivating a settled disposition against satisfaction — reviewing every answer he gave and asking how it could have been better.
Content
Secular stagnation and the neutral interest rate
Secular stagnation — the idea that advanced economies face a structural shortfall in aggregate demand (overall spending) that keeps growth below potential even at near-zero interest rates — is Summers’s central macro diagnosis. The mechanism runs as follows: a high savings propensity combined with weak investment appetite means the real interest rate (the rate after stripping out inflation) needed to balance the economy has fallen, perhaps to negative territory. Central banks can only reduce rates so far; once they hit zero, the usual lever breaks. Summers contends the Federal Reserve has underestimated how far the neutral rate has shifted, and has therefore tightened policy — raised rates — sooner than warranted, producing chronic below-target inflation.
He cites structural drivers that are likely to persist: rising life expectancy pushing up precautionary saving; increased economic insecurity; greater inequality concentrating income in hands with a lower propensity to spend; and cheaper capital goods (his iPhone example — you can buy an enormous amount of computing power for very little money), which compress investment in dollar terms even when investment in real terms is decent.
The Fed’s inflation target in practice
Summers distinguishes between the 2 per cent target as a ceiling — the way the Fed in practice appears to treat it — and as a symmetric midpoint, the way the Fed officially describes it. A genuinely symmetric target would imply willingness to run above 2 per cent in a strong economy to compensate for years spent below it; price-level targeting (think of it as a target that keeps a running tally) or nominal GDP targeting would achieve this mechanically. Yet through nine years of recovery, with unemployment at 4.3 per cent, no FOMC dot projected above 2 per cent has ever been published. Summers’s instinct is towards genuine symmetry and, tentatively, nominal GDP targeting — a framework under which slower growth automatically tolerates higher inflation.
Labour unions beyond the wage premium
Cowen puts the empirical case against unions sceptically: the union wage premium has fallen to perhaps 7–8 per cent, and after accounting for higher prices and reduced demand for labour, the net gain may be 3–4 per cent. Summers accepts the arithmetic but widens the frame. Union workplaces show lower employee turnover (a direct measure of job satisfaction), better safety records, and more structured channels for addressing discrimination complaints. Separate from these direct benefits, stronger unions have historically provided political support for Social Security and healthcare — public goods that raise welfare broadly.
He also challenges the premise of a level playing field. Where an employer fires an organiser, the remedy is back pay minus interim earnings, recoverable only after years of litigation — an attractive cost-benefit for the employer. His analogy: he does not know exactly how much weight he should lose, but he is confident about the direction. He is similarly confident that more bargaining power for workers is the right direction, even if the optimal magnitude is uncertain.
Capital taxation and the failure of infinite-elasticity models
The 1980s theoretical literature — some of it from Summers himself — generated results suggesting the optimal long-run tax on capital income is zero: in a steady state, capital should accumulate until even wages are higher, and taxing the return on capital merely reduces that accumulation. Summers now repudiates the empirical premise: the savings rate does not, in practice, respond strongly to the rate of return. Real interest rates have varied substantially without producing the predicted swings in savings behaviour. He adds that some of the original mathematical results were subsequently shown to be wrong.
On practical grounds: much capital income reflects monopoly profits (rents rather than returns to genuine productive risk-taking); realised capital gains are already under-taxed, and unrealised gains often escape tax entirely at death; and the income is highly concentrated. His conclusion is that a comprehensive income tax — one that taxes capital and labour at similar rates — is a better approximation of the right policy than a zero-capital-tax regime.
Monopoly, market structure, and the limits of antitrust
Summers is genuinely uncertain about monopoly’s economic significance, and says so. Higher concentration ratios and profit shares, combined with low investment and low interest rates, are consistent with increased market power — the picture fits. But he is wary of the inference. Apple Pay makes Apple larger while simultaneously subjecting the financial industry to outside competition; whether that is more or less competitive overall is not obvious.
He identifies two selective areas of concern: hospital and healthcare-network consolidation in regional markets, where the move from several competitors to one dominant provider is concrete and harmful; and information-intensive platforms (Facebook, Google), where the issues are privacy and network effects rather than classic price-fixing. He is not sure antitrust is the right tool for the latter: the products are given away for free, which makes standard consumer-welfare analysis awkward. His tentative view is that antitrust enforcement has more likely been too lax than too tight over the past decade, without endorsing the stronger claims one hears about a new Standard Oil era.
China’s debt and the airport analogy
Asked whether China is approaching a discrete financial crisis, Summers argues that much of Chinese debt is explicitly or implicitly government guaranteed, and that a rapidly growing government’s fiscal capacity is large. Western analysis that treats Chinese corporate and local-government debt as analytically similar to private Western debt probably overstates the crisis risk.
He offers what he calls the airport analogy: building infrastructure ahead of demand looks foolish if growth disappoints and visionary if growth materialises. Dulles Airport, built in a relative wilderness, was derided for a decade before Northern Virginia’s technology cluster vindicated it. China’s infrastructure programme is subject to the same retrospective ambiguity. Summers can readily imagine a smooth deceleration rather than a financial event, while not dismissing the possibility of a 1997-style crisis as negligible.
Mexico and the middle trap
Mexico, Summers agrees, has been a persistent puzzle: sound macroeconomic management and NAFTA membership since 1995, yet growth stuck at 2–2.5 per cent. He points to two factors. First, rule-of-law and security problems — the risk of an evolution towards Colombia-style instability — are probably underestimated by outside observers. Second, Mexico is caught in what he describes as ‘the same problem as Michigan’s’: neither able to leverage the global economy (like the winners who build cross-border supply chains and sell globally) nor simple enough to be pulled along by it (like the hundreds of millions lifted out of poverty in China or India). Mexico and similar middle-income countries can neither compete on Chinese labour costs nor command the premium that advanced manufacturing commands.
Mentoring as mutual collaboration
Summers reframes mentoring: he does not think of himself as a generous patron conferring wisdom on the young, but as someone who has consistently sought out the ablest people he could find to work with, and has got an enormous amount out of every relationship. Sheryl Sandberg made him a better Treasury Secretary. Alan Krueger, Ed Balls, Natasha Sarin — in each case the relationship has been reciprocal. His method is direct: he tells the truth about what was done better and what was done worse, and expects the same in return.
The production function against complacency
Cowen puts it to Summers that his answers at 62 are in some ways richer than they were at 52, and asks what explains that. Summers offers two answers. The first is the habit of surrounding himself with outstanding young collaborators who ask questions that keep him on his toes. The second is a standing disposition against complacency: on the plane home after any interview or talk, he reviews which questions he could have answered better and what he would say next time. The refusal to be satisfied — with himself, with institutions, with any answer that is merely adequate — is his description of the engine.
Related
- Larry Summers — speaker
- Tyler Cowen — host
- Joe Studwell — fellow Conversations with Tyler guest; development economics conversation that extends some of the secular-stagnation themes into emerging markets