Steve Fazzari on Stimulus and Keynes

Episode Summary

Executive Summary: Russ Roberts and Steve Fazzari debate the 2009 stimulus, Keynesian aggregate-demand theory, and why the recession’s severity was so hard to diagnose in real time. They agree the crisis exposed limits of forecasting and macroeconomics, while Fazzari defends stimulus as helpful but insufficient and Roberts presses on measurement, allocation, and the difficulty of separating true effects from counterfactuals.

Main Topics: Assessment of the 2009 stimulus (Priority: 5/5): Fazzari argues the stimulus had the intended directional effect and likely prevented a deeper, more persistent downturn, but was too small and was oversold politically as a quick fix. How to know whether stimulus worked (Priority: 5/5): Both speakers emphasize the counterfactual problem: the central difficulty is proving what would have happened without the policy, making ex post evaluation highly uncertain. Keynesian aggregate demand and stimulus logic (Priority: 5/5): They walk through the textbook Keynesian case that insufficient demand can cause involuntary unemployment, and that demand can be raised through money transfers or government spending. Allocation versus aggregate demand (Priority: 4/5): Roberts questions whether broad stimulus is misdirected when unemployment is concentrated in sectors and regions; Fazzari agrees targeting matters but says spillovers and flexibility can still make non-targeted stimulus useful. Labor-market frictions and unemployment (Priority: 4/5): The discussion turns to why unemployment behaves differently across groups and over time, highlighting mismatch, sectoral shifts, imperfect information, and the difficulty of understanding job creation/destruction dynamics. Limits of economics as a science (Priority: 4/5): Both suggest macroeconomics often resembles historical judgment more than hard science, because models are underdetermined by the evidence and economists must weigh theory, data, and narrative. Post-crisis changes in macroeconomics (Priority: 4/5): Fazzari expects more attention to financial instability, moral hazard, uncertainty, and the interface between finance and macro, as the crisis exposed blind spots in pre-2008 models.

Key Arguments: Fazzari contends the stimulus likely reduced the depth of the recession and prevented an even worse outcome, even though it did not restore robust growth quickly. Roberts argues that the main defense of stimulus—"things were worse than we thought"—is difficult to verify independently and risks circular reasoning. Both agree that forecasts from 2008–2009 were badly wrong, suggesting the economics profession underestimated the crisis’s severity. Fazzari says Keynesian logic does not require demand to come from private buyers; government spending can also mobilize idle resources. Roberts distinguishes between spending that creates value and spending merely to boost nominal demand, stressing that allocation matters and that some stimulus may raise prices rather than output. Fazzari concedes that targeted stimulus would be better in theory, but notes political constraints and spillovers make exact targeting difficult. Both stress that labor markets are highly dynamic: millions of jobs are created and destroyed each period, so net job numbers hide substantial churn. Fazzari predicts macroeconomics will place more weight on financial instability, behavioral limits, and the interaction between finance and macro after the crisis.

Data Points: Stimulus package size: $700+ billion initially; later discussed as about $800-$862 billion - Size of the 2009 fiscal stimulus package under debate Predicted unemployment without stimulus: 8.5% - Forecast cited as the administration’s expected peak unemployment rate absent stimulus Observed unemployment: over 10% - Actual unemployment rate exceeded forecasts after the stimulus and crisis Unemployment among college-educated workers: about 5% - Roberts references differential labor-market outcomes by education level Unemployment among non-college workers: about 15% - Roberts contrasts this with the college-educated rate to show uneven labor-market damage Net job creation in one recent month: 103,000 - Used to illustrate weak labor-market recovery despite ongoing job churn Time horizon referenced for recovery expectations: within a year or so - Fazzari describes the optimism of policymakers that the economy would rebound quickly after stimulus

Pivotal Quotes: "what is the counterfactual?" — Steve Fazzari: Explaining why evaluating stimulus is inherently difficult in economics "I think there will be more interest in these issues, and I hope there will be." — Steve Fazzari: On how the financial crisis will reshape macroeconomics and the profession "the economy needed in the past couple of years and still needs more demand stimulus." — Steve Fazzari: Fazzari’s basic policy conclusion about the recession "I think we're more in the world of the historian than the scientific researchers." — Russ Roberts: Describing the epistemic limits of macroeconomic debate and policy evaluation

Implications: The episode suggests future macroeconomics will be more cautious, more finance-aware, and more humble about prediction. For policymakers, it underscores that stimulus may help, but design, targeting, and uncertainty about effects remain major challenges.