The Fiscal Stimulus Illusion — Research Note #2

The Financial View
6 days ago
5 min read

RESEARCH NOTES · SERIES I · GLOBAL MACRO

April 2026 · Original Thesis

Every time an economy slows down, two ideas resurface. The first — covered in Note #1 — is the currency play: weaken the exchange rate, make exports cheap, call it competitiveness policy. The second follows immediately after: deploy a massive fiscal stimulus package, call it investment, and claim it will pay for itself. Both share the same DNA. Both are fast, visible, and politically comfortable. And both have the same fundamental problem: the historical evidence does not support the story politicians sell.

This note makes a precise argument. Not that government spending is always harmful, or that fiscal policy has no legitimate role. The argument is that large-scale undirected fiscal stimulus, deployed as a response to slowdowns without structural reform, consistently underdelivers its promised multiplier, crowds out private investment, adds to long-run debt burdens, and functions primarily as political cover for avoiding the hard work that actually produces durable growth.

I. The Multiplier Problem

The intellectual case for fiscal stimulus depends on the fiscal multiplier: for every dollar the government spends, GDP rises by more than one dollar. IMF research in the early 2010s — frequently cited by stimulus advocates — estimated multipliers of 1.5x to 1.8x during downturns. The textbook mechanism was compelling enough that it survived repeated scrutiny.

The problem arrived when empirical evidence was collected after the policy was deployed at scale. A comprehensive review of OECD economies covering 1970 to 2014 found average spending multipliers below 1.0 in normal conditions. Post-2008 IMF follow-up research found that earlier estimates had been systematically overstated. More damaging: NBER analysis found that a substantial share of stimulus spending leaked immediately into imports, debt repayment, and asset price inflation rather than circulating through the productive economy. The mechanism worked in the model. In practice, the plumbing was different.

II. Japan's Thirty-Year Test

If you want a clean long-run empirical test of fiscal stimulus as policy, Japan is the experiment that runs itself. Between 1990 and 2020, Japan deployed approximately 30 major fiscal stimulus packages. Government gross debt rose from 65% of GDP in 1990 to over 260% by 2023 — the highest among all developed economies. Real GDP growth over the same period averaged approximately 0.8% per year.

The Japan case is uniquely damning not because a single large package failed, but because the repeated application of the same logic — growth disappoints, so spend more — produced precisely the outcome fiscal stimulus was supposed to prevent: stagnation, rising debt, and an economy structurally dependent on government support rather than private-sector productivity. The zombie company phenomenon — firms sustained by fiscal transfer and cheap credit rather than genuine competitiveness — is the supply-side legacy of three decades of demand-side stimulus.

Abenomics crystallised the pattern. The yen depreciated roughly 22% against the dollar between late 2012 and mid-2013. Trade deficits continued to grow for two years after the depreciation. Export volumes recovered far less than models predicted. What rose was corporate margins and equity prices — a windfall for shareholders, not a structural improvement in the economy's competitive position.

Japan's fiscal stimulus programmes were vast in scale and persistent in application. The outcome was not growth. It was the institutionalisation of the conditions that made growth impossible.

III. The COVID Natural Experiment

The 2020-2022 period offered the most compressed cross-country fiscal experiment in modern history. Four major economies deployed unprecedented stimulus in the same timeframe, using broadly similar policy instruments. The comparative data is instructive.

The United States deployed approximately 27% of GDP in combined fiscal support. Real GDP recovered to pre-pandemic levels by Q2 2021. That recovery was real. But it was accompanied by inflation of 9.1% by June 2022 — the highest in four decades — which eroded real wage gains and transferred wealth from wage-earners to asset-holders.

The United Kingdom deployed approximately 19% of GDP. By 2023, UK real wages remained below 2021 levels in inflation-adjusted terms — meaning the stimulus that was meant to protect household living standards had, via the inflation it generated, actively undermined them. Peak inflation: 11.1% — the highest in the G7.

Japan, notably, deployed substantially less fiscal support per capita than either the US or UK. Its real GDP recovery was not materially worse. The implied marginal fiscal multiplier — the additional growth generated by the incremental spending — appears remarkably small.

IV. The Real Purpose of Fiscal Stimulus

This is the part of the analysis that most commentary avoids, but it is the most analytically important. Governments do not deploy large fiscal stimulus packages because the empirical multiplier evidence tells them to. They deploy them because they are politically optimal: immediate in effect, visible in delivery, and diffuse in cost.

The inflation bill from a large stimulus arrives slowly. The debt stock accumulates quietly on a balance sheet most voters never examine. The productivity investment that was not made — because public debt crowded out private capital — is a counterfactual. The labour market reform that was postponed because the stimulus bought another year of tolerable unemployment figures — that cost never has a line item.

What stimulus consistently delays is the structural work: labour market reform, regulatory rationalisation, skills investment, export sector development, pension sustainability. These produce genuine long-run growth. They are also politically costly, slow-moving, and resistant to deadline-driven announcement cycles. Stimulus is their substitute — not their complement.

V. When It Has Worked

The historical record is not uniformly negative, and intellectual honesty requires acknowledging the exceptions. Post-WWII infrastructure investment in the United States built genuine productivity-enhancing capacity while private demand was organically recovering. South Korea and Taiwan's state-directed industrial investment in the 1970s and 1980s was conditional on export competitiveness — firms that did not perform globally lost support. Germany's post-reunification fiscal programme produced lasting convergence, but required twenty years and came paired with significant labour market reform.

The pattern in successful cases is consistent: targeted, conditional, pairing public investment with real structural reform rather than substituting for it. That profile — targeted and conditional — describes almost none of the large-scale stimulus packages deployed by advanced economies since 2008.

Conclusion: Read the Signal Behind the Spending

When a government announces a major new fiscal stimulus package, the instinct is to ask how large it is. The more useful analytical question is: what structural problem is this spending designed to avoid?

When the announcement comes with multiplier projections of 1.5x or higher, long-term debt sustainability assurances, and promises that this time is different — look for what reform was not announced on the same day. That omission is usually the more important signal.

The real message of large fiscal stimulus is almost never 'growth is coming.' The real message, if you read the underlying incentive structure, is that structural reform has been deferred. The spending is the signal. What is absent from the spending is the story.

That is the lens that produces useful analysis. Not the size of the package. But what the package is a substitute for.