Here’s how fiscal prudence could foster India’s emergence as an innovation-driven economy

He died in 1971, the year India ran an enormous fiscal deficit to finance the Bangladesh war. The two are unconnected. But the question his interlocutors were really asking was: Can a fiscally stretched economy afford to innovate? It turns out to be the wrong question. The right one runs the other way.

A paper published this year in Oxford Economic Papers by Can Sever of the International Monetary Fund (IMF) titled ‘Government Debt and Innovation-Led Growth’ offers a formulation of why.

The mechanism is not a ‘crowding out’ in the simple textbook sense of higher interest rates displacing private investment. Research and development (R&D) investment is a real option. However, it is irreversible, distant in its returns and acutely sensitive to variance in future payoffs.

When government debt rises, policy uncertainty rises with it. It translates to distortionary taxes to service public debt and sovereign risk premiums feeding through to corporate borrowing costs, apart from tough questions about future spending priorities.

For an R&D-intensive company weighing whether to build a laboratory or a production line, elevated debt raises the option value of waiting. And waiting, in the field of innovation, often means not doing it at all.

Sever tests this using a variant of the identification strategy proposed by Raghuram Rajan and Luigi Zingales in their 1998 paper on financial dependence and growth.

The logic Sever follows is that if debt suppresses R&D, industries that are inherently more R&D-intensive (by virtue of their underlying technology) should suffer disproportionately when government debt is high.

Using 36 advanced economies and four decades of industry-level manufacturing data, he finds precisely this. Local projection methods show the correlation deepening over the span of a decade. Schumpeterian hysteresis, one might call it. Projects that are held off are not simply delayed. Stocks of knowledge erode. Teams disperse.

I ran a version of this analysis for India using data from IMF’s Global Debt Database, World Bank development indicators and the Annual Survey of Industries done by India’s statistics ministry. There are four findings.

First, on just a one-year horizon, the debt-R&D channel is not statistically significant. This mirrors Sever’s own finding that emerging markets as a group show no short-run sensitivity. There can be structural reasons for this. Indian manufacturing firms in chemicals, pharmaceuticals and capital goods have faced several other constraints; these could relate to infrastructure, intellectual property enforcement or talent availability. They dominate policy uncertainty in the short run.

Second, on a 3-to-10-year horizon, the picture reverses. The interaction between R&D intensity and government debt becomes statistically significant at year three and then deepens monotonically thereafter. The ex-ante cost is invisible. The ex-post cost is large. This is important. It means that the standard one-year fiscal incidence analysis systematically underestimates what elevated debt does to our innovation trajectory.

Third, India’s R&D-intensive industries have outperformed others regardless of the debt burden. Chemicals, pharmaceuticals, electronics and machinery have all grown faster than low-tech manufacturing across every sub-period in the sample. Demand, the export market’s pull and industrial policy all partly share credit for this. But it does not mean these industries are debt-insensitive. It means their headwinds have been outweighed by tailwinds.

Fourth, general government debt surged during covid and remains elevated but under control. Given the three-year lag, the medium-term innovation cost of that fiscal-enlargement episode still lies ahead of us, going by our current trajectory.

This brings one to ‘revdis’ or freebies. The competitive populism now endemic across state governments is typically analysed as a fiscal problem. But the Sever framework, as applied to India, suggests a second-order effect that is distinct and potentially more durable. Fiscal transfers of this kind are financed through some combination of higher debt or compressed capital expenditure. Capital expenditure includes public R&D infrastructure.

The cost of a loan waiver is not immediate; it materializes when firms choose not to invest. It shows up in factories not built, R&D not undertaken and a knowledge economy that remains perpetually deferred.

Sarabhai’s question has found its data. But this is not a problem for the Union government alone to solve. As our fiscal architecture is federal, so are its failings. When states borrow to distribute rather than to invest and when competitive populism crowds out capital expenditure across multiple tiers of government simultaneously, the effect is cumulative.

In all fairness, the Centre over the past couple of years has sought to give an impetus to R&D. But India must shift away from its current equilibrium across states. The costs imposed by large fiscal deficits will determine whether the next generation creates technologies or imports them. Posterity, which has no vote in either state or national elections, will bear the burden of both.

The author is a public policy professional.