Policy

This paper evaluates whether India’s foreign exchange reserves, now exceeding $700 billion, are economically optimal rather than merely adequate. Using quarterly data from 2000Q1 to 2025Q4, we estimate a long-run reserve demand equation via Dynamic OLS and calculate the welfare-maximizing reserve level using the Jeanne and Rancière framework, calibrated with crisis probabilities from an asymmetric complementary log-log model. We find that India’s actual reserves have exceeded the optimal benchmark in every sample quarter, with excess holdings reaching $145.02 billion by 2025, or $190.70 billion once the Global Financial Safety Net is incorporated. We further show that a two-tranche reserve management framework, separating a core liquidity buffer from a dedicated sovereign wealth fund, would have generated $344.58 billion in cumulative wealth over 2004 to 2025 against $140.80 billion under the status quo, at an annual fiscal cost of holding reserves of roughly $15 billion in 2025 alone. These findings suggest that India’s reserve accumulation strategy, while historically prudent, now imposes a quantifiable and avoidable fiscal drag that structural reform could substantially mitigate.

As India’s working age population expands, the need to create productive and future ready jobs at scale is pertinent. This report maps employment growth in India between 2018 and 2025 using the Periodic Labor Force Survey with an eye towards illuminating occupation-level job growth patterns, the changing skill profile of the Indian workforce and exposure to AI developments. Overall, we find that employment has expanded since 2018, though much of the increase was in agricultural self-employment and low-skill nonfarm occupations. Our analysis provides evidence for job polarization - job growth was concentrated in low-skilled occupations coupled with modest growth in some high-skilled professional roles, with relatively no growth in medium-skill jobs. High-skilled, cognitive occupations are most exposed to emerging AI applications. So far, employment growth has been pronounced at both ends of the AI-exposure range. Manual, low-skill jobs, which are least-exposed to AI grew tremendously while high-skilled job grew to a lesser extent, resulting in the missing middle and echoing the broader polarization of the labor market. These findings draw policy attention towards the need for industry-relevant skills that enhance productivity and boost the AI-adaptability. Building data infrastructure that captures evolving worker and job characteristics is important to produce timely and relevant evidence that supports policy in times of technological change.

In February 2026, the Ministry of Statistics and Programme Implementation (MoSPI) released a comprehensive revision of India’s national accounts, shifting the base year from 2011–12 to 2022–23 alongside significant methodological reforms. This paper provides a detailed comparative analysis of the old and new GDP series, examining the 3.08% decline in nominal GDP for FY26 and tracing it to improved informal sector measurement (ASUSE), the adoption of double deflation, expanded administrative data sources (GST, LLP filings), and the introduction of Supply–Use Tables. Using implied inflation across sectors as a diagnostic, we decompose sectoral growth into contributions from better data collection and improved deflation methodology — finding that the tertiary sector experienced the most significant downward revision in nominal levels, while manufacturing saw real growth revised sharply upward from 4.5% to 9.3% in FY25, driven primarily by the shift to double deflation. Quarterly data further reveal a more optimistic trajectory for manufacturing than annual figures suggest.

When a government breaks its own fiscal rules to survive a crisis, what makes anyone believe it will go back? For most emerging and developing economies (EMDEs), the answer is: nothing does. Of the 54 EMDEs with an active escape clause heading into COVID-19, 81% had no legally binding path back to fiscal discipline, just a promise to eventually sort it out. This paper tests whether writing that promise into law actually matters. Using IMF Fiscal Rules Dataset coding locked in before the pandemic, we compare countries with a codified correction mechanism to those with one in name only, and find that codified countries clawed back 2.7 percentage points more of their primary balance from the COVID trough than otherwise-similar countries without one, while a mechanism with no specified return path buys nothing i.e. its effect is statistically indistinguishable from zero. The ad vantage doesn’t come from harsher post-crisis austerity, since expenditure, revenue, and debt adjustments show no effect, but from before the crisis hit: codified countries sim ply ran larger surpluses going in. A probit model adds a troubling wrinkle — codification tracks income, not need, so the countries least likely to write the rule down are the ones most exposed to the next shock. 

The US–Israel military operation against Iran in February 2026, which severely disrupted oil traffic through the Strait of Hormuz, drove the Indian Basket crude oil price from US$69 to US$117 per barrel within weeks, a shock of 69 percent. This paper estimates the headline CPI inflation impact of this shock under a range of sustained year-average crude price trajectories and government pass-through stances on petrol and diesel. We extend the Tomar (2019) two-channel decomposition by computing fuel-specific Leontief shares from the 131-sector Chadha et al. (2020) Input–Output table, applying the framework separately to LPG, petrol, and diesel. Across four sustained year-average crude scenarios for 2026–27 (US$100, US$117, US$120, and US$125 per barrel), we find that the LPG revision alone with petrol and diesel frozen adds only 19–23 basis points to year 2026–27 CPI. Partial pass-through of 20–50 percent on petrol and diesel keeps headline CPI within the RBI’s 6 percent upper tolerance band at all four crude levels, but at full passthrough, every scenario at US$117 and above breaches the ceiling, with year 2026– 27 CPI rising to 7.21–7.92 percent. A stress-test scenario at US$150 sustained, in which even 50 percent pass-through breaches the ceiling, is presented in the appendix.

The US–Israel military operation against Iran (February 2026) and Iran’s retaliatory activities including closure of the Strait of Hormuz have triggered a large energy supply shock for India, the world’s third-largest oil importer. This paper provides a bottom-up fiscal accounting of the shock’s impact on the Indian central government’s budget for FY 2025–26 and FY 2026–27. We trace the transmission through five channels: higher fertiliser subsidies (costlier domestic gas, costlier urea imports, costlier DAP imports), higher LPG subsidies, foregone excise rev- enue from the Special Additional Excise Duty (SAED) cut on petrol and diesel, a modest customs duty gain, and foregone petroleum-sector dividends. Under our primary scenario (war through June 2026), the additional fiscal deficit in FY 2026– 27 is Rs. 74,087 crore, pushing the deficit from 4.30% to 4.49% of GDP. In the worst case (war through December), the deficit reaches 4.99% of GDP, a deviation of 0.69 percentage points from the Budget target.