India's official GDP figures have misrepresented the trajectory of the world's fifth-largest economy for the better part of two decades, according to a major new working paper published by the Peterson Institute for International Economics (PIIE). It finds that India overstated annual growth by up to two percentage points after 2011 — and understated it during the boom years of the 2000s.
The paper, India's 20 Years of GDP Misestimation: New Evidence, authored by economists Abhishek Anand, Josh Felman, and Arvind Subramanian, finds that India's statistical methodology both understated growth during a genuine boom and overstated it during a subsequent slowdown — conjuring a picture of steady, uninterrupted expansion that never quite existed. What acquires significance to the paper is, Subramaniam was Chief Economic Advisor to the Government of India, having served from 16 October 2014 to 20 June 2018.
The researchers estimate that growth between 2005 and 2011 was underestimated by roughly one to one-and-a-half percentage points per year, while growth between 2012 and 2023 was overestimated by approximately one-and-a-half to two percentage points annually. Correcting for both errors, the paper concludes that India's economy grew at an average of four to four-and-a-half per cent per year over the post-2011 period — not the six per cent reported in official statistics.
The findings arrive shortly after the Indian government, in February 2026, introduced a revised GDP series following an extensive consultative process — an exercise the authors describe as commendable. The paper does not assess the quality of the new numbers, which require time to evaluate, but it provides a benchmark of re-estimated historical growth rates against which the revised methodology can be measured.
A Boom Erased, a Slowdown Concealed
At the heart of the paper is a striking divergence between official GDP data and a range of macro-economic indicators — exports, bank credit, industrial production, electricity consumption, direct tax revenues, and corporate sales — that should, in theory, move broadly in line with aggregate output.
According to official figures, growth eased only modestly from an annual average of 6.9 per cent in 2005–11 to around 6.1 per cent in 2012–24. The macro indicators tell an altogether different story. Real credit growth fell from 15.6 per cent to 5.6 per cent a year. Real exports dropped from 13.9 per cent to 5.4 per cent. The index of industrial production plummeted from 16.1 per cent to 2.9 per cent. Direct tax revenues slumped from 13 per cent to 7 per cent annually.
The authors demonstrate that correlations between these indicators and GDP were strong under India's older national accounts methodology — in use until 2015 — but broke down or weakened significantly after the January 2015 revision was introduced. This pattern holds both within India over time and in cross-country comparisons: India became a clear outlier after the methodology changed, with reported growth running well above what its own macro indicators, and the experience of comparable economies, would predict.
Three data points illustrate the divergence with particular sharpness. In the demonetisation year of 2016, real corporate sales grew at 1.4 per cent while official Gross Value Added (GVA) growth was recorded at 8.0 per cent. In 2019, when a credit crisis triggered by nonbank financial institutions gripped the economy, real sales fell 4.5 per cent — yet official GVA still registered growth of 3.0 per cent. In 2024, real sales rose 2.2 per cent against official GVA growth of 6.4 per cent.
Two Methodological Failures
The paper identifies two principal sources of the misestimation: inappropriate data sources and inappropriate price deflators.
The first problem concerns the informal sector. Unorganised enterprises account for roughly 44 per cent of GVA across the Indian economy — a majority share in agriculture, construction, trade and repair services, hotels and restaurants, and real estate. Because hard data on this vast sector are scarce, India's 2015 methodology proxied informal activity using formal sector indicators, on the assumption that the two move together. This assumption was broadly defensible until three major shocks — demonetisation in 2016, the rollout of the Goods and Services Tax, and the Covid-19 pandemic — hit the informal economy far harder than the formal one.
Survey data from the National Sample Survey show that from 2010 to 2015, formal and informal sector revenue growth were essentially identical. After the shocks, they diverged sharply: annual nominal growth averaged ten per cent in the formal sector but only 6.8 per cent in the informal sector — a gap of 3.3 percentage points every year. By continuing to use the formal sector as a proxy after this divergence opened up, the statistical methodology effectively overstated the performance of the much larger informal economy.
The second problem concerns the deflators used to strip out price effects and arrive at real GDP. Under the 2015 revision, the bulk of GVA was estimated by deflating nominal values rather than measuring volumes directly. The deflators chosen were largely drawn from the Wholesale Price Index — an index that weights commodities heavily and, as the paper demonstrates, follows oil prices closely. When oil and other commodity prices fell relative to consumer prices, as they did persistently from 2011 onwards, the WPI-based deflators understated actual inflation in the economy. The result was that real growth was systematically overstated.
The paper is careful to note that the absence of double deflation — often cited as the main flaw in the 2015 series — was a comparatively minor problem. The real damage was done by the choice of wrong deflators. Calculations show that between 2011 and September 2025, annual WPI growth averaged 2.2 percentage points less than CPI growth, pushing the GDP deflator 1.4 percentage points lower than consumer price inflation on average. The bias, the authors conclude, was systemic rather than cyclical.
The Backcasting Controversy
The paper also revisits a contested episode from 2018, when India's statistical agency applied its 2015 methodology retroactively to the boom years of 2005–11 in a process called backcasting. The exercise, the authors note, was announced not by the official statistics body but by Niti Aayog — and its findings contradicted those of a government-appointed expert committee led by economist Sudipto Mundle, whose estimates had been published just months earlier.
The Mundle Committee had concluded that the older GDP series had understated growth during 2005–11, revising annual estimates upward by an average of 0.4 percentage points. The official backcasting, by contrast, revised growth downward — producing a gap of nearly 1.7 percentage points between the two sets of figures. The consequence was profound: under the official backcasting, the economy's boom years appear only marginally stronger than the subsequent decade, flattening what was in reality a period of exceptional dynamism.
The paper's analysis finds that correlations between corporate sales and GVA are stronger under the Mundle Committee estimates than under the official backcasted series, lending further support to the view that the official figures understated the boom.
To produce revised estimates, the authors make two key adjustments to the 2015 methodology for the period 2011–23. They replace WPI-based deflators with CPI-aligned indices in manufacturing, trade and repair services, hotels, construction, and financial services. And they replace formal sector proxies for the informal economy with data drawn directly from the NSS surveys of unincorporated enterprises.
The result is a significant downward revision. Against an official GVA growth estimate of 5.9 per cent per year for 2011–23, the authors' revised figures range between 4.0 and 4.4 per cent, depending on assumptions about sectors not covered by the informal enterprise surveys. The deflator correction alone accounts for one percentage point of the gap; adjustments to informal sector data account for a further 0.4 to 0.8 percentage points.
The overestimation compounds over time. The paper calculates that as of 2025, the level of real GDP is overstated by approximately 22 per cent, and the level of real consumption by about 31 per cent. India's economy is thus materially smaller, and average living standards substantially lower, than official statistics indicate.
Cross-country regression analysis corroborates these findings. In a sample of 51 comparable economies for 2012–24, India emerges as a positive outlier with growth running around 2.4 percentage points above what its own macro indicators would predict — a result that is statistically significant at the one per cent level. The same approach applied to the 1995–2011 period, using the old methodology, produces no such outlier status for India.
Policy Costs and the Tax Puzzle
The authors point towards the real-world consequences of the misestimation. When GDP data signalled that the economy was strong during periods when it was actually weak, they argue, businesses were prone to misinvest, households to overspend, and the central bank to maintain unnecessarily tight monetary policy. Inaccurate numbers also blunted the urgency of economic reform: why change course when official figures showed world-beating growth?
The paper also addresses what might appear to be a contradiction in its argument: if growth has been overstated, why have tax revenues risen so sharply in recent years? The answer, the authors suggest, lies not in the overall pace of expansion but in its composition. Post-Covid growth was concentrated in sectors — Global Capability Centres, financial markets, and premium real estate — that generated disproportionately large tax liabilities for higher-income individuals. The share of personal income reported by individuals earning more than Rs 1 crore (approximately USD 110,000) nearly doubled in two years. As the newly affluent spent on heavily taxed luxury goods such as SUVs, GST revenues also surged. Tax collections boomed because of the skewed, K-shaped character of the recovery, not because the economy as a whole was expanding rapidly.
A Rewritten History
Perhaps the paper's starkest conclusion is historiographical. The Indian economy did not, as official figures imply, enjoy two decades of broadly stable growth in the six-to-seven per cent range. It boomed between roughly 2003 and 2010 — a period of genuine dynamism that the post-2015 methodology has understated. It then slowed sharply as a succession of shocks — the global financial crisis, the Twin Balance Sheet crisis, demonetisation, GST, the nonbank credit freeze, and Covid — took their toll. The 2015 methodology effectively erased the boom and concealed the slowdown, replacing both with a flat line.
The paper notes that even with revised estimates, India remains among the seven or eight fastest-growing major economies in the world for the 2011–23 period — behind Ireland, China, and Bangladesh on official figures, but still a strong performer by international standards. That achievement, the authors suggest, does not need to be inflated.
The February 2026 revision to India's national accounts, the paper concludes, represents a genuine attempt to address longstanding methodological weaknesses. Whether it succeeds in restoring accuracy to the historical record will become clearer only as the new methodology is used to produce a full back series — and as researchers are able to scrutinise the resulting numbers in the years ahead.
Comments