The Numbers Game, Navigating the Contentious Landscape of India’s Economic Data

The Numbers Game, Navigating the Contentious Landscape of India’s Economic Data

In an era defined by data-driven decisions, the veracity of a nation’s core economic statistics is foundational to its governance, market confidence, and global credibility. For India, the world’s fastest-growing large economy, this foundation has been the subject of intense, often fractious, debate. As economist and former Monetary Policy Committee member Ashima Goyal outlines, criticisms of India’s national accounts data—particularly following the methodological rebasing to 2011-12—have become a persistent feature of economic discourse. These debates are poised to intensify as India undertakes another major base-year revision. Goyal’s analysis provides a crucial framework for separating substantive critique from ideological noise, categorizing criticisms into four distinct types, and defending the evolution of statistical systems as not only necessary but also robust. The central contention is clear: while India’s data architecture is imperfect and must continuously improve, the most vocal criticisms often stem from a resistance to change, selective interpretation, or motivated reasoning rather than from a constructive engagement with the complexities of measuring a vast, informal, and rapidly transforming economy.

The Four Faces of Criticism: A Taxonomy

Goyal’s taxonomy is instructive for anyone navigating the data debate:

1. Status Quoism: Resistance rooted in familiarity with old methods or investments in past datasets. Critics in this camp lament the loss of long-term comparability and often distrust new, more comprehensive systems.

2. Selective Data Use: The practice of cherry-picking specific data points or periods to support a preconceived narrative—for instance, highlighting quarters where growth appears inflated while ignoring those where it may be understated.

3. Allegations of Systematic Bias and Motive: The most serious and politically charged critiques, which impute deliberate manipulation or systematic overestimation of growth to the statistical agencies and the government.

4. Constructive, Feasible Suggestions: The only category Goyal argues deserves serious attention. This involves acknowledging the inherent difficulties of measurement and proposing specific, actionable improvements to methodology, coverage, and transparency.

The public debate is dominated by the first three, which generate more heat than light. The fourth, though less sensational, is where the true progress of a statistical system resides.

Defending Evolution: The Case for Methodological Upgrades

A primary target of “status quo” critics has been the shift in data sources for the corporate sector. The move from the Annual Survey of Industries (ASI), which sampled about 2,500 factories, to the Ministry of Corporate Affairs’ MCA-21 database, encompassing over 3.5 lakh firms, was a seismic shift. Critics argued the MCA database included inactive or “dummy” firms. However, as Goyal counters, clinging to a tiny, potentially unrepresentative sample in an economy that now boasts over 2 million active firms is untenable. The MCA-21 database, based on statutory filings, offers near-universal coverage of the formal corporate sector. While it required (and underwent) rigorous data cleaning, it represents a monumental leap towards capturing the true scale and dynamism of Indian corporate activity. Resistance to this change often reflects a discomfort with abandoning familiar, if limited, tools for more powerful but complex new ones.

Similarly, the transition from the infrequent and delayed Employment-Unemployment Surveys to the high-frequency Periodic Labour Force Survey (PLFS) was initially criticized for a perceived loss of depth. Yet, the PLFS has proven transformative. For the first time, India has quarterly, nationally representative data on employment, unemployment, and labour force participation. This is indispensable for timely macroeconomic and social policy formulation. Critics who once opposed the PLFS now routinely utilize its rich data, a testament to its value. A proposed shift to monthly frequency will likely face similar initial resistance before becoming the new standard.

The Informal Sector Conundrum: Estimation vs. Omission

The most technically complex and politically sensitive criticism revolves around the estimation of the informal sector, which constitutes a massive share of India’s economy and employment. Critics allege that the national accounts systematically overestimate growth because the formal sector data (which is solid) is used to extrapolate trends for the informal sector (which may be stagnating or shrinking). They argue this creates an automatic upward bias.

Goyal’s rebuttal is nuanced and points to the sophistication embedded in the methodology. She explains that the production approach to GDP—considered more reliable—is used as the “controlling total.” The expenditure approach serves as a cross-check. Crucially, the methodology for estimating the unorganized sector did not simply assume it grew in lockstep with the formal sector. In the 2011-12 rebasing, the method evolved from using a uniform Gross Value Added (GVA) per worker to assigning different weights based on labour skills and types of activity within informal manufacturing. This introduces necessary differentiation.

Furthermore, the expenditure side of the accounts uses commodity flow data—tracking the actual movement of goods—which inherently captures consumption by the informal sector. The difference between the production and expenditure estimates is recorded as a “discrepancy.” As Goyal highlights, this discrepancy has fluctuated between positive and negative over recent years. From the first quarter of FY2021 to the second quarter of FY2026, the cumulative discrepancy was negative to the tune of nearly ₹8.86 lakh crore. A negative discrepancy means the expenditure-side estimate was higher than the production-side estimate and was statistically adjusted downward to match. This cumulative negative value suggests a systematic underestimation by the production approach, not an overestimation. This single fact severely undermines the argument for a consistent upward bias.

Finally, household consumption expenditure—the largest component of GDP and the one most reflective of the informal economy—is measured as a residual after accounting for better-measured government and corporate spending. By construction, residual estimates are prone to being underestimates, not overestimates. This further challenges the narrative of inflated growth numbers.

Selective Spotlighting and the Inflation Deflator Debate

Goyal provides a compelling case study in how data can be selectively weaponized. Critics often pounced on periods where the GDP growth rate was higher than the GVA growth rate, attributing the difference to a spike in “net taxes” (taxes minus subsidies) and arguing the underlying growth was weaker. However, in subsequent quarters (Q1 and Q2 of FY26), when net taxes fell below their long-term average, both GDP and GVA showed robust and aligned high growth. The critics’ narrative shifted, but their spotlight did not retrospectively illuminate the earlier, contradictory evidence.

The debate then pivoted to the GDP deflator—the price index used to convert nominal GDP into real, inflation-adjusted GDP. The argument posited that if the deflator underestimated true inflation, then the resulting real GDP growth would be artificially high. This is a technically valid concern. However, Goyal notes the inconsistency: these same critics did not allege an overestimation of inflation during the high-inflation periods of 2020-2023, which would have implied that real growth was underestimated at the time. The selective application of skepticism reveals a motivated pattern.

She also references the IMF’s “Report on Observance of Standards and Codes” (ROSC), which gave India’s national accounts a ‘C’ grade. Critics highlighted this relentlessly. However, they conveniently glossed over the fact that the IMF gave India a ‘B’ for its price statistics (like the deflators). Since the overestimation argument hinges critically on errors in inflation measurement, the stronger grade for price data weakens the core critique.

The Path Forward: Embracing Complexity and Championing Constructive Critique

The fundamental truth, as Goyal underscores, is that measuring the GDP of a $4 trillion economy with a massive informal sector, undergoing digital and structural shifts, is an extraordinarily complex task. It is a blend of direct measurement, sophisticated estimation, and unavoidable statistical reconciliation. No system is perfect, and India’s statistical establishment, embodied by the National Statistical Office (NSO), operates under immense pressure and scrutiny.

The constructive way forward, which aligns with Goyal’s fourth category of criticism, involves:

1. Continuous Methodological Transparency: The NSO must continue to enhance the documentation and public explanation of its methods, assumptions, and source data. Detailed methodological papers, released alongside major revisions, can pre-empt misunderstanding and build trust.

2. Investing in Direct Measurement: While estimation is necessary, there is no substitute for expanding direct data collection. Strengthening surveys like the PLFS, the Annual Survey of Unincorporated Enterprises (ASUSE), and the revamped Household Consumption Expenditure Survey is paramount. The upcoming base year revision is an opportunity to incorporate these improved datasets.

3. Leveraging Big Data and Technology: India’s digital infrastructure—GST filings, UPI transactions, e-way bills—generates terabytes of real-time, high-frequency data. Statisticians must be empowered to creatively integrate these new data lakes into their estimation frameworks, not as replacements, but as powerful cross-validation tools.

4. Fostering an Independent, Expert Dialogue: The debate must move from the op-ed pages and social media spats to sustained, technical dialogue within a community of professional statisticians, economists, and sectoral experts. Bodies like the Advisory Committee on National Accounts Statistics should be platforms for rigorous, non-partisan scrutiny.

5. Global Benchmarking and Peer Review: Engaging in continuous peer review with international agencies and statistical bodies of other large, complex economies can help identify best practices and blind spots.

Conclusion: Data as a Public Good, Not a Political Football

The intense debate over India’s economic data is, in one sense, a sign of democratic vitality—people care about the country’s economic narrative. However, when criticism descends into motivated reasoning, selective analysis, or ad hominem attacks on institutions, it corrodes a vital public good: trusted information.

Ashima Goyal’s analysis serves as a necessary corrective. It urges a more discerning consumption of data criticism, distinguishing between the noise of status quoism and selective outrage, and the signal of feasible, constructive suggestions. It defends the necessary evolution of statistical systems in a changing economy, while firmly rebutting the most persistent allegations of systematic bias with countervailing evidence from the accounts themselves.

As India stands on the cusp of another major data revision, the goal should not be to produce numbers that satisfy a particular political narrative, but to refine a system that, despite its inevitable “warts,” strives relentlessly for accuracy, transparency, and relevance. The true measure of success will be a statistical framework robust enough to withstand both the complexities of the Indian economy and the scrutiny of a vibrant, demanding democracy. In this endeavor, constructive critics are essential partners; the rest are merely commentators in a noisy, unhelpful debate.

Q&A Section

Q1: According to Ashima Goyal, what are the four types of criticisms leveled against India’s economic data, and which does she consider valid?
A1: Ashima Goyal categorizes the criticisms into four types:

1. Status Quoism: Resistance to methodological changes due to familiarity with old systems.

2. Selective Data Use: Cherry-picking data points or periods to support a preconceived narrative.

3. Allegations of Systematic Bias and Motive: Accusing statistical agencies of deliberate manipulation for political or economic reasons.

4. Constructive, Feasible Suggestions: Acknowledging measurement complexities and proposing specific improvements.
   Goyal argues that only the fourth category—constructive, feasible suggestions—deserves to be taken seriously, as the first three are often driven by resistance, selective reasoning, or imputed motives rather than a genuine engagement with methodological challenges.

Q2: How does Goyal counter the criticism that India’s GDP overestimates growth by extrapolating formal sector performance to the informal sector?
A2: Goyal provides a multi-layered rebuttal:

• Methodological Nuance: She clarifies that the 2011-12 rebasing moved away from assuming uniform informal sector growth. It began using different weights for various labour skills and activities within informal manufacturing, introducing necessary differentiation.

• The Discrepancy Evidence: Crucially, she points to the statistical “discrepancy” between the production and expenditure approaches to GDP. From Q1 FY21 to Q2 FY26, the cumulative discrepancy was negative (≈ -₹8.86 lakh crore). A negative discrepancy means the expenditure-side estimate (which uses commodity flow data capturing informal activity) was higher than the production-side estimate and was adjusted downward. This suggests a possible underestimation, not overestimation, of production.

• Residual Measurement: Household consumption, key to the informal economy, is measured as a residual, a method prone to underestimation, not overestimation.

Q3: What was the significance of shifting from the ASI to the MCA-21 database for corporate sector data, and why was it criticized?
A3: The shift replaced the sample-based Annual Survey of Industries (ASI—~2,500 firms) with the Ministry of Corporate Affairs’ near-universal MCA-21 database (over 3.5 lakh firms). This was a major upgrade to capture the expanded formal corporate sector. It was criticized on grounds that the MCA database included inactive or “dummy” firms and disrupted long-term comparability. Goyal counters this “status quoism” by arguing that a tiny, potentially unrepresentative sample is inadequate for an economy with over 2 million active firms. The MCA data, based on statutory filings, was cleaned and matured, offering a far more comprehensive and realistic picture of corporate activity.

Q4: How does the “selective data use” critique manifest in debates about GDP versus GVA growth and the GDP deflator?
A4: Critics selectively highlight periods that fit their narrative. For example:

• GDP vs. GVA: When GDP growth exceeded GVA growth due to a high “net taxes” component in some quarters (like Q3 FY24), critics argued true (GVA) growth was lower. However, they ignored subsequent quarters (Q1 & Q2 FY26) where net taxes were low yet both GDP and GVA showed robust, aligned high growth.

• GDP Deflator: Critics argue that an underestimated deflator inflates real GDP growth. Yet, as Goyal notes, they did not allege an overestimation of inflation during the high-inflation period of 2020-23, which would have implied real growth was underestimated then. This selective application of skepticism reveals a pattern of motivated reasoning rather than consistent methodological concern.

Q5: What does Goyal propose as the constructive path forward for India’s statistical system?
A5: While defending the current system, Goyal implicitly advocates for a path focused on continuous improvement rooted in her fourth category of critique. This involves:

• Enhancing Transparency: Better documentation and explanation of methodologies and source data.

• Strengthening Direct Data Collection: Investing in foundational surveys (e.g., employment, unincorporated enterprises, consumption) to reduce reliance on estimation.

• Integrating New Data Sources: Creatively leveraging big data from GST, UPI, and other digital pipelines for cross-validation and richer insights.

• Promoting Expert Dialogue: Moving debates into technical forums with professional statisticians and economists for non-partisan scrutiny.

• Engaging in Global Peer Review: Benchmarking against international standards and learning from other large, complex economies.