ExplainSpeaking: Understanding the ‘discrepancies’ in India’s new GDP data

*Dear Readers,*

On February 27, India’s Ministry of Statistics and Programme Implementation (MoSPI) brought out a new series of data for the country’s economic output. Any country’s economic output is most broadly mapped by a measure called the Gross Domestic Product or GDP. Simply put, the GDP gives the market value of all final (as opposed to intermediate) goods and services produced within the geographical boundaries in a year.

This is one of the most significant updates because the GDP data — both its level and its growth rate from one year to another — forms the basis of all policymaking in the country.

While there were many new technicalities in the new GDP series, the main change was the change in “Base Year”. The base year, as the name suggests, forms the base for comparison in the coming years. Before the new series, 2011-12 was being used as the base year. This meant the goods and services produced in 2011-12 and the prices at which they were sold and bought in that year formed the basis for all future analysis.

But, as would be plain to most readers, every few years India produces and buys a different mix of things and their prices also change, and as such, comparing to an “outdated” base year can throw up data that is misleading. For instance, why calculate GDP by looking at the number of typewriters being produced and their prices, when almost the whole economy has shifted to computers? It is for this reason that it is best to update the base year from time to time.

Before 2011-12, India used 2004-05 as the base year, and before that, 1999-2000 was the base year. Now, 2022-23 of Financial Year 2023 (FY23) is the new base year; this is the eighth such revision in independent India’s history.

Problems with the old GDP series

The outgoing GDP series, which was adopted in January 2015, created a lot of controversy.

For many critics, the GDP series with the base year of 2011-12 overstated India’s GDP growth. Those who had worked on this series countered by claiming that it was no worse than any previous GDP series; indeed, much better than any in the past. Here’s an explainer from 2019 when debates about the quality of GDP data really gained momentum after Arvind Subramanian, who had been the Chief Economic Advisor for the government between 2014 and 2018, argued that GDP data was painting a flawed picture of the economy.

Over time, the chorus grew with more and more people pointing to India’s GDP data not passing, as it were, the smell test. For instance, in the post-pandemic period, real GDP growth rates (that is, GDP growth rate after one takes away the effect of inflation in prices) were considered by many as too high.

For instance, according to the outgoing GDP series, in the current financial year (2025-26), the nominal growth of GDP is 8%, and the real growth (the one after subtracting the rate of inflation) is 7.4%. Many argued that the inflation rate they faced was far more than 0.6%, thus making this measurement of GDP growth hard to believe. In turn, these issues also raised questions about the quality of India’s inflation data.

There was another, even bigger problem undermining the credibility of India’s GDP data. This is called “discrepancies”. Ironically enough, “discrepancies” is an actual sub-head used by MoSPI when it calculates GDP data (see the screenshot of any official release).

So, what are ‘discrepancies’ in GDP data?

There are two main ways to calculate India’s economic output.

One way is to look at everything India produces in a year and add up all the monetary “value” created in a year. This is typically captured by a measure called the Gross Value Added (GVA), and it looks at what value was produced in different sectors of the economy.

The other way to look at the same economy is to add up all the money spent by different people or entities (be it individuals, governments, or business houses) in the economy. This is typically called the Gross Domestic Product or GDP.

The two variables are connected thus:

GDP = GVA + Net Indirect Taxes Tax govt levies on different goods minus the subsidies govt provides for the production of different goods

*Tax govt levies on different goods minus the subsidies govt provides for the production of different goods

In theory, the two calculations should yield the same economic output. “But,” as MoSPI states in its FAQs, “often these two numbers don’t match exactly. This small difference is called the ‘statistical discrepancy’. It happens because some data, especially on the spending side, is not available, or reported late.”

Imagine a scenario where a car was produced sometime during the year. The value added in producing the car gets captured in the production-side measure of economic output, called GVA. But suppose, for reasons entirely above-board, when one looks at the data from the consumption (or expenditure) side, it is not clear who bought that car, nor is it clear if that car added to the stock of existing unsold inventory.

Perhaps, in a few months or a year, after the financial year has ended, this information will be available, but at the end of the financial year, there will be a mismatch between the production side estimate and the expenditure side estimate of the same year’s economic output. To bridge the artificial statistical gap, MoSPI added a sort of dummy component called “discrepancies”.

It is important to note here that the production side estimates are given primacy, and the “discrepancies” are added (or subtracted) from the expenditure side estimates. But high levels of “discrepancies” can undermine the credibility of data, as indeed they did for the outgoing GDP series.

What were the levels of discrepancies in the old series?

CHART 1 shows the level of discrepancies as a percentage of real GDP in the outgoing series of data (with the Base Year of 2011-12). The lower the levels of discrepancies, the higher the credibility of the data.

As can be seen, there are many years with high levels of discrepancies. Pronab Sen, the former Chief Statistician of India and former Secretary of MoSPI, says that ideally, this ratio should not exceed 2% (the yellow zone). The blue zone is the 1% level.

How have things changed in the new GDP series?

If one looks at the new GDP data with the Base Year of 2022-23, there are some curious bits. But before that, here are the different components of GDP.

Typically, in India, from an accounting standpoint, there are three main creators of GDP:

1. 1.The money spent by Indians in their individual capacity. Technically, it is called the Private Final Consumption Expenditure or PFCE. This is the biggest contributor to India’s GDP; broadly speaking, almost 60% of India’s total GDP comes from this kind of spending.

1. 1.The money spent by businesses and firms towards creating new productive capacities. Think of money spent towards making new factories or buying new laptops for one’s employees for office use. This is called the Gross Fixed Capital Accumulation or GFCF, and this has typically accounted for almost 30% of India’s GDP in a year. This category also includes the money spent by governments towards creating productive assets, such as roads, etc.

1. 1.The third biggest engine is the money that the government spends on its daily functioning, such as salaries, pensions and fuel, etc. This is called the Government Final Consumption Expenditure of GFCE, and it accounts for almost the remaining 10% of India’s GDP.

There are other sub-heads too, like Net Exports (the net effect of what Indians spend on imports over what they earn from exports), but more often than not, this figure is negative because Indians import more than what they export. Then stands Change in Stocks, which maps the change in unsold inventory. There is also Valuables, which notes the change in the valuations of valuables (like precious metals) in the country.

And, of course, there are “discrepancies”.

Now, see TABLE 1 that lays out “real” GDP data (both level and growth rate) for 2023-24 (or FY24).

While the overall real GDP growth rate is 7.2%, the three main individual components (accounting for 99% of the total real GDP) have grown at only 5.7%. Clearly, something else is growing at a much faster rate to pull up the overall real GDP growth to 7.2%.

As it turns out, the “discrepancies” have grown from zero in FY23 to over Rs 1 lakh crore in FY24. Change in stocks, too, has registered a 116% increase.

The data is even sharper in FY25 (see TABLE 2).

The overall real GDP has grown by 7.1% according to the new GDP series, but the three main components that account for 98% of all the real GDP have only grown by 6.1%.

What about “discrepancies”? They have grown by 230% in FY25 (over their FY24 level); that’s almost Rs 3.5 lakh crore. The discrepancies in the current financial year (FY26) are pegged at Rs 4.9 lakh crore.

Simply put, while India has just updated its GDP series, discrepancies (as a percentage of real GDP) are rising again (see CHART 2).

Why is this happening?

Part of the problem is that MoSPI may not have all the data to bring down discrepancies by matching total production with total consumption. This is done with the use of something called the “Supply and Use Tables”.

As he was unveiling the new GDP series, the current MoSPI Secretary, Subhash Garg, assured that by the time the final sets of GDP data are provided for any financial year, discrepancies will be brought down to a minimum. To be sure, the data for FY25 is the “First Revised Estimate”, and the data for FY26 is the “Second Advance Estimate”. These tables will go through a few more revisions.

The other bit of the problem, points out Pronab Sen, has to do with moving away from the Base Year, which is 2022-23. To be sure, the discrepancies are zero when one looks at nominal GDP data in the new series. But the percentages start climbing up (CHART 2) when one looks at real GDP data.

That’s because, as one moves away from the base year, the quality of price information worsens, explains Sen. That, in turn, implies that the old problem of poor quality of deflators (the inflation rate used to arrive at real GDP) comes into the picture. Again, Garg had informed that MoSPI is now using 600 odd deflators in the new series (instead of 180 deflators) to improve the calculation of real GDP.

So, what does this say about GDP data?

Typically, it is real GDP growth that drives conversations. That growth rate was being questioned on account of high levels of “discrepancies”. As this analysis of the new series of GDP data shows, the level of discrepancies is again rising. It is also a possibility that, in time with better data, these discrepancies will be resolved and minimised. However, this shows how estimating India’s GDP is not a straightforward exercise.

Data on what India produces might be easier to capture, but data on what India spends on (the bit that GDP per se tells about) is far more complicated. Sen explains that typically, one looks at production data and then tries to “allocate” it to some entity on the consumption side. But, economy-wide exact details of consumption are not available.

“The real hard data that you have on the expenditure side is about what the government spent, the exports and the imports, and the data on corporate investments (from their balance sheets). Expenditure towards consumption and investment by households (families and individuals) is not available in granular detail. For instance, Household Consumption Expenditure Surveys that MoSPI uses are surveys (or a sample, and not Census), and as such, they can only provide ratios, but cannot be used for measuring levels,” says Sen.

Does India’s real growth rate feel genuine to you? What can be done to improve the credibility of India’s national income accounts?

*Share your view and queries at udit.misra@expressindia.com*

*Take care,*

*Udit*

Udit Misra is Senior Associate Editor at The Indian Express. Misra has reported on the Indian economy and policy landscape for the past two decades. He holds a Master’s degree in Economics from the Delhi School of Economics and is a Chevening South Asia Journalism Fellow from the University of Westminster. Misra is known for explanatory journalism and is a trusted voice among readers not just for simplifying complex economic concepts but also making sense of economic news both in India and abroad. Professional Focus He writes three regular columns for the publication. ExplainSpeaking: A weekly explanatory column that answers the most important questions surrounding the economic and policy developments. GDP (Graphs, Data, Perspectives): Another weekly column that uses interesting charts and data to provide perspective on an issue dominating the news during the week. Book, Line & Thinker: A fortnightly column that for reviewing books, both new and old. Recent Notable Articles (Late 2025) His recent work focuses heavily on the weakening Indian Rupee, the global impact of U.S. economic policy under Donald Trump, and long-term domestic growth projections: Currency and Macroeconomics: "GDP: Anatomy of rupee weakness against the dollar" (Dec 19, 2025) — Investigating why the Rupee remains weak despite India's status as a fast-growing economy. "GDP: Amid the rupee's fall, how investors are shunning the Indian economy" (Dec 5, 2025). "Nobel Prize in Economic Sciences 2025: How the winners explained economic growth" (Oct 13, 2025). Global Geopolitics and Trade: "Has the US already lost to China? Trump's policies and the shifting global order" (Dec 8, 2025). "The Great Sanctions Hack: Why economic sanctions don't work the way we expect" (Nov 23, 2025) — Based on former RBI Governor Urjit Patel's new book. "ExplainSpeaking: How Trump's tariffs have run into an affordability crisis" (Nov 20, 2025). Domestic Policy and Data: "GDP: New labour codes and opportunity for India's weakest states" (Nov 28, 2025). "ExplainSpeaking | Piyush Goyal says India will be a $30 trillion economy in 25 years: Decoding the projections" (Oct 30, 2025) — A critical look at the feasibility of high-growth targets. "GDP: Examining latest GST collections, and where different states stand" (Nov 7, 2025). International Economic Comparisons: "GDP: What ails Germany, world's third-largest economy, and how it could grow" (Nov 14, 2025). "On the loss of Europe's competitive edge" (Oct 17, 2025). Signature Style Udit Misra is known his calm, data-driven, explanation-first economics journalism. He avoids ideological posturing, and writes with the aim of raising the standard of public discourse by providing readers with clarity and understanding of the ground realities. You can follow him on X (formerly Twitter) at @ieuditmisra ... Read More