The government’s recent gross domestic product (GDP) estimates have, to put it mildly, been mired in controversy. Fortunately for all of us, this year’s Economic Survey presciently realized the GDP data would generate a lot of heat, which is why it included a little box with the neat title, Clarifying in Advance Possible Misinterpretations in GDP-demonetisation Effects.

The Economic Survey’s note tells us how we should estimate the impact of demonetisation on the economy from the GDP numbers. It says the Central Statistics Office’s (CSO) first advance estimate of growth for 2016-17, which didn’t take into account the impact of demonetisation, would be a good benchmark. It adds, “An even better counterfactual for comparison would be the level of nominal rather than real GDP growth. After all, demonetisation is mostly a nominal demand shock, so its effect in the first instance will be on nominal magnitudes… Therefore, the most appropriate gauge of demonetisation would be to compare actual nominal GDP growth—or the Survey’s estimate of it—with the counterfactual nominal GDP growth without demonetisation. According to the CSO, this counterfactual is 11.9%, while the Survey’s estimate is around 11N (11.25) percent."

**ALSO READ** | **What explains 7% GDP growth despite Modi’s demonetisation drive?**

What impact did the Economic Survey believe demonetisation would have on GDP growth? It said, “Given the uncertainty, we provide a range: a N (one-fourth) percentage point to 1 percentage point reduction in nominal GDP growth relative to the baseline of 11N (11.25) percent." In other words, the government’s chief economist expected nominal GDP growth for 2016-17 to be 10.25-11%.

Well, nominal GDP growth for 2016-17, according to the CSO’s second advance estimates, came in at 11.5%. Instead of being lower than the chief economic adviser’s pre-demonetisation estimate of 11.25%, growth post-demonetisation was actually higher. Simply put, if we apply the method advocated by the Economic Survey, demonetisation seems to have boosted GDP growth, rather than disrupt it—a conclusion that seems to call for a willing suspension of disbelief.

Going by the CSO’s estimates, however, the first estimate was for 11.9% and the second was for 11.5%—suggesting, at first blush, that demonetisation shaved 0.4 percentage points off nominal growth.

**ALSO READ** | **What analysts, economists think of India’s GDP growth rate at 7%**

But unfortunately, it’s not that simple. What the CSO has gone and done is to revise the 2015-16 GDP numbers, so the first advance estimates’ growth rates for 2016-17 are not comparable with those given by the second estimates.

For instance, if we kept the GDP for 2015-16 given in the first estimate, which is what we should do if we want to compare the first and second estimates, we get a nominal GDP growth rate of 12.3% for 2016-17. That’s above the 11.9% first estimate by the CSO, suggesting that, going by the Economic Survey’s logic, demonetisation boosted GDP growth by 0.4 percentage points. Surreal, but nice.

Perhaps we shouldn’t go by the growth percentages, riddled as they are with base effects. Instead, let’s look at the absolute numbers of nominal GDP. The first advance estimates put it at Rs151.93 trillion and the second estimates at Rs152.51 trillion. That’s an increase of 0.38%. Once again, we come to the conclusion that demonetisation raised nominal growth by around 0.4%.

**ALSO READ** | **GDP growth was creditless, moneyless, as well as cashless, in December quarter**

How did we get this 0.38% increase? If we compare the expenditure side data for the first and second estimates, we find that private consumption has gone down 1.5%, government consumption is lower by 8%, while gross fixed capital formation has been revised up by 1.4%. Lamentably, the biggest increase has been in the “discrepancies" number, which is up a humongous 70.9%. Ignoring that, the next biggest upward revision is in inventories or change in stocks, up 19.7%. Chart 1 has the details.

What we can glean from these numbers, once again delicately averting our eyes from “discrepancies", is that demonetisation seems to have had some negative impact on consumption and inventory accumulation has boosted the GDP figures. This fits in snugly with the increase in inventory seen in the quarterly corporate financials.

**ALSO READ** | **Demonetisation’s impact on the corporate sector**

But given these huge discrepancies, the expenditure side numbers do seem to call for spadefuls of salt. What, instead, do the figures for gross value added (GVA) say?

Well, the first estimates had pegged nominal GVA growth at 10.8%, revised down to 10.4% in the second estimate. That 0.4 percentage point revision seems to have cropped up again, this time on the downward side. But look a little closely and you find the GVA data for the previous year has been revised down while computing the second estimate. If we keep the original numbers for 2015-16, as we should for comparing the change in growth rates, nominal GVA growth according to the second estimate now comes to 11.98%, well above the 10.8% growth in the first advance estimate. We are once again forced to conclude that demonetisation has boosted growth, this time in gross value-added.

Which sectors contributed the most to the increase in growth estimates? Manufacturing accounted for a large chunk, very likely due to higher inventories and higher commodity prices (Chart 2). Agricultural production was the second biggest contributor.

For how demonetisation affected GVA growth, see **here**.

In short, using the method recommended by the government’s Economic Survey to gauge the impact of demonetisation on growth, we find from the CSO’s data that demonetisation has actually boosted both GDP and GVA nominal growth this fiscal year.

That conclusion seems to cry out for this bit of doggerel: Oh what a tangled web we weave/When GDP estimates we do believe.

Manas Chakravarty looks at trends and issues in the financial markets.

Respond to this column at manas.c@livemint.com.

To read more Capital Account columns, click **here**.