Economic growth estimates are a trap, not to be taken at face value

GDP revisions aren't unique to India. Internationally too, early estimates suffer from limited data availability, an issue that's addressed by successive revisions.
For representational purpose (Photo | IANS)
For representational purpose (Photo | IANS)

Economic growth estimates are a trap. For a given financial year, the government issues a salad of seven statistics in all, but the actual state of the economy becomes clear and certain only if you wait a tireless three years.

First advance estimates are trickier among all and rarely tell you the whole truth, simply because they are tentative and liable to change. Put another way, these estimates are much like statisticians playing a hunch based on instinct and intuition. We've hauled in the latest GDP data for interrogation, and here's what we found.

First, the FY21 numbers. Two years ago, we were told that the Covid-19 pandemic has cut the Indian economy fiercer than a Samurai blade, hemorrhaging Rs 11.26 lakh crore off our national output. The country suffered the sharpest contraction ever at 7.7%. Turns out, they are inaccurate. Thanks to subsequent data revisions, the contraction is now placed at a relatively respectable 5.8%, which means, only Rs 8.5 lakh crore worth of output was hacked off our economy, far lower than the dreaded Rs 11.26 lakh crore.

But wait. This isn't the final estimate, which will be out only next February and chances are, the true extent of the economic scarring could be lower than Rs 8.5 lakh crore. Even if it remains unchanged, there's no reason to worry. As Chief Economic Advisor Dr V Anantha Nageshwaran noted last month, we're over the hump having covered the lost ground with the economy adding about Rs 12 lakh crore worth of output in FY22.

One may argue this is a simplified view and instead one must consider the opportunity loss. The RBI did some calculations last year and concluded that our economic scarring was double the government's estimate at Rs 19.1 lakh crore. In a further spirit-amputating tone, it observed that it would 13 years for us to recover fully. The RBI's understrappers arrived at this number by recreating a growth path using the past trend of 6.6% and then estimating the GDP level assuming a growth of 7.5% for the next 13 years.

Given that we are no longer projecting a 7.5% growth in the medium-term, it's likely that recovering the estimated output loss may take even longer. Meanwhile, last week's GDP data release also has another shocker. The third revised estimates or the final estimates for FY20 indeed confirm that the slowdown -- which the government was in denial for most part of the year -- was acute than initially imagined. If the first advance estimates pegged FY20 real GDP growth rate at 5% or Rs 147.78 lakh crore, the final numbers show that the economy grew at a molasses-like 3.9% with output settling at Rs 145.34 lakh crore. That's a significant Rs 2.4 lakh crore overestimation.

In fact, this base effect led to subsequent corrections in both FY21 and FY22 data. For FY22, the latest projection of 9% growth rate is lower than the first advance estimate's 9.2%, but in absolute numbers, the national output is expected to be higher at Rs 149 lakh crore as against the previous estimate of Rs 147 lakh crore. This is only the first revised estimate and the final number is expected two years from now. The correction in FY22 base is also influencing the current year's estimates, which means, the estimated slowdown in Q3, FY23 at 4.4% may well be seen as written in water.

GDP revisions aren't unique to India. Internationally too, early estimates suffer from limited data availability, an issue that's addressed by successive revisions in the form of second advance estimates, provisional estimates, and three revised estimates.

According to a 2018 RBI study, on most occasions, initial data releases either underestimate or overestimate real GVA and GDP growth. This is mainly because firmer data are captured in successive rounds of revisions accompanied with gradual increase in data coverage. Interestingly, the study observed a bias each time the growth cycle ‘turns.’

There were substantial upward revisions during growth years, and downward revisions during slowdown and financial crises. For instance, advance estimates for FY10 pegged real GVA growth at 7.2%, while it settled at 8.6%, an underestimation of 140 bps. On the other hand, in FY09, real GDP growth was revised downwards from 7.1% to 3.9%, translating to an overestimation of 320 bps.

The bottom line is, not every GDP estimate should be taken at face value.

Related Stories

No stories found.
The New Indian Express
www.newindianexpress.com