Coincidentally, RBI has taken the tone of doing everything for growth and hence the inflation number may not matter in the broader scheme of things for the MPC.
The shutdown announced by the government in March and the subsequent extensions over the last three months have made gathering statistics very difficult. The government organisations that collect such data find it hard to do so as the sources have been partially functioning and getting quotation on prices or production is not complete. Therefore, when the data is released on inflation or industrial output, there is a caveat put that the data would be subject to revision.
Various organisations have their forecasts on how India’s GDP growth would look like for FY21. The first one from the IMF put it at 1.9% in FY21, which would be higher than even China. This was in early April. The UN had earlier mentioned that India would be a shining star this year along with China. This was the period when India had very low infected cases and, prima facie, it looked like that the 21-day lockdown would have achieved the goal of containment and that there would be no extensions. However, there have been several extensions with modifications on what can be permitted with different laws across states, which, in a way, has not made business operations easy.
This has changed the way one has looked at growth prospects for India, and forecasts brought out by various investment banks, commercial banks, research agencies and scholars came out with numbers ranging from (-)1% to (-)25%. Such variation can baffle the reader because one is left wondering how these numbers are arrived at. Normally, forecasters vary by 0.5-1% point on either side of an average. But such variation is quite remarkable and deserves explanation.
Any econometric model used is based on certain explanatory variables that are defined numerically. But today the main factor guiding the equation/s is a black swan event of a shutdown with an undefined tenure which cannot be numerically expressed and has no precedent. Therefore, no model can throw up a GDP number as all the so-called independent variables that are being used have been assumed to take a certain shape. Hence, depending on what the estimator or forecaster assumes to hold for the year, a different GDP growth number is deduced. Even if a sectoral approach is taken, can one be confident of construction getting back to normal in September or could it be December. For hotels and travel, would it be Q3 or Q4 or maybe not even this year? Hence, the crux is the set of assumptions being made on how soon various components of GDP start moving during the year.
Roughly speaking, one knows that the GDP number for the year is split almost evenly across the four quarters ranging from 22-23% to 25-26%. The first quarter is a virtual washout with few exceptions in the non-services sector. But from Q2 onwards, the assumption made would be critical. The IMF had spoken of the second half being the inflection point and RBI also hinted at this period being different. However, observing how things have worked so far, the virus affliction numbers are rising and while the government is opening the economy, one is never sure if it will be shut again after three months if the situation escalates—where there is a fair possibility. Therefore, this entire exercise is akin to the analogy of monkeys throwing darts as any number can turn out to be true. A forecast based on a model that says 20% fall or gut feeling of 10% or a plain shoulder shrug of, say, 5% decline is at the same level!
The other problem is with the numbers being brought out by the CSO or the Office of the Economic Adviser. If data collection is incomplete, there may actually be no point in disseminating the same as it can only mislead given that there can be substantial revisions in the future. Even under normal conditions, the provisional figures change when the actual numbers flow with a gap of 2-3 months. The race to bring out provisional numbers by a fixed date runs this risk of interpretation and can also affect monetary policy decisions if it pertains to prices, which will be the case this year as CPI data is only partial.
Coincidentally, RBI has taken the tone of doing everything for growth and hence the inflation number may not matter in the broader scheme of things for the MPC. But the CPI indices for this year will influence those of FY22, which may then be the signpost for the MPC.
The challenge is to make sure that the data that is brought out when things get to normalise is correct as both the output numbers and prices get into all economic ratios whether it is fiscal deficit or outstanding debt or CAD etc. There is a likelihood of these numbers changing significantly as data comes in. While output data is based on what is reported by individual units which if missed in April or May can finally be reckoned by March when the accounts are closed, the same does not hold for price data which will be lost. Averaging of prices will address partly the issue but cannot correct the same. Currently, too, there has been a tendency to understate food inflation as the lockdown has increased prices paid by households which is not captured adequately due to the system of obtaining quotes from specific touchpoints. Extreme shortages have made consumers pay more than has been reported.
The problem with data interpretation would carry forward for the next year too, where most estimates are in the higher range of GDP growth being above 8%. While this would be a statistical possibility, it would not reflect real growth as for the first time the country would be coming out from a negative GDP growth number in FY21. Therefore, data disseminated in FY22 would also carry the risk of overstating the true position.
The way out is to eschew relatively high frequency data this year until there are signs of normalisation to ensure that the right signals are given. Price deflators would be probably more relevant in stating the true inflation situation. The proxies used for the unorganised sector would be probably even more chaotic as the MSME business has been completed distorted. From the point of view of analysts, monetary data could be indicative proxies—though even here with the plethora of relaxation made by RBI, growth in credit may have its limits when used for interpreting real economic activity.
The author is chief economist, CARE Ratings. Views are personal