Coronavirus | IIP and CPI become a casualty when data is having its moment

Representative image

Biostatistics and data science are having their moments. Rummaging through the mountains of data that COVID-19 is throwing up is turning out to be a crucial determinant in mankind’s ability to win its battle against the virus.

Paradoxically, however, the pandemic’s spread, enforced lockdowns, and physical distancing, have made some commonly put out high-frequency data the casualty.

In an ordinary pre-COVID-19 world these should have been par for the course. Not any longer. Or so it appears.

How can researchers collate and collect data on price trends and factory production, when factories and shops have remained shuttered?  How do researchers step out of their homes when everyone has been asked to remain confined to their homes?

India’s incomplete factory output and retail price data that the National Statistics Office (NSO) released on June 12 needs to be seen in this light.

The NSO data showed that India’s factory output measured by the index of industrial production (IIP) contracted by a record 55.5 per cent in April.

This was on expected lines. The entire country was in a forced lockdown during the entire month. With no production taking place, output will only contract sharply compared to the previous year.

But there’s a caveat that the NSO has flagged. These numbers were based on partial information and is not comparable with previous months and that the figures will be revised later.

Here’s what the NSO said. “In view of the preventive measures and announcement of nationwide lockdown by the government to contain spread of covid-19 pandemic, majority of the industrial sector establishments were not operating from the end of March. This has had an impact on the items being produced by the establishments during the month of April where a number of responding units have reported nil production. Consequently, it is not appropriate to compare the IIP (Index of Industrial Production) of April 2020 with earlier months, and users may like to observe the changes in IIP in the following months. These quick estimates will undergo revision in subsequent releases as per the revision policy of IIP."

The statistics department also did not release retail inflation data but said food inflation in May slowed to 9.69 percent from 10.5 percent in the previous month.

The government gathers and organises IIP data through 16 data source agencies located in various central ministries and departments. These, in turn, collect data from their primary sources — the factories.

The frame for coverage of units is decided by the source agencies. For IIP, both large and medium factories are covered.

The source agencies also decide the sample size for data collection. Fourteen such source agencies compile the output data for the manufacturing sector's data, which accounts for 77.6 percent of IIP.

These source agencies are located in various ministries, departments and the government’s subordinate offices. For instance, for the mining sector, the Indian Bureau of Mines, collects data on metallic and non-metallic minerals, and compiles the index.

Likewise, for the electricity sector, the central electricity authority collects and organises the monthly power production data.

These organisations collect the data through various methods including mail, questionnaires, and paper forms at the level of individual establishments.

Importantly, the source agencies report to the NSO only the item-wise total production of all the units and not the production of individual factories and establishments.

Similarly for price data that is usually collected from selected 1,114 urban markets and selected 1,181 villages through personal visits by field staff.

According to the NSO, in view of the lockdown “the price collection of Consumer Price Index (CPI) through personal visits of price collectors was suspended with effect from March 19”.

It said that for May, the data has been collected, to the extent feasible, through personal visits of field staff, telephone calls from the designated outlets in select markets, and information collected during the personal purchase of field staff for the items transacted from neighborhood outlets.

The importance of high quality, credible data in policy making cannot be overemphasised.  All the more so in times of the COVID-19 pandemic that has left a shattering trail of devastation across the economy.

Any repair task begins with identifying the problem and the extent of the damage. The former is a matter of prima facie evidence, the latter is a function of sound data.

For instance, IIP and retail inflation, measured by the consumer price index (CPI), are the two main markers that the Reserve Bank of India (RBI) tracks to gauge the economy’s health. In fact, the RBI is now mandated, under law, to keep the retail inflation within the two to six percent range, with the CPI serving as the primary guide for interest rate related decisions.

But what if the government says the CPI data is due for revisions and the latest data is imperfect?

The economy’s rapidity of revival will critically depend on policy agility and nimbleness that should be guided by quick and reliable time series data.

Otherwise, policy responses will amount to exercises similar to shooting a target in the dark that is also rapidly moving.