Why national data will remain contested SUBHOMOY BHATTACHARJEE

The expansion and transformation of the Indian economy has altered the scale of measurement and the way data is gathered

Afew years ago, then chairman of the National Statistical Commission (NSC) Pronab Sen released the results of the sixth national Economic Census to a minuscule crowd of reporters in central Delhi. Last week, when the vice chairman of Niti Aayog Rajiv Kumar called a press conference to contest an employment report prepared by the National Sample Survey Office (NSSO), it was a media event that was posted on YouTube.

The firestorm over the latter report, reported by

Business Standard (the government maintains it is just a draft report), continues. National statistics are now contested territory and the battle over them will intensify in the years to come.

The Indian state prints statistical reports on a wide range of topics, mostly economic. Including monthly and annual ones, they number about 30 in a year, plus a few sample surveys. But as more such data has come under scrutiny the focus has turned to how it is collected.

The 4,800-odd officers of the Indian Statistical Service and the subordinate statistical services have a long history of compiling the data. The Indian Statistical Institute was set up in 1932, but the first synchronous census was conducted even earlier, in 1881. That makes India one of the oldest nations to set up a robust statistical machinery.

The Central Statistics Office coordinates these activities. It compiles the National Accounts from which the GDP data emerge. The index of industrial production, consumer price indices, human development statistics including gender statistics, the annual survey of industries and the economic census are all its domain. Assisting it is the NSSO, which carries out sampling on an all-India basis to throw up estimates for several of those data. Just as it did through the Periodic Labour Force Survey from April 1, 2017, which showed unemployment at a 45-year high for 2017-18, suggesting that demonetisation and the advanced deadline to introduce the Goods and Services Tax had taken its toll, a fact picked on gleefully by the opposition.

There is an interesting pattern to the recent datar elated controversy. Apart from employment data, the government also criticised the first series of revised GDP data on a new base year when it was published in 2015 for favouring the second term of the United Progressive Alliance. Things got complicated in 2018, when an NSC committee, which sets standards for government statistical work, calculated a back series with the new base year to show GDP growth through the UPA decades as being better than earlier reported. The statistics ministry rechecked the data and published a revised back-series months later that put the UPA’s performance in less flattering light, which the Congress party called “a joke”.

Beyond politics, most of the swings in the data have a lot to do with the huge changes sweeping through the Indian economy. “The challenge of measuring the economy has deepened,” says N R Bhanumurthy, National Institute of Public Finance and Policy professor and member of one of the key statistical committees.

From 2000, when it was $462 billion economy, India became a $2.13 trillion beast in 2018. Sectors within have also waxed and waned in importance. Imagine trying to measure a cycle running on a country lane and continuing to track it as it metamorphoses into a vehicle travelling on an expressway.

This is at the heart of the challenge India’s statisticians face. In recent years, almost every one of the periodic statistics of the economy has had to change both the scale of measurement and the way the data is gathered. The ministry of statistics has commissioned the seventh economic census, the results of which are due towards the end of 2019. As its name implies it is a sort of national business register that lists every sort of economic enterprise in the country from the smallest to the largest. It too could stoke controversy. Why?

Started in 1977, the database has had to be updated frequently. The fifth census was in 2005 and the sixth was published in 2014. Already several changes have become necessary since then. The number of virtual enterprises for instance, have shot up since then along with the number of people employed in them.

To get round these problems, the ministry of statistics has roped CSC eGovernance Services India Limited, under the ministry of information technology, to do the enumeration. This is the first time the ministry has asked an agency outside its own workforce to do data collection work, except for the population census. This is significant.

About 15,00,000 enumerators are being trained to gather the data for the economic census. According to Dinesh Tyagi, CEO of CSC e- Governance Services, the same force could also be used for the population census, which, if everything runs to plan, could be held every two years instead of 10 years at present.

The ministry also signed up with National Council for Applied Economic Research in January 2019 to obtain new data technologies and build capacity. For instance, the Urban Frame Survey, the bedrock of blocks on maps that statisticians use to select their samples, has been digitised since November 2017. It is now based on satellite imagery, in collaboration with the National Remote Sensing Centre.

All these changes mean more granular measurement instead of broadbrush estimates. Some of this has already happened like the replacement of the manufacturing sectorbased Annual Survey of Industries with the MCA-21 data compiled by the ministry of company affairs. The new sets of data tumbling out don't discredit the older ones; they are simply measuring a vastly different sized economy. But governments will need to be prepared to respond.