Improve data collection

The government has in a welcome move decided to overhaul the economic data-gathering processes in the country. Five panels have been set up to review the gross domestic product (GDP). These panels will review data collection processes in the Central Statistics Office (CSO), the National Sample Survey Office (NSSO) and the Directorate General of Commercial Intelligence and Statistics (DGCIS). These panels will also suggest mechanisms to ensure data integrity and disaggregation by industry and region. A committee led by Ravindra Dholakia, also a member of the Monetary Policy Committee, will oversee the process. This move could not have come at a more appropriate time since data inconsistencies and contradictions are rampant and can mislead policy makers. In addition, inconsistencies reflect poorly on the credibility of the methodology and, ultimately, on the government. This problem has become particularly acute since many key statistical indicators have been reworked and re-based relatively recently.

For instance, the GDP calculation methodology has changed and the base year has changed from 2004-05 to 2011-12. The new computation uses market prices while the earlier method used factor cost. The new GDP calculations added roughly 200 basis points to growth since the recalculated GDP growth for 2013-14 stood at 6.9 per cent (later again revised downward to 6.6 per cent), whereas the earlier growth rate was five per cent. The GDP is supposedly continuing to grow at over seven per cent, but this is at odds with multiple other indicators. For example, a low index of industrial production (IIP) indicates weak manufacturing. Other data such as low bank credit, high non-performing asset (NPA) levels, low corporate earnings, debt restructurings, stalled projects, etc., all suggest the growth rate is an overestimate. Chief Economic Adviser Arvind Subramanian and former Reserve Bank of India (RBI) Governor Raghuram Rajan have publicly pointed out many inconsistencies. The CSO is now recalculating the GDP for 2004-05 to 2011-12 using the new method. The comparative details should throw some light on changes. 

Obviously, an understanding of the data is vital. If the economy is really growing at 7.5 per cent or better, is there any need for a fiscal stimulus, or monetary easing? To take another example of data-driven policy, the RBI now uses the consumer price index (CPI) as its key inflation indicator as against the wholesale price index (WPI) that was used earlier. But especially over the past two years when this change has happened there has been a wide divergence between the two indices. WPI inflation was in negative territory for six successive quarters starting October-December 2014 while CPI inflation stayed at above six per cent. Not surprisingly, this led to the government and the RBI combating two hugely different problems.

China has faced similar inconsistencies in data collection. Chinese Premier Li Keqiang trusts data such as electricity consumption, railway freight and bank loans far more than GDP estimates. But efforts to develop Keqiang-style indices using Indian data have thrown up more contradictions. A certain amount of contradiction may be inherent in a continent-sized economy with major regional differentials, but the inconsistencies are large enough to merit an overhaul of processes and methodology. The sooner the better.