As institutions change, so does data credibility
An economy’s resilience and sustainability is best measured through its institutional strength
The official back data on India’s gross domestic product (GDP), released by the Central Statistical Office (CSO) on Wednesday, runs the risk of denting the market’s trust and conviction in official data released by government agencies. The new data release contradicts the earlier findings of a committee set up by National Statistical Commission to develop a methodology for deriving back data by linking the old series with the new base year of 2011-12. Whenever the base for national accounts is updated—as it should be to incorporate conceptual changes, statistical changes (such as methodological upgrades) or results of new surveys conducted closer to the base year—the old series is routinely linked to the new series for providing comparable data which can then be used by economists, academicians and analysts for creating a viable and comparable time-series. That is now a task riddled with occupational hazards that could throw up numerous contradictory outcomes. The segue from an old base year to a new base year should have ideally been smooth, as it has in the past; unfortunately, it has now become hostage to partisan electoral politics.
An economy’s resilience and sustainability is best measured through its institutional strength. And from the credibility of the data it discloses. By releasing data that competes with an earlier study without providing a robust theoretical reasoning to underpin that change, the CSO and the government think tank Niti Aayog may have undermined markets’ faith in official data. The new set of numbers strives to prove that growth under the previous government, led by former prime minister Manmohan Singh, was not as high as was estimated by the National Statistical Commission’s committee. Observers cannot be faulted for thinking that the CSO’s new series—which rejects the commission’s earlier findings without actually explicitly stating so—may have been tasked with neutralizing the earlier data set’s threat to the 2014 election campaign’s claim that the economy had stagnated during 2009-14. The new set also attempts to alter the perception that peak GDP growth occurred between the years 2004-05 and 2007-08; it creates an enabling environment to prove that economic growth under Prime Minister Narendra Modi has been comparable with the previous regime.
Analysts will have to wait a bit longer to also figure out changes made at the granular level. For example, the increased share of primary (which includes production of raw material, such as mining output, and basic foods) and secondary (manufacturing, processing and construction) sectors in the new GDP series, and a corresponding lower share of the tertiary sector, has seemingly contributed to the lower GDP growth data for the back years. It is thus moot whether there is methodological integrity to the entire exercise and whether the new results are the outcome of factors that are not entirely statistical. As economics Nobel laureate Ronald Coase famously said: If you torture data long enough, it will confess to anything.
Please click here to read more.
Livemint.com, 30 November, 2018, https://www.livemint.com/Opinion/pgeW67c4HhgYMUWZnoICnK/Opinion--As-institutions-change-so-does-data-credibility.html