More than 19 months after the statistical tumult began over India’s GDP, the Narendra Modi government is finally trying to fix the mess.

India’s GDP numbers have been receiving flak since a new calculation was introduced in January 2015. Overnight, India’s GDP growth accelerated from 4.7% to 6.9%. It didn’t make sense to many, including the country’s chief economic advisor, Arvind Subramanian, or the former Reserve Bank of India governor, Raghuram Rajan.

The Indian government has now formed five committees that will review data used for major economic indicators like GDP, index of industrial production and inflation, the Economic Times reported on October 18. The primary aim is to make the data collection process more transparent.

The committees will also release data that is segregated according to industries and geographies, which is currently unavailable.

“These committees are expected to cover the requirement of statistics for estimation of GDP, data governance for quality, timelines and credibility of collected data and derived estimates, provide for data integrity and audit trials of a National Statistical System,” the ministry of statistics and programme implementation said in directive, according to the Economic Times.

Plugging the holes

The need to clear the ambiguity stems from the diverging trends marked by various indicators. For instance, the ground reality in Asia’s third-largest economy didn’t exactly mirror the spectacular GDP growth in the last year. While GDP accelerated, making India one of the fastest-growing economies in the world, industrial production numbers were muted. Private investment in the country also remains subdued and until May 2016, exports fell for 18 straight months.

The situation got so bad that even the Chinese – widely acknowledged as masterful economic data manipulators – accused India of cooking up its GDP numbers. The newly appointed committees will take stock of the data collected and clean up processes.

Currently there are different departments that compute various indicators, often leading to discrepancies. For instance, the department that computes the GDP doesn’t collect data on foreign trade. The committees will streamline the functioning of three main statistics departments that oversee the computation of critical economic indicators:

  • Central Statistics Office: The CSO was established by the government in May 1951, and since then has been responsible for maintaining the national accounts of the country, among other statistics. The organisation is the main hub of economic data collection and analysis, and various departments and state governments disseminate data to the CSO for further analysis. Apart from the GDP, it also computes the IIP and consumer price inflation indices, as well as compiles social sector statistics.
  • National Sample Survey Organisation: The NSSO was set up in 1950, and conducts nation-wide surveys about socio-economic indicators. The main aim of these surveys is to aid policy-making. For example, the NSSO collects data on sanitary conditions, drinking water access, education, employment, and wages. The surveys typically take years to complete, and hence the data is available with a time lag.
  • Directorate General of Commercial Trade and Statistics: This department is a part of the ministry of commerce and maintains data about India’s foreign trade – imports and exports. It also collects India’s country-specific trade data.

“There are significant holes in the Indian economic data with a lack of regular and consistent state level output and industry level data," said Rajiv Biswas, Asia-Pacific chief economist at IHS Markit, a consultancy. "As multinational corporations are increasingly looking to invest in India, they are seeking good quality state-level data for India but the available data is weak and very lagged in its publication.”

For example, Biswas said, the IIP data is volume-based indices using information from surveys that is collected through a variety of agencies and “are a dog’s breakfast of inaccurate and misreported data.”

“As a result the industrial production data is unreliable – it is a case of garbage in, garbage out,” Biswas explained. “An additional problem is that the industrial production data uses a very dated base year of 2004-'05 while the GDP data is using a 2011-'12 base year.”

This article first appeared on Quartz.