Opinion
The IIP is broken, here’s how to fix it
Summary
- The decline in the IIP’s reliability post-liberalization is part of a broader erosion in the state’s capacity to track private sector growth.
India’s statistics ministry generates only one high-frequency gauge of economic activity. And that lone barometer, the index of industrial production (IIP), is completely broken. In a 30 August research note, Credit Suisse analysts Neelkanth Mishra, Abhay Khaitan and Prateek Ancha show that the IIP has consistently understated growth in industrial output over the past decade.
They are the latest in a long line of analysts who have complained about IIP’s unreliability. In mid-2010, analysts were gobsmacked to find that astronomic growth rates in the production of alarm clocks and insulated cables were spiking up IIP growth. In 2018, my former colleague, Manas Chakravarty, wrote a tongue-in-cheek piece about indigestion driving up India’s industrial growth. At that time, the reported production of digestive enzymes and antacids had reached stratospheric levels.
You might also like
Are we ready for the switch from FASTags to GPS?
TPG to sell Manipal Health stake, promoter buyback plan fails
For sugar stocks, ethanol prices must hit sweet spot
Debt funds roiled, but fund houses are okay
Complaints on the IIP have been voiced across governments, and have persisted despite revisions in methodology. The statistical establishment has consistently failed to fix the index despite it being the most-studied statistical product in the country.
The Rangarajan Commission—set up to review the statistical system more than two decades ago—was perhaps the first official panel to flag serious discrepancies in IIP data in its 2001 report. The Committee on Financial Sector Assessment, set up jointly by the Union government and Reserve Bank of India (RBI) in the mid-2000s, reiterated those concerns. In 2011, the National Statistical Commission (NSC) appointed former National Sample Survey chief N.S. Sastry to conduct an independent statistical audit of the IIP. Sastry found several problems both in the collection and validation of data, and suggested remedies.
The audit report was supposed to serve as a template for other audits. But fate had other plans. Sastry’s recommendations were largely ignored. No other statistical audit has ever been attempted in India since then.
As complaints persisted, the parliamentary standing committee on finance asked for another review in 2012. Many of the problems identified by the reviewer, former RBI statistician R.B. Barman, would not have arisen had Sastry’s recommendations been implemented faithfully. In 2014, a panel led by Saumitra Chaudhuri submitted another set of recommendations.
But the complaints persisted. The NSC-appointed committee on real sector statistics headed by Sudipto Mundle reiterated concerns about the IIP’s unreliability in its 2018 report. Like Mishra and his colleagues, this committee also argued that the IIP was under-stating industrial growth. A 2020 working paper by economists Radhika Pandey, Amey Sapre and Pramod Sinha reported bizarre growth rates in several constituents of the index: from solar equipment to digestive enzymes.
The decline in the IIP’s reliability in the post-liberalization era is part of a broader erosion in the state’s capacity to track private sector growth. The dismantling of the licence-permit raj unshackled private entrepreneurs, helping them grow their businesses. At the same time, it created a vacuum in industrial data. In the absence of compensatory investments in the statistical system, the Indian state has struggled to size up the private corporate sector since then.
The IIP is based on data from a sample of industrial firms that is supposed to be representative of the entire universe of industrial firms. Unfortunately, there is no database that captures the universe of firms adequately. The existing databases—from the economic census to ASI to MCA-21—are all flawed in their own ways. Three decades after liberalization, we still lack a comprehensive and dynamic firm-level database that meets the economy’s basic statistical requirements.
Ineffective statistical leadership has compounded problems. Raw data for the IIP comes from source agencies spread across ministries. The statistics ministry fails to coordinate effectively to get accurate data for the respective sectors.
Statisticians are supposed to validate raw data before accepting it as meaningful ‘statistics’. When they fail to do that, you get unreliable indices such as the IIP. If the validation principles recommended by Sastry or Barman had been followed, the bizarre results reported by Pandey and her co-authors would have been detected internally.
Several panels have suggested that each monthly release of the index should include the number of sampling and non-responding units for each item. But the statistics ministry seems keen to limit disclosures.
The Rangarajan Commission had suggested the excise database be used for the IIP. In the same vein, the Mundle committee recommended exploring the GST database, which can help create a comprehensive business register. An accurate and dynamic business register can help generate high-frequency indicators for industrial and service sectors, and serve as a universal frame for conducting firm-level surveys.
New databases can help fix the index but these need to be tested before being plugged into the statistical system. The ministry needs to validate such databases and release the validated data for public feedback before deploying them for statistical purposes.
Once India’s statistical governance and disclosure norms improve, improvements in statistical products will follow..
Pramit Bhattacharya is a Chennai-based journalist. His Twitter handle is pramit_b
Elsewhere in Mint
In Opinion, Ashok Haldia tells how EY could shake up accounting. Narayan Ramachandran argues why G7 price cap on Russian oil will fail. Long Story narrates how Maruti Suzuki sped past Hyundai.
