Home / Opinion / The way forward for out-of-sync IIP

The Index of Industrial Production (IIP) is a short-term indicator of manufacturing performance. Being a volume index, IIP measures proportionate changes in the quantities of a specified set of manufactured, mining and electricity products over two periods of time. Changes are weighted by their economic importance in one or other, or both, periods. IIP is considered to be a lead indicator and is used as the core ingredient in the compilation of annual and quarterly national accounts. It differs from its corresponding gross domestic product (GDP) component as it does not represent the universe. Based on a large statistically valid sample and due to the law of large numbers, deviations are expected to be in a narrow band. However, we now see divergence in the growth rates of manufacturing measured in terms of IIP and GDP (see chart).

A study of the earlier series of IIP also indicates that IIP suffers from a downward bias in manufacturing growth. Further, weaker association in manufacturing growth between IIP and GDP is seen beyond the immediate years of release of the IIP series.

Representativeness of IIP has been a matter of debate in the post-industrial deregulation and reforms era, particularly after 1991. Though the submission of production returns was mandatory, a general perception was that the response had deteriorated. Further, weaker association in later years may be because the primary data that is used for computing the index became poorer in quality and probably scarcer in quantity arising partly from substantial non-response.

IIP, based on fixed weights, both at broad manufacturing groups and at item-level derived from the Annual Survey of Industries (ASI)/National Accounts Statistics, became dated, as the exercise for changing the base year itself commenced after nearly a decade and it took around five years before a new series was actually released.

IIP uses hybrid frame of units. For nearly 50% of the IIP weight, the frame of units is fixed. IIP in these products captures only the change in the production of the units in the frame. This leads to a downward bias because it ignores the new units which come into production.

Another issue in IIP is the unit of measurement. Since the index is a fixed base, real growth for most of the items is measured in terms of quantities such as tonnage and numbers. Only 54 items with a weight of 11.5% in IIP (15.3% of manufacturing) are measured in value terms. But the problem in items that are measured in terms of quantity such as locomotives, wagons, commercial vehicles, cars, drillers and loaders is that it bundles the commodities of different values into a single unit.

The downward bias in IIP was earlier ignored as complete information on manufacturing used to take considerable time. ASI takes two-three years for summary results. National Sample Survey Organisation’s (NSSO) unorganized sector survey is a five-yearly exercise. The ASI and NSSO surveys have been mutually exclusive and non-overlapping and cover the entire manufacturing sector, but they are delayed, leading to dependency in the earlier series of GDP.

The new series of GDP with 2011-12 as its base, however, divides manufacturing as originating from the corporate sector and the household sector. The corporate sector data is now available more comprehensively from the MCA 21 database. Further, as against the share of the unregistered sector in manufacturing value added of over 30% during 2008-11 (old GDP series with 2004-05 as base), the share of the household sector in value added from manufacturing is only 8.6% (2011-14).

Currently, data for the manufacturing subset of IIP is sourced from 16 source agencies. Fifteen of these agencies are commodity-specific. But direct sourcing of data from companies may not only be cost-effective but also homogeneous. The arm’s length relation between data users and data generator will ensure that this is not contaminated with the users’ bias.

Wider differences in the results of the manufacturing sector as observed from IIP and GDP distort expectations. Most of the apprehensions about the new GDP series came about because of a wider discrepancy between the two sources of manufacturing data. While earlier there were limited options, the availability of MCA 21 data sets and their mandatory filing provide a viable option to address this expectation distortion.

In the post-liberalization period, the use of IIP was restricted to a tracking variable for quarterly and annual GDP series. The Department of Industrial Policy & Promotion rarely uses the information for any policy corrections. Since the MCA data is available on a quarterly basis and that being coterminous with the quarterly GDP estimates, it is time that we discontinued with the IIP series which is at odds with industrial reality and so much at variance with other information.

Rajarangamani Gopalan and M.C. Singhi are, respectively, former secretary and economic adviser to the government of India.

Comments are welcome at

Subscribe to Mint Newsletters
* Enter a valid email
* Thank you for subscribing to our newsletter.

Never miss a story! Stay connected and informed with Mint. Download our App Now!!

Edit Profile
My ReadsRedeem a Gift CardLogout