GDP / National Accounts Revised Series with 2011-12 as base year
The structure of economic activities changes over time due to changes in structure of production and demand in the economy. On production side, the production pattern changes along with changes in technology and innovations in the system and in this process some production become obsolete and other comes in vogue. While on demand side, the consumption pattern also changes over time. The changes in relative prices stimulate changes in the consumption and production choices. Therefore, to account for these structural changes and to update the prices, the rebasing exercise is needed after a certain period. The exercise of rebasing national accounts brings up a fresh lot of information about the changes in economic structure of the economy, along with switching over to new base prices. This also helps in judging the size of the economy, correction of biases and looking afresh at the relative importance of sectors in the economy.
The recent introduction of new series of national accounts by Central Statistics Office (CSO) revised the base for National Accounts Statistics to 2011-12 from 2004-05 , which was last set in January, 2010. Along with revision of base, a number of methodological changes have also been made. Despite its comprehensiveness, the Press Note puzzled more people than it explained to; the user being surprised by unexpected moves of growth numbers. Many being apprehensive, warranted for a cautious approach to use new numbers for policy purpose. The confusion prevails on-which numbers should be used, old or new? This note tries to describe the changes made in the new series and why growth numbers for some of sectors, especially manufacturing, witness unexpected changes.
The new series of national accounts is an improvement upon old (base:2004-05) in terms of its comprehensive coverage of Corporate Sector and Government Activities and incorporation recent data generated through National Sample Surveys. It also brings up some change in methods of evaluation, approaches to account economic activities, introduced new concepts and incorporation of new classifications. Originally the base revision was due for 2009-10 but it was postponed due to the global financial crisis. In this base revision to 2011-12 prices, recent data sources such as NSSO Employment-Unemployment Survey 2011-12, Unincorporated Enterprise Survey 2010-11, Household Consumer Expenditure Survey 2011-12 etc. have been used in the estimation.
Earlier "GDP at factor cost" was known as simply the "GDP" in India. It is nothing but sum of the factor costs incurred during the process of turning out economy's output for the concerned year. Thus, it is a compilation of wages, interests salaries, profits etc. This concept - GDP at factor cost - used to be expressed both in constant prices (with 2004-05 prices as the base year prices) and current prices. For most purposes, including academic works, GDP at factor cost in constant prices was used as "GDP". Further by adding net indirect taxes (ie. product taxes - product subsidies), GDP at market prices were also reported in the National Account Statistics. (For details of the calculations one can see the brochure issued by CSO, Ministry of Statistics and programme Implementation)
In the revised series, as is the practice internationally, industry-wise estimates are presented as Gross Value Added (GVA) at basic prices, while "GDP at market prices" will be referred to as "GDP". GVA at basic prices can be referred to as GVA at producer price and GDP at market price as GDP at buyer price. Estimates of GVA at factor cost (earlier called GDP at factor cost) can be compiled by using the estimates of GVA at basic prices and production taxes less subsidies. It would result in effect on size of GVA compared to GDP at factor cost, which may be different for different sector. For example, net production tax being positive in manufacturing would result in higher GVA than GDP in the sector. New growth figures for GVA at Basic prices would also carry an impression of tax and subsidies which was not the case in GDP at factor cost. The production tax has been distinguished from product tax as the first is independent of quantity produced while the second varies with it. Similar distinction is also made between production and product subsidies.
for eg. Production Taxes - Land Revenues, Stamps and Registration fees and Tax on profession etc.
Production Subsidies - Subsidies to Railways, Input subsidies to farmers, Subsidies to village and small industries, Administrative subsidies to corporations or cooperatives, etc.
Product taxes or subsidies are paid or received on per unit of product. Some examples are:
Product Taxes: Excise Tax, Sales tax, Service Tax and Import and Export duties etc.
Product Subsidies: Food, Petroleum and fertilizer subsidies, Interest subsidies given to farmers, households etc through banks, Subsidies for providing insurance to households at lower rates etc.
GDP at market prices which is henceforth be referred as GDP, can be computed by adding net of product tax and product subsidies in GVA at basic prices.
Gross Value Added (GVA) at basic prices = compensation of employees + operating surplus/mixed income + consumption of fixed capital (CFC) or depreciation + Production taxes - Production subsidies
Value added is calculated as Output - intermediate consumption
GVA at factor cost (earlier referred to as GDP at factor cost) = GVA at basic prices – (Production taxes - Production subsidies)
Gross Domestic Product (GDP) = Σ GVA at basic prices + Product taxes - Product subsidies
(Note that it is not production taxes / subsidies but product taxes and subsidies)
In manufacturing many argue that the output was falling and then how come the new series shows that manufacturing sector was in fact performing not that bad. It was because although the output was stagnant or less but the value addition was better off. GDP is a measure of value added, it’s not about output. It’s the case of output being stagnant but value-addition is going up.
Earlier the sectoral manufacturing data value addition was sourced from the RBI Industrial Outlook Survey conducted on an quarterly basis; but now with the Ministry of Corporate Affairs making it obligatory on the part of the companies registered under the Companies Act for online reporting, the MCA 21 database has been used for the manufacturing sector value added. The MCA database as on date covers 5 lakh companies and is fairly representative of the universe. The RBI surveys are small in size and not much reliable for the sectoral analysis. Further, the manufacturing value added was calculated from ASI Annual data and extrapolated using IIP for the intervening period. The limitation with this data was that the ASI and IIP are establishment based data while the MCA database goes beyond establishment based value addition and also incorporates data on brand pricing ,marketing etc i.e. includes allied activities which were earlier outside the purview of manufacturing value added. Further the corporate segment manufacturing coverage accounts for almost 66-70 percent of the manufacturing sector.
Incorporation of National Industrial Classification - 2008 (NIC-2008 classifications) for industries is possibly one reason for the adjustments in the activities for an industry. The number of industries has increased from eight in old series to eleven in new series, the additional three industries reclassified within service sector named as “Transport, storage, communication & services related to broadcasting”, “Real estate, ownership of dwelling & professional services”, and “other services”. The description of industries has also been changed for example earlier “Community services etc.” become “Public administration and defence”.
The new, “Effective Labour Input Method” is adopted for Unincorporated Manufacturing & Services Enterprises for compiling the estimates for unorgansied non-agriculture sector. This method assigns due weights to different types of workers based on productivity and skills, unlike the earlier method which assumed equal value addition of each worker, irrespective of their skills and productivity. The adoption of new method is likely result in better estimates of value addition in the unorgansied non-agriculture sector.
In the Trade sector the gross value added was earlier calculated using the gross trading index this has been replaced by the index derived from sales tax collections. The Gross Trading Income (GTI) index tracked the growth in volume of tradable goods, in the economy, derived from current estimates of production in agriculture and manufacturing. The underlying assumption was that value added is strongly correlated with the physical volume of goods available for trade. This is a reasonable assumption in short intervals of time; however, when projections are extended over long periods of time, errors build up. This is because in addition to physical volume, value added also depends on levels of intermediation between the producer and consumers; changes in underlying quality of goods; and changes in marketing practises, for instance bundling higher quality value added services with goods like warranties etc. and so on. In the current new series, in addition to the updated surveys, this has also been partly corrected by changing the underlying indicator from a volume indicator to one based on value, namely sales tax collections. Since sales taxes are value based, growth in this indicator captures the underlying growth better in value added. Consequent to this and incorporation of new survey results, the 2011-12 estimates in the new series are less than those in the old series as can be seen from the table below.This would also change in the future when the GST is introduced and the Ministry of Finance is able to collect GST sector wise by invoking the Collection of Statistics Act, 2008.
Then the new series data collected from local bodies is also used and the coverage is 60 percent.
The ‘valuables’ segment, which basically comprises of gold and jewellery which is an important component of capital formation , was treated as consumption. In new series valuables are combined into household savings and, therefore, consumption has come down and savings have gone up accordingly. The new GDP numbers will be liable to changes in future based on change in base year of IIP WPI and CPI series. These are important indices which play a pivotal role when computing GDP at constant and current prices. Based on revisions of base year of these indices, GDP growth rates may change.
Differences in statistics with old and new base:
|GDP at factor cost at current prices (at 2004-05 prices)|
|GDP in Rs. Crore||GDP Growth rates in %|
|2011-12||2012-13||2013-14 (PE)||2014-15 (AE)||2011-12||2012-13||2013-14 (PE)||2014-15 (AE)|
|a. GDP at factor cost||8391691||9388876||10472807||15.8||11.9||11.5|
|b. net indirect taxes||618031||724405||882266||15.5||17.2||21.8|
|c. GDP at market prices (c= a+b)||9009722||10113281||11355073||15.7||12.2||12.3|
|As per New Series estimates||2011-12||2012-13||2013-14||2014-15 (AE)||2011-12||2012-13||2013-14||2014-15 (AE)|
|GDP at factor cost at 2011-12 prices||8206398||9263138||10487074||11702988||12.88%||13.21%||11.59%|
|GDP at market prices 2011-12 prices||8832012||9988540||11345056||12653762||13.09%||13.58%||11.54%|
|GDP at factor cost at constant prices (at 2004-05 prices)|
|GDP in Rs. Crore||GDP Growth rates in %|
|2011-12||2012-13||2013-14 (PE)||2014-15 (AE)||2011-12||2012-13||2013-14 (PE)||2014-15 (AE)|
|a. GDP at factor cost||5247530||5482111||5741791||6.7||4.5||4.7|
|b. net indirect taxes||385520||417736||454051||6.0||8.4||8.7|
|c. GDP at market prices (c= a+b)||5633050||5899847||6195842||6.6||4.7||5.0|
|As per New Series estimates (NS)||2011-12||2012-13||2013-14||2014-15 (AE)||2011-12||2012-13||2013-14||2014-15 (AE)|
|GDP at factor cost at 2011-12 prices||8206398||8609516||9178444||9865247||4.91%||6.61%||7.48%|
|GDP at market prices 2011-12 prices||8832012||9280803||9921106||10656925||5.08%||6.90%||7.42%|
|Source: Summary of macro economic aggregates published by CSO. Press releases of CSO dated 30 January 2015 and 9 February 2015|
The growth of GVA at constant prices reported higher at 4.9 % in 2012-13 with new base (2011-12) against 4.5% growth of GDP at factor cost at old base (2004-05). In 2013-14, the growth of GVA at new base is reported at 6.6% which was 4.7% at old base. The growth of GVA for manufacturing have shown noticeable rise in 2012-13 and 2013-14, which is updated to 6.2% and 5.3% at new base year compared to earlier 1.1% and -0.7% at old base year respectively. The sectoral share of manufacturing has also substantially increased in 2012-13 and 2013-14, as reported at 17.9% at new base compared to 14.1% at old base in 2011-12. These differences are surely on account of better coverage and changed methodologies and possibly change in concepts and classifications. It is crucial to know how much difference these factors has made to GVA and in which sector.
Note: Estimates for the earlier series (2004-05 series) have been derived from GVA at factor cost, while estimates for the 2011-12 series have been derived from GVA at basic prices.
The growth at constant prices, computed either using new base or old base prices, should not differ if other thing remains the same. In other words, merely a shift in base year cannot be accounted for the differences in growth at constant prices calculated using new or old base. With this logic, the differences in the sectoral growth rates calculated at constant prices of new and old base is plotted in the graph above. These differences inter-alia shows the under or overestimation bias in the growth of the economy, assuming that estimates in new series are better representation of the activities. In overall, the growth of the economy was underestimated with about two percentage point in 2013-14 in the old series. Although the reasons for this bias are not restricted only to structural changes, while the other factors such as conceptual changes, improved valuation methods, better coverage etc. have also been responsible. Among the sectors, the growth of ‘mining & quarrying’, ‘Manufacturing’, ‘trade, repair and hotels’, ‘transport, storage, etc’, and ‘community services etc’, appears heavily underestimated, while the growth in ‘Agriculture and allied’ and ‘Financial, real estate & business services’, was overestimated in old series.
Further, it can be seen that growth in current prices in 2014-15 is estimated at 11.5% as against 13.6% in 2013-14, though real (constant price) growth in 2014-15 is 7.4% as against 6.9% in 2013-14. i.e., the trend in growth rate shown by current price is exactly the opposite of the trend shown by the real prices. This is reported by MOSPI as due to the hugely differential rates of inflation. For instance, there has been a reduction in the growth in the underlying price indices, Wholesale price index (WPI) and Consumer Price Index (CPI) in 2014-15 as compared to the corresponding growth in 2013-14. WPI and CPI, increased by 3.4% and 6.0% in 2014-15, as compared to 6.0% and 9.5% in 2013-14. Consequently, the GDP deflator (implicit price deflator for GDP which is a measure of the level of prices of all new, domestically produced, final goods and services in an economy; it can be broadly thought of as GDP at current price / GDP at constant price) increased by 3.8% in 2014-15 as against 6.2% in 2013-14, leading to diverging trends. (Growth in GDP can come from either change/growth in prices (P) or from changes in quantity (Q) of output. Increase in P for 2013-14 was far greater than increase P for 2014-15 and it outweighed the changes in growth in output (Q) to give it a negative direction.)
The reasons for the rise in growth for manufacturing sector at new base are structural as well as change in compilation methodology. The methodological changes includes the change in approach, better coverage, use of new valuation methods and introduction of new concepts. Some of these highlighted in the Press Note are as follows:
- The shift from Establishment approach to Enterprise approach: The establishment approach used in Annual Survey of Industries did not capture the activities of a unit other than manufacturing. Whereas, an enterprise along with its manufacturing activities is also engaged in activities other than manufacturing such as ancillary activities etc. Now, in new approach, the activities of a manufacturing company other than manufacturing are accounted in manufacturing sector. The enterprise approach is facilitated by MCA 21 data with Ministry of Corporate Affairs. These changes possibly have increased the coverage of registered sector of manufacturing.
- Incorporation of findings of NSSO Surveys: The details of new NSS Surveys viz. Unincorporated Enterprises Surveys (2010-11) and Employment & Unemployed Survey, 2011-12 are now available, therefore incorporated in the new series. The updates are an improvement in the representation of activities in the unorganized manufacturing sector.
- The change in Labour input Method: The new series has switched over to “Effective Labour Input Method” for Unincorporated Manufacturing & Services Enterprises. Earlier method was assigning equal weights to all types of worker, while the new method assigns different weight for workers as per their productivity.
- The inclusion of production tax less subsidies: The net of production tax and production subsidies is positive in manufacturing, while it is inter-alia negative in ‘agriculture and allied’ and ‘Electricity, gas etc’. Therefore, the positive net production tax would increase the size of GVA in the sector in absolute and relative to other sectors. Moreover, any change, including change in policy, if alters the lump sum production tax and subsidies then this may also likely to reflect in the growth rates in the sector.
In sum, one can say, the vast difference in the new series figures is not just because of updation of the database or change in methodology but more so because of the change in data source. Trend analysis not possible right now but can be tried by working out the difference in the ASI value added and the MCA database, which is left to the academicians.
The new GDP numbers will be liable to changes in future based on change in base year of IIP WPI and CPI series. These are important indices which play a pivotal role when computing GDP at constant and current prices. Based on revisions of base year of these indices, GDP growth rates may change.
Also see the FAQ released by Ministry of Statistics and Programme Implementation on the matter.
Gross Domestic perplexity, Arvind Virmani, Indian Express, 13 June 2016
- Jitender Singh, IES(2008) with inputs from Ms A Srija, IES(1996)
- Email- firstname.lastname@example.org, email@example.com
- Authors acknowledge with thanks the comments received from Mr. Shirke Shrinivas Vijay (ISS 2011) at the draft stage.