For a comprehensive inflation index

For a comprehensive inflation index

There is an urgent need to establish a robust data collection mechanism for services, like we have for goods. Without it, the methodological error in inflation measurement will persist, say Rashmi Rastogi & Tanu M Goyal


THE Wholesale Price Index (WPI) is a broad-based measure of inflation , widely used for policy making in India. The index, however, has been acknowledged as incomplete since it includes only goods. In order to give a complete picture of inflation, the need for a services-based index has already been recognised among policymakers . It is important to note that the Producer Price Index (PPI) in the US, the near counterpart of WPI in India, introduced services in 1995 and has been regularly updated by the inclusion of new services. Against this backdrop, a team at ICRIER constructed a preliminary service-based inflation index which illustrates considerable variation with existing measures of inflation. The policy implication is clear. We need to fast-track a services database to prepare the basis for a comprehensive inflation index. This would dovetail precisely with the governments declared intention of launching a monthly measure of inflation. It is also an opportune time for the new government to correct this historical anomaly in inflation measurement.
Inflation is between 0 and 1% at present , but last year it was different. The highest recorded year-on-year inflation as measured by the WPI was in August 2008 at 12.81%. A number of economists however, felt that the actual increase was not as severe as recorded by the WPI. This is because it did not account for the price of services, which remained more or less constant over that period.
There are several reasons that prompt inclusion of services in the price index. First, over the years, share of services in the GDP has increased gradually, services now account for a little more than 60% of GDP. Second is the increasing contribution of services to international trade. Indias trade in services as a percentage of GDP stood at around 3.37% in 1990 which increased to around 15.37% in 2007. Third, services contribute to the government indirect tax collection via the service tax and this share has been steadily rising. Finally and most significantly, rising share of services in private final consumption expenditure (PFCE) mandates their inclusion in the basket. In 2007-08 , one third of the total PFCE was on services . It has gradually increased over the past few years from about 26% in 1990 to around 33% in 2007-08 . Non-inclusion of services therefore does not give a representative picture of inflation.
While the government has recognised the need for a comprehensive index including services, the proposal has been gathering dust in the ministry, due perhaps in large part to the immense resources it will need. The benefits that such an index will bring to bear on policy formulation, especially inflation targeting, will however more than justify its construction.
Our attempt to address this important issue revolved around creating a services index and merging it with the existing WPI to arrive at a composite price index. We recognise its infirmities, but take heart in the fact that our composite index usually predicts inflation to be in between the CPI (IW) and WPI, a dominant view amongst economists in government.
The services index has been constructed using weighted average method, where weights are assigned according to contribution of each service to PFCE in 2006-07 . Due to non-availability of data on certain services, the index includes services which are part of CPI (IW) and CPI (UNME). Services included in the index are housing, medical care, education, recreation and amusement and transport & communication.
IN ADDITION, prime lending rate, a proxy variable of financial institutions is also included. In order to obtain a comprehensive measure of inflation, the constructed service index is combined with WPI. Weighted average method was used to construct Composite Price Index where weights to services (33%) and WPI (67%) are based on consumer expenditure on goods and services, according to NSS 63rd round.
Inflation in composite price index shown in accompanying graph represents a comprehensive measure of inflation which is equivalent to national income deflator comprising goods as well as services. Divergence in two price indices Service Index and WPI explains the behaviour of inflation in the Composite Price Index. The reason for divergence in the period April 05 to December 05 was due to very high consumer expenditure on housing rentals and transportation & communication. Average inflation in housing rentals reached to 14.6% and in transport & communication inflation was 7.1%. During the period from March 08 to December 08, the Indian economy was influenced by external factors that resulted in the increase in WPI fuel and WPI primary. Average inflation in WPI primary and WPI fuel was 10.8% and 10.9%, respectively. Inflation in WPI primary goods touched to 12.4% in October 08 and WPI fuel reached to 17.2% in July 08 while the prices of services did not show much increase in the corresponding period.
This picture can be readily observed in the graph. The time period between September 06-May 07 and March 08-February 09, average inflation in service index is 4.6% and 4.1%, respectively, while average inflation in WPI is 5.9% and 8.9%. Low inflation in services index pushed down inflation in composite price index, i.e., 5.9% average inflation in September 06-May 07 and 7.3% in March 08-February 09. On the other hand, when the average inflation in service index is higher, i.e., 6.6% between April 05-December 05 and lower WPI inflation 4.6% pushed down the inflation in composite price index to 5.2%.
The results clearly suggest the need for a comprehensive price index that includes services in the measure of inflation. ICRIERs endeavour reinforces what we have known for a long time now, i.e., the current baskets in the WPI and CPI are obsolete and need to be updated and new weights assigned in the index. It is, therefore , recommended to establish a robust data collection mechanism for services, like we have for goods. Without it, the methodological error in inflation measurement will persist.


(Rashmi Rastogi is a research associate and
Tanu M Goyal is a research assistant