In recent weeks, the Wholesale Price Index (WPI) has fallen for the first time in over 30 years. This was described in many places, on television and in print, as inflation “enters negative territory”, making it sound like the Pakistani army entering Swat! Unnecessarily convoluted modern verbiage, or if you prefer, too much jargon, to describe the simple phenomenon of falling prices or deflation (Inflation is the percent change in the level of the price index).
Policymakers have been opaque or inaccurate in their responses. The Reserve Bank of India’s (RBI) governor has been quoted as saying that “India doesn’t suffer from any demand contraction, therefore there should be no concern about negative inflation” and that “negative inflation should be seen more as a statistical feature rather than a structural attribute of the economy.”
Also See Fuel Versus Manufacturing Inflation (Graphics)
Many others have also pointed to the statistical feature known as the “base effect”, for the fuel component of the WPI, as the reason for falling prices now. The base effect refers to the temporary rise (or fall) in inflation due to a sudden, large fall (or rise) in the price level last year (period) at the same time. Hence current inflation is not a good indicator of underlying inflation. A strong base effect implies that the inflation rate over a long span will be negatively correlated with the last period’s inflation rate.
In early June 2008, when the administered price of petroleum products were raised, the fuel component of the WPI jumped and the WPI for that week also rose sharply into double digits (see graphics). The finance ministry called a conference to discuss these adverse developments. Subsequently, fuel prices were lowered in early November. Hence, “fuel group inflation”, which had risen from 7.8% to 16.2% last year, has fallen from—6.7% to—12.8% this year around that time (see graphics). Hence the base effect is very much present in recent data for fuel inflation.
The above may seem to suggest that the milder WPI deflation is also due to the base effect. Outright deflation, as is occurring in many countries since last year, would be cause for alarm. It is harder for the economic system to cope with deflation than mild to even moderate inflation, for reasons that Keynes first explained eloquently ages ago. Hence, the statements by officials playing down deflation are not surprising.
The base effect in response to the fuel group price decline is glaring. However, it does not follow that we should attribute the recent WPI fall to the base effect. The reason for this is straightforward. The price level (and thus the inflation rate also) is a weighted average of the price levels of the underlying subgroups: primary, fuel and manufacturing. Manufacturing is the biggest subgroup—it comprises the lion’s share, about two-thirds of the WPI (see table). Thus most of the movements in the WPI and its inflation rate closely follow those in manufacturing. Hence it is unlikely, although not impossible, that deflation can result from the drop in fuel prices. Rather, manufacturing inflation must also be low.
Further scrutiny of the data reveals this to be the case. The approximate inflation contribution in percentage points for each of the subgroups can be calculated by multiplying the inflation rate for each subgroup by its weight (A technical clarification: The exact inflation contributions, which will precisely sum to the inflation rate, need to be calculated differently). For the last three weeks reported in the table, the contribution of manufacturing to inflation has been close to zero, while that of fuel obviously negative.
Can it be the case that changes in fuel prices simultaneously feed into manufacturing, so that fuel prices are the underlying causal factor, even though the contribution of manufacturing in an accounting sense will invariably be greater? To some extent, yes. When fuel inflation jumped last summer from 7.8% to 16.2%, manufacturing moved up from 9.6% to 10.3%, not a small move. However, there have been moderately big moves in total inflation without fuel changing, reflecting independent moves in manufacturing inflation, e.g. a 105 basis points rise in early April 2009 followed by a 125 basis points drop the next week, despite the WPI for the fuel group being rather flat then for five weeks in a row.
Overall for the last 10 months, manufacturing, and thus total WPI inflation, has been falling fairly steadily due to the global recession under way. The table shows the base effect impact of changes in the administered price of fuel. From mild “deflation” in the summer of 2007, the fuel group inflation rate zoomed in 2008 and fuel deflation has returned in 2009. However, manufacturing inflation has been falling steadily since the world economy collapsed last September.
Our conclusion that the deflation reflects the weak economy and not the fuel base effect does not imply that the RBI should push rates closer to zero, as other central banks have done. After all, the more crucial measure of inflation, the Consumer Price Index is running close to 10%. RBI is paying some heed to that now.
Monetary policy is very hard to conduct when food inflation is high, possibly due to ongoing rural fiscal stimulus, while the WPI is falling. However, a correct interpretation of the data is a prerequisite for suitable measures. This article is meant merely to interpret recent data. To reiterate, the falling WPI should not be attributed mainly to the base effect for fuel prices. Instead, it is due to the weakness in manufacturing.
Graphics by Ahmed Raza Khan / Mint
Vivek Moorthy is a professor of economics at IIM Bangalore. Comment at firstname.lastname@example.org