Any trained economist uses the word deflation with caution and will almost certainly define it as a sustained decline in prices, coupled with demand contraction. Apart from consumption and investment drops, which can spiral downwards, deflation has perverse redistributive effects. For the week ending March 14, point-to-point WPI (wholesale price index) shows inflation of 0.27%. That?s still a positive rate of inflation. However, there is perennial confusion between a rate of inflation declining and prices coming down. Consequently, people complain that though the rate of inflation is declining according to government figures, prices increase. Of course, they do. Prices decline only when the rate of inflation becomes negative. A decline in the rate of inflation only means that prices are increasing at a slower rate. There is an additional problem with government figures, or any figures for that matter. Perceptions of inflation are invariably based on some notion of which commodities are important, that is, weights attached to these. After all, prices move in different directions and to gauge what is happening overall, an index is needed. That means choosing a basket of commodities, attaching weights to these, accepting a formula for aggregation and collecting price data. Consequently, inflation by WPI is one thing. Inflation by any of the three CPIs (consumer price index) is another. And the GDP deflator, used to convert nominal GDP to real GDP, is another.

There are problems with each of those four issues mentioned in constructing WPI. This government set up a committee to revamp WPI and recommendations were submitted. They haven?t been implemented. So we are stuck with a WPI that has problems. However, that doesn?t mean a government deliberately falsifies data. Perceptions about high inflation are primarily based on food inflation. But the weight of primary articles in WPI is only 22.02% (some food articles also figure under the manufactured category). If 70% of my expenditure basket consists of food (quite likely for the poor), obviously an index with 22% weight will under-play my perceived inflation. Incidentally, WPI was last revised on April 1, 2000. Had earlier WPI continued, weight of primary articles would have been 32.3% and inflation, even by WPI, would have been higher today. There are (or were) four different CPIs ? UNME (urban non-manual employees), IW (industrial workers), AL (agricultural labourers) and RL (rural labourers). CPI (UNME) has been discontinued, because there will eventually be a CPI (Urban). However, until CPI (Urban) surfaces, CPI (UNME) continues to be computed. This has 47.13% weight for food, beverages and tobacco and is last available for January 2009, when it showed 14.0% inflation for food, beverages and tobacco and 10.4% overall.

CPI-IW is also last available for January 2009 and shows inflation of 10.45%. I don?t quite understand what?s happening to CPI (AL) and CPI (RL). They are probably headed for merger and will become something like CPI (Rural). That combined index is last available for February 2009 and shows an inflation rate of 10.79%. If the four (or three, or two) CPIs are showing inflation rates in excess of 10%, why are we talking about deflation? That doesn?t mean inflation measured by any of the CPIs isn?t declining. It is declining and is probably headed for the 5-6% range, depending on which CPI one uses. Even then, any talk of deflation is unwarranted. The problem is we don?t discuss CPI. We don?t discuss CPI, because it comes with a time-lag of more than a month. WPI now has an effective time-lag of 12 days. WPI isn?t what one should use to discuss inflation and policy related to inflation. However, because of delivery problems with CPI, we are reduced to this wholesale price idiocy (WPI).

It actually gets worse. Because for deflation to occur, decline in prices must be sustained. But that 0.27% is a point-to-point figure, a comparison of the level of the index on March 14, 2009 with that on March 14, 2008. That?s a function of the level of the index in the base period. A higher base last year will reduce inflation this year. And a lower base last year will increase inflation this year, even though longer-term trends might be the same. If one tracks the all-commodity monthly trends last year, one finds a significant increase in March, followed by another one in June. Then one had significant drops in November and December. Therefore, point-to-point WPI-based inflation is bound to be low in March 2009 and given underlying longer-term inflation trends, will also be low in June and immediately thereafter, perhaps increasing again beyond November. March 14, 2009 shows an increase of 0.27%. It is even possible for the inflation rate to temporarily turn negative. However, that won?t last and is little more than a statistical anomaly. We can use the word disinflation for this. Deflation is hardly the word, with its strong negative connotations. On a longer term annualised basis, WPI-based inflation is probably running at around 3%, with a GDP deflator of around 4%.

Beyond price index numbers, is there any evidence of demand contraction, other than what has happened through the export front? Not really, after that initial shock of three months or so. The January 2009 IIP is 0.5% lower than the January 2008 one, but it isn?t as if the economy isn?t growing. 5.3% growth in Q3 of 2008-09 may be low growth and no one may believe the government when it talks about 7.1% (right up to the precise decimal point) in 2008-09 and now even in 2009-10, but there is growth. Steel and cement seem to have perked up in January 2009.

With three fiscal stimuli, loosening of monetary policy, some cushion provided by agriculture and entrepreneurial strengths, there is no demand contraction. Investment demand may take some time to recover, but consumption demand should pick up fairly soon. Therefore, use of the deflation word is irresponsible. All it really proves is that we should stop paying any attention to WPI and derive a better measure. The author is a noted economist