Will CPI stand up please?4 min read . Updated: 01 Apr 2009, 11:55 AM IST
Will CPI stand up please?
Will CPI stand up please?
Dangerous deflation is the main concern right now. Following last week’s report of 0.27% Wholesale Price Index (WPI) inflation, a negative value is possible now for the first time in ages. Slipping below zero into deflation would put enormous pressure to cut interest rates and ease monetary policy further.
Simultaneously, people are complaining about the high cost of living—food prices, specifically, and Consumer Price Index (CPI) inflation has been well above 10%. To assess the conflicting data, some economic fallacies first need to be sorted out. People’s real income is what they can actually buy. This real income is their money income (what they receive) divided by the price level. The inflation rate is the percentage change in the price level over the specified period. As long as there is inflation, even if it is falling, the price level will be higher than in the last period. So at a given money income, real income falls.
Also See The Consumer Price Perplex (Graphic)
The above point needs emphasis since many ostensibly knowledgeable people do not realize this. Unfortunately, whenever inflation falls, much economic commentary suggests that lower inflation implies relief for the consumer. Not so. If money incomes are adjusted to increase with past inflation—automatically or otherwise—then a fall in inflation will raise real income. But this is generally not the case, as in much of the informal sector. This point is critical to assessing the impact of falling inflation.
A related issue is the well-known base effect, which implies that a sudden jump in the price level leads to a corresponding jump in the inflation rate, which will get reversed in the next period. Hence, a central bank should not respond to a temporary rise or fall in inflation that is likely to get reversed in the next corresponding period. Instead, it should focus on the average inflation rate of two or more periods (see “Dissecting our inflation targets’, Vivek Moorthy, Mint, 29 July 2008).
However, right now neither of these considerations apply. The inflation measure most relevant for the public—the various CPI indices (for Industrial Workers, Agricultural Labour, Rural Labour) are running very high, above 10% in recent months. Let us specifically look at the most widely used CPI, that for industrial workers. A year ago, it was around 5.5%, not unusually low. The current high 10% CPI inflation thus cannot be due to a base effect. Neither is the very low current WPI inflation due to the base effect in reverse, since early last year WPI was around 5% also, not unusually high.
The high CPI inflation is getting less official attention than the low WPI. Unlike the prolific economist Surjit Bhalla (who switches between WPI or CPI, depending on which is lower and thus suits his purpose), the government always looks at, and reacts to, WPI. The finance ministry expressed great concern in press statements last July when WPI rose above 12% due to the petrol price hike, and the Reserve Bank of India (RBI) raised rates. The official focus on WPI is very consistent, even if it is misplaced.
Why does the government ignore CPI and focus on WPI is a question that an increasing number of people have been asking. We suspect the WPI fixation is partly because the quality of the CPI data is awful in some crucial ways. CPI for a given period is released one to two months later than WPI, but such a short lag is not that critical. The problem is with the data itself. This can be seen in the accompanying table, which lists the value of CPI for Industrial Workers from November 2007, and the corresponding inflation rate. For three of the last four months (October, November and January), the CPI value is 148. Further, the price level was repeatedly 134 in the corresponding months a year ago. You may blink or run your eyes in disbelief, but when you look at the data again, the same price level is reported for these months. Hence, reported inflation is 10.45% for three of the last four months.
Efforts to improve data have largely gone into producing quarterly gross domestic product data and conforming with the International Monetary Fund’s Special Data Dissemination Standard. Instead, policymakers should have tried to ensure something more fundamental in this regard—a reasonably good measure of the cost of living. This should be feasible—inflation data are collected from a small sample of the population. It is not such a difficult task, compared with tracking and tagging the whole billion plus population with ID cards etc. Once a month it should be possible to compile and compute CPI, even if a few sub-items do get repeated, as happens for WPI. Accuracy in data is more important that its frequent release and speedy dissemination.
Much anecdotal and other evidence suggests that rural India is booming, and hence food prices are rising and CPI inflation is high. A very recent Mint article (“Karnataka sees 50% increase in the price of rice", 24 March) provides several reasons why rice prices are high and rising in Karnataka. Rising incomes is one of the factors. Nevertheless, the 10%-plus CPI inflation for three of the last four months, when consumer prices are falling all over the world, probably reflects the fact that data was just not updated.
Actual CPI inflation may be high and might warrant a tight policy stance, despite the weak industrial sector. But when WPI is close to touching zero, 10% CPI inflation is most likely too high. CPI should be first made to stand up, so that its inflation rate can be calculated.
Graphics by Ahmed Raza Khan / Mint
Vivek Moorthy is a professor of economics at IIM Bangalore. Comments are welcome at firstname.lastname@example.org