The Thursday syndrome


Indian Express, 21 August 2008


The sharp increase in inflation in recent weeks has come as a surprise to both the government and observers of the Indian economy. Better methods for carefully watching inflation numbers could help give us early warnings of inflation. A more careful analysis reveals that early warings of inflation were available in January 2008.

The inflation number most watched and commented upon in India is the year-on-year inflation rate based on the Wholesale Price Index. There has been a debate about how the WPI is not revised on time, or on how the weights do not reflect the consumption or production basket. However, as the WPI is a measure that is announced every week it continues to serve as the headline inflation rate in India.

What does the year-on-year rate measure? Standing in August 2008, an increase of 10 percent year-on-year in the price of oil, for example, means that if in August 2007, the price of oil was Rs 100 per kg, then today the price of oil is Rs 110 per kg. This could have been because the price rose from Rs 100 to Rs 150 in December 2007, and since then declined to Rs 100 in February 2008, and has stayed steady since March 2008.

Alternatively, it could be that the price of oil remained steady at Rs 100 per kg till March 2007 and has since been steadily rising, and is now up to Rs 110. The two scenarios say very different things about the inflation process. The first one suggests that there was a sharp spike in prices late last year, and we don't need to take any anti-iflationary measures today. The second, however, suggests that we should take worry as prices seem to be gettting higher and higher.

When we focus on inflation on a year-on-year basis, as we tend to do in India, we miss out on the inflationary process described above. Also, as the clamour and the drum-beats increase, policy could, erroneously, respond to price rise that has actually already been brought under control. Or, it could mean that when prices start moving on an upward path month after month, the year-on-year rate will miss the trend.

The way this problem has been addressed in developed countries is to focus on monthly changes. There is, however, one problem with monthly changes. If there is seasonality in the data, it might suggest there is a rise in prices and we might start taking policy action, but that would be a mistake. In other words, when the mango season ends, like is happening now, and the price of mangoes goes up, even doubles, or triples, there is no need to panic. This is merely seasonal. The price of mangoes will come down when the next mango season comes. So while economists like to focus on monthly inflation, they do so after seasonally adjusting the data. All this seems fairly obvious. Some of the simplest techniques available to do this are nearly a 100 years old. The latest techniques are ubiqutious today, software packages to apply them are available for free download on the internet and any economist with a training in statistics and with a view to understanding the process of inflation in India can do so with fairly little effort. Even though economists disagree on the nitty-gritty details of various techniques, various methods give similar results and serve the broad purpose.

To make the job of economists easier, statistical departments in advanced countries release seasonally adjusted data in addition the raw data. These numbers are then picked up by the media, the central bank for policy analysis and by the government for it communication with policy makers.

Despite the obvious advantages of watching seasonally adjusted monthly inflation data which have been pointed out by Indian economists for many years, it is unfortunate that the Government of India still does not produce and release seasonally adjusted data. In a recent study at NIPFP we find evidence of the crucial early warnings which would have emanated from monthly data. For example, inflationary pressures were visible from January 2008 when the data for December 2007 was released. While the headline inflation numbers jumped up only in March, the trend was visible from December 2007. Inflationary pressure continued to grow after that. If monthly data had been tracked, the December data (that is out by mid-January) would have rung alarm bells. This would have stopped RBI on its tracks in its purchase of USD 21 billion in the spot and forward markets in January 2008. It would then perhaps also not have bought another USD 11.16 billion till the end of May 2008. This not only pushed the rupee to weaken, making the rupee price of global commodities even higer, it also added to money supply.

There is also a risk of missing the good news with year-on-year data. In the mid 1990s when seasonally adjusted monthly inflation was already showing signs of prices having being brought under control, year-on-year inflation continued to be high. The RBI, unaware of this good news, kept on tightening monetary policy. This appears to have kept monetary policy tight for too long.

As long as the focus of monetary policy was on keeping exports competitive through manipulating the rupee dollar rate, the RBI could conduct monetary policy without the analytical arsenal normally used by modern central banks. However, as the Raghuram Rajan and Percy Mistry reports point out, today the contradictions between a policy that focuses on the exchange rate and a policy that focuses on reducing inflation has become so sharp that the RBI needs to reform itself into a central bank that focuses on inflation. In this setting, the intellectual tools for understanding the inflation process have to be sharpened. Among other techniques, an improvement in the measurement of inflation and in understanding the inflationary process must be an imporatant element of the reform undertaken. The CSO can contribute to this process by releasing improved inflation data and tracking monthly changes in prices.


Back up to Ila Patnaik's media page
Back up to Ila Patnaik's home page