Carlos Vegh and Guillermo Vuletin have an article Overcoming the fear of free falling: Monetary policy graduation in emerging markets, in `The role of Central Banks in financial stability: How has it changed?’, Federal Reserve Bank of Chicago, 2012. They have drawn on this to write a column on voxEU titled Graduation from monetary policy procyclicality.
In an ideal world, monetary policy should stabilise business cycle conditions. When times are good, the central bank should raise rates, thus reining in a boom. When times are bad, the central bank should cut rates. As an example of how this might not arise, recall that on 16 Jan 1998, in what was arguably a pretty bad time for the Indian economy, RBI raised rates by 200 basis points. Economists have a delicate and damning phrase for monetary policy that fuels a boom and exacerbates a bust: it is termed “procyclical” monetary policy.
Vegh and Vuletin construct a measure of monetary policy procyclicality : the correlation between the cyclical component of the short-term interest rate and GDP. This is computed for a large number of countries for the 1960-1999 period. Here is the result:
The bars in black are advanced economies and the bars in yellow are developing countries. There is a striking pattern: All the countries with a negative correlations — i.e. interest rates are raised in bad times — are developing countries. This is a striking demonstration of the faulty monetary policy frameworks that are found in developing countries: Every country which suffers from procyclicality in this period is a developing country.
India fares pretty badly in this list: Starting from the bottom, we have Uruguay (Rank 1 from the bottom), Chile, Mexico, Venezuela, Gambia, and then India (Rank 6 from the bottom).
Things got better after 1999. Vegh and Vuletin repeat the analysis for 2000-2009 and find that India did much better. The correlation swung to a positive value. India moved up, to the middle of the distribution. You could find one developed country — Japan — which did worse than India in this period.
In my assessment, there are two elements to this story: (a) Is there room to manoeuvre for monetary policy and (b) Is the monetary policy process properly constructed? The first is largely a question about exchange rate flexibility. If the central bank pursues exchange rate goals, this uses up the instrument of monetary policy. The second is about how monetary policy is conducted.
The figure above shows how India’s exchange rate regime evolved towards flexibility. The structural break dates (23 May 2003 and 23 March 2007) are computed using the methodology of Zeileis, Shah, Patnaik, 2010. For 4.74 years, we had INR/USD volatility of 2.31% per year. On 23 May 2003, the contradictions of this regime became unbearable, and the exchange rate regime changed: volatility jumped up to 3.93% per year: a rough doubling. On 23 March 2007, the contradictions of this regime became unbearable, and the exchange rate regime changed: volatility jumped up to 9.05% per year: another rough doubling.
From 23 March 2007 onwards, we have finished 5.43 years — the longest single period out of the three shown here — in this zone of high exchange rate flexibility. On the subject of the Indian exchange rate regime, you may like to read this post.
I believe these changes have substantially, though not completely, freed monetary policy of the burden of pursuing exchange rate goals. This is one half of the story of Indian monetary policy reform. The second half is that of setting up a sound monetary policy process. Now that you have a lever of setting the short rate in your hands, what would you like to do with it? The first stage is a relatively easy and nihilistic one: it requires getting out of trading in the currency market. By 23 March 2007, this was completed. The second stage is harder: it requires institution building. We have not yet begun on this phase.
This increase in exchange rate flexibility is consistent with the Vegh & Vuletin calculation which shows that the procyclicality of Indian monetary policy was reduced in the 2000-2009 period.