For a long time, India’s central bank and its monetary policy framework were a mystery. A year ago, that changed, with India formally adopting “flexible-inflation targeting" (FIT) in June 2016. A key feature of FIT is that monetary policy has an explicit inflation target in the long-term but medium-term inflation “projections" become the intermediate target. Thus, the success of FIT depends heavily on the accuracy of medium-term inflation forecasts.

Working towards a reliable inflation-forecasting system, the Reserve Bank of India (RBI) introduced a suite of models called the forecasting and policy analysis system (FPAS). Central to the FPAS is the quarterly projection model (QPM), a forward-looking model to assess the medium-term path of the economy. It relies on dynamic stochastic general equilibrium (DSGE), a model based on the principles of New-Keynesian (NK) economics. NK economics has become popular with most central banks around the world.

Structured around the concept of a small, open-economy, the model consists of four equations, each related to a key endogenous variable: the IS curve (output gap), the Phillips curve (inflation), smoothed-Taylor rule (short-term interest rate), and uncovered interest rate parity (exchange rate). In a customized version, called the production QPM, the RBI has tried to capture certain characteristics of the Indian economy, such as monetary transmission mechanism, supply shocks to food prices and the role of monetary policy credibility in setting expectations.

In spite of having such a seemingly “sophisticated" model at its disposal, the RBI’s performance on accurate inflation forecasting leaves much to be desired. In December 2016, RBI’s monetary policy committee (MPC) forecast that headline inflation in India in January-March would average 5%, with an upside bias. The actual inflation turned out to be 3.6%, a massive 140 basis points (one basis point is one-hundredth of a percentage point) lower than the forecast and the highest percentage forecasting error in RBI’s history. CPI (consumer price index) inflation in June eased to 1.54%, a record low, while the RBI continues to forecast an average of 2.75% for the first half of FY18.

So, what in the world is going on? The answer may lie in the flaws of the core theoretical framework, the NK DSGE model. DSGE has come under attack from eminent economists like Paul Krugman and John Cochrane for its failure to predict the global financial crisis. The model has been criticized for relying on rational expectations, assuming complete markets, ignoring the financial sector and shockingly, for creating models which are “money-neutral", i.e. monetary models without money. To counter the Lucas critique and build models with micro-foundations, DSGE reduces the complex economy down to simplistic and hypothetical actors, such as a representative firm and a representative household. It also relies deeply on unobservable variables such as the output gap and equilibrium real interest rate.

Such a model is unlikely to capture the complexities of the modern world. A working paper released in 2011 by the Federal Reserve Board, Washington DC echoed this view when it found that “the benchmark estimated DSGE model forecasted inflation and gross domestic product (GDP) growth very poorly". It is no surprise, then, that Wall Street has largely ignored DSGE-based models.

So, what can the RBI do? Two things come to mind. First, improve the DSGE framework by relaxing unrealistic assumptions and incorporating more features of the Indian economy, such as a large informal sector and an all-pervading shadow economy. Second, explore alternative approaches which make realistic assumptions and incorporate the complexity of the real world.

Two such approaches are emerging in literature. The first, Post-Keynesian (PK) economics, has its origins in the General Theory of Employment, Interest and Money of John Maynard Keynes and is closest in spirit to Keynes’ original work, according to historian Robert Skidelsky. PK economics acknowledges the central role of money, banking and debt in the modern economy. A testimonial to this approach is the work of Hyman Minsky, the American economist whose “financial instability hypothesis" was largely ignored for decades, till the “Minsky Moment" came in 2007 during the subprime crisis in the US.

The second, complexity economics, uses computational rather than mathematical analysis to model the economy as a constantly evolving set of institutions, agents and technological innovations. It abandons micro-foundations to focus on emergent phenomena and assumes that the economy is not necessarily in equilibrium. The Economic Complexity Index developed using this approach is a more accurate predictor of GDP per capita than many competing growth theories.

RBI’s highly capable economists can develop parallel models by applying alternative approaches to the Indian context. When the different models agree on the assessment of the economic situation and policy recommendations, the RBI can be confident about its policy actions. When the models are in violent disagreement, the RBI can proceed with caution. The US Federal Reserve may not have been blindsided by the crisis if it had Minskian economists in its team.

Pluralism is critical for all social sciences and economics is no exception. Harvard economist Dani Rodrik writes in his book Economics Rules that economics is a collection of models, and diverse situations call for different models. “Too often, economists mistake a model for the model that applies everywhere and at all times", he says.

The RBI should realize that DSGE is a model, not the model. Exploring multiple schools of thought can only improve the accuracy and effectiveness of India’s monetary policy framework.

Rohan Chinchwadkar is an assistant professor at IIM Trichy

Close