Economics is a science of thinking in terms of models joined to the art of choosing models which are relevant to the contemporary world. It is compelled to be this, because, unlike the typical natural science, the material to which it is applied is, in too many respects, not homogeneous through time.
— John Maynard Keynes, letter to Roy Harrod, 4 July 1938
Almost 20 years ago, Richard Clarida, Jordi Galí and Mark Gertler published a seminal article in the Journal of Economic Literature entitled “The Science of Monetary Policy: A New Keynesian Perspective,” which quickly became required reading for students of monetary economics and aspiring central bankers alike. The trio summarized the theoretical macroeconomic framework for analyzing inflation targeting and interest rate rules that has been utilized by many central banks around the world over the past few decades.
While monetary policy making had traditionally been seen as mostly an art (sometimes a dark one) that was typically practiced by bankers, the scientific approach became more and more influential, culminating in the rise of leading academic economists like Ben Bernanke and Janet Yellen to the helm of the Federal Reserve. However, like it or not, art may now be making a comeback in monetary policy, and partly at the expense of science. Here’s why.
First, in the standard New Keynesian model, the only instrument of monetary policy is the nominal interest rate. Quantities such as the central bank balance sheet, credit or monetary aggregates play no role. Since the global financial crisis, however, central banks routinely use their balance sheet as an additional instrument of monetary policy (e.g., by purchasing vast sums of sovereign or other bonds, a practice termed quantitative easing or QE), which is difficult to incorporate into the standard model. Hence Ben Bernanke famously quipped in 2014 that “the problem with QE is it works in practice, but it doesn’t work in theory.”
While there is some agreement among economists that QE worked, serious questions remain about exactly how. With economists unsure about how to explain and model QE, central bankers using the balance sheet thus require some artistic skills in calibrating the right dose of quantitative easing or tightening.
Second, a key ingredient of the New Keynesian model is the relationship between unemployment and inflation, the so-called Phillips curve. It posits that falling unemployment leads to rising inflation. However, empirically this link has become weaker recently (the Phillips curve has flattened) for reasons that economists keep debating. Whatever the main reason may be – lower productivity growth, demographic changes, hidden labor market slack or, as Alan Krueger argued in his lunch talk at the recent Jackson Hole conference, rising monopsony power by employers due to rising firm concentration and collusion that reduces workers’ wage bargaining power – a flat Phillips curve hugely complicates a central bank’s ability to steer inflation.
Third, in the New Keynesian model, changes in nominal interest rates have real effects because prices are assumed to be “sticky,” i.e., they change only slowly. However, as a paper by Alberto Cavallo also presented at Jackson Hole shows, the advent of Amazon and, more broadly, e-commerce has led to more frequent price adjustments by retailers, making prices less sticky. In the logic of the New Keynesian model, more flexible prices would reduce the impact of a given change in nominal interest rates on the real economy.
Fourth, Cavallo’s paper also argues that the growth of e-commerce has increased the pass-through of changes of cost factors, such as fuel prices and exchange rates, to consumer prices as retailers are quicker to pass on cost shocks into online prices. Central banks may therefore have a weaker grip on inflation, especially in the short run, as it will tend to be driven more by short-run forces outside their control. This means that central banks either have to adjust interest rates more frequently and more aggressively if they want to control inflation over shorter time horizons, or they will have to accept larger and more frequent deviations from their inflation targets, which may harm their credibility.
Fifth, Fed chair Jerome Powell in his introductory remarks at Jackson Hole went to great lengths to emphasize the large uncertainty about the true levels of the “stars” that have been guiding central banks and are key variables in the standard model. These are the neutral interest rate (r-star), the natural rate of unemployment (u-star) and the growth rate of potential output (y-star). In Powell’s own words, “Guiding policy by the stars in practice … has been quite challenging of late because our best assessments of the location of the stars have been changing significantly.”
Re-examining the framework
The bottom line is that the advent of quantitative easing and important structural changes in the economy have made the traditional science of monetary policy somewhat less useful for monetary policymakers. Some of the challenges noted above, e.g., the uncertainty about the exact location of the “stars,” relate to parameter uncertainty but don’t put the framework in question. Yet other challenges, such as prices becoming less sticky or the Phillips curve becoming entirely flat, question the usefulness of the model framework itself.
To be sure, the science of monetary policy is evolving and actively looking for ways to adapt the mainstream model or come up with new frameworks. In the meantime, however, policymakers will likely (have to) rely less on models and simple policy rules and become more “artistic” in reading and responding to economic and financial signals. Put differently, monetary policy is likely to become both less “boring” (remember Mervyn King’s quip?) and, unless you have a deep understanding of art, more unpredictable – yet another reason to brace yourself for more volatility!
A version of this material was circulated in Sunday Signposts on 26 August 2018.