The dilemmas of monetary policy

For some central bankers, inflation is a mystery. Dan Tarullo, after finishing his term in 2017 as a member of the U.S. Federal Reserve Board, gave a speech at the Brookings Institution entitled “Monetary Policy Without a Working Theory of Inflation.” The main message: having been a central banker, he doubted that there was a theory of inflation that would serve to guide, in real time, monetary policy decisions. After all, he stated, the inflation models commonly used in macroeconomic analysis are based on two variables that are not observable, the output gap and inflation expectations. To reduce this uncertainty, Dan Tarullo recommended, among other things, increasing the professional and academic diversity of the members of the Federal Reserve Board.

During the recent inflationary episode, many central bankers also expressed their perplexity about inflation. For example, the accounts of the ECB’s February 2022 meeting expressed doubts about the ability of existing macroeconomic models to explain the evolution of inflation, and the need to explore other sources of information, such as surveys and “the opinion of experts”.

But what are “the opinions of experts” based on, if not on models, whether quantitative or qualitative? Perhaps what the ECB Governing Council members meant was the need to expand the variety and type of models used to evaluate and predict inflation. What is clear is that any opinion worth entertaining, internally consistent and well-reasoned, must have a model, whatever it may be, behind it. For example, those who believed that the inflationary shock would be permanent were implicitly saying that inflation expectations would become unanchored and a price-wage spiral would follow. Which hasn’t happened, and explains the failure of their prediction. A broader vision, incorporating the magnitude and breadth of the supply disturbances, served to understand the transitory nature of the episode and delivered a better forecast.

This debate is reminiscent of the parable of the hedgehog and the fox, by the German philosopher Isaiah Berlin. Berlin divides intellectuals into two types: hedgehogs, who interpret reality through a strong idea — for example, Plato, Nietzsche or Proust — and foxes, who gather information quickly from a wide variety of sources — for example, Aristotle, Shakespeare or Goethe. From this distinction, Philip Tetlock, professor of political psychology at the University of Pennsylvania, analyzed the behavior of a large number of experts in different areas, including economics and politics, classifying them as hedgehogs or foxes, and came to the following conclusion: most experts, especially the most famous ones, behave like hedgehogs, interpreting the world through a great unifying idea. But he also came to the more interesting conclusion, perhaps counterintuitive but empirically verified, that the more famous the experts — which typically implied a more accentuated hedgehog characteristic — the worse their forecasts. Foxes, typically less famous but more agile in adapting their mental models to changing circumstances, were better forecasters. The reason was that while the hedgehogs focused mostly on defending their grand theories and gaining notoriety, the foxes tried to get their forecasts right.

The world we face requires an abundance of foxes at the helm of monetary policy. Unlike the decades before the pandemic, when economic fluctuations were primarily variations in demand around a virtually static supply, now both demand and supply fluctuate, nationally and internationally. The challenge of climate change, the objectives of energy independence, and the need to strengthen national security in a world in geostrategic transition, are generating changes in the structure of the global economy that require complex analyses. Agility and a variety of knowledge will be essential to make correct monetary policy decisions. For example, understanding financial, energy and commodities markets, to be able to anticipate changes in prices not discounted by the markets. Or the evolution of artificial intelligence and its impact on productivity, labor markets, and inflation. Or geostrategic trends and their impact on business strategies for managing global supply chains and pricing decisions. The intersection of all these factors increases the difficulty of predicting economic growth and inflation, and requires the development of new analytical tools.

This rapidly changing world requires effective risk assessment processes, which must include the analysis of alternative scenarios. The acceleration of interest rate increases over the last two years was due to the need to avoid the most acute risk — the development of uncontrolled inflation — even at the risk of causing a greater economic slowdown. As the inflationary risk fades, the risk assessment must be rebalanced and monetary policy can put more focus on stabilizing growth. It also must avoid a return to the pre-Covid period of excessively low inflation — even though the inflation target is symmetrical, conservative central bankers remain more reluctant to accept upside risks than downside risks to inflation.

Effective and proactive communication with governments and civil society will be essential to reinforce the independence of monetary policy in the face of potentially divergent objectives. In the U.S., for example, there are already calls for interest rates to remain low to reduce the cost of the investments necessary to confront climate change risks. Central banks will suffer losses in their income statements at some point. These losses, derived from the fall in value of their bond holdings as interest rates rise, are paradoxically the result of the success of the monetary policy measures adopted to counteract the effects of the pandemic, and if these measures had not been adopted the losses in terms of growth would have been much greater. Price increases hurt, without a doubt. And central bankers will have to explain that they have done everything possible to avoid them with the information that was available in real time. Generating trust while at the same time communicating the existing uncertainty is a very difficult task, but necessary.

Monetary policy decisions, like investment decisions, require forecasting the future in conditions of uncertainty. This uncertainty, already high when Dan Tarullo doubted the macroeconomic models of the time, will increase in coming decades, and will require central bankers with a fox mentality to stabilize inflation so that the economy can develop its full potential. Happy holidays and a prosperous 2024!