“The only function of economic forecasting is to make astrology look respectable,” the economist Ezra Solomon once said. It looks even less respectable today, as practitioners attempt to demystify the fallout from multiple shocks including the pandemic, the war in Ukraine, and the shifting geopolitical landscape.
Andrew Bailey, governor of the Bank of England, knows this only too well. On Thursday as the central bank raised interest rates by 25 basis points, he faced a host of questions over the bank’s large upward change to its growth and inflation projections. Economic forecasting is indeed an inexact science; but hefty revisions do little to reassure the public that central bankers know what they are doing.
It is not just the BoE that has been off the mark. The US Federal Reserve and the European Central Bank have both erred in their inflation forecasting too, particularly since the onset of the pandemic. The IMF has also chopped and changed its recent growth outlooks. Given the role that forecasts play in informing the decisions of investors, households, and policymakers, accuracy is important. For central bankers a record of decent forecasting is also vital to build credibility, particularly as higher trust can help anchor inflation expectations.
Recent criticism of erroneous forecasts needs to be tempered. Economists have faced an unusually uncertain world since 2020. They have had to take positions on epidemiology, war scenarios, supply chain shifts, and rapidly evolving domestic and international policies. Another element is a limited public understanding of what forecasts actually represent. They are conditioned on judgments made at a particular point in time. As new data comes in, those judgments need to be recalibrated. They should be treated as indicative, not as gospel truth.
Economists still have some questions to answer. Central bankers were arguably too slow to raise rates when inflation picked-up in 2021, clinging to the notion that price pressures were “transitory”. There were misjudgments over the effects of fiscal stimulus, how stable inflation expectations were, and the damage from the pandemic to supply, which contributed to monetary policymakers falling behind on inflation. The preceding decade of low inflation may have also lulled them into a false sense of security. Errors have been made during the rate-raising cycle too. Both the BoE and ECB failed to grasp just how sticky food inflation would be.
Forecasting can be augmented in a few ways. Recent shocks underscore the importance of drawing on expertise beyond the economics profession. Economists should also continue to explore how advances in big data, machine learning and AI can offer opportunities to improve economic analysis and better model complexity.
Improving how forecasts are communicated, particularly in times of uncertainty, is essential. Showing projections under different scenarios can help deepen understanding of the range of possible outcomes. Likewise, although central banks have attempted to highlight the probability distributions around their forecasts, they could be conveyed in a more accessible way.
Above all economists need to be clearer in setting out how and why their key judgments have changed between forecasts — and how that affects the numbers. A greater effort to show how they reach their projections can go some way towards building trust and normalising an understanding of economic forecasts as reference points rather than foresight. After all, John Maynard Keynes, the famous economist, is often credited — perhaps wrongly — as asking: “when the facts change, I change my mind. What do you do?”
