After many years of low inflation, consumer prices rose rapidly in 2021 and 2022 and inflation reached the highest levels in over 30 years. The inflation target, like the Riksbank's the forecasting ability, was tested to an extent not seen since the introduction of the target. Many countries experienced similar developments (see Figure 1).^[3] The rise in inflation was preceded by an unusual combination of events affecting economies around the world: first, the pandemic, which led to severe supply disruptions and pent-up consumer demand, large fiscal and monetary stimulus, and then the war in Ukraine, which led to large increases in energy and food prices, among other things. Central banks did not anticipate that inflation would rise so quickly and inflation forecasts consequently showed large forecasting errors. The ability of central banks to forecast inflation has therefore been severely criticised. ^[4] See for example Bordo and Levy (2022), House of Representatives (2022), Levy (2023), Australian Government (2023) and House of Commons (2023). An exception is Holm and Martinsen (2023) who praised Norges Bank for its inflation management. Their assessment was that the decisions were generally well balanced and timely.

Figure 1. Inflation in Sweden and nine other countries or currency areas Annual percentage change

At the end of 2022, the Centre for Monetary Policy and Financial Stability (CeMoF) at Stockholm University was commissioned by the Riksdag Committee on Finance to review monetary policy in Sweden in 2022. The evaluation (Hassler, Krusell and Seim, 2023) argues that the Riksbank should have increased its preparedness when inflation started to rise in December 2020 in the United States and when the Bank of England started to raise its policy rate in December 2021.^[5] See page 7 of Hassler, Krusell and Seim (2023).

In this Commentary, we take up the evaluation's observations and proposal to compare the development of inflation in Sweden with that of the rest of the world and we examine how good, or bad, ten different central banks including the Riksbank were at forecasting their target variables when the development of inflation was completely different from that of previous decades.^[6] We study how well the central banks forecast their target variables, as this makes the forecasts comparable insofar as it is these variables that the central banks are asked, or have themselves chosen, to stabilise. The question we ask is whether the forecasting errors differ greatly between central banks under these very challenging circumstances and how well the Riksbank fares in such a comparison.^[7] The different target variables are affected to different degrees by, for example, energy prices and interest expenditure. However, the comparison partially takes this fact into account, as we normalise the forecasting errors by how much the target variables have varied during the period under evaluation. Central banks make their forecasts under different conditions. However, we refrain from fully considering these differences here, leaving it for further studies. For example, we disregard the fact that central banks have chosen to condition their forecasts on different assumptions for interest rates, exchange rates, oil prices and so on. Another aspect that we do not consider is that central banks make different numbers of forecasts per year. The appendix, which shows the inflation forecasts of all ten central banks, shows this clearly. It can also be noted that some central banks worked with scenarios and not forecasts in 2020. For example, the Reserve Bank of New Zealand published three different scenarios between March 2020 and January 2021. We also disregard this in our Commentary. .