Following the 2022 inflation surge, this thesis examines whether the Swedish central bank's (the Riksbank) monetary policy reaction function changed. A Taylor rule in gap form, with the inflation gap and the unemployment gap as regressors, is estimated with ordinary least squares on monthly Swedish data for January 2010 to December 2024. Both gaps are interacted with dummy variables for three regimes: 2010-2014, 2015-2021 and 2022-2024. This allows the regime-specific reaction coefficients to be compared with the low interest rate period as the baseline. Inference is based on heteroskedasticity- and autocorrelation-consistent standard errors, and model stability is examined with cointegration and structural break tests. The results reject parameter stability. The most significant structural break is dated to September 2022, and the regime coefficients differ significantly. During 2015-2021, the response to the inflation and unemployment gaps was in practice zero. This is expected when the policy rule is inactive at the lower bound. During 2022-2024, the contemporaneous response to the Consumer Price Index with a fixed interest rate (CPIF) gap is negative, reflecting a monetary policy lag, since the interest rate peak came after the inflation peak. Instead, the response to core inflation (CPIF excluding energy) is positive and significantly stronger than in the baseline period, although the data cannot establish that it exceeds one. The main conclusion is that the reaction function has changed, and any assessment of the Riksbank's crisis policy depends heavily on the chosen inflation metric. The results are robust to excluding the pandemic period, to an alternative break point and to lagged regressors, but sensitive to how the gap measure is constructed. This is discussed as an area for improvement together with real-time data and dynamic specification.