Unveiling inflation: Oil shocks, supply chain pressures, and expectations

Aastveit, Knut Are; Bjørnland, Hilde Christiane; Cross, Jamie; Olsen, Helene Working Paper Unveiling inflation: Oil shocks, supply chain pressures, and expectations Working Paper, No. 12/2024 Provided in Cooperation with: Norges Bank, Oslo Suggested Citation: Aastveit, Knut Are; Bjørnland, Hilde Christiane; Cross, Jamie; Olsen, Helene (2024) : Unveiling inflation: Oil shocks, supply chain pressures, and expectations, Working Paper, No. 12/2024, ISBN 978-82-8379-332-1, Norges Bank, Oslo, https://hdl.handle.net/11250/3176403 This Version is available at: https://hdl.handle.net/10419/322330 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/4.0/ Working Paper Unveiling inflation: Oil Shocks, Supply Chain Pressures, and Expectations Norges Bank Research Authors : Knut Are A astveit Hilde C. Bjørnland Jamie L. Cross Helene Olsen Kalstad K eywords Inflation Expectations, Inflation Pass -through, Oil Prices, Supply Chain Pressures. 12 | 2024 Norges Bank Working Paper Working papers fra Norges Bank, fra 1992/1 til 2009/2 kan bestilles på epost: [email protected] Fra 1999 og senere er publikasjonene tilgjengelige på www.norges-bank.no Working papers inneholder forskningsarbeider og utredninger som vanligvis ikke har fått sin endelige form. Hensikten er blant annet at forfatteren kan motta kommentarer fra kolleger og andre interesserte. Synspunkter og konklusjoner i arbeidene står for forfatternes regning. Working papers from Norges Bank, from 1992/1 to 2009/2 can be ordered by e-mail: [email protected] Working papers from 1999 onwards are available on www.norges-bank.no Norges Bank’s working papers present research projects and reports (not usually in their final form) and are intended inter alia to enable the author to benefit from the comments of colleagues and other interested parties. Views and conclusions expressed in working papers are the responsibility of the authors alone. ISSN 1502-8143 (online) ISBN 978-82-8379-332-1 (online) Unveiling inflation: Oil Shocks, Supply Chain Pressures, and Expectations∗ Knut Are Aastveit‡, Hilde C. Bjørnland†, Jamie L. Cross§and Helene Olsen Kalstad¶ July 4, 2024 Abstract: This paper demonstrates that inflation expectations have acted as significant amplifiers of recent global demand and supply shocks, thereby playing a crucial role in maintaining inflation at relatively high levels. This finding is established by applying a structural vector autoregression model that includes various shocks to global demand and supply, along with domestic inflation and inflation expectations for six economies: the United States, Canada, New Zealand, the Euro area, the United Kingdom, and Norway. We begin by documenting that global demand and supply shocks in the oil market, as well as global supply chain disruptions, have been major drivers of the recent inflation surge in all these economies. Subsequently, through various counterfactual and conditional forecasting exercises, we demonstrate that inflation expectations generally amplify the transmission of global shocks to inflation and have played a critical role in sustaining elevated inflation rates in recent years, particularly in the United States, Canada, and New Zealand. JEL classification codes: E31, C11, C32, D84, Q41, Q43. Keywords: Inflation Expectations, Inflation Pass-through, Oil Prices, Supply Chain Pressures. ∗This paper should not be reported as representing the views of Norges Bank. The views expressed are those of the authors and do not necessarily reflect those of the Norges Bank. We thank Christiane Baumeister, Roberto Casarin, Francesco Furlanetto, Domenico Giannone, Ivan Petrella, Giorgio Primiceri, Francesco Ravazzolo, Benjamin Wong and conference participants at the EIA 2024 virtual financial market workshop, the Women in Macroeconomic Workshop II in Melbourne, the SNDE 2024 Symposium in Padova, the FEEM 2024 workshop on Energy Transition in Milan, the Dolomiti Macro Meeting 2024 in Kastelruth and seminar participant at Norges Bank for useful comments and suggestions. This paper is part of the research activities at the Centre for Applied Macro and commodity Prices (CAMP) at the BI Norwegian Business School. ‡Norges Bank and BI Norwegian Business School. Email: knut-are.aastv[email protected] †BI Norwegian Business School and Norges Bank. Email: hilde.c.b[email protected] §Melbourne Business School, University of Melbourne. Email: [email protected] ¶BI Norwegian Business School. Email: helene.o.k[email protected] 1 1 Introduction After more than a decade of low inflation in most advanced economies, global inflation rates surged unexpectedly in 2021 and have remained elevated since. This recent escalation occurred against a backdrop of significant global supply-side factors triggered by the COVID- 19 pandemic and the Russian invasion of Ukraine, both of which have exerted sharp upward pressure on prices. Recent research has also highlighted the impact of various supply factors for the current inflation surge, see for instance Ascari, Bonam, and Smadu (2024), Benigno, di Giovanni, Groen, and Noble (2022), Celasun, Hansen, Mineshima, Spector, and Zhou (2022), Crump, Eusepi, Giannoni, and S¸ahin (2024) and Bai, Fern´andez-Villaverde, Li, and Zanetti (2024) for the role of global supply chain pressures, or Baumeister (2023), Casoli, Manera, and Valenti (Casoli et al.), Gagliardone and Gertler (2023), and Bernanke and Blanchard (2024a,b) for the significance of energy market shocks.1Furthermore, the recovery in demand following the pandemic-induced economic downturn has also likely contributed to upward price pressures, as emphasized by Ascari, Bonomolo, Hoeberichts, and Trezzi (2023), Ball, Leigh, and Mishra (2022), Benigno and Eggertsson (2023), Bergholt, Canova, Furlanetto, Maffei-Faccioli, and Ulvedal (2024), Eickmeier and Hofmann (2022), Giannone and Primiceri (2024) and Rubbo (2023) among others. Beyond these supply and demand factors, the role of inflation expectations in exacerbating price pressures is critical. As economic agents anticipate higher future prices, their behavior changes in ways that can drive prices up further. For instance, in a model with imperfect information and bounded rationality, Beaudry, Hou, and Portier (2024) show that when supply shocks affect many sectors, agents infer that the common component of inflation has increased, which drives persistent inflation dynamics through their effect on expectations. This self-fulfilling prophecy can entrench inflation expectations, making it more challenging for policymakers to stabilize prices. 1See also: Aastveit, Bjørnland, and Cross (2023) for a structural analysis of the role of oil market shocks for inflation prior to the COVID-19 pandemic and Garratt and Petrella (2022) for a recent analysis of predictive ability of commodity prices for inflation pre-covid. 2 In this paper, we propose a novel empirical framework to investigate the drivers of inflation by combining shocks to global demand and supply with domestic inflation expectations. By examining these elements together, we contribute to a more nuanced understanding of the inflationary dynamics in the post-pandemic world. Understanding the role of inflation expectations is also vital for developing effective policy responses, as policymakers need to both manage and anchor inflation expectations to prevent an ever more prolonged period of high inflation.2 To address these issues, we employ a structural Bayesian VAR model based on the global oil market VAR model proposed by Baumeister and Hamilton (2019), extended to incorporate a local inflation block as in Aastveit et al. (2023), and global supply chain pressures measured by the Federal Reserve Bank of New York (Benigno et al.,2022). Our analysis focuses on the recent inflation surge and the role of inflation expectations across six economies: the United States (US), Canada, the United Kingdom (UK), the Euro area, Norway, and New Zealand. This approach allows us to examine the responses of both expected and actual inflation across different geographical regions to various shocks, including global oil supply shocks, global supply chain shocks, global activity shocks, and oil demand shocks. Our study yields four main findings. First, using data from 1998Q1 to 2023Q4, we find that inflation and inflation expectations in all six economies respond significantly to most of the identified shocks, although with some heterogeneity. Notably, for global activity, oil demand and oil supply shocks, the strongest responses are observed in the US and the weakest in Norway. Responses to global supply chain shocks are relatively consistent across economies. However, unlike responses to the demand and oil shocks, the peak impact on both inflation and inflation expectations occurs with a delay of about two years after the global supply chain shock. Second, a historical decomposition analysis for the entire period analyzed reveals that 2Using different frameworks, Conrad, Enders, and Glas (2022), Neri (2023), and Ascari and Fosso (2024) independently study inflation expectations in pre-pandemic samples, while Weber, Gorodnichenko, and Coibion (2023) use micro-data on daily shopping and expected inflation to show an increase in disagreement about inflation expectations among U.S. households during the pandemic. 3 demand shocks have been the primary driver of inflation in most economies. However, during the recent inflation spike from 2020 to 2021, supply shocks have become increasingly important. In fact, all of the global supply chain shocks, oil supply shocks and global demand shocks have contributed significantly to the recent inflation surge. Third, using counterfactual exercises where we hold inflation expectations constant, we demonstrate that inflation expectations generally amplify the transmission of global shocks to inflation. The effect is most pronounced in the US. Zooming in on the period since 2021, we show that inflation would have been considerably lower in all six economies had inflation expectations been held steady at their 2021Q1 levels. Thus, our exercise underscores that inflation expectations have played a critical role in sustaining elevated inflation rates in recent years, particularly in the US, Canada, and New Zealand. Finally, reduced-form conditional forecasts also provide supporting evidence that postpandemic inflation would have been substantially lower than observed if inflation expectations had been held constant throughout this period. Interestingly, we also find that if agents had perfect foresight then inflation expectations would better predict the inflation surge in all economies except Norway. Taken together, these results indicate that policymakers should continue to prioritize the management of inflation expectations in order to mitigate their amplifying effects on inflation. Additionally, strengthening the resilience of supply chains and fostering energy independence can reduce vulnerability to supply shocks and support long-term price stability. Our paper contributes to the recent debate on the drivers of the post-pandemic inflation surge. While most other recent studies tend to focus either on the importance of a single shock (see, e.g., Benigno et al. (2022) and Celasun et al. (2022)) or on distinguishing between aggregate demand and aggregate supply shocks as drivers (see, e.g., Eickmeier and Hofmann (2022), Bergholt et al. (2024), and Giannone and Primiceri (2024)), we examine the importance of various types of global demand and global supply shocks and their interaction with 4 domestic (country-specific) inflation expectations.3Applied to six advanced economies, we find that inflation expectations act as a significant amplifier of global shocks and have played a crucial role in maintaining inflation at relatively high levels. This mechanism has not been emphasized much in explaining the recent inflation surge. The paper is organized as follows. In Section 2 we present the data and the empirical methodology. Section 3 presents the empirical results related to the identified shocks, while in Section 4 we detail the drivers of inflation during the pandemic and post-Covid period. Section 5 concludes. 2 Empirical Methodology To analyze the interaction between global supply and demand shocks and domestic inflation expectations, we employ a structural Bayesian VAR model based on the global oil market VAR model proposed by Baumeister and Hamilton (2019), extended to incorporate a local inflation block as in Aastveit et al. (2023), and global supply chain pressures measured by the Federal Reserve Bank of New York (Benigno et al.,2022). Details on the model, estimation and data are provided below. 2.1 Structural VAR Model Let yt= (y0′ t,yπ′ t,yG′ t)′denote an (n0+nπ+ 1) ×1 vector where y0 tis an n0×1 vector of variables associated with the global market for crude oil, yπ tis an nπ×1 vector of countryspecific inflation variables, and yG tcontains the GSCPI variable. Following Baumeister and Hamilton (2019), we specify y0 t= (qt, yt, pt, it)′, in which qtis the percentage change in global crude oil production, ytis the percentage change in global real economic activity (here proxied by the world industrial production index as in Baumeister and Hamilton (2019)), ptis the percentage change in the global real price of oil, and itis the observable change in above- 3One notable exception is Ascari et al. (2023), they examine the impact of global supply chain pressures for the Euro area, but also include other types of shocks, such as a demand shock and an oil shock. 5 ground global crude oil inventories as a percent of the previous month’s world production. The global market for crude oil is then modeled through the following five equations: qt=c1+αq,ppt+b′ 1xt−1+u∗ 1t,(1) yt=c2+αy,ppt+b′ 2xt−1+u∗ 2t,(2) qt=c3+βq,ppt+βq,yyt+i∗ t+b′ 3xt−1+u∗ 3t,(3) i∗ t=c4+ Ψ∗ 1qt+ Ψ∗ 2yt+ Ψ∗ 3pt+b′ 4xt−1+u∗ 4t,(4) it=χi∗ t+et,(5) where cj, j = 1,2,3,4 are intercept terms, and xt−1=y′ t−1,...,y′ t−4′is a vector of lagged observations of the variables over the past year. Equation (1) is the oil supply curve in which αq,p is the short-run price elasticity of supply and u∗ 1tis an oil supply shock. Equation (2) specifies that global real economic activity depends contemporaneously on oil prices through αy,p, and u∗ 2tis an global activity shock associated with changes in the global business cycle. Equation (3) is the oil demand curve in which βq,p is the short-run price elasticity of demand, βq,y captures contemporaneous links with global real economic activity, i∗ tdenotes the change in global crude-oil inventories as a percent of the previous month’s world production, and u∗ 3tis a oil demand shock. Equation (4) models the dynamics in above-ground crude oil inventories, which depends contemporaneously on oil production, global real economic activity, and the price of oil, respectively through Ψ1, Ψ2, and Ψ3, and u∗ 4tan inventory demand shock. Finally, equation (5) accounts for measurement-error in the available data on global crude-oil inventories, in which χ < 1 captures the fact that observable inventories are a proportion of the true total quantity of inventories, and etcaptures the measurement error between true and measured inventories. This specification assumes that oil market is contemporaneously effected by variables in y0′ t, but can only be effected by the inflation and GSCPI variables with a lag of one quarter. This is a reasonable assumption, as the dynamics of the oil market are primarily driven by imme- 6 of inflation expectations in amplifying the effects of the shock for inflation across the whole sample. Having focused on average responses, we will detail the drivers of inflation in the post-covid period in Section 4. 3.1 Posterior Distributions Figure 1 shows the posterior distributions of the contemporaneous coefficients in the inflation block and GSCPI for the US. The lines represent the prior distributions, and the histograms are the complete set of retained draws from the posterior distribution after burn-in. Figure 1: Prior (solid red curves) and posterior (purple histograms) distributions of structural parameters in the inflation and GSCPI blocks for the United States In most cases we find that the posterior distributions have shifted far away from the prior distributions, and are relatively concentrated at specific values. This suggests a high 13 degree of likelihood information about such parameters is present in the data, however, some parameters exhibit substantial uncertainty as seen from the spread of their posterior distributions. Looking first at the inflation expectation equation (top row), we find that the posterior distributions have a wider range than those found in Aastveit et al. (2023). Since their investigation ended in December 2019, this result suggests that inflation expectations have become more sensitive to events in the oil market since 2020. Moreover, in addition to these results, we find that accounting for the global supply chain index is more important for inflation expectations than actual inflation, as evidenced by the posterior distribution for the former being shifted further to the right than the latter. Finally, the posterior distributions for parameters in the global supply chain equation (bottom row) are predominantly close to zero. This indicates that the supply chain index is not significantly influenced by the international oil market or domestic inflation conditions in the US. Similar figures for the five other economies are placed in the Online Appendix. Comparing posterior distributions across economies reveals that inflation expectations in the US are more sensitive to global variables than those in other economies. For instance, variations in global oil prices, supply chain pressures, and real global economic activity economic exert a more pronounced influence on the inflation expectations of US consumers. This suggests that US households are more attuned to international economic factors when formulating their inflation expectations. 3.2 The effect of oil market and global supply chain shocks Figure 2 graphs the impulse responses to four of the most important shocks identified: oil supply, global activity, oil demand and global supply chain for inflation expectations in the US, Canada, New Zealand, the Euro area, the UK and Norway. Figure 3 shows the effect of the same four shocks for inflation across the same six economies. In the figures, we have normalized the oil market shocks so that each increases the oil price with 10% on impact. The GSCPI shock is normalized so it increases GSCPI with one standard deviation on impact. 14 Figure 2: Impulse responses for Inflation Expectations. The three oil market shocks: oil supply, global activity and oil demand, have been normalized to elicit a 10% increase in the real price of oil, while the global supply chain shock is normalised to increase GSCPI with a standard deviation. The posterior median is shown in boldface and the shaded area is the 95% joint credible set obtained from the posterior distribution of 100,000 structural model. The countries under study are the US, Canada (CA), New Zealand (NZ), the Euro area (EU), the UK and Norway (NO). Starting with the effect of oil supply shocks, Figure 2 illustrates that in the US, a positive oil supply shock initially increases inflation expectations, but this initial increase quickly dies out. Conversely, the impact on actual US inflation is more persistent, lasting for approximately one year, c.f. Figure 3. These responses may suggest that while households initially anticipate an increase in inflation, the actual effects on the broader economy take some time 15 to fully materialize. In the other economies, the oil supply shock also raises expected inflation, but by less than for the US, and the effect on inflation is not significant in the Euro area, the UK and Norway. This suggests that while households might briefly adjust their expectations, the shock does not leave a lasting imprint on the actual inflation rate in the European economies. Figure 3: Impulse responses for Inflation. The three oil market shocks: oil supply, global activity and oil demand, have been normalized to elicit a 10% increase in the real price of oil, while the global supply chain shock is normalised to increase GSCPI with a standard deviation. The posterior median is shown in boldface and the shaded area is the 95% joint credible set obtained from the posterior distribution of 100,000 structural models. The countries under study are the US, Canada (CA), New Zealand (NZ), the Euro area (EU), the UK and Norway (NO). 16 As with the oil supply shock, the global activity shock has a stronger effect on US inflation variables, increasing expected inflation for approximately one year, c.f. Figure 2. This response is mirrored in actual inflation, c.f. Figure 3, which experiences an increase that is more persistent and larger in magnitude than the response of inflation expectations. The inflation rates in Canada, the Euro area, New Zealand and the UK also demonstrate a significant response to global activity shocks, with both expected and actual inflation increasing. This indicates that households swiftly adjust their expectations, reflecting the real impact of changes in economic activity on prices. In Norway, the response of inflation expectations to the global activity shock is milder, and we find no significant effect on actual inflation. The oil demand shocks have similar effects as the global activity shocks, although somewhat more muted response in both expected and actual inflation. Again, the response is strongest in the US, while for Norway and the UK, the effect for expected inflation is barely significant, resulting in a more muted response for actual inflation. For all economies, we find that inflation reacts to the global supply chain shock with a lag of 1-2 quarters and the effect peaks after about 2 years, c.f. Figure 3. Ascari et al. (2024) also find similar results for the Euro area using a SVAR with different identifying restrictions. This dynamics is consistent with the fact that it takes time for supply chain issues to materialize to the real economy. After this, inflation increases gradually in all economies, exhibiting a more persistent response than to the oil market shocks, with the increase lasting for 2 to 3 years. Although the shock increases inflation in all four economies, the increase is the lowest in Norway compared to the others. Interestingly, we find that inflation expectations increase more rapidly and significantly in most economies, suggesting they can play a role in amplifying the effects of global supply chain shocks, c.f. Figure 2. To dig more into the role of inflation expectation for transmitting the shocks, Figure 4 presents a comparative analysis of the median inflation impulse response functions from 17 Figure 4: Actual and counterfactual: Holding inflation expectations constant. The countries under study are the US, Canada (CA), New Zealand (NZ), the Euro area (EU), the UK and Norway (NO). two scenarios: one actual and one counterfactual.4In the counterfactual scenario, inflation expectations are held constant, allowing the figure to specifically illustrate the effects of shocks on inflation. This setup helps isolate the direct impact of oil price and global supply chain shocks on inflation by removing any influence from changing inflation expectations. The observed differences between actual and counterfactual IRFs indicate the extent to which inflation expectations themselves propagate the inflationary effects of oil price and GSCPI 4A similar exercise was performed in Wong (2015) and Aastveit et al. (2023) analysing the effect of oil shocks on the US economy. 18 shocks. As with Figures 2 and 3, each shock modeled is assumed to increase the real price of oil by 10 percent, while a global supply chain shock increases GSCPI with a standard deviation. The figure shows that the effects of the oil market shocks on actual inflation in the US would have been lower, but of similar sign, had inflation expectations been constant throughout the sample. Even more notable is the results for the global supply chain shock, where we find that had inflation expectations been constant, the supply chain shock would not have had significant effects on US inflation. Similar, although somewhat smaller, effects are found for Canada and New Zealand, while for the three European economies (Norway, the Euro area and UK), inflation expectations mostly matters in the propagation of the global demand shocks. Summing up, we have found that inflation and inflation expectations in all six economies respond significantly to most of the identified shocks, although with some variation across economies. Notably, for global activity, oil demand and oil supply shocks, the strongest responses are observed in the US and the weakest in the European economies. All economies respond significantly to the global supply chain shocks, but the responses are much more delayed compared to the responses of the the oil market demand and supply shocks. 4 What drove inflation during Covid and beyond? We now turn to analyze the driving forces of the recent inflation surge over the period 2020Q1-2023Q4. Three exercises are conducted. First, we use historical decomposition to identify the main drivers of inflation. Second, we examine the role of inflation expectations as a driver of actual inflation using a structural counterfactual exercise. Finally, we use (reduced-form) conditional forecasts to investigate the role of expectations in shaping the predicted path of inflation during this period given pre-pandemic information. 19 4.1 Historical Decomposition We first analyze the driving forces behind the recent increase in inflation by examining the historical decomposition for inflation from 2020Q1-2023Q4 in Figure 5. For clarity, we have aggregated the shocks into three categories: global supply (oil supply and global supply chain shocks), global demand (global economic activity and oil demand shocks), and domestic shocks (domestic shocks to inflation and expected inflation). In the Online Appendix, we provide the historical decomposition for all shocks over the entire sample period. Figure 5: Historical Decomposition CPI (demeaned) from 2020. The countries under study are the US, Canada (CA), New Zealand (NZ), the Euro area (EU), the UK and Norway (NO). 20 Figure 5 illustrates that both global demand and supply shocks have contributed to the recent inflation surge. First, as the pandemic hit the world in early 2020, global demand shocks contributed to reducing inflation across the board. During 2020, global demand pressures worked to push inflation rates down in all six economies. Conversely, domestic inflationary shocks worked to increase inflation, especially in the US. Although the specific domestic shocks are not explicitly identified, they correspond well with the expansionary policies implemented in most economies at the time, suggesting domestic demand shocks may have also been important in this period. Second, from 2021, inflation starts to increase, driven by a mix of global demand, supply, and domestic shocks. For the US, approximately one-third of the increased inflation can be attributed to global supply shocks, another third to global demand, and the remaining third to domestic shocks. For the other economies, the proportions vary slightly, with the Euro area and the UK showing the most influence from global demand, and New Zealand showing the least influence from global demand. Third, the decline in inflation from 2022 to 2023 is primarily due to a reduction in domestic and global demand shocks across all economies, although supply factors continue to contribute to increased inflation rates. 4.2 Counterfactual Exercise To understand the role of inflation expectations in a historical context, we perform a counterfactual exercise. In the exercise, we generate a sequence of inflation expectations shocks that mute the inflation expectations response after shocks to the other variables in the system, thus holding inflation expectations flat since 2021Q1. Compared to Figure 4, where we shut down inflation expectations to examine impulse responses, this exercise investigates how inflation would have evolved in the pandemic period had inflation expectations not responded to any of the structural shocks in this time. The results are shown in Figure 6. The general result is that inflation would have been 21 lower if inflation expectations were unchanged during this period. However, there is notable heterogeneity in the magnitude of these differences across the economies. Figure 6: Counterfactual exercise - holding inflation expectations constant from 2021Q1. The countries under study are the US, Canada (CA), New Zealand (NZ), the Euro area (EU), the UK and Norway (NO). In the US, Canada and New Zealand, we find that inflation would have been much more stable, and even reduced, if inflation expectations were fixed during this period. In contrast, inflation in the Euro area, UK, and Norway would have still increased, but not as much as if inflation expectations had changed during this period. This result is also broadly consistent with the results in Section 3.2, where we find that inflation expectations matter less in the 22 —Online Appendix— Not for publication A Priors and Posteriors Figure 8: Canada: Prior (solid red curves) and posterior (purple histograms) distributions of structural parameters in the inflation and GSCPI blocks. 29 Figure 9: New Zealand: Prior (solid red curves) and posterior (purple histograms) distributions of structural parameters in the inflation and GSCPI blocks. Figure 10: Euro area: Prior (solid red curves) and posterior (purple histograms) distributions of structural parameters in the inflation and GSCPI blocks. 30 Figure 11: UK: Prior (solid red curves) and posterior (purple histograms) distributions of structural parameters in the inflation and GSCPI blocks. Figure 12: Norway: Prior (solid red curves) and posterior (purple histograms) distributions of structural parameters in the inflation and GSCPI blocks. 31 B Historical Decomposition Figure 13: US: Historical decomposition for CPI (demeaned) full sample Figure 14: Canada: Historical decomposition for CPI (demeaned) full sample 32 Figure 15: New Zealand: Historical decomposition for CPI (demeaned) full sample Figure 16: Euro area: Historical decomposition for CPI (demeaned) full sample 33 Figure 17: UK: Historical decomposition for CPI (demeaned) full sample Figure 18: Norway: Historical decomposition for CPI (demeaned) full sample 34 C Conditional Forecasts Figure 19: US: Coditional forecasts in the pandemic period. Figure 20: Canada: Conditional forecasts in the pandemic period. 35 Figure 21: New Zealand: Conditional forecasts in the pandemic period. Figure 22: Euro area: Conditional forecasts in the pandemic period. 36 Figure 23: UK: Conditional forecasts in the pandemic period. Figure 24: Norway: Conditional forecasts in the pandemic period. 37