Long-term inflation expectations and monetary policy in the euro area before the pandemic✰
Introduction
Long-term inflation expectations play a central role in monetary policy. Their anchoring to the inflation target is a necessary condition for central banks to maintain price stability, as it prevents temporary shocks from having long-lasting effects on inflation. Large and persistent inflation surprises can lead investors, firms and households to revise their long-term inflation expectations, which can become de-anchored from the central bank's target and affect inflation.
Motivated by these considerations, this paper quantifies the macroeconomic effects of shocks to long-term expectations and assesses the role of monetary policy in their transmission. The paper contributes to the literature in two ways. First, it traces out the responses of standard macroeconomic variables to shocks to (survey-based) long-term inflation expectations in a VAR for the euro area identified with sign and narrative restrictions and unveil the transmission mechanism by using local projections (Jordá, 2005). Second, it takes an off-the-shelf New Keynesian (NK) model, in which agents’ revise the estimate of the inflation target in response to changes in inflation, to show that it can qualitatively match the responses in the data when a shock to the perceived inflation target occurs. In both the VAR and the NK model, the paper highlights the role of monetary policy.
Negative shocks to long-term inflation expectations identified in the VAR exerted a downward pressure on euro-area inflation in 2013–14 and then in 2019. Had the European Central Bank (ECB) not responded to the negative shocks to inflation expectations by loosening monetary policy, output would have contracted and inflation would have declined by more. The NK model can replicate the VAR impulse responses and counterfactuals. After a negative shock to firms’ estimate of the inflation target, they revise downward the prices of their goods through their indexation to the target estimate, raising aggregate demand. The central bank responds to the fall in inflation by lowering the policy rate, but less than expected inflation. The increase in the real policy rate lowers aggregate demand. The net effect on output is a mild increase. If the central bank does not respond to the shock to firms’ target estimate, the ex-ante real policy rate increases by more, causing a larger decline in inflation and a contraction of output.
The analysis focuses on the euro area for several reasons. First, there is evidence that long-term expectations have de-anchored from the ECB's inflation target (Łyziak and Paloviita, 2017; Corsello et al., 2021). Second, until the review of the monetary policy strategy in 2021, the inflation target of the ECB was vaguely defined as “below, but close to, 2 per cent”, in contrast with most of the central banks in advanced economies that had a 2 per cent target. I exploit this feature to identify shocks to long-term inflation expectations as exogenous changes in the perception of the ECB’ inflation target. Third, since 2013, survey- and market-based indicators of long-term inflation expectations in the euro area have been systematically below those in the U.S, which have remained very close to 2 per cent. Fourth, inflation in the euro area between 2013 and 2019 was, on average, 1 per cent, compared with 1.6 in the U.S.
Related literature. – This paper relates to a number of theoretical and empirical contributions on the effects of shocks to inflation expectations and to the inflation target. Arias et al. (2016) show that a decline in long-term inflation expectations can be contractionary if the real interest rate increases because the policy rate is at its effective lower bound (ELB). The asymmetry posed by the ELB generates an asymmetry in the probability distribution of output, which becomes skewed towards negative outcomes. My results on the role of monetary policy are in line with the findings in Arias et al. (2016). Carvalho et al. (2022) show, using a small-scale model with endogenous long-term inflation expectations, that large and persistent forecast errors lead firms to doubt a constant inflation target and to switch to a constant gain expectation formation mechanism to estimate the target. The authors, however, do not evaluate the macroeconomic effects of changes in long-term expectations. Ireland (2007) estimates a NK model to infer about the Federal Reserve's inflation target. Exogenous increases in the target lead to a permanent increase in both inflation and the nominal short-term interest rate. De Michelis and Iacoviello (2016) use a VAR to quantify the impact of changes in the inflation target of the Bank of Japan. The authors set up a theoretical model, in which agents gradually update their estimate of the target, to account for the estimated response of nominal and real variables to the identified shocks to the inflation target.
To the best of my knowledge, very few VAR-based studies have assessed the macroeconomic impact of changes in long-term (survey-based) inflation expectations. Diegel and Nautz (2021) studies the role of long-term expectations in the monetary policy transmission mechanism using a VAR estimated on U.S. data. My paper is the first focusing on the euro area. Unlike Diegel and Nautz (2021), and De Michelis and Iacoviello (2016), who focus, respectively, on the U.S. and Japan, I study the euro area case, which is an interesting case as I have argued above. Compared with these studies, I do not assume that changes in long-term inflation expectations are due to exogenous variations in the inflation target, but rather to forecasters’ revisions of their estimates of the vaguely defined ECB's inflation target.
Outline. – The remainder of the article is organized as follows. Section 2 describes the VAR and the identification of the shocks. Section 3 presents VAR results on the macroeconomic effects and the transmission mechanism of shocks to long-term inflation expectations. Section 4 studies the role of monetary policy using structural counterfactuals. Section 5 assesses the robustness of the results. Section 6 uses a small-scale model to rationalize the empirical findings and Section 7 concludes.
Section snippets
The effects of changes in long-term expectations: a VAR approach
The empirical analysis is based on the assumption that movements in long-term inflation expectations reflect changes in forecasters’ estimates of the ECB's inflation target. I justify this assumption by resorting on the results of special rounds of the SPF. In Section 2.1, I describe the survey and the key characteristics of the long-term inflation expectations, focusing on the central tendency as well as the probability distribution.
Empirical results
Inference is based on 50,000 draws from the posterior distribution of the reduced form parameters of the VAR and 100 draws from the unitary sphere for each draw from the posterior. I discard around 22,400 draws (out of 50,000) from the posterior distribution of the VAR, as the maximum eigenvalue of the associated companion matrix implies explosive dynamics. About 10,700 draws, out of a total of 10 million, are retained for inference, on average 0.22 draws (out of 100) from the unitary sphere
The role of monetary policy: a counterfactual analysis
In this section, I assess the role of monetary policy in the transmission of shocks to long-term inflation expectations. I first carry out a counterfactual using the impulse responses of the VAR (Section 4.1). I then quantify the impact of the ECB's monetary policy response to the fall in long-term inflation expectations on inflation and real GDP growth in two periods in which expectations fell substantially: 2013–14 and 2019 (Section 4.2).
Robustness analysis
The robustness of the results with the VAR is tested along the following dimensions: its specification, the inclusion of a linear trend; the estimation period; the use of core consumer prices; the use of narrative restrictions; the addition of restrictions on the variance decomposition.13 I also check for the possibility
Rationalising the VAR evidence: a small scale new Keynesian model
In this section, I present the model that I used to rationalize the VAR evidence and to conduct counterfactuals to highlight the role of monetary policy. I do not aim at showing that the VAR can approximate well the data generating process of a linearized model.
The economy is populated by a representative household, a representative finished-goods producing firm, a continuum of intermediate-goods producing firms and a central bank. The central bank sets the policy rate according to a
Concluding remarks
In this paper, I have used data for the euro-area to quantify the macroeconomic consequences of shocks to (survey-based) long-term inflation expectations from the ECB's inflation target and to assess the role of its monetary policy in the transmission of these shocks.
A decline in long-term inflation expectations, which is not countered by monetary policy, can lead to a contraction in economic activity and to larger negative effects on inflation, compared with the case in which the central bank
References (48)
- et al.
Measuring euro area monetary policy
J. Monet. Econ.
(2019) - et al.
The macroeconomic risks of undesirably low inflation
Eur. Econ. Rev.
(2016) - et al.
Misspecifications in vector autoregressions and their effects on impulse responses and variance decompositions
J. Econom.
(1993) - et al.
Monetary disturbances matter for business fluctuations in the G-7
J. Monet. Econ.
(2002) - et al.
Business cycle measurement with some theory
J. Monet. Econ.
(2011) - et al.
Anchored or de-anchored? That is the question
Eur. J. Polit. Econ.
(2021) - et al.
Long-term inflation expectations and the transmission of monetary policy shocks: evidence from a SVAR analysis
J. Econ. Dyn. Control
(2021) Measuring the stance of monetary policy in zero lower bound environments
Econ. Lett.
(2013)- et al.
Modest policy interventions
J. Monet. Econ.
(2003) - et al.
Anchoring of inflation expectations in the euro area: recent evidence based on survey data
Eur. J. Polit. Econ.
(2017)
Disinflationary shocks and inflation target uncertainty
Econ. Lett.
Bayesian skepticism on unit root econometrics
J.Econ. Dynamics Control
Discretion versus policy rules in practice
Carnegie-Rochester Conference Series on Public Policy
What are the effects of monetary policy on output? Results from an agnostic identification procedure
J. Monet. Econ.
Inflation expectations and the pass-through of oil prices
Rev. Econ. Stat.
Sentiments
Econometrica
Quantifying confidence
Econometrica
Narrative sign restrictions for SVARs
Am. Econ. Rev.
Inference based on Structural Vector Autoregressions identified with sign and zero restrictions: theory and applications
Econometrica
The perils of nominal targets
Rev. Econ. Stud.
News, noise, and fluctuations: an empirical exploration
Am. Econ. Rev.
Information, animal spirits, and the meaning of innovations in consumer confidence
Am. Econ. Rev.
Unconventional monetary policy and the Great Recession: estimating the macroeconomic effects of a spread compression at the zero lower bound
Int. J. Central Bank.
Missing disinflation and missing inflation: a VAR perspective
Int. J. Central Bank.
Cited by (1)
The Ecb's Time-Varying Policy Trade-Off
2024, SSRN
- ✰
Economic Outlook and Monetary Policy Directorate, Banca d'Italia. The views expressed in this article are those of the author alone and do not necessarily represent the position of Banca d'Italia or the Eurosystem. I would like to thank Fabio Canova, Matteo Iacoviello, Martin Ellison, Òscar Jordá, Marianna Riggi, Francesco Corsello, Alex Tagliabracci, Eva Ortega, Fabrizio Venditti, Alessandro Notarpietro and Tiziano Ropele for their useful comments. I am particularly grateful to Marco Bernardini and Antonio Conti for also helping me with the WinRats code. I also thank the participants at the conference “Advances in Business Cycle Analysis, Structural modelling and VAR Estimation” in honour of Fabio Canova and held in Hydra (Greece) on 23–24 October 2021; see the webpage: https://sites.google.com/view/fabio-canova-homepage/home/conference-in-my-honour. I am deeply indebted to Fabio for being my Ph.D. supervisor at “Pompeu Fabra”. I thank Valentina Schirosi for proofreading the text. For more details on the robustness of the results, see the working paper version at https://www.bancaditalia.it/pubblicazioni/temi-discussione/2021/2021–1357/index.html.