key: cord-1053629-dumuimbz
authors: Moessner, Richhild
title: Determinants of Inflation Expectations in the Euro Area
date: 2022-04-08
journal: Intereconomics
DOI: 10.1007/s10272-022-1037-6
sha: c76e0a97ddc70b469620555e0b11dec2cd35cd58
doc_id: 1053629
cord_uid: dumuimbz

Euro area inflation has been rising strongly in the wake of the COVID-19 pandemic, giving rise to concerns that there could be second-round effects, with higher inflation leading to higher inflation expectations, which in turn lead to higher inflation. This could result in more persistent rises in inflation.

Forum vanced economies, and on testing the implications of the theoretical literature and evaluating the degree of anchoring of expectations. 1 The Eurosystem work stream on infl ation expectations (Eurosystem work stream on infl ation expectations, 2021) has recently studied the determinants of infl ation expectations in the euro area within the ECB's monetary policy strategy review, but without considering the eff ects of exchange rate changes. They fi nd that short-term euro area infl ation expectations have been aff ected by oil prices, monetary policy shocks and central bank infl ation projections. Galati et al. (2018) also fi nd that short-term infl ation expectations in the euro area have been aff ected by oil prices. Household infl ation expectations in the United States and Japan responded to changes in food and oil prices (Ueda, 2010) . Gerlach et al. (2011) fi nd that shortterm infl ation expectations in major advanced and emerging economies have been aff ected by food, energy and core consumer price infl ation. Patra and Ray (2010) fi nd that lagged infl ation, movements in food and fuel prices and the output gap are the main determinants of shortterm infl ation expectations in India. They note that the scarce literature on the determinants of infl ation expectations in emerging economies has mainly focused on target credibility or the role of fi scal expectations (Minella et al., 2003; Celasun et al., 2004) .

Euro area infl ation has been rising strongly in the wake of the COVID-19 pandemic, giving rise to concerns that there could be second-round eff ects, with higher infl ation leading to higher infl ation expectations, which in turn lead to higher infl ation. This could result in more persistent rises in infl ation. It is therefore important to study the drivers of infl ation expectations.

Our paper contributes to a better understanding of the determinants of short-term infl ation expectations in the euro area. As possible determinants of infl ation expectations, we consider food and energy prices, both countryspecifi c measures in the form of components of the consumer price index (CPI), and global food commodity prices and oil prices. We also consider the eff ects of changes in exchange rates and global freight prices. Moreover, we include the output gap as an explanatory variable.

Infl ation expectations matter for infl ation in economic models commonly used for monetary policy analysis (Clarida et al., 1999; Smets, 2003; Woodford, 2003; Levin and Moessner, 2005; Moessner, 2021) . The recent debate on monetary policy frameworks, including average infl ation targeting, focused on the role of infl ation expectations, e.g. in the monetary policy strategy reviews of the US Federal Reserve and the European Central Bank (ECB) (Powell and Wessel, 2020 ; Eurosystem work stream on infl ation expectations, 2021). The decisions by the Federal Reserve to adopt an average infl ation targeting framework in 2020, and by the ECB to change to a symmetric infl ation target of 2% in 2021, were partly based on the role of infl ation expectations. Policymakers also considered infl ation expectations to be important for economic outcomes and monetary policy decisions in the past, despite diffi culties in identifying their determinants: Former chair of the Federal Reserve Alan Greenspan noted, "I am not saying what that [infl ation expectations] is a function of. We know it's a very diffi cult issue, but that is the key variable. It's important, but just because we can't make a judgment as to what these driving forces are in an econometric sense doesn't mean that it's not real" (Federal Open Market Committee, 1994; Coibion et al., 2018) .

The literature on infl ation expectation formation mechanisms has been reviewed in Coibion et al. (2018) , who argue that survey-based expectations should be included more systematically in macroeconomic studies. Shiller (1978) noted the importance of studying the determination of infl ation expectations. A recent review of the literature on the determinants of infl ation expectations in advanced and emerging economies and new evidence can be found in Kose et al. (2020) . They conclude that empirical studies on infl ation expectations have mostly focused on ad-Richhild Moessner, Bank for International Settlements, Basel, Switzerland; and CESifo, Munich, Germany.

We consider the following 16 euro area member countries: Austria, Belgium, Estonia, Finland, France, Germany, Greece, Ireland, Italy, Latvia, Lithuania, the Netherlands, Portugal, Slovakia, Slovenia and Spain. The sample period is from 2000Q1 to 2021Q1 at quarterly frequency.

We study the determinants of infl ation expectations by estimating the following dynamic fi xed eff ects panel regressions, using a panel of 16 euro area member countries: 

where π e it denotes next-year CPI infl ation expectations from Consensus Economics surveys in percent; π it food CPI and π it energy CPI denote year-on-year CPI food price infl ation and CPI energy price infl ation, respectively, in percent, in country i at time t; outputgap it denotes the output gap; ΔNEER it is the quarter-on-quarter change in the nominal eff ective exchange rate in percent, calculated from the log change in the nominal eff ective exchange rate, with an increase indicating an appreciation of the domestic currency; π t commodity are year-on-year changes in global commodity prices; fi nally, α i are country fi xed eff ects to control for observed and unobserved country heterogeneity. We use robust standard errors clustered at the country level.

Equation (1) is our baseline specifi cation and the results are shown in column I of Table 1 . We fi nd that countryspecifi c food CPI infl ation has a signifi cant positive eff ect on professionals' survey-based infl ation expectations. A ten percentage point increase in food CPI infl ation leads to an increase in infl ation expectations of around 0.5 percentage points. By contrast, energy CPI infl ation has no signifi cant eff ect. We also fi nd that depreciations of the domestic exchange rate in nominal eff ective terms have a signifi cant positive eff ect on infl ation expectations. A 10% depreciation over the quarter of the domestic exchange rate in nominal eff ective terms leads to an increase in infl ation expectations of around 0.7 percentage points. Moreover, we fi nd that the output gap has signifi cant positive eff ects on infl ation expectations. These eff ects are in addition to those of lagged infl ation expectations. The infl ation expectations are highly persistent, with a coefficient of around 0.7 on lagged infl ation expectations.

For robustness, we also estimate versions of equation (1) where we add core CPI infl ation, π it core CPI , and changes (year-on-year) in global freight prices, π t freight . These re-This paper analyses the determinants of survey-based short-term infl ation expectations of professionals in the euro area, using dynamic cross-country panel estimation for 16 euro area member countries from the fi rst quarter of 2000 to the fi rst quarter of 2021. We use survey-based CPI infl ation expectations of professionals, since they are available on a comparable basis for the countries in our sample, and since they are not distorted by risk and liquidity premia, in contrast to fi nancial market-based measures. 2 We fi nd that country-specifi c food CPI infl ation has a signifi cant positive eff ect on professionals' survey-based infl ation expectations in the euro area. This eff ect is larger than that of energy CPI infl ation and of oil and global food commodity prices. We also fi nd that depreciations of the nominal eff ective exchange rate lead to signifi cantly higher infl ation expectations. Moreover, infl ation expectations in the euro area are persistent and the output gap has signifi cant positive eff ects.

We use data on professionals' survey-based CPI shortterm infl ation expectations. These are taken from Consensus Economics surveys for next-year CPI infl ation expectations.

Data on headline consumer price indices (CPI) comes from Datastream and national sources. Data on food CPI indices, energy CPI indices and core CPI indices are based on data from the OECD, national data and Bank for International Settlements (BIS) estimations. Core CPI infl ation is defi ned as excluding food and energy. Brent oil prices (quarterly averages, US dollar per barrel) are from Datastream. Global commodity prices are taken as the IMF all commodity price index. Global food commodity prices are taken as the UN FAO food nominal price index. As a measure of global freight prices we use quarterly averages of the Baltic Dry Index, as reported daily by the Baltic Exchange in London, from Datastream (in US dollar per points). This index provides a benchmark for the price of moving the major raw materials by sea.

Data on output gaps (as a percentage of potential GDP) was obtained from the OECD, and is linearly interpolated from annual data. Nominal eff ective exchange rate indices (broad indices, quarterly average) are from the BIS, with an increase refl ecting an appreciation of the domestic currency. l l a f

Here, π t food denotes changes (year-on-year) in global food commodity prices, and π t oil denotes changes (year-onyear) in oil prices in percent. The results of equation (2) are shown in column IV. We fi nd that the eff ect of global food prices is positive and signifi cant, but smaller than that of country-specifi c food CPI infl ation. Oil prices also have a small signifi cant positive eff ect on infl ation expectations. Again, exchange rate depreciations and the output gap have signifi cant positive eff ects on infl ation expectations of a very similar magnitude to those for the baseline specifi cation in equation (1).

We also estimate a version of equation (2) where we consider changes in global commodity prices instead of oil prices and global food prices. These results are shown in column V of Table 1 . We fi nd again that global commodity prices have a signifi cant positive eff ect on infl ation expectations, with a similar magnitude to the baseline specifi cation shown in column I.

This paper analysed the determinants of short-term infl ation expectations in the euro area based on surveys of professionals, using dynamic cross-country panel estimation for 16 euro area member countries. We fi nd that country-specifi c food CPI infl ation has a signifi cant positive eff ect on professionals' survey-based infl ation expectations. A ten percentage point increase in food CPI infl ation leads to an increase in infl ation expectations of around 0.5 percentage points. This eff ect is larger than that of global food prices. Oil prices and global commodity prices also have a signifi cant positive eff ect.

We also fi nd that depreciations of the domestic exchange rate in nominal eff ective terms lead to signifi cantly higher infl ation expectations. A 10% depreciation over a quarter of the domestic exchange rate in nominal eff ective terms leads to an increase in infl ation expectations of around 0.7 percentage points. Moreover, infl ation expectations in the euro area are persistent and positively aff ected by the output gap.

These results on the drivers of infl ation expectations are relevant for monetary policy, since euro area infl ation has been rising strongly in the wake of the COVID-19 pandemic. Additionally, there are concerns of possible second-sults are shown in columns II and III of Table 1 , respectively. We fi nd that core CPI infl ation has no explanatory power over and above the output gap. Moreover, the effects of changes in global freight costs are insignifi cant. The eff ects of exchange rate depreciations, food CPI infl ation and the output gap remain signifi cant and of similar magnitude as in the baseline specifi cation of column I.

We also estimate the following specifi cation, where we use changes in global food commodity prices and in oil Table 1 Short-term infl ation expectations Note: Fixed eff ects panel estimation; sample period: 2000Q1-2021Q1. Robust standard errors clustered at the country level. ***/**/* denote statistical signifi cance at the 1%/5%/10% confi dence level, respectively. Forum round eff ects, with higher infl ation leading to higher infl ation expectations, which in turn lead to higher infl ation. This could result in more persistent infl ation increases. Such second-round eff ects would make it more diffi cult and costly for monetary policy to bring infl ation back down to target.

Obstacles to Disinfl ation: What is the Role of Fiscal Expectations?

The Science of Monetary Policy: A New Keynesian Perspective

The Formation of Expectations, Infl ation, and the Phillips Curve

Infl ation expectations and their role in Eurosystem forecasting

Transcript for meeting of 5-6

Defl ation risk in the euro area and central bank credibility

Did the crisis aff ect infl ation expectations?

Infl ation expectations and the great recession

Exchange rate passthrough: what has changed since the crisis?

Infl ation Expectations: Review and Evidence

Infl ation persistence and monetary policy design: an overview

Infl ation Targeting in Brazil: Constructing Credibility under Exchange Rate Volatility

Eff ects of infl ation expectations on infl ation, CESifo Working Paper

Infl ation Expectations and Monetary Policy in India: An Empirical Exploration

What are infl ation expectations? Why do they matter?

Rational Expectations and the Dynamic Structure of Macroeconomic Models: A Critical Review

Maintaining price stability: how long is the medium term?

Determinants of households' infl ation expectations in Japan and the United States