Research

Working papers

State-Dependent Central Bank Communication with Heterogeneous Beliefs Banque de France Working paper 875 (Submitted AER)

This paper studies the optimal disclosure strategy of a Sender who wishes to influence heterogeneous Receivers’ expectations through public disclosure. I introduce heterogeneous priors in a standard bayesian persuasion model a la Gentzkow and Kamenica (2011) and characterize how the optimal disclosure depends on the heterogeneity of beliefs. I show that heterogeneity matters in two ways: (i) it is optimal to be send signals with positive error probabilities, contrarily to the homogeneous belief case; (ii) higher dispersion in beliefs brings the information authority to choose lower error probabilities. I apply the framework to a Central Bank communication problem in which the policy maker communicates about aggregate demand to distort firms’ investment decisions. I test the model’s predictions empirically by showing that FOMC unemployment rate forecasts are systematically biased in opposite directions in recessions and expansions, and the less so the higher the private sector disagreement is, in line with the model’s predictions.

Replicability of a journal: a comprehensive overview (joint with H. Kingi, F. Stanchi and L.Vilhuber) Banque de France Working paper 853 (R&R CJE) Slides [SUERF Polcy Brief]

Replication, reproduction, and falsification of published articles are an important part of the scientific endeavor, and have been widely discussed. The lessons learned from earlier replication exercises such as (Dewald et al, 1986) lead several journals to implement code and data depository requirements. These should have lead to improved reproducibility of journal archives, since a good part of the failure to replicate is due to poor availability of replication materials (McCullough et al, 2006). We set out to test the latter assumption. We assessed all articles published in the American Economic Journal: Applied Economics (AEJ:AE) as to the reproducibility of their computational results, and tested the reproducibility for all those with some data available.

Measuring and Comparing Consumption Inequality in France and the United States (joint with A.Accardo, C.Jude and A.Penalver) Banque de France Working paper 904

This paper constructs an annual dataset of consumption by income quintiles for France since 1995 in order to make a granular comparison of consumption inequality with the United States. We first match consumption data with the national accounts and then use a Kalman filter to interpolate missing observations so as to obtain yearly data. We validate this technique by applying it to a US dataset with pseudo-missing data comparable to our French data. Comparing consumption and income inequality across the two countries, we find consumption inequality to be higher in the US and overall not to mirror income inequality. Finally, based on the weights of the different consumption items, we construct annual price indices by quintile. We find that dispersion in inflation across income groups is less marked in France than the United States.

Econometric Analysis of Regime Switches and of Fiscal Multipliers OFCE Working paper 2014-01

In this paper, I extend the technique of vector autoregressions (VARs) to account for the possibility of a state-dependent fiscal multipliers for France, Germany, Italy and the United States. I estimate a non linear smooth transition vector autoregression, following Auerbach and Gorodnichenko (2012a). My results suggest that the output multiplier of government purchases is significantly higher in recessions than expansions.

Work in progress

The Signaling Effect of State-Dependent Central Bank Communication

In this paper, I document the state-dependent topic composition of FOMC minutes using natural language processing techniques. I then propose a model in which the state-dependent reporting of the central bank impacts agents’ expectations about both the state of the economy and the interest rate, therefore creating a ``signaling effect’’ about an interest rate change. Taking into account this signaling effect of the topic composition, I derive the central bank’s optimal state contingent communication policy.

Central Bank Information Shocks: Evidence from Speeches} (joint with P.Hubert and M.Le)

We first quantify the contribution of speeches to markets’ movements over the inter-meeting periods for the ECB and the Fed. We identify the speeches that act like monetary policy shocks, through textual analysis. We then estimate the macroeconomic effects of such shocks.

The Effect of Monetary Policy on Consumption Inequality (joint with A.Accardo, C.Jude and A.Penalver)

Based on the methodology and data set developed in Accardo et al. (2022), we estimate the impact of monetary policy shocks on consumption inequality for France and the United States.

Are Internal and External Communication Strategies of Central Banks Consistent? (with K.Istrefi and G.Sestieri)

We document how the minutes and statements’ composition for respectively the Federal Reserve, ECB and BoE vary with their information set. Through the lens of a model of state-dependent communication, we show that the Fed behaves optimally with its statements and minutes’ coverage of inflation. We then explore how markets react on days for which internal (minutes) and external (statements) communications do not coincide, testing if they internalize the communication strategy of the central bank.

Media Coverage of Central Banks’ Monetary Policy: Heterogeneity and its Implication (with K.Istrefi, B.Sagna and G.Sestieri)

We document the media coverage of three main central banks, in terms of time-variability, homogeneity across newspapers and across countries. Building a model making explicit the editorial role of newspapers, we study the impact on expectations and implications for the macroeconomy.

State-Dependent Central Bank Communication: a DSGE approach

In a model where the policy maker choose both the communication and interest rate instrument, I study how the central bank should adapt its communication according to the state of the economy (realization of the fundamental). To do so, this paper extends the state-dependent communication model as developed in Herbert (2022).

Older work

The Pass-through of the Sovereign Debt Crisis to Bank Lending in the Eurozone: a Narrative Approach

Layoffs and Wage Rigidity (joint with C.Huckfeldt)

Blog, Policy

Why can sovereign and corporate borrowers borrow at negative rate? Banque de France Econ Notepad 212, with Adrian Penalver and Maria-Sole Pagliari [Link]

Economists: not enough transparency? Evidence from a reproducibility exercise of a top economic journal. SUERF Policy Brief 288 (joint with L.Vilhuber) [Link]

Towards more clarity in central bank communication (joint with K.Istrefi and G.Sestieri)

Discussions

Reactions of household inflation expectations to a symmetric inflation target and high inflation (Galati et al., 2022), ESCB Research Cluster 2022, Slides

Good Policy or Learning Evolution? A Markov-Switching approach to Understanding the Determinants of Fed policy (Best and Hur 2022), WEAI, Slides

Fiscal Policy and Inflation Expectations (Mello et Ponce 2022), CEMLA-ECB-NY FED conference Slides

Consumer Behavior at Low and Negative Interest Rates: Evidence for a Savings’ Reversal (Felici et al. 2021), ESCB Research Cluster 2021 Slides

Reading Between the Line: Using Text Analysis to Estimate the Loss Function of the ECB (Paloviita et al. 2020) - ESCB Research Cluster 2020 Slides

Fireside Chats: Communication and Consumers’ Expectations in the Great Depression (Mathieu Pedemonte), Columbia University - Young Economist Symposium, 2019 Slides

Money Flow in a Dynamic Economy (Larry Marsh), Midwest Economic Association, 2019 Slides

Measuring ECB’s Communication: a Media-based Automated Approach (Francesco Pesci), Conference on the New Challenges of Central Bank Communication, UC Louvain, 2018 Slides