Demographics, Monetary Policy, and the Zero Lower Bound

• Published date: 01 October 2023
• Author: MARCIN BIELECKI,MICHAŁ BRZOZA‐BRZEZINA,MARCIN KOLASA
• Date: 01 October 2023
• DOI: http://doi.org/10.1111/jmcb.12972

DOI: 10.1111/jmcb.12972

MARCIN BIELECKI

MICHAŁ BRZOZA-BRZEZINA

MARCIN KOLASA

Demographics, Monetary Policy, and the Zero

Lower Bound

We developa New Keynesian life-cycle model to assess the importance of

population aging for monetary policy. The model successfully matches the

age proles of consumption-savings decisions made by European house-

holds. It implies that demographic trends contribute signicantly to the de-

cline of the natural rate of interest (NRI) and potential output growth, and

exacerbate the risk of hitting the zero lower bound (ZLB), given the current

ination targets. Under a realistic assumption that the central bank updates

its estimates of the NRI only with some lag, aging may additionally lead to a

sizable and persistent deationary bias, elevating the ZLB risk evenfurther.

JEL codes: E31, E52, J11

Keywords: aging, life-cycle models, monetary policy, zero lowerbound,

imperfect information

L     longevity have be-

come standard features of most developed and several developing economies.

We would like to thank Janusz Jabłonowski for sharing his estimates of the life-cycle proles from

the Household Finance and Consumption Survey (HFCS), and Stephen Cecchetti and Jacek Suda for use-

ful discussions. Comments received at the Chief Economists Meeting at the Bank of England, NBP-NBU

Conference in Kyiv,Dynare conference in Tokyo, RCEA conference in Warsaw, Computing in Economics

and Finance conference in Milan, Society for Economic Dynamics Annual Meeting in Mexico City, Mid-

west Macroeconomic Meeting in Nashville, and seminars at Deutsche Bundesbank, Lietuvos Bankas, and

Narodowy Bank Polski are gratefully acknowledged as well. Most of this paper was written when Marcin

Bielecki and Marcin Kolasa were workingat Narodowy Bank Polski. The views expressed herein are those

of the authors and not necessarily those of Narodowy Bank Polski, European Central Bank or International

Monetary Fund.

M B is an Assistant Professor at the Facultyof Economic Sciences, University of Warsaw

and an Economist at the Forecasting and Policy Modelling Division, Directorate General Economics,

European CentralBank (E-mail: m.p.bielecki@uw.edu.pl). M B-B is a Professorat

the Department of Quantitative Economics, SGH WarsawSchool of Economics and an Economic Advisor

at Narodowy Bank Polski (E-mail: michal.brzoza-brzezina@nbp.pl). M K is a Professor at

the Department of Quantitative Economics, SGH Warsaw School of Economics and a Senior Financial

Sector Expert at the Monetary and Capital Markets Department, International Monetary Fund (E-mail:

mkolas@sgh.waw.pl).

Received March 26, 2020; and accepted in revised form August 23, 2021.

Journal of Money, Credit and Banking, Vol. 55, No. 7 (October 2023)

© 2022 The Ohio State University.

1858 :MONEY,CREDIT AND BANKING

Recently, these demographic trends have also become of interest to monetary eco-

nomics, mainly because of their impact on the natural rate of interest (NRI) and po-

tential output, that is, two key variables affecting stabilization policies implemented

by the centralbanks.

The goal of this paper is to offer a quantitative assessment of how demographics

affects the ability of monetary authorities to stabilize the economy. To this end, we

use a full-scale life-cycle framework pioneered by Auerbach and Kotlikoff (1987).

Our detailed setup allows us to incorporate granular data on fertility and mortality

rates as well as age proles of labor income and labor supply, estimated using the

Household Finance and Consumption Survey (HFCS) for the euro area countries,

and to closely match the resulting life-cycle patterns of consumption-savings deci-

sions to these data. Even more importantly, we extend the life-cycle framework for

New Keynesian features. This allows us to offer a novel, quantitative analysis of the

consequences of demographic trends for monetary policy,going beyond the effects of

aging on the NRI and potential output, which were already discussed by the extant lit-

erature that we summarize later. In particular,we can calculate how the probability of

hitting the zero lower bound (ZLB) has evolved(and will evolve) due to demographic

changes and its possible interactions with the pension system design. Furthermore, we

show what the declining NRI and potential output imply for monetary policy under

a realistic scenario where the central bank faces imperfect information and estimates

these unobservable variables in real time. To achieve this, we additionally extend our

framework in a way that mimics central bank learning about the actual levels of the

NRI and potential output.

Our model parameterization and quantitative simulations are focused on the euro

area, which, similarly to many other developed economies, faces a continuing in-

crease in the old-age dependency ratio due to low fertility and falling mortality. We

nd that, despite its glacial speed, the impact of aging on the economy can be sub-

stantial from the monetary policy perspective. In particular, given the history and

currently available projections, the NRI in the euro area is projected to decline due

to demographic forces by more than 1.5 percentage points between 1980 and 2030.

The growth rate of potential output per capita declines by 0.6 percentage points over

the same period. Updating the estimates of these two latent variables by the cen-

tral bank in real time makes the demographic transition neutral for price stability, at

least when the ZLB constraint can be ignored. However, if the monetary authority

learns about the impact of demographic processes on the NRI and potential output

only slowly over time, the outcome is a prolonged period of below-target ination.

This deationary bias may be sizable—over 0.3 percentage points for several years.

Furthermore, according to our simulations, demographic developments signicantly

increase the probability of monetary policy being constrained by the ZLB. Even if the

NRI and potential output are perfectly observed by the monetary authority,the annual

probability of hitting the constraint increases from just below 1% in the 1980s to al-

most 4% in 2030. These numbers accumulate to a dramatic increase of the chance

of hitting the ZLB over longer horizons: the probability rises from less than 3%

for the whole 1980s to almost 30% for the 2020s. Under the imperfect information