Output Comovement and Inflation Dynamics in a Two‐Sector Model with Durable Goods: The Role of Sticky Information and Heterogeneous Factor Markets

• Published date: 01 February 2022
• Author: TOMIYUKI KITAMURA,TAMON TAKAMURA
• Date: 01 February 2022
• DOI: http://doi.org/10.1111/jmcb.12800

DOI: 10.1111/jmcb.12800

TOMIYUKI KITAMURA

TAMON TAKAMURA

Output Comovement and Ination Dynamics in a

Two-Sector Model with Durable Goods: The Role

of Sticky Information and Heterogeneous Factor

Markets

In the United States, the ination and output of durable and nondurable

goods respond to a monetary policy shock in the same direction with a delay.

However, the existing NewKeynesian dynamic stochastic general equilib-

rium models that generate the positive output comovement cannot explain

this delayed response in sectoral ination. We show that adding sticky in-

formation to both goods along with heterogeneous factors of production can

explain the observed patterns in sectoral ination and output. Moreover, in

line with recent empirical ndings, the estimated information stickiness is

larger for housing than for nondurable goods and services.

JEL codes: E31, E32, E52

Keywords: information rigidity, production factor immobility, sectoral

comovement, ination, durability, monetary policy

S  G R,  has been a renewed

interest in studying the role of durable goods (e.g., housing) in explaining economic

booms and slow recoveries. In order to understand how these sectors affect business

cycles and the monetary policy transmission mechanism, researchers have developed

The previous version of the paper has been circulated as “Can Sticky Information Solve the Negative

Comovement Problem in a Two-SectorModel with Durable Goods?” The authors would like to thank the

editor Pok-sang Lam and two anonymous referees for valuable comments. Weare also indebted to Jason

Allen, Gino Cateau, Geoffery Dunbar, Bill Dupor, Paul Evans, StefanoGnocchi, Yuriy Gorodnichenko,

Aubhik Khan, Christien Mitchell, Wataru Miyamoto, Yaz Terajima, and Julia Thomas, as well as seminar

participants at Midwest Macroeconomics Meetings and Econometrics Society Asian Meetings for helpful

comments and discussions. Views expressed in this paper are solely those of the authors and may differ

from ofcial views of the Bank of Canada and the Bank of Japan. No responsibility for them should be

attributed to the Bank of Canada or the Bank of Japan.

T K is a Director, Bank of Japan (E-mail: tomiyuki.kitamura@boj.or.jp). T

T is a Senior Economist, Bank of Canada (E-mail: ttakamura@bankofcanada.ca).

Received February 14, 2018; and accepted in revised form January 15, 2021.

Journal of Money, Credit and Banking, Vol. 54, No. 1 (February 2022)

© 2021 The Ohio State University

314 :MONEY,CREDIT AND BANKING

dynamic stochastic general equilibrium (DSGE) models that explicitly take into ac-

count the production and consumption of durable goods and housing. In the context

of New Keynesian models, a good deal of work has analyzed howto replicate the ob-

served response patterns in sectoral output following a monetary policy shock. This

analysis is based on the recognition that a plain-vanilla, two-sector sticky-price model

with durable goods exhibits a counterfactual, negative output comovement,as shown

by Barsky, House, and Kimball (BHK 2007). These studies show that frictions such

as sticky wages, input–output linkages, and borrowing constraints, among others,

can restore the positive output comovement pattern observed in the data. However,

none of the existing models can simultaneously account for the delayed responses

in sectoral ination and the positive comovement of sectoral output, both of which

are salient features observed in a vector-autoregression (VAR) analysis. As we show

in Section 1 of our paper, building on the work of Erceg and Levin (2005, 2006),

the ination and output of residential investment, durable goods, and nondurable

goods plus services exhibit hump-shaped responses, reaching their peak with a

delay.

To ll this gap in the literature, this paper analyzes whether a two-sector DSGE

model, featuring sticky information of the type considered by Mankiw and Reis

(2002) in both types of goods and rm-specic labor and capital studied in Wood-

ford (2003), is capable of explaining the ination dynamics at the sectoral level

in addition to the comovement of output. As shown by Mankiw and Reis (2002,

2006), sticky information generates procyclical and delayed responses in the aggre-

gate output and ination following a monetary policy shock in a one-sector model.

However, it is not clear whether information stickiness ensures the sectoral ina-

tion responds with a delay and the output of different types of goods comoves pos-

itively. This is because information stickiness in each sector may have additional

impacts on the sectoral price adjustments, which, in turn, could affect the sub-

stitution between different types of goods. Thus, our paper examines how sticky

information interacts with heterogeneous production factors to generate the slug-

gish dissemination of monetary policy effects observed in the sectoral ination and

output.

In addition, to evaluate how well our model can quantitativelyexplain responses in

the data, we estimate sectoral output and ination responses to an unexpected mon-

etary shock using VAR. We then apply the minimum-distance estimation method on

the VAR responses following Christiano, Eichenbaum, and Evans (2005). Speci-

cally, we estimate the degree of information stickiness in each sector, to minimize

the distance between the model-implied and VAR impulse responses to a monetary

policy shock. Target variables matched in this exercise are residential investment,

the output of nondurable goods plus services, and the ination rates of these two

variables. Following the argument in BHK (2007), we treat residential investment as

exibly priced durable goods, given the fact that house prices are negotiable.

The choice of sticky information in our model is supported by a growing number

of empirical studies documenting the prevalence of information rigidities in various