Industrial structure and the probability of crisis: Stability is not resilience*

• Date: 01 January 2019
• Author: Dongyeol Lee,Hyunjoon Lim
• Published date: 01 January 2019
• DOI: http://doi.org/10.1002/ijfe.1658

Received: 14 February 2018 Revised: 11 August 2018 Accepted: 9 September 2018

DOI: 10.1002/ijfe.1658

RESEARCH ARTICLE

Industrial structure and the probability of crisis: Stability is

not resilience*

Dongyeol Lee1,2 Hyunjoon Lim2

1International Monetary Fund,

Wash ingto n, DC

2The Bank of Korea, Seoul, South Korea

Correspondence

Dongyeol Lee, International Monetary

Fund, 700 19th St NW,Washington, DC

20431, USA.

Email: leedongr@gmail.com

JEL Classification: F33, G01, L16

Abstract

We utilize country-level data to investigate the empirical linkage between an

economy's industrial structure and the probability of a banking crisis. This paper

shows that a higher share of the service sector tends to significantly increase the

risk of a banking crisis. We also explore the potential channels through which

the industrial structure may affect an economy's vulnerabilityto external shocks.

The result of the analysis suggests that a higher share of the service sector sub-

stantially increases vulnerability to a banking crisis through deterioration in

profits and worsening of the current account balance. Our study provides some

implications for the recent acceleration of deindustrialization in several emerg-

ing and advanced economies: for example, the industrial policy should call for

a more cautious approach to reduce the potential risks of deindustrialization

(increase in crisis vulnerability).

KEYWORDS

banking crisis, deindustrialization, industrial structure

1INTRODUCTION

There is a broad consensus among economists that an

increasing share of services in an economy should gen-

erally reduce its output volatility at the cost of slowing

growth. The growing interest in this topic has been under-

pinned by the dramatic decline in output volatility in

most industrialized economies since the 1980s, which has

been termed “the Great Moderation”. A substantial body

of literature has documented the Great Moderation (e.g.,

Blanchard & Simon, 2001; Kim & Nelson, 1999; McConnell

& Perez-Quiros, 2000; Stock & Watson, 2003).1Decom-

posing output by one-digit industry, Eggers and Ioannides

(2006) find that compositional shifts accounted for a lit-

tle less than half the sharp decline in the US volatil-

ity after the mid-1980s, mostly representing the decline

of manufacturing. Many studies have used theoretical

*The views expressed in this paper are those of the authors, and do not

necessarily represent those of the International Monetary Fund or the

Bank of Korea.

models incorporating multiple sectors to show that a ris-

ing share of the service sector reduces output volatility

(e.g., Da-Rocha & Restuccia, 2006; Duarte & Restuccia,

2010; Moro, 2012; Ngouana, 2013). It is intuitively plau-

sible that, given a lower volatility of labor productivity in

services than in manufacturing, a higher share of the ser-

vice sector will help an economy to reduce the volatility

of the aggregate output. In line with this finding, policy-

makers in some countries that are highly dependent on

manufacturing, including China, Korea, Japan and many

European countries, have viewed fostering the service sec-

tor and further reinforcing its competitiveness as a priority

for achieving sustainable growth.2

The crises in recent years, however, highlight that a ris-

ing share of the service sector may have divergent and

potentially critical implications. Specifically,this shift may

cause escalated vulnerability in an economy while still

helping reduce the economic volatility. A statement that

the growing service sector simultaneously causes lower

volatility and escalates vulnerability may initially appear

212 © 2018 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/ijfe IntJ Fin Econ. 2019;24:212–226.