Multiple Structural Breaks and Inflation Persistence: Evidence from China

• DOI: http://doi.org/10.1111/asej.12044
• Date: 01 March 2015
• Author: Tianfeng Li,June Wei
• Published date: 01 March 2015

Multiple Structural Breaks and Inflation

Persistence: Evidence from China*

Tianfeng Li and June Wei

Received 18 July 2013; accepted 7 November 2014

The degree of inflation persistence fluctuates more in developing countries than in

industrialized countries. This is because there are more economic reforms in

developing countries than in industrialized countries. Hence, it is important to

study the inflation persistence in developing countries. The present paper investi-

gates the pattern of inflation persistence in China, a developing and transitional

economy. Specifically, the paper studies statistically the number of structural

breaks in China’s inflation persistence based on the monthly retail price index

(MRPI) and the quarterly retail price index (QRPI) inflation series from 1983 to

2011. The findings show that there are five and three structural breaks in the MRPI

and QRPI inflation persistence, respectively. The present paper also exposes a high

degree of persistence over the whole sample period and a slight decline in the level

of the persistence since 1994. Furthermore, the persistence of two inflation series

does not change monotonously but with obvious cyclical patterns. Using these

derived estimated structural breakpoints, likely causes of the breakpoints in infla-

tion persistence are investigated. Finally, theoretical and practical implications are

discussed.

Keywords: economic reforms, fiscal policy, inflation persistence, monetary policy,

multiple structural breaks.

JEL classification codes: C22, E31, E52, E58, P21.

doi: 10.1111/asej.12044

I. Introduction

To keep inflation at low and stable levels much attention has been paid to better

comprehending the properties of inflation dynamics. The consensus is that

medium-term and long-term inflation is entirely a monetary phenomenon;

however, considerable controversy still exists regarding the related aspects of

inflation persistence. Among these aspects is the extent to which inflation

sluggishly converges after a shock. Hence, a better understanding of inflation

* Li (corresponding author): Department of Finance and Banking, School of Business, The

University of Shanghai for Science and Technology, Shanghai 200093, China, and Askew School of

Public Administration and Policy, The Florida State University, Tallahassee, Florida 32306-2250,

USA. Email: tl14c@my.fsu.edu; litianfeng@gmail.com. Wei: Department of Management/MIS,

College of Business, The University of West Florida, Pensacola, FL 32514, USA. This study is

supported by a grant from University of Shanghai for Science and Technology (No: 1F-13-303-002)

to Tianfeng Li.

bs_bs_banner

Asian Economic Journal 2015, Vol. 29 No. 1, 1–20 1

© 2015 The Authors

Asian Economic Journal © 2015 East Asian Economic Association and Wiley Publishing Pty Ltd

persistence is of great importance in the identification of underlying inflationary

pressures and in setting monetary policy for monetary authorities.

Only a few published studies consider inflation persistence in developing

countries, including China (e.g. Gottschalk, 2003; Lu and Zhang, 2003; Balcilar,

2004; Önder, 2009; Zhang and Clovis, 2009 and 2010; Chiquiar et al., 2010;

Zhang, 2011), with most of the existing literature focusing on industrialized

countries such as the USA, the UK and Canada (e.g. Nelson and Plosser, 1982;

Fuhrer and Moore, 1995; Cogley and Sargent, 2001; Benati and Kapetanios,

2003; Levin and Piger, 2004; Sheedy, 2010). As a large developing and transi-

tional economy, China has experienced low as well as high inflation alternately

since 1978, and at the same time it has undertaken frequent economic reforms

addressing price, state-owned enterprise, and financial and fiscal regimes. Infla-

tion persistence is more likely in China than in industrialized countries because of

the impact of the large-scale economic reforms, as implied by the well-known

‘Lucas critique’ (Lucas, 1976). Hence, it is worth systematically investigating the

sizes and changes in inflation persistence and exploring the underlying causes of

the changes in China.

Even though many studies have been conducted to empirically assess inflation

persistence, the results appear to be quite divergent, especially when the structural

breaks are taken into account in the estimation of persistence. Indeed, the incor-

poration of structural breaks has resulted in lower persistence estimates (e.g.

Taylor, 2000; Levin and Piger, 2004; Batini, 2006). However, disregarding the

breaks could lead to high persistence and induce inaccurate estimation on the long

memory behavior in the conditional volatility (e.g. Diebold, 1986; Lamoureux

and Lastrapes, 1990; Granger and Hyung, 2004).

Because of the importance of structural changes, a growing body of research

has been devoted to developing powerful tests for structural changes in various

modeling contexts. These studies developed in two directions: one is from a priori

break date to an unknown break date, and the other is from a one break model to

a multiple-break model.1Specifically, early work tested structural shifts for a

known single break date (see Quandt, 1958; Chow, 1960). The research pro-

ceeded to the modeling where this break date is unknown. Quandt (1960)

extended Chow’s test by proposing likelihood ratio test statistics for an unknown

change point. This approach is referred to as the supremum (max) test. As similar

cases, the most important contributions are those of Andrews (1993) and

Andrews and Ploberger (1994), who developed the Wald and Lagrange multiplier

test and exponential and average tests, respectively. The development of these

tests for one unknown structural change in the data led to increasing attention

being paid to the case of multiple structural breaks (e.g. Yao, 1988; Yin, 1988;

Yao and Au, 1989; Liu et al., 1997). Bai and Perron (1998) consider the estima-

tion of multiple structural shifts in a linear model estimated by least squares, and

propose a selection procedure based on a sequence of tests to detect multiple

1 See Jouini and Boutahar (2005) and Eksi (2009) for a more detailed discussion.

ASIAN ECONOMIC JOURNAL 2

© 2015 The Authors

Asian Economic Journal © 2015 East Asian Economic Association and Wiley Publishing Pty Ltd