Is Export Spillover Localized in China

• DOI: http://doi.org/10.1111/rode.12080
• Author: Min Shao,Ligang Song,Qun Bao
• Published date: 01 May 2014
• Date: 01 May 2014

Is Export Spillover Localized in China

Qun Bao, Min Shao, and Ligang Song*

Abstract

The hypothesis of localized export spillover has been widely supported by the previous studies. Based on

Chinese firm-level data of more than 47,000 firms from 2000 to 2006, this paper investigates whether export

spillover is localized in China. Drawing on the idea of revealed comparative advantage, six export spillover

indicators are constructed to capture both the within- and between- provinces, cities as well as industries

spillover effects. The multinomial logit model estimation results show that there is no sign of localized

spillover effects despite the fact that domestic firms gain from export spillovers due to geographic and

industrial agglomeration effects. This is especially true for those new entrants to export markets. The

finding indicates that there may be export congestions resulting from over-agglomeration of local exporters

due to the rapid export expansion in China.

1. Introduction

One popular viewpoint about the export spillover effect is the “localized spillover.”

The hypothesis states that spillover effect is larger if local concentration of economic

activity is higher. It is Jaffe et al. (1993) who first present an economic analysis that

supports geographic localization of knowledge spillovers. Such a localized knowledge

spillover effect is further confirmed by many others (Rosenthal and Strange, 2003;

Orlando, 2004). The same logic applies in the study of export spillover and technology

spillover. In their seminal paper, Aitken and Harrison (1999) argue that foreign inves-

tors may generate positive technology spillover only for plants nearby. While technol-

ogy spillover occurs, the benefits are likely to be captured first by neighboring

domestic firms, and then gradually spread to more distant firms. The localized export

spillover hypothesis has been widely supported (Aitken et al., 1997; Girma and

Wakelin, 2002; Kneller and Pisu, 2007; Koenig et al., 2010).

The core of the localized spillover hypothesis is that with geographic distance

increasing, learning and imitation cost will become higher, causing a negative rela-

tionship between geographic distance and spillover effects. Such localized spillover

hypothesis seems to be quite reasonable. For example, it may be easier for domestic

firms to conduct face-to-face contact with the multinationals locating nearby, and thus

have better access to export information about the international markets. Domestic

firms near the multinationals also have geographic advantages in learning the multi-

nationals’ superior production or management techniques, which in turn enable them

to compete more successfully on export markets.

* Bao: Nankai University, Tianjin, China. Tel: +86-22-23509982; Fax: +86-22-23503700; E-mail: baoqun@

yeah.net. Shao: Nankai University, Tianjin, China. Song: Crawford School of Public Policy, Australian

National University, Canberra, Australia. The authors acknowledge financial support by the China’s

National Science Fund (71103100; 71203107), the Excellent Scholar Research Program (NCET-11-0248),

and National Social Science Fund Large Grant (12&ZD087), the Key Project of Key Research Base in

Humanities and Social Science of Ministry of Education (10JJD790016; 11JJD790024).

Review of Development Economics, 18(2), 218–230, 2014

DOI:10.1111/rode.12080

© 2014 John Wiley & Sons Ltd