Spatial Spillovers from Foreign Direct Investment: Evidence from the Yangtze River Delta in China

• Date: 01 March 2015
• Author: Yoshihiro Hashiguchi,Kiyoyasu Tanaka
• Published date: 01 March 2015
• DOI: http://doi.org/10.1111/cwe.12106

40 China & World Economy / 40–60, Vol. 23, No. 2, 2015

©2015 Institute of World Economics and Politics, Chinese Academy of Social Sciences

Spatial Spillovers from Foreign Direct Investment:

Evidence from the Yangtze River Delta in China

Kiyoyasu Tanaka, Yoshihiro Hashiguchi*

Abstract

This paper examines the spatial externality from foreign direct investment on domestic

firms. Using Chinese firm-level data for 2004, and after accounting for an endogeneity

problem, we find that foreign firms generate a significantly positive spillover effect on the

regional productivity of domestic firms in similar counties and industries. Estimating a

spatial-autoregressive model, we further show that such local spillovers could transmit to

domestic firms in other counties and industries through interactions among domestic firms.

However, these spatial multiplier effects decline with distance, thereby reducing the foreign

direct investment spillover effects for domestic firms in distant locations.

Key words: China, foreign direct investment, productivity, spatial spillovers

JEL codes: C21, F21, F23, R12, R58

I. Introduction

The past decades have seen considerable interest among academics and policy-makers

regarding inward foreign direct investment (FDI) and potential spillovers from foreign to

domestic firms. A number of developing countries have attempted to attract foreign investors

by giving them preferential investment incentives. The expectation is that the local firms

can absorb advanced technology brought in by multinational firms, which provides a

justification for policies to promote FDI. Motivated by these reasons, a large number of

studies have estimated FDI spillovers. However, the evidence so far has been quite mixed,

and the existing evidence is inconclusive as to whether governments should promote FDI

for local industry development (G

ö

rg and Greenaway, 2004).1

To reconcile the mixed evidence, recent studies have examined the geographic scope

*Kiyoyasu Tanaka (corresponding author), Research Fellow, Inter-Disciplinary Studies Center, Institute

of Developing Economies, Chiba, Japan. Email: kiyoyasu_tanaka@ide.go.jp; Yoshihiro Hashiguchi,

Research Fellow, Development Studies Center, Institute of Developing Economies, Chiba, Japan. Email:

yoshihiro_hashiguchi@ide.go.jp.

1Havranek and Irsova (2011) conducted a meta-analysis to show that the average spillover to suppliers

is quantitatively large, which favors vertical FDI spillovers.

41

Spatial Spillovers from Foreign Direct Investment

©2015 Institute of World Economics and Politics, Chinese Academy of Social Sciences

of FDI spillovers because knowledge transfers tend to be localized over space (Jaffe et al.,

1993; Audretsch and Feldman, 2004). Specifically, prior studies such as Barrios et al. (2006),

Girma and Wakelin (2007), Halpern and Murak

ö

zy (2007) and Jordaan (2009) examine whether

local firms benefit more from foreign firms in nearby locations than those in distant locations.

These studies suggest that spillover effects are regional in space and are stronger for the

local firms that are closer to foreign firms.2

We extend the literature by investigating whether foreign firms generate additional

positive spillovers over space through interactions of local firms. If domestic firms in one

region benefit directly from foreign firms in the same region, these domestic firms could

produce additional knowledge transfers to other domestic firms in neighboring regions.

Indeed, it has long been argued that firms and workers in the agglomerated area benefit

from productivity advantages generated by agglomeration economies through more efficient

sharing of local suppliers, better matching between employers and workers, and technology

or knowledge spillovers among firms and workers (Duranton and Puga, 2004). Through

spatial interactions between domestic firms in proximate regions, original spillovers from

FDI in one region may diffuse across other domestic firms in distant locations.

Meanwhile, the extent of such spatial spillovers depends on two forces. The distance

would gradually dissipate knowledge transfers and communication among local firms

because the quality of knowledge flow decays over distance (Keller, 2002). By contrast, a

spatial concentration of FDI activity could strengthen the density of knowledge flows,

which, in turn, mitigates the decay process of spatial spillovers. The net spatial externality

from FDI will be determined by the spatial distance of knowledge flows and the magnitude

of FDI activity. While the above related studies find a geographic localization of FDI

spillovers, our hypothesis highlights that local spillovers could magnify over space through

interactions between local firms. To the best of our knowledge, such a hypothesis has not

been subject to formal analysis.

We seek to quantify spatial spillovers from FDI using a firm-level dataset on Chinese

manufacturing firms in the Yangtze River Delta area: Shanghai, Jiangsu, Zhejiang and Anhui.

As is well known, coastal regions in China have received a substantial amount of FDI

inflows and achieved higher economic growth compared with inland regions (Xu and

Sheng, 2012). Foreign investors have been attracted not only for the regions’ natural

advantages, such as access to marine ports, but for fiscal incentives provided by local

governments. Nevertheless, the geographic concentration of FDI varies significantly even

2Another branch of related studies, including Koenig et al. (2010) and Bao et al. (2014), examines

whether spillovers from exporters to non-exporters are localized. The former study finds evidence of

localized export spillovers in France whereas the latter shows non-localized spillovers in China.