International R&D Spillovers and Business Service Innovation

• Date: 01 December 2016
• Published date: 01 December 2016
• Author: Neil Foster‐McGregor,Robert Stehrer,Johannes Pöschl
• DOI: http://doi.org/10.1111/twec.12379

International R&D Spillovers and Business

Service Innovation

Johannes P€

oschl

1

, Neil Foster-McGregor

2

and Robert Stehrer

1

1

wiiw, Vienna, Austria and

2

UNU-MERIT, Maastricht, The Netherlands

1. INTRODUCTION

SERVICES are the largest sector in most developed economies and provide a major share

of intermediate inputs into the manufacturing sector. They are often relatively R&D

intensive and are becoming increasingly traded across international borders. According to the

WTO, trade in services accounted for around 20 per cent of world trade in goods and services

in 2007, a share that increases to around 50 per cent if transactions are measured in terms of

direct and indirect value-added content (Francois and Hoekman, 2010). Services also provide

more than 30 per cent of the intermediate inputs of the manufacturing sector in the EU and

occupy a central position with respect to innovation. In the US, the business services sector

alone is twice the size of the manufacturing sector and accounts for 25 per cent of total US

employment (Jensen, 2011).

While the importance of services for the manufacturing sector is widely acknowledged,

studies considering trade-related spillovers have to date concentrated on trade in goods, with

the services sector largely neglected as a source of spillovers. The main aim of this paper was

to quantify the productivity changes in the manufacturing sector resulting from innovation in

research and knowledge-intensive, high-technology service industries. The focus of this study

lies on two R&D-intensive service industries that include telecommunications and computer

activities. The annual R&D growth rates of these two sectors for the countries and time per-

iod covered in our analysis were significantly higher than the rates recorded in the manufac-

turing sectors, with these two industries having tremendously changed the way business is

done in recent decades. Computers have become ever faster, enabling the use of powerful

software that supports further production automatisation as well as administrative processes

(e.g. Enterprise Resource Planning Software). Another major impact of these service indus-

tries in the last two decades has been on communication, both within the firm and between

the firm and the outside world. The Internet as a new and versatile form of communication

and mobile telephony are two media, without which current production techniques are almost

inconceivable. These advances have lowered trade transaction costs and facilitated a further

unbundling of production activities, leading to what Grossman and Rossi-Hansberg (2008)

refer to as trade in tasks. It is questionable whether the emergence of global value chains

could have happened to the observed extent without software enabling firms to communicate

and transfer information without delay. Moreover, machine-to-machine communication with

suppliers and customers has greatly improved just-in-time resource planning. This creation of

The authors are grateful to Alejandro Cunat, Klaus Gugler, Richard Kneller, Bart Los and an anony-

mous reviewer for comments. Furthermore, they want to thank the whole WIOD team for the great

collaboration in the construction of the database. The paper was produced in the framework of WIOD

(http://www.wiod.org/), an international economic research project focusing on the analysis of interna-

tional input–output linkages. The project is funded by the EU Seventh Framework Programme for socio-

economic sciences and humanities.

©2016 John Wiley & Sons Ltd 2025

The World Economy (2016)

doi: 10.1111/twec.12379

The World Economy

intelligent networks along the entire value chain is often referred to as the Fourth Industrial

Revolution.

Based upon these observations, we hypothesise that the productivity of a manufacturing

industry depends not only upon its own R&D stock but also on technological improvements

of these service industries (as well as other manufacturing industries). These improvements

are available to each industry in the form of intermediate products. Our results support recent

evidence of positive productivity effects from international manufacturing spillovers. The

results also highlight the positive and significant spillover effects from knowledge-intensive,

high-technology service industries. Relative to the productivity effect of the industry’s own

R&D, the size of the effects is about one-third for the high-technology service industries and

two-thirds for the foreign own-industry spillovers.

The remainder of the paper is set out as follows: Section 2 gives a brief overview of the

literature in this field; Section 3 describes the related empirical specification; Section 4 pro-

vides information on the data used in the analysis; Section 5 presents the main results; Sec-

tion 6 discusses the results from a model extension and presents a number of robustness tests;

and Section 7 concludes.

2. LITERATURE OVERVIEW

Since the 1970s, a large empirical literature has developed analysing the extent of technol-

ogy spillovers across firms, industries and countries. Early studies at the industry level such

as Terleckyj (1974), Griliches (1979) and Scherer (1982), which tended to use information

from input–output tables to measure the extent of linkages between industries, generally came

to the conclusion that an industry’s technology generation has a significant impact upon the

productivity of other industries.

Following these earlier studies at the industry level and the development of open economy

endogenous growth theories (e.g. Grossman and Helpman, 1991), a strand of literature looke d

to estimate the importance of technology spillovers across countries, beginning with the semi-

nal study of Coe and Helpman (1995). Coe and Helpman (1995) in their study examine the

impact of international R&D spillovers for 22 OECD countries. They construct a measure of

the stock of foreign knowledge that is available to each importing country by weighting the

R&D stocks of its trade partners by import shares. A measure of multifactor productivity

(MFP) is then regressed on both the domestic and foreign R&D stocks, with the results sug-

gesting that both are important sources of productivity growth.

Due to data limitations, few existing studies to date have been able to analyse international

spillovers at the industry level adequately, exceptions include Keller (2002), Schiff and Wang

(2006) and Wang (2007). Keller (2002) was the first to incorporate the analysis of domestic

spillovers between manufacturing industries into an international setting. He constructs two

international spillover components. The first measures the extent of spillovers from the same

industry in other countries and is constructed using trade in advanced intermediate goods,

while the second captures spillovers from other foreign industries for which he additionally

uses information from the import input–output table. Keller’s findings suggest that the produc-

tivity effects from international and domestic spillovers are substantial and as large as the

effects of the industries’ own R&D efforts. Schiff and Wang (2006) consider the impact of

North–South R&D spillovers at the industry level and find that spillovers are significant, with

North–North spillovers found to be stronger than South–South spillovers. Wang (2007) shows

©2016 John Wiley & Sons Ltd

2026 J. P€

OSCHL, N. FOSTER-MCGREGOR AND R. STEHRER