Productivity Breakdown of the Information Technology Industries across Countries Benjamin B. M. Shao Wesley S. Shu Arizona State University San Diego State University Ben.Shao@asu.edu wshu@mail.sdsu.edu Abstract In this paper, we measure productivity growth of the information technology (IT) industries in fourteen OECD countries over the thirteen-year period of 1978 through 1990. The IT industries are the providers of IT capital goods, and this macro-level analysis seeks to find out how efficiently IT capital goods are produced. The basic unit of analysis employed is the Malmquist Total Factor Productivity (TFP) index. The Malmquist TFP index is next decomposed into three constituent elements accounting for different sources of productivity growth: technological progress, efficiency change, and the effects of economies of scale. The approach of measurement is based the concept of distance functions and employs the non-parametric frontier method of data envelopment analysis (DEA). Our results indicate that each country’s IT industry manifests its own particular patterns in various performance measures. Among the fourteen countries examined, ten had witnessed productivity growth in their IT industries. Overall, these IT industries are found more productive than other industries when compared with previous research. Further analyses reveal that most of productivity growth measured is due to technological progress. Efficiency change is found to exert a relatively small positive effect on the productivity growth. Moreover, the change of scale economies unfavorably affects productivity for most countries. Finally, practical implications for formulating IT policy are drawn from our results for further discussions. 1. Introduction The impact of information technology (IT) is a research topic that has received increased attention from researchers during the past two decades. Most of the research has been focusing on the effects of IT adoption on individual, organizational and national performance, often measured in terms of productivity, efficiency, profitability, quality, variety, etc. (e.g., [2, 8] for individual-level studies; [14, 28] for firm-level studies; [10, 11] for country-level studies). In comparison with the abundance of research activities conducted on the values of IT adoption and usage, few efforts have been devoted to examining the accumulation of IT capital, the foundation on which such IT’s contributions are based. It is, therefore, our intention to investigate the economics of IT from a different perspective. Instead of focusing on the benefits of IT usage, we intend to examine the production of IT capital goods in a cross-country context. The significance of IT production by IT industries cannot be downplayed, considering that IT spending over the last three decades has increased more than 40 times and, in the present technology-intensive era, IT is regarded as one of the major drivers responsible for the recent U.S. productivity growth between 1995 and 2000. In the current study, we employ the non-parametric frontier method of data envelopment analysis (DEA) to compute the Malmquist Total Factor Productivity (TFP) indexes for a sample of fourteen OECD countries’ IT industries. The technique used allows us to further decompose the Malmquist TFP index into three components: (a) shifts in production technology, (b) pure changes in technical efficiency, and (c) effects of economies of scale. This paper seeks to explore whether IT capital goods are produced efficiently in these OECD countries and, if so, what are the respective sources for their productivity growth. The decomposition of the Malmquist TFP index is informative: the component of technological progress represents the degree of innovation potential, and the component of pure technical efficiency change reflects the potential of an IT industry to catch up with other leading benchmarks [1]. The rest of the paper proceeds as follows. Section 2 reviews the relevant literature on IT’s productivity gains and explains the connection between the current study and the prior research. Section 3 sets the modeling stage with the specification of our DEA models for calculating the Malmquist TFP index and its three constituent components. In Section 4, the country-level panel data set used for our study is discussed, and the empirical results are presented. Section 5 discusses the policy implications for the development of a country’s IT industry. Finally, Section 6 concludes the paper. 2. IT’s contributions The productivity contributions of IT can be discussed from three different perspectives. The first and the most Proceedings of the 36th Hawaii International Conference on System Sciences - 2003 0-7695-1874-5/03 $17.00 (C) 2003 IEEE 1