Archived — Working Papers
Information identified as archived on the Web is for reference, research or recordkeeping purposes. It has not been altered or updated after the date of archiving. Web pages that are archived on the Web are not subject to the Government of Canada Web Standards. As per the Communications Policy of the Government of Canada, you can request alternate formats on the "Contact Us" page.
No. 20: Information Technology and Labour Productivity Growth: An Empirical Analysis for Canada and the United States
by Surendra Gera, Wulong Gu and Frank C. Lee, Micro-Economic Policy Analysis, Industry Canada, March 1998
Over the past two decades, the use of information technology (IT) has become more intensive in most industrialized economies, as spending on IT-related goods has increased dramatically. All sectors of the economy are experiencing significant changes in the way goods and services are produced and delivered as a result of the increased diffusion and use of information technologies. In particular, many services such as finance, insurance, and real estate; wholesale and retail trade; and communication and business services have emerged as major users of these technologies. Two inter-related forces have contributed to these developments. First, communications and information processing costs have fallen dramatically — they are now 1/10000 of what they were just 20 years ago — and this has spurred and deepened globalization. Second, globalization, in turn, has advanced technological change by intensifying competition and expediting the diffusion of technology through international trade and foreign direct investment (FDI).
In a parallel development, overall productivity growth in OECD economies has slowed significantly since the early 1970s. The decline has been especially noticeable in the service sector, which consumes over 80 percent of IT goods. This has raised numerous questions about the implications of IT investments for productivity growth.
On a theoretical front, new growth theory predicts that physical investments should have a greater impact on productivity growth than traditional growth accounting would suggest, due to positive externalities associated with such activities. In the contributions of Romer (1986), and Grossman and Helpman (1991), these externalities arise because of "knowledge spillovers" — increases in physical investments of profit-seeking firms contribute to the general stock of knowledge upon which subsequent firms can build. In De Long and Summers (1991), investment externalities arise as a result of the "learning by doing effect" — workers and managers learn new skills and more efficient methods of production by using newly installed equipment. These models suggest that the IT sector, which has been one of the most technologically dynamic sectors of the economy over the last 20 years, is likely to have a greater impact on productivity growth than other sectors.
Until recently, however, there has been little empirical evidence that IT capital has contributed to increases in output and productivity growth, and this has led to a heated debate about the "productivity paradox" (see, for example, Brynjolfsson, 1992, Meijl, 1995, and the next section of this paper for a review). In this connection, two recent studies carried out on U.S. data deserve specific mention. Berndt and Morrison (1995) examined the impact of investments in high-tech office and information technology capital on productivity growth across two-digit manufacturing industries from 1968 through 1986. They found that increases in high-tech investments are negatively correlated to multi-factor productivity growth and they tend to be labour-using. However, they did find some evidence that industries with a higher proportion of high-tech capital had higher measures of economic performance. In contrast, Brynjolfsson and Hitt (1995) argued that IT has become a productive investment for many firms. Using data from a large number of firms over the 1988–1992 period, they found that while "firm effects" may account for as much as one-half of the productivity benefits imputed to IT in earlier studies, the elasticity of IT remains positive and statistically significant.
None of these studies took account of domestic and international R&D spillovers from the IT sector when analysing the relationship between IT investment and labour productivity growth. Omitting this variable for a small open economy such as Canada's, which relies a great deal on international trade and FDI, could potentially bias the results. Indeed, Bernstein (1996a, b) confirmed the importance of international spillovers to Canada. His study found that many Canadian industries benefited from R&D carried out in the United States. Some recent studies have also found evidence of domestic and international R&D spillovers from the IT sector (see for example, Bernstein, 1996b; Meijl, 1995, and Sakurai et al., 1996).
In the spirit of these studies, we derive an empirical framework that allows us to estimate the relationship between IT investment and labour productivity growth for Canada and the United States. The framework also takes into account R&D spillovers from IT and non-IT sectors from foreign and domestic sources.
The aim of this paper is twofold:
- to examine the relationship between IT investment and labour productivity growth; and
- to examine the importance for labour productivity growth of domestic and international R&D spillovers embodied in IT goods.
The paper examines these issues using a set of OECD industry databases for Canada and the United States. We find strong support for the proposition that IT investments and international R&D spillovers, particularly those embodied in imports of IT goods, contribute to higher labour productivity growth in Canadian industries. However, the results are generally less robust for the United States.
The paper proceeds with a brief review of earlier literature. In Section 3, we describe the empirical model used in our estimation. Section 4 presents a general overview of the data and general trends. In Section 5, we present regression results for both Canada and the United States. Finally, in the last section, we offer some conclusions and discuss policy issues raised by the analysis.
- Date modified: