The objective of this paper is to assess the impact of information and communications technology (ict) on economic performance at the sectoral level in Belgium over the period 1990-2000. The growth accounting approach used in the framework of the neoclassical growth theory for the study of the sources of economic growth will be adopted here in order to quantify the impact of ict use on output and labour productivity growth. Since annual data on ict capital stock are not readily available, we use data from a number of sources to construct this indicator at the sector level for Belgium over the period 1990-2000. Our findings should indicate (i) to which extent ict contributed to output and labour growth at the sectoral level in Belgium in the 1990s and (ii) whether industries making intensive use of ict performed better then non-intensive ict ones over the same period.
Some remarks are made below about (i) the sectoral orientation of our study, (ii) the focus being on the use of ict rather than its production and (iii) the implications of the utilization of the growth accounting approach as far as the implications of the New Economy on growth dynamics are concerned. We believe that these remarks will clarify the aim of our study and justify, therefore, the methods and data used in this paper.
First, it is commonly believed that a country need not produce ict goods in order to benefit from the positive effects of the new technologies on its growth. These goods can simply be imported from abroad and used by firms in the importing country. So, the limited share in gdp of sectors producing ict goods should not be an obstacle for the new technologies to exert a positive impact on economic performance, nor for being able to quantify this impact.
This is of course not to deny that the existence of an ict-producing sector can exert positive effects on economic performance. Besides the positive impact of supplier-user relationships on the efficiency of the innovation and production process of ict goods and on their subsequent diffusion rate, high rates of tfp growth in ict-producing industries contribute directly to the growth at the aggregate level. This last point however, concerns less a small open economy like Belgium when it comes to assess the impact of ict on its economy. Moreover, it is of rather limited interest when it comes to assess the impact of ict utilization on economic performance.
The approach adopted in our paper to quantify the impact of ict use on economic performance at the sectoral level is similar to those used in other studies: first, we quantify the contribution of ict capital stock to output and labour productivity growth and compare it with the contribution of other factors of production. Second, we identify sectors that are intensive users of ict assets and check whether the contribution of ict to economic performance is more important in those sectors than in sectors that are non-intensive users of ict.
In some studies, the impact of ict use on growth is assessed on the basis of the contribution of ict-using sectors to aggregate growth. However, the link between the extent to which ict contributes to growth within an industry – which is the aim of our analysis – and the contribution of ict-intensive industries to aggregate growth seems best to be an approximate one. Indeed, besides ict use, this contribution depends on many other factors including the evolution of the share of ict-intensive industries in gdp. Besides, it is extremely difficult to disentangle the role of new technologies from other factors in this contribution to aggregate growth. Therefore, we decided not to use this aforementioned method in our paper.
Secondly, the growth accounting approach used in this paper in order to decompose output and labour productivity growth into their components has some limitations. Its successful implementation requires the imposition of assumptions of constant returns to scale and competitive output and input markets. More fundamentally, externalities due to the utilization of a production factor can not be accounted for by this method. This is a major caveat as far as measuring the impact of new technologies on growth is concerned. Indeed, in the framework of the growth accounting approach, contribution of ict use to growth is explained solely by rapid ict capital accumulation, itself being the result of profit-maximizing response of producers to rapidly falling quality-adjusted prices of ict equipment. These falling prices reflect better performance and new characteristics of ict goods and are the result of rapid technological progress in ict-producing sectors, and especially in the computer equipment industry. They can be seen as “pecuniary externalities” and integrated easily in the traditional neoclassical framework if hedonic prices for ict capital goods are available.
This framework, however, provides no special role for ict capital in the growth process: its contribution to economic growth is simply function of its rate of accumulation like any other input. Consequently, it might explain the strong labour productivity growth observed in industrialized economies in the 1990s but hardly Total Factor Productivity (tfp) growth. We can expect, however, that the resurgence of a new techno-economic paradigm such as the New Economy should influence economic growth in more fundamental ways through such channels as ict-related production spillovers or network effects. These “non-pecuniary externalities” would exert an impact on growth through their impact on tfp growth. Such non-traditional effects can not be accounted for within the framework of the growth decomposition method used in paper. Hence, an assumption of non-existence of spillovers or externalities linked to ict-use is made implicitly each and every time this method is used in order to assess the contribution of ict to economic growth.
Thirdly, we use industry-level data in our study in order to measure the contribution of ict capital to economic performance. This down-top approach has a number of advantages when it comes carry out a study aimed at measuring the impact of new technologies on growth. Indeed, since both output/productivity growth and ict intensity vary to a large extent among industries, a possible link between these variables might be lost if data at the aggregate level are used in the analysis. According to Stiroh (2002), this is one reason why “earlier research on the economic impact of ict at disaggregated levels typically found a substantial impact, why earlier aggregate studies did not (p. 6)”. The loss of heterogeneity when data at higher levels of aggregation are used in the analysis– implying that ict-intensive industries are combined with less intensive ones – might obscure the productive impact of ict. These considerations motivated the use of data at the sector-level in our analysis of the impact of ict on output and productivity growth in Belgium.
The rest of the paper is organized as follows. Section 2 introduces the methodology and data used in our study in order to measure the contribution of ict to output and labour productivity growth. Section 3 presents growth accounting estimates at the sector level while section 4 compares findings obtained for ict-intensive industries with those obtained for non-intensive ones. Section 5 concludes the paper.