Structural Change and Economic Growth. A Longitudinal and Cross-Country Study

Abstract

The main purpose of the present study is to contribute for a deeper understanding of the growth process of the Portuguese economy over the last three decades, by explicitly taking into account the relationship between changes occurring at the industry level of the economy and overall macroeconomic changes. Although a few studies have already addressed the matter for the Portuguese case, a number of important issues relating structural transformation, technology and economic growth remained unexplored, and it is our purpose to fill this gap by considering the neo-Schumpeterian stream of research as the main theoretical frame of analysis. After comprehensively surveying the relevant literature on the field, a preliminary assessment of the relationship between technology, structural change and the macroeconomic performance of the Portuguese economy is undertaken using shift-share analysis. This technique is applied considering total factor productivity growth, and employing different levels of breakdown of economic activity, which include the division of industries according to their skills and innovativeness potential. The impact of Verdoorn effects is also acknowledged. The inclusion of capital in the measurement of productivity growth reveals that the performance of the Portuguese economy was globally mediocre in the period under scrutiny, which was characterised by very slow rates of TFP growth. The results show furthermore that most of the (low) productivity gains came from the shift of labour and capital resources across sectors, rather than from intra-productivity gains. Structural change gains arose, however, in a context of relatively slow change in the broad Portuguese economic structure, which maintained a strong bias towards traditional and low-skilled activities. The latter part of the thesis is dedicated to the investigation of the benefits in terms of productivity growth arising from an increase in the relative importance of technologically dynamic industries. This is done using panel data regression methods and analysing the Portuguese case with reference to a number of other countries that presented similar structural characteristics in the late 1970s, but which have experienced widely different growth trajectories since then. The results provide empirical support to the hypothesis according to which substantial benefits have accrued to countries that successfully changed their structure towards more technologically advanced industries. Moreover, the results lend some support to the view that ICT-related industries are strategic branches of economic activity, but only when producing industries are considered. This accentuates the fact that most spillovers from advanced industries, and particularly ICT producing industries are local and national in character.