Performance Of Technology Diffusion:a Cross-country Comparisons Study

Many studies have deemed variations in total factor productivity{TFP) as the reason why countries differ in GDP per capita. In fact, TFP can be decomposed into two components, one relates to technology while the other does not. Compared with less developed countries (LDCs), the level of technology used in developed country (DCs) is usually higher; in other words, the technologies employed in DCs are closer to the frontier. Using technical indicators or technical achievement index, the extant literature suggests technology convergence appears in DCs and LDCs. Taking EATs and BRICs as examples of fast-growing economies, this dissertation tries to evaluate their technological improvement in terms of their performance in technology diffusion. And the main contents of this study are divided into three parts.Part Ⅰ is devoted to comparing various quantitative indexes used in technology diffusion literature, their specifications, and applications. From the perspecitive of extended marginal performance, a direct measure is chosen to measure technology diffusion. In addition, adoption lags are used to assess technology diffusion. Such an angle enables us to do empirical analysis because data is readily available; moreover, it is easier to appraise from the study how per capita income has affected technology adoption costs of the enterprises.Part Ⅱ introduces a one-sector neoclassical growth model. In this model, technological advances are captured by a monopolistically competitive intermediate goods market. Intermediate goods producers use the new technology and bear its cost, and gain monopoly profits. Decisions whether to adopt the technology or not impinge on the equilibrium path of productivity; also, technology that incurs higher costs would reduce the output of the final product, hence hinders technology absorptive capacity. The higher the cost of the new technology, the longer it takes for the enterprise to adopt it. The defining characteristic of this model is:technology adoption cost serves a causal link between productivity and technology adoption lag. If technology adoption lag can capture the performance of technology diffusion, then we can construct an econometric model to identify it. In Part Ⅲ, using a panel data set, a nonlinear econometric model is established according to the structural model presented in Part II. Nonlinear Least Squares (NLS) estimation is employed to do this cross-country analysis on technology diffusion. The technologies of interest are three General Purpose Technologies (GPTs):electric power, personal computers, and Internet. The sample countries are APEC (including EATs and BRICs) members.How fast a country adopts and a GPT diffuses is closely related to the country’s GDP per capita. After analyzing the introduction and absorption of a specific technology in a country, we can see that on the firm level obstacles encountered in absorbing this technology would increase the cost of adopting another technology in the future. In this study, it is proved that EATs countries improved their technological absorption capacity through reforms; the resultant change is that, while their adoption of new technologies accelerated, they experienced rapid economic growth as well. It is also found that if the time lag adopting an early technology is large, technology invented at a later time could be quickly adopted. This can be explained by, to some extent, globalization and information technology, which have hastened the technical information exchanges among countries, facilitated procuring advanced technology from DCs, therefore contribute to a reduced cross-country technology gap.