Performance Of Technology Diffusion: A Cross-country Comparisons Study

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