At the heart of most explanations for the non-convergence in international incomes across countries is the concept of technological change. Improvements in technology, whether through invention of new techniques or through the adoption of better technologies that have been invented elsewhere, are central to the process of growth and development. Any barriers that prevent such improvements are then the focus of theories that attempt to explain why poor countries have failed to catch-up with their rich counterparts, or why inequalities can exist within a country or region. A large body of literature therefore focuses on barriers to technology adoption. See, for example, Parente and Prescott (1994), Greenwood and Yorokoglu (1997) and Leung and Tse (2001), in which the barriers take the form of costs incurred in the adoption of technologies. In some cases, this cost is of an implicit, “learning-bydoing” type (as in Khan and Ravikumar, 2002) and in others is of a pecuniary or contractual type (as in Acemoglu et al. 2007). At the empirical level there is evidence of delays in adoption and diffusion of new technologies; Comin and Hobijn (2010), for instance, suggest that there is an average lag of 45 years before a newly invented technology is fully adopted across countries. In particular, the pattern of technology diffusion involves invention and early adoption in advanced economies, followed by “trickle-down” diffusion in economically lagging, developing countries (see Comin and Hobijn 2004). Empirical studies also suggest different rates of technology adoption as a source of productivity differences and inequalities within countries (see Chanda and Dalgaard, 2008). A new and growing body of literature, not entirely unrelated to the above-mentioned adoption-cost related studies, stresses the notion of “appropriate technology” as an underlying rationale for the slow diffusion of technologies, and the resultant productivity differences across countries. The aim of this study is to examine the implications of this idea, which suggests that a technology may not be “appropriate” in a country if the conditions that are needed for the realization of its potential level of productivity are not met. In Basu and Weil’s (1998) model, for example, the barrier to technology adoption arises due to the localized nature of learning-by-doing. Specifically, a follower country can adopt a leading country’s technology only if the capital intensity of the new technology falls in a range that is close to the capital intensity of existing technologies in the follower country. In Acemoglu and Zilibotti (2001) the reason for productivity differences occurring when the same technology is used in different locations (e.g. in developed vs developing economies) is attributed to skill shortages in the developing economies. This suggests a ‘skill bias’ (which may be a low-skill or highskill bias) in the choice of technology, which may explain the slow diffusion of the capital and skill intensive technologies in the developing world (see Caselli and Coleman, 2006).