Chapter one presents a critical and detailed review of the stochastic frontier methodology from a macro-data perspective.  The advantages over the standard growth accounting approach are emphasised, and the main features of the translog production function, used throughout the thesis, are discussed. Chapter two uses the stochastic frontier approach to estimate different specifications of the production function, technological catch-up (efficiency improvements) and technological change (shifts in the production frontier) for 57 developing countries over the period 1960-1990.  It is well known that alternative specifications of the production function lead to ambiguous empirical evidence for competing theories of economic growth (Durlauf and Quah 1999).  Therefore, tests are performed to find the specification in line with the data under analysis.  Then the important issue of the role of human capital in the process of economic growth is also investigated, since it is not yet unambiguously determined (Islam 1995, p.1154).  Chapter three analyses the results based on Model 4* (Chapter 2) in more detail to provide a consistent decomposition of output growth.  The evolution of the entire distribution of the growth and productivity sources is analysed and a formal test for assessing the importance of growth factors is performed. With respect to regression analysis, this approach is likely to be more informative (Quah, 1996a,b, 1997).  The base of both the test and the visual analysis is the non-parametric kernel density estimator. The findings in the previous chapters motivate Chapter 4 of the thesis, which further explores the relative importance of FDI, imports of capital goods and human capital accumulation in the development process.