| This dissertation offers a more rigorous empirical examination of competing endogenous growth models, and in doing so moves away from the convergence literature and tests support for endogenous growth directly using a single country framework - in this case Canada. This approach avoids the lofty assumption that an identical production function exists across nations, or the mixing together of economies at vastly different stages of development.;An explicit production theoretic framework that captures some of the more common drivers found in the economic growth literature - private capital, public capital, human capital, and research and development capital - is presented. The results suggest that the use of a broad concept of capital is important in overcoming diminishing returns. In particular, Canada's long-run production technology appears best described as exhibiting non-diminishing returns to reproducible inputs, a sufficient condition for endogenous growth.;Due to the non-stationarity of most macroeconomic time series, the vast majority of previous studies have used a first differenced specification, but at the detriment of losing important long-run information embodied in the levels data. More importantly, estimating in first differences is inappropriate if the model has an error-correction representation, and could affect inferences regarding the long-run causality between output and its determinants. Here, I use cointegration and systems-based approaches in addition to traditional first differences. The results suggest that the cointegration specification is appropriate.;Close attention is paid to data quality and model specification, two major shortcomings in much of the existing literature. The database used here incorporates important methodological changes to Canada's System of National Accounts undertaken in 1997 to align with international standards. The five human capital measures, constructed from Census of Canada benchmarks and enrollment rates, represent an improved and comprehensive source of information on Canada's stock of human capital, and allow for a detailed examination of how different levels of educational attainment interact with growth. Most of the series extend back 76 years, providing an examination of the interaction of investment and growth over a much longer time span than has usually been attempted. |
