We study the effects of openness, trade orientation, and human capital on total factor productivity for a pooled sample of developed and developing countries. Total factor productivity emerges from a parsimonious specification of the aggregate production function. Potential determinants of total factor productivity include measures of openness, trade orientation, and human capital. Higher openness benefits total factor productivity. Outward-oriented countries experience higher total factor productivity, over and above the positive effect of openness. Human capital generally contributes positively to total factor productivity. In poor countries, however, human capital interacts with openness to achieve a positive effect.
Students of trade theory and policy have since the time of Adam Smith debated whether openness and trade liberalization provide the necessary ingredients for economic growth. Edwards (1993) describes the ebb and flow of this debate during the latter half of the twentieth century. Various protectionist theories captured the major attention of trade policy makers after World War II. During the last two decades, however, a growing body of empirical evidence has legitimized the role of market-oriented reforms and trade liberalization. Now, the International Monetary Fund and the World Bank make market-oriented reforms and trade liberalization a condition for financial aid.
The effect of openness and trade liberalization on economic growth remains a highly contentious issue, however.1 Trade and exchange rate regimes interact with other economic and non-economic factors to affect changes in real per capita income.2 Larger trade implies greater openness that facilitates the economy's adoption of more efficient techniques of production, leading to faster growth of total factor productivity and, hence, real per capita income.3 The expansion of exports relaxes the foreign exchange constraint and allows for larger imports of key inputs in the production process. Finally, improvements in the terms of trade can exogenously increase output.
The empirical tests of the effects of openness and trade orientation on economic growth (e.g., Dollar, 1992, Sachs and Warner, 1995 and Edwards, 1998) typically employ cross-section analysis. Edwards (1993) argues that “A more precise answer to this general question (how openness and trade orientation affect output growth) would require more detailed analysis relying, at least in part, on time series data, …” (p. 1385, parentheses added) and to examining, among other things, “… the robustness of specific results …” (p. 1390). We employ pooled cross-section, time-series data with robustness analysis to improve the likelihood of uncovering important links between openness and trade orientation and total factor productivity.4
The reemergence of the importance of growth theory has also refocused much of the debate toward how public policy can affect economic growth. The standard neo-classical growth models [i.e., the Solow (1956) model, Ramsey's (1928) optimal growth model, or Samuelson's (1958) overlapping generations model, and their descendants] have been challenged by the literature on endogenous growth. The neo-classical paradigm considers technological change as an exogenous process whereas the endogenous growth literature makes this process endogenous, looking for possible driving forces.
Mankiw (1995) recently discusses three practical empirical problems associated with the neo-classical approach along with a proposed solution — the model cannot explain the observed differences in real income per capita across countries, the observed rate of convergence across countries, or the observed differences in real rates of return across countries. He modifies one parameter, the returns to capital, by defining capital to include both physical and human capital, and argues that this one change goes a long way toward allowing the neo-classical model to explain these three empirical regularities
We study the effects of openness, trade orientation, and human capital on total factor productivity for a pooled cross-section, time-series sample of developed and developing countries. We estimate total factor productivity from a parsimonious specification of the aggregate production function involving output per worker, capital per worker, and the labor force, both with and without the stock of human capital. Then, we search for possible determinants of total factor productivity, with special emphasis on trade variables.
We find that opening the economy to trade generally benefits total factor productivity. Opening the economy to trade means increasing exports to GDP, improving the terms of trade, and lowering the real value of the domestic currency. Moreover, the stock of human capital contributes positively to total factor productivity in many, but not all, specifications. That is, human capital has a negative effect on total factor productivity in high-income countries and a positive effect in middle-income countries. The effect of human capital on total factor productivity in low-income countries moves from negative to positive as the country moves from a low to a higher level of openness. As a result, counterproductive effects emerge from human capital investment within the low-income countries unless the country achieves a certain level of openness.
