Institutions, Entrepreneurship, and Regional Differences in Economic Growth1

Abstract

(ProQuest: ... denotes formula omitted.)IntroductionThe question of why some areas are rich and some are poor has been at the center of economics since Adam Smith ([1776] 1998) first published his Inquiry into the Nature and Causes of the Wealth of Nations. In his analysis, Smith focused the division of labor and how the division of labor was limited by the size (or 'extent' as he termed it) of the market. Larger markets lead to an increase in the division of labor and thus higher productivity. Higher productivity, in turn, leads to economic progress directly by increasing wages and indirectly through freeing up scarce resources for other uses.Ricardo (1817), however, focused attention back on the role that inputs such as land, labor, and capital played in economic growth. The creation of macroeconomic statistics in the early twentieth century led economists to focus on aggregate theories of growth that could explain this newly developed macroeconomic data. Solow (1956) developed a simple growth model where economic output was simply a mathematical function of capital and labor inputs [Y =/(K,L)] based on neoclassical theory that, when tested empirically, fit the available U.S. data quite well. The Solow model was the dominant theory of economic growth from the time of its creation until the 1980s and is still heavily used in many graduate macroeconomics classes. While this model has been augmented to sometimes include measures of technology and human capital quality, it fundamentally ignores the institutional arguments made by Adam Smith. In the Solow growth model, these complex institutional structures are simply represented by the functional form of the model, f(*).During the 1980s new data sets were created that contained macroeconomic data on a large number of countries over an extended period of time (see, for example, Summers and Heston (1991)). The creation of these data sets allowed economists for the first time to test whether per-capita incomes across countries were converging to equality-a key prediction of the neoclassical growth model (Romer, 1994). Subsequent research on the question of convergence has shown that there is no clear tendency for poor regions to grow faster than wealthier regions (Romer, 1994) although some research does show that regions within a country do converge, albeit slowly (Barro and Sala-i-Martin, 1992; Holtz-Eakin, 1993). At best, convergence is a slow and discontinuous process (Martin and Sunley, 1998). The finding that convergence sometimes happens slowly within a country (or a set of similar countries) has led to the idea of conditional convergence, where convergence happens conditional on regions having similar properties.The failure to find convergence in cross-country regressions was problematic since the Solow model was the model of economic growth at the time and had a strong influence on public policy.9 It could not, however, explain key features of the real world, such as persistent differences in income levels across countries. Neoclassical growth theory could also not explain the relationship between entrepreneurship and economic growth since the level of innovation was determined exogenous to the system. Even models like Barro and Sala-i-Martin's (1992) that relax the neoclassical assumption of uniform technology across space have no explanation for why technological innovation might vary from place to place. Out of this disenchantment came endogenous growth theory, which relaxed the neoclassical assumption of exogenous technological change and the non-excludability of technology.Unfortunately, endogenous growth theory cannot explain divergent levels of income across countries or the rapid development of countries like South Korea (Parente, 2001). From the standpoint of public policy, the failure of endogenous growth models to provide an explanation for varying levels of economic development is troubling because the sources of growth in endogenous growth models (such as the percentage of GDP spent on research and development) may not be the route to development. …