Do all countries follow the same growth process?

Abstract

We estimate a finite mixture model in which countries are sorted into groups based on the similarity of the conditional distributions of their growth rates. We strongly reject the hypothesis that all countries follow a common growth process in favor of a model in which there are two classes of countries, each with its own distinct growth process. Group membership does not conform to the usual categories used to control for parameter heterogeneity such as region or income. However, we find strong evidence that one country characteristic that helps to sort countries into different regimes is the quality of institutions, specifically, the degree of law and order. Once institutional features of the economy are controlled for, we find no evidence that geographic characteristics play a role in determining the country groupings.JEL Codes: O11, O17

Human Capital, Social Capabilities and Economic Growth

“…The effect of human capital on growth is conditioned by living conditions: when living conditions are favorable, the contribution of human capital to growth is amplified. In Owen et al (2009) analyses of a sample of developed and developing countries, it was shown that countries do not necessarily follow similar growth paths; rather, countries can be sorted into categories, each with its own unique growth processes. Two of the largest sources of heterogeneity in this respect are economic complexity and the quality of institutions.…”

Caught in a Poverty Trap? Testing for Single vs. Multiple Equilibrium Models of Growth

“…These papers have convincingly demonstrated that aggregate production functions differ significantly across groups, and that per capita income and literacy in 1960 can predict which group a country will belong to over the next 40 years. However, as both Durlauf and Johnson (1995) and Owen et al (2009) admit, the evidence from this line of inquiry is consistent with both models of multiple steady states and models of unified growth and thus cannot constitute a test of the poverty traps hypothesis. 5,6 Poverty trap models of the second type admit per capita income as a singlevalued function of country characteristics, but explicitly hypothesize bidirectional causality.…”

“…A poverty trap of this nature would thus imply that the relationship between current observable parameters and current income per capita is a function of initial conditions. A group of papers has used regression tree analysis or threshold estimation to sort countries into groups such that countries within the same group exhibit a common aggregate production function (e.g., Durlauf and Johnson 1995;Hansen 2000;Durlauf et al 2001;Papageorgiou 2002;Canova 2004;Owen et al 2009;Sirimaneetham and Temple 2009). These papers have convincingly demonstrated that aggregate production functions differ significantly across groups, and that per capita income and literacy in 1960 can predict which group a country will belong to over the next 40 years.…”

Education, Inequality, and Development in a Dual Economy

“…where the first term of the numerator of the RHS equals w ht + ϕ w mt > 0 from (12) and (13).Since the LHS decreases with ϕ and the RHS and the denominator of the RHS increase with ϕ, the numerator of the RHS increases with B t . Thus, the numerator of the RHS of (37) is positive at B t = 0 and is increasing in B t .…”