Author Aubourg, Rene Wiener
Title Two essays on the relationship between economic growth and carbon dioxide emissions using multivariate contingency approach and panel cointegration analysis
book jacket
Descript 94 p
Note Source: Dissertation Abstracts International, Volume: 64-11, Section: B, page: 5423
Chairs: David Good; Kerry Krutilla
Thesis (Ph.D.)--Indiana University, 2003
The objective of this two-essay dissertation is to investigate whether an environmental Kuznets curve (EKC) exists for CO2 emissions in Latin America and the Caribbean. In Essay 1, we use a multivariate contingency framework that allows for some degree of EKC homogeneity conditional upon the quality of a country's political institutions and the severity of its external debt. To test this hypothesis, we extend the basic reduced form EKC equation to incorporate specific interaction effects between political institutions, the external debt and the income level so that the turning point is no longer a constant but a function of several variables
We find strong support for the presence of different EKCs conditional upon political institutions and the debt burden with most turning points falling within the sample range. Countries with little democracy have higher turning points than highly democratic ones and are more environmentally harmful. More debt increases CO2 emissions and delays the EKC turning point. Furthermore, economic growth tends to be environmentally benign at high democracy levels but harmful at low democracy levels. We discuss the significance of these results in the context of policy complementarity and debt reduction programs
In Essay 2, we test the validity of the EKC hypothesis for CO2 emissions in the context of non-stationary variables using the same data set and a variant of the multivariate contingency model used in Essay 1. After applying standard panel unit roots and cointegration techniques, we find strong evidence of non-stationarity among variables as well as cointegration on the basis of the EKC relationship with all turning points falling outside sample range. Interestingly enough, there is conflicting evidence of cointegration using the basic EKC framework with only GDP and GDP squared as regressors. We discuss the policy implications of our findings in regard to three issues: the cost associated with ignoring stationarity, policy complementarity, and possible misspecification of the basic EKC framework
School code: 0093
DDC
Host Item Dissertation Abstracts International 64-11B
Subject Environmental Sciences
Economics, General
Statistics
0768
0501
0463
Alt Author Indiana University