Faculté des sciences économiques et sociales

Essays on investments and environment : a spatial econometrics perspective

Monteiro, José-Antonio ; Grether, Jean-Marie (Dir.)

Thèse de doctorat : Université de Neuchâtel, 2010 ; 2201.

This PhD dissertation investigates the links between foreign direct investment (FDI), pollution and environmental policies in an interdependent world. To tackle the issue of spatial dependence, I propose to apply new spatial estimators. The thesis consists of four papers. The first chapter, entitled Spatial Dynamic Panel and System GMM: a Monte-Carlo Investigation, investigates the... Plus

Ajouter à la liste personnelle
    Summary
    This PhD dissertation investigates the links between foreign direct investment (FDI), pollution and environmental policies in an interdependent world. To tackle the issue of spatial dependence, I propose to apply new spatial estimators.
    The thesis consists of four papers. The first chapter, entitled Spatial Dynamic Panel and System GMM: a Monte-Carlo Investigation, investigates the finite sample properties of estimators for spatial dynamic panel models in the presence of several endogenous variables. I show that in order to account for the endogeneity of several covariates and get unbiased estimates, you should apply the system generalized moments method (GMM) estimator instead of traditional spatial dynamic estimators. From a practical point of view, system GMM is computationally less demanding and easier to adapt to unbalanced panel data.
    In the second paper, Pollution Havens: a Spatial Vector Autoregression Approach (VAR), I investigate the mutual relationship between FDI, pollution and trade openness for 141 countries during the nineties. After applying second generation panel unit root tests, I estimate a trivariate spatial VAR model where a shock occurring in a given country can potentially affect the economic conditions in the neighboring countries. Based on generalized empirical likelihood estimations, the results highlight a reverse causality between FDI and pollution emission, suggesting the existence of a potential endogenous pollution haven effect to low income countries in particular. In addition, spatial spillovers play a key role in the linkage between FDI, trade openness and SO2 emission and highlight the strategic nature of these variables. Multilateral and regional cooperation between developed, emerging and developing countries should be extended to capitalize on these spillovers, address potential issues and seize new opportunities.
    In line with the previous study, in the third paper, entitled Complex FDI and Environmental Regulation: the Role of Spatial Dependence, I investigate if differences in environmental regulations influenced OECD’s FDI allocation during 1981-2000 taking into account “third-country” effects in a multi-country setting. The findings, based on system GMM estimates of a spatial dynamic gravity model, confirm the existence of a negative relationship between FDI and environmental stringency, once I correct for endogeneity and spatial dependence. In addition, the evidence of positive “third-country” effects for FDI suggests the prevalence of highly complex vertically integrated FDI from OECD countries to developing economies. Multinationals tend to allocate each part of the production process in different countries to ensure cost minimization. Emerging and developing countries, which might be tempted to use environmental regulation as an instrument to attract FDI, will only attract the most polluting part of the production process. This is not the best way to ensure sustainable economic growth, at least on the long run.
    In the last chapter called Unequal Diffusion of Eco-labels: a Spatial Econometric Approach, I analyze the decision to introduce an eco-labelling scheme through a heteroskedastic Bayesian spatial probit. This framework allows the government’s decision to introduce an eco-label program to be influenced by the decision of the neighboring countries. I propose to estimate this model by implementing a new algorithm with higher mixing and converge. Based on a sample, including 141 developing and developed economies, I show that the probability for a country to introduce an eco-labelling scheme depends on the eco-label programs adopted by countries which are spatially close or sharing a strong trade intensity relationship with each other. These results explain why developing countries, which are standards takers, have naturally been put at a disadvantage in terms of eco-labelling adoption. Therefore, eco-label programs should be as transparent as possible and rely on standard harmonization and mutual recognition to remove any potential technical trade barriers and encourage the participation of developing countries.