Analysing Linkages between Climate Change, Economic Development, and Civil War : An Econometric Analysis
This thesis analysed linkages between climate change, economic development and the incidence of civil conflict from 1981 to 2015. This dissertation consists of four empirical papers contributing to a better understanding of the complex interrelation between climate change, economic development and civil conflict. To analyse quantitatively these linkages, we use new cross-country panel data from 1981 to 2015. We include control variables for good governance and democracy as well as a food production index to identify potential mechanisms of climate change impacts on conflict. We test new measurements of changes in rainfall and temperature and compare the effects of shocks to the growth effects. We estimate the impact of lagged conflicts on the incidence of civil conflict with the inclusion of a lag-dependent variable in the econometric estimation. We take into account potential heterogeneity by testing for different impact mechanisms of climate change on conflict in both studied regions (sub-Saharan Africa compared to the Middle East and North Africa). Finally, we test different measurements of climate change (continuous growth rate versus extra ordinary shocks in climate variables, in addition to annual levels of rainfall). In addition, we consider the role of economic policies in avoiding new conflicts and triggering peace if climate shocks cannot be avoided. Several main conclusions are drawn based on the results of this thesis: (1) Weather variables are good instruments not only for economic growth rate but also for domestic food production. The main pattern found in this research on the climate change effects is that temperature contributes to climate impacts more than rainfall impacts on economic development. The main effect of climate comes from the temperature growth effects and it is not extreme shocks that drive economic declines, which indicates that the climate rather operates in a non-linear process. (2) Beyond the economic growth rate, climate change also impacts civil war via domestic food production. A 10% reduction in economic growth or domestic food production leads to 1.2% and 1.59% increase in the probability of civil conflict, respectively. There appears to be evidence of different impact mechanisms of both indicators on the incidence of civil conflict, in different regions. (3) A direct impact of climate is identified on the incidence of civil conflict. Rainfall growth has a direct significant impact on reducing the likelihood of civil conflict, this finding is robust to the inclusion of several model specifications. (4) The inclusion of the impact of reverse conflict is essential in order not to overestimate the impact of climatic and economic factors on conflict. (5) The level of democracy of a political system and good governance are important control variables. (6) Proper economic policies that can accelerate economic growth and food policies that enhance food security seem to play a significant role towards peace. However, applying an econometric “state-of-the-art” approach to control for potential endogeneity, unobserved heterogeneity and dynamic specification problems, still results in some limitations, i.e., sensitivity of estimation results with regard to used measurement concepts of central climate variables. Furthermore, applied econometric approaches are limited to identify causal relationships. Therefore, future research is needed to address these problems. In this regard, we consider the combination of econometric and modelling approaches in a Bayesian estimation framework as a promising approach.