The causal effects of the Indian Ocean tsunami and armed conflict on Aceh’s economic development

Abstract

This PhD thesis investigates the causal long-term economic effects of the Indian Ocean Tsunami and the armed conflict in Aceh, Indonesia (chapters 2, 3 and 4). It also contains an analysis of land use change and the consequences for soil-organic carbon (SOC) in Eastern Panama that is unrelated to previous chapters. Chapter 2 stands at the core of my PhD thesis; it is the equivalent of a job market paper.
 In chapter 1, I provide an introduction to and summary of my PhD thesis. In particular, I describe why I believe that I make original contributions to knowledge that are significant and rigorous.
 In chapter 2, I carry out a quasi-experimental analysis investigating the causal effects of Tsunami flooding on long-term per capita economic output. The existing literature suggests that natural disasters are growth depressing in the short-term, and in the longterm, natural disasters either cause a continued shortfall of economic output, or an eventual convergence to the pre-disaster counterfactual trend. I picked the Indian Ocean Tsunami in Aceh as a case study for this PhD thesis, because I posit that if there is one case for which there is evidence that goes against the conventional wisdom, namely in the form of increased economic output in the long run, it probably is Aceh. The reason why I expect to see creative destruction is that Aceh received a windfall of aid and was the stage of the largest reconstruction effort the developing world has ever seen. I conclude that natural disasters are not necessarily the cause of output reductions and that they can be windows of opportunity for the economy.
 In chapter 3, I investigate the reasons behind the creative destruction, and take a closer look at different sectors and subcomponents of the economy. I examine three channels through which the Tsunami may have affected per capita economic output. First, I find that the Tsunami causally accelerated the structural transformation process, a process through which people and the economy move out of agriculture, and into more productive sectors such as services. Second, I show that the Tsunami brought with it a windfall of aid and other funds, which allowed for a building back better of physical capital and increased capital formation. Third, I show that aggregate private consumption not only was smoothed in a reaction to the Tsunami, but even boosted to sustainably higher levels, compared to the no-Tsunami counterfactual.
 In chapter 4, I investigate whether the 30 years long armed conflict in Aceh left any negative economic legacy effects, once the fighting stopped and the peace agreement was signed. The separatist war took a toll on the Acehnese economy. Even though the conflict has ended, did the negative economic effects also end? Aceh’s economy has higher per capita growth rates in times of peace than in times of war, which can be either a sign of a peace dividend or creative destruction from the Tsunami. But does the armed conflict leave a negative legacy for future growth rates, even after peace has officially been declared? I find that that peacetime growth rates are negatively affected by the wartime conflict intensity. Using violence data on the incidence of killings, injuries, and other ‘measurable human suffering’, I assess whether districts that were heavily affected by armed conflict grew systematically differently from those that were spared from the brunt of the violence. I find that there are severe negative economic legacy effects of violence, and the more violence occurred in a district during the separatist war, the slower it was growing during times of peace.
 Chapter 5, topically unrelated to the previous chapters, is looking at land use change in Eastern Panama and the consequences for soil organic carbon (SOC). In this chapter, I compare SOC concentrations of primary forests to two competing land use alternatives: Forest-to-pasture conversion for cattle grazing versus indigenous forest-to-crop conversion. I find that both land use changes reduce SOC concentrations significantly, yet the pasture land use has lower levels of SOC than indigenous crop cultivation. The soil carbon levels of secondary forests are not statistically different from primary forests, implying that the forest conversions are reversible, in terms of their impact on SOC, which suggests that allowing secondary forests to re-grow in former cultivated areas in the Eastern part of Panama holds promise for climate change mitigation.
 In the concluding chapter 6, I present a summary of the main findings and an outline for future research.