Sammendrag:

There is an increasing focus on involving business in development work. This thesis uses the case of StatoilHydro in Brazil. I show that StatoilHydro has substantial financial and political support from the Norwegian Government and is promoted as an agent of Norway s energy and climate initiatives in Brazil despite their different interests. Based on a framework for studying TNCs power, I analyze whether the governmental decision to promote StatoilHydro in such a way is due to the company s power resources and power activities. First, I argue that StatoilHydro evidently has strong power resources because of its governmental ownership, its control of the research agenda and its technological power. Second, StatoilHydro holds and uses its highly developed power activities towards the whole of society, not just politicians. I argue that lobbying is not as important a channel for power as StatoilHydro s more invisible power, such as network power, power of argumentation and persuasion, CSR and PPP. In both dimensions, discursive power stands out as an important, and at the same time an underestimated source of power. I would like to mention three interesting findings. First, I argue that the promotion of StatoilHydro as an NOC is misleading and has improved StatoilHydro s market share in Brazil. I suggest that the definition of energy companies should be more nuanced than today s NOC-IOC classification. Second, StatoilHydro promotes its climate profile and work on Carbon Capture and Storage (CCS) in Brazil although CCS is not relevant for its activities in Brazil. Therefore, when dealing with CCS it is important to separate between power generation and exploitation, oil and gas, and onshore and offshore production. Therefore, StatoilHydro s promotion of CCS should be seen as `greenwashing´. Third, the Government has no focus on studying the power or influence from TNCs. If incorporation of TNCs should be positive for the total of Norwegian development work the Government has to be more aware of TNCs power and how it can render possible influence through invisible mechanisms.