E-government is being promoted by international agencies and G8 nations as a means to obtain efficiency, accountability and transparency in the governance of economically less developed countries. In particular, the model for good governance is the one advocated by new public management: the minimal, service-delivery state. The paper shows, first through the case of Jordan, how e-government is difficult to implement, given the characteristics of the local administration, the socioeconomic context and the dynamics of the technological infrastructure. On the basis of such evidence, it asks more generally whether the marketization of the state, embedded in e-government, makes sense as the paramount approach to improve democracy and foster development. It turns out that the transformation of citizens into customers is problematic, and the correlation between good governance and minimal state with development can hardly be demonstrated historically.If such failures are both pointed out by institutional economics theory and by current practice, the paper explores possible reasons why these projects continue to attract development aid funds. Specifically, the paper puts forward a new interpretation centred on the newly established link between aid and security. In this light, e-government appears to be one of the new tools for the rich metropolitan states to govern "at a distance" (through sophisticated methodologies and technologies) the potentially dangerous, weak, borderland states. But such an approach, as many ICT fixes for the private sector have shown, may also fail and backfire: new ICT applications can drift away from the set targets and global, durable disorder within and between states may sustain intact. New research approaches leading to new practices are desperately needed.