This paper investigates the effects of aid fragmentation on democratization. Aid fragmentation, i.e. the provision of aid by a large number of donors with similar market shares, has been shown to impact negatively on governance in the recipient country. But aid fragmentation also makes coordination among donors more difficult, frustrating attempts of implementing political conditionalities. Thus, fragmented overall aid is expected to decrease leverage on inducing political reform in recipient countries. I argue, however, that fragmentation of democracy aid has the opposing effect. Democracy aid is geared towards multiple actors on the recipient side, reducing the transaction costs compared to aid geared exclusively towards the national executive. At the same time, coherent democracy aid would offer only a limited set of options towards solving the complex process of creating viable political institutions. Since a beneficial institutional layout is difficult to define ex ante and from the outside, coordinated donor engagement enforced by conditionalities is likely to lead to the imposition of institutional blue-prints that do not reflect the needs of the people. In a more fragmented aid setting, multiple recipient actors can choose from various types of assistance offered, leading to institutional arrangements that better coincide with local needs. I expect this effect to differ conditional on the level of democracy in the recipient country, because developed party systems and an organized civil society that is able to benefit from a diverse donor environment are more common in less autocratic states. These propositions are tested with data covering the years 1990 to 2009. I apply an error correction model (ECM) to model the temporal dynamics, quantile regression to examine the conditional effects and Granger causality tests. The preliminary empirical evidence suggests that fragmented democracy aid has a positive impact on democracy, whereas fragmentation of non-democracy aid has a negative impact.
