It is widely acknowledged that top-down support is essential for bottom-up participatory projects to be effectively implemented at scale. However, which level of government, national or sub-national, should be given the responsibility to implement such projects is an open question, with wide variations in practice. This paper analyzes qualitative and quantitative data from a natural experiment in the state of Rajasthan in India, where a large national flagship project that mobilized women into self-help groups for micro-credit and created a women's network for other development activities was implemented in two different ways. Some sub-regions were given to the state government of Rajasthan to manage, while the Government of India centrally managed other sub-regions. The study finds that the nature of top-down management had a large bearing on the nature and quality of local-level facilitation. Centrally and locally managed facilitators formed several groups with similar financial performance. But centrally managed facilitators formed groups that were less likely to engage in collective action, be politically active, and engage with other civil society organizations. These results raise important questions on how responsibilities for participatory development projects should be devolved, and how the nature of management affects the sustainability of bottom-up interventions.