This paper considers a particular debate in the scholarly literature on ethnicity in the United States regarding ethnic entrepreneurship which has come to be known as the ‘ethnic enclave economy hypothesis’. According to this hypothesis, ethnic enterprises and their workers benefit from clustering. The current consensus seems to be that the hypothesis is both redundant and misguided. In denying this critique, this paper draws on Lefebvre's theorisation of space and on industrial cluster and actor network theories to argue that the dominant interpretation of the notion of an enclave has been misconceived. Hence the need is to begin to theorise what ethnic enclave economies really are, beyond the spatial metaphor in which the hypothesis was first grounded, in order to interrogate generative processes of ethnic business formation.