This paper uses a historical natural experiment – the opening of the Suez Canal – to investigate the relationship between geography and the formation of institutions. While the conventional view is that good geography (commodity endowment) inevitably favours the creation of extractive institutions, we discover that a second aspect of geography – location – may in fact encourage the establishment of non-extractive institutions when rent extraction by elites depends on the productivity of non-elites. Specifically, we find that entrepôt colonies (Hong Kong and the Straits Settlements) received larger public investments in the post-Suez period than resource colonies (British India, Ceylon, and West Africa), after accounting for year effects and permanent differences across colonies. We demonstrate, using supplementary data, that the entrepôt colonies' locational advantage, coupled with their lack of extractable resources, plays a key role in explaining our empirical findings.