The “canonical” literature on national identity (Gellner, Anderson, etc.) has stressed the imaginary, often strategic (and even functionalist) nature of modern national identity. After pointing to its theoretical and empirical shortcomings, I suggest that nationhood, as understood today, was linked to an emancipatory project – the construction of a “liberal” order – that was universally attractive. In those states where a “liberal” or “bourgeois” revolution succeeded, the old world of monarchical courtiers, corporate interests, and social estates disappeared, replaced by an abstract society of politically equal individuals (as defined in Cochin, Furet), and political borders became congruent with national identity. By contrast, in those states where the liberal revolution failed (or happened late in time), the preservation of spatially-defined barriers (inequalities) led to a break between the (old) center and one or more national peripheries or communities. I test the hypothesis by looking at the evolution of the political identity of Jews across American and European countries over the 19th century, complemented with a regression discontinuity design that exploits the differential political treatment of Polish and Russian Jews under Tsarism.