David Chilosi (King’s College London)

joint with S. Nikolić

The “border effect” is an old puzzle in trade economics: crossing a border increases price gaps and reduces trade by much more than would be expected on the basis of observable determinants of trade costs. A number of recent papers have offered a novel solution: ethnic diversity undermines trust and increases transaction costs. These studies take ethnic groups as exogenous. The assumption of fixed ethnic boundaries agrees with idea that culture is defined by its historical persistence but has yet to be subjected to systematic empirical testing. Our paper asks if, to what extent, and why the effect of ethnic ties on trade costs changed over time with a new approach and new data. We look at the decades before and after the creation of a paradigmatic multi-national state, Yugoslavia, and draw on a panel of over 550,000 observations. We find that religious differences explained a large share of the border effect in pre-unification Yugoslavia, but their influence vanished over time. The decline is visible also in city-pairs that were initially in the same state and began about twelve years before the unification of Yugoslavia, suggesting that nation-building, rather than a weakening incentive to rely on private order institutions, was the main factor.

VIEW PAPER