Cletus Coughlin and Dennis Novy are concerned with the famous
border effect in trade, the empirical finding that trade within countries
sizably exceeds the countries’ cross border trade, after holding constant
for other determinants of trade such as the economic size of the trading
partners and the distance between them. Specifically, Coughlin and Novy
are interested in the relative magnitude of the estimated effect. To analyze
this issue, they combine data sets that allow them to jointly analyze trade
within US states, trade between US states and the international trade of
US states. Interestingly, they find that the international border effect
(that adds to the effect of crossing a state border) is smaller than the
state border effect itself, reflecting the strong local concentration of
economic activity