While market integration typically depends on the level of technology and infrastructure, economists have shown that borders affect prices and trade flows (p.117). Administrative borders are strongly trade diverting but what about other types of borders such as the ethno-linguistic ones? The authors study the Austro-Hungarian Empire as it represents a rather unique example of multi-ethnic polity not divided by national borders (p.118).
Borders before the borders
They use the evidence provided by the grain prices in 20 cities of the Empire between the 1870s and 1914. The censuses undertaken every decade account for the composition of these urban centres’ population linguistic composition (p.123). It comes out that, while over the whole empire during the period grain prices did converge, they did so much faster between cities that would belong to the same country after 1918. In other words “integration became progressively asymmetric since the mid 1880s” (p.124).
Surprisingly, a border effect is observable even before the establishment of administrative borders. Noticeably, this factor cannot be observed before the 1880s (p.126). This effect cannot either be explained by geographical distance (which influence on grain price decreases over the period; p.127). On the other hand, infrastructures were particularly improved along the lines of the future post-1918 national borders, but this factor only had a marginal impact on grain price. In the same way, physical geography (e.g. Carpathian Mountains), did not significantly affect trade (p.128). Finally, the degree of integration with neighbouring countries (Germany, Russia) did not have a sizeable impact either (p.129).
“Overall, we conclude that ethno-linguistic network […] probably almost fully explain the emergence of border effects along the future political borders across the Habsburg Empire. […] To put a number on that effect […] doubling the ‘matching probability’ between any two cities in the sample reduces the price dispersion between them by more than a 50% reduction in the railway distance between the two cities” (p.130). According to the authors, the mutually excluding process going along nation-building originated such trend (p.132), language barriers as such don’t qualify as a cause since it can’t explain the increase of the border effect. Economic proto-nationalism explains the bias observed above in market integration within the Empire (p.133).
Ethno-linguistic barriers thus took over transport cost as the major cause of dividedness in the Empire, it is only when the possibility to integrate arises that the refusal thereof really matters. On the other hand, the post-1918 partition of the Empire appears less arbitrary than previously assumed, and the new administrative barriers may have had less of a negative impact as well since trade-diverting line existed before hand (p.134).