Stefan Nikolic (p)

Abstract
Increasingly scholars stress that national borders and ethnic diversity persistently increase transaction costs and thus impede market integration. The so-called “border effect” is an old puzzle in trade economics: crossing a border increases price gaps by much more than would be expected on the basis of observable determinants of trade costs (Engel and Rogers 1994). The recent “cultural turn” in economics offers a potential solution: ethnic diversity undermines trust and increases transaction costs (see Alesina and Ferrara 2005 for a literature review of the economics of ethnicity).
To analyse the role of political and ethnic borders in the integration of urban markets, we exploit border changes that accompanied the creation of a paradigmatic multi-ethnic state, Yugoslavia. The possibility of exploiting changes in political borders distinguish this study from the other econometric analyses of the effects of ethnicity on market integration.
We are in the process of completing the data-set. The final panel analysis will be based on a comprehensive data set containing yearly wholesale prices of grains, before and after the First World War, from period statistics for 173 markets spread across Yugoslavia, yielding over 500,000 observations.
We have ran an exploratory analysis based on a subset of observations and controls: political borders, internal borders, religious similarity, river access, sea access, rail access, and city-pair log distance. Our dependent variable is the absolute value of natural logarithm of the price ratios between cities.
Our preliminary results show that political but not ethnic borders fragmented the Yugoslav market. This would imply that the Yugoslav cities had a lot to gain from unifying and little to lose. These results sit better with the findings of scholars stressing that the effect of ethnicity on trust costs is historically contingent (e.g. Schulze and Wolf 2008) than of those who stress the deep roots of ethnic conflict (e.g. Voigtländer and Voth 2012).

Note: See also extended abstract.

Thomas De Graaff (p)

By applying a Bayesian multilevel gravity model, this paper revisits the impact of home-ownership and social renting rates on intercity migration in the Netherlands. Where most of the extant (economic) literatures focuses on using fixed effects for cities of origin and destination, I adopt a Bayesian multilevel approach to model the cities’ varying effects. For this application, this approach has two main advantages. It (i) allows for simultaneous
estimation of city specific varying effects and the effects of city specific home-ownership and social renting rates on migration flows. Moreover, it allows (ii) for prediction of migration flows between cities both in- and out-of-sample. The results show that home-ownership rates decrease migration flows significantly with an elasticity close to -1. Municipal social renting rates have negative impacts as well, but their elasticities are close to zero however. The latter results are robust to the year of observations, alternative specifications (including household characteristics), and the use of interaction effects. I conclude this paper by predicting changes in all in- and out-going migration flows in Amsterdam driven by an changes in the housing market structure.

Luisa Schneider (p), Matthias Wrede

This paper studies the effects of multiple increases of the German Real Estate Transfer Tax (RETT) on spatial mobility decisions of individuals. We look at RETT hikes introduced by a federal reform in 2006 and exploit the staggered introduction of these hikes and the different levels of the tax across the 16 federal states of Germany.
We use longitudinal individual-level data from the German Socio Economic Panel (SOEP) to evaluate the impact on moving probabilities and tenure choice of homeowners and renters. Preliminary results of our competing risks hazard model show that a higher tax rate lowers the probability of moving into ownership, both for previous owners and renters. We nd no statistical significant effects on moves from own to rent or rent to rent.
Our analysis using survey data will be extended by the evaluation of administrative regional data from the Federal Statistical Office. We will observe aggregate relocations on the district level along with regional controls like unemployment rates, GDP on the county level and land tax levels. We will investigate the effect of the RETT on in- and outmoves
in a Diff-in-Diff setting where we will compare border-districts in high-tax states to those in low-tax states.