Uncovering the Sources of Geographic Market Segmentation: Evidence from the EU and the US

Non-Technical Summary

There is a large body of scientific work that argues that integration of separate geographic markets increases aggregate welfare. Historically, increased market integration has happened through reductions in transport costs and decreased cross-border trade frictions, such as tariffs or regulatory barriers. Given the potential gains of market integration, assessing whether markets are geographically integrated or segmented and the sources of lingering market segmentation remain issues of central importance to international economics.

This raises an important question: When do we say that markets, countries in our case, are geographically integrated or segmented? Put differently, what measure should we use to make this judgement? One strand of literature assesses whether prices of identical products differ significantly more between than within countries. If prices differ more between than within countries, countries are said to be geographically segmented. While looking at prices of the same product seems a very intuitive strategy, it does not consider that the set of products that is commonly available across markets might be very small to begin with. That is, considering whether the set of products is available in both markets is also relevant to geographic market segmentation. Another strand of literature looks at whether trade shares discontinuously fall at borders. Differences in trade shares may indeed capture both differences in prices and product availability but may also stem from cross-country differences in consumer preferences.

In this paper, we develop an integrated framework to assess the presence of cross-border geographic market segmentation and uncover its sources by measuring the importance of both price and product availability differences as manifestations of cross-border market segmentation. To this end, we propose a two-step approach. In the first step, we develop a cost-of-living framework and shows that price and product availability differences can be measured in a common unit, separately from consumer taste. The cost-living framework closely relates to the Consumer Price indices published by central banks and governments. In the second step, we derive testable conditions that compare the price and product availability differences between and within countries. We show that these conditions are sufficient to detect whether markets are geographically segmented in a large class of theoretical models typically used in international economics.

Implementing our framework yields three main insights. First, cost-of-living differences are roughly 2.5 times larger between than within EU countries. In contrast, cost-of-living differences are only marginally larger between US states compared to within US states. Our decomposition shows that taste differences account for a little over 80% of the variation in cost-of-living differences within EU countries and across US states. This illustrates that it is quantitatively important to control for taste differences when assessing the presence of cross-border market segmentation.

Second, both price and product availability differences are significantly higher and economically much more important between than within EU countries. While price and product availability differences are also significantly higher between than within US states, the differences are quantitatively small. Our findings on the EU trade frictions are particularly noteworthy, since we focus our European analysis on a subset of Western EU countries that are arguably more integrated.

Third, product availability differences between European countries quantitatively dominate price differences. In particular, price differences are around 10 percentage points larger between than within EU countries. In contrast, differences in product availability are around 30% larger between EU than within countries. Hence, in terms of cost-of-living differences, this suggests that cross-border segmentation is three times more important through fixed trade frictions than through variable trade frictions, even though the latter has received the most attention in the literature.

Read Working Paper 013