Measuring border effects in European cross-border regions

ABSTRACT

This paper presents a new methodology to measure border effects from a different perspective with respect to the standard gravitational approach. The methodology proposed measures supply-side border effects by identifying two types of limits produced by the border to the productive system: inefficiency in exploiting local resources (efficiency needs) and scarce endowment of resources (endowment needs), the former calling for intervention on resource governance, the latter requiring new investment. The methodology, applied to the European Union’s Cross Border Cooperation Program regions, suggests a stronger presence of efficiency needs with respect to endowment ones.

KEYWORDS:

• cross-border regions
• border effect
• European Union
• cross-border cooperation
• regional growth

JEL:

• R11
• R12
• R58

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

SUPPLEMENTAL DATA

Supplemental data for this article can be accessed at https://doi.org/10.1080/00343404.2017.1364843.

ORCiD

Roberta Capello http://orcid.org/0000-0003-0438-6900

Andrea Caragliu http://orcid.org/0000-0003-0865-3404

Ugo Fratesi http://orcid.org/0000-0002-0755-460X

Notes

1. In fact, this literature defines border effects as ‘the downward impact of national boundaries on the volume of trade, i.e., that two different countries trade far less with each other than do two locations in the same country, after controlling for factors such as income, alternative trading opportunities, and distance’ (Evans, 2003, p. 1291).

2. Among the few relevant exceptions, Chandra, Head, and Tappata (2014) assess the strength of border effects on individual travel decisions, while Persyn and Torfs (2015) model the intensity of commuting across Belgian municipalities to measure the intensity of regional border effects.

3. Border regions have been identified in the study entitled ‘Collecting Solid Evidence to Assess the Needs to be Addressed by Interreg Cross-Border Cooperation Programmes’, Framework Contracts 2014CE16BAT010/2014CE16BAT011/2014CE16BAT012 (Service Request 2015CE160AT044), in which the authors were partners.

4. That is, border regions comprise 220 million inhabitants and make up for almost €6 nominal billion (out of 520 million inhabitants and almost €14 billion respectively). Taken together, these figures suggest that border regions are also on average as productive as the average EU region – if anything, per capita value added in border regions is slightly higher than in the rest of Europe.

5. This methodology implicitly assumes that border and non-border regions follow similar development models, i.e., the strategic importance of a resource on average hold also for border regions. While this could be disputable, we are confident that this assumption can be accepted since the within-border regions heterogeneity makes the latter diverse enough not to allow the identification of a border regions-specific development model. This is also reflected in a non-significant estimate of the border dummy in equation (1), which suggests that border regions grow exactly like non-border ones.

6. Robustness checks on a longer period, namely 2006–13, were run. Signs and significance of the different regressors remained stable.

7. Since NUTS-3 classifications have been frequently updated in EUROSTAT, a remarkable effort in data harmonization has taken place.

8. The same analysis has also been performed excluding from the sample the regions hosting large cities to test whether the results are due to the presence of cities. The results remain unaffected.

9. Additionally, accessibility is measured in terms of the population that can be reached from the region, and when the region has a coastline, nobody lives on the other side of the border by definition.

10. Analytically, the dependent variable is calculated as (GDP₂₀₁₃ – GDP₂₀₀₈)/GDP₂₀₀₈.

11. For a discussion of the identification of agglomerated, urban and rural regions, see Capello and Chizzolini (2008).

12. For ease of presentation of the results, the usual statistics (R² and joint F-test of significance) are reported in Table A4 in Appendix A in the supplemental data online.

13. The of equation (2) is not significantly different from the of equation (1).

14. A second graphical example is provided and commented on in Appendix A in the supplemental data online.