![Sample our Area Studies journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5922/TOC-Subject-Sample-2021-Area-Studies.jpg"})
ABSTRACT
This study explores the role of stabilization and association agreements (SAAs), bilateral investment treaties (BITs) and free trade agreements (FTAs) in the Western Balkans and their impact on the region’s inward foreign direct investment (FDI) and exports, with primary interest in the effects on intra-regional FDI and trade. Results show that BITs were not related to intra-regional FDI, nor to the FDI from other countries. The Central European Free Trade Agreement (CEFTA) contributed to increased intra-regional trade. Finally, the SAAs turn out to be highly significant for FDI from the EU to signee countries, and their exports to the EU.
Disclosure Statement
No potential conflict of interest was reported by the authors.
Supplementary Material
Supplemental data for this article can be accessed on the publisher’s website.
Notes
1. The cutoff date for data and information used is June 2019.
2. This agreement is usually referred to as “new” CEFTA, or ‘CEFTA 2006ʹ. The “old” CEFTA was started in 1992 with the members comprising the Central and Eastern European (CEE) countries that joined the EU in 2004, 2007 and 2013. The first members were Poland, Hungary and Czechoslovakia (later Czech Republic and Slovak Republic), while Slovenia joined in 1996, Romania in 1997, Bulgaria in 1999, and Croatia in 2003. These countries left CEFTA as they joined the EU. This indicates how far Western Balkan countries were lagging behind in terms of economic integration initiatives.
3. As found by Desbordes and Vicard (Citation2009) in their analysis of the quality of political relations between host and home countries and FDI, the BITs are more effective in promoting FDI between countries with comparatively tense political relationships, which has been the case in the Western Balkans.
5. See e.g. Blonigen and Piger (Citation2014) for further insights into determinants of FDI, and Fugazza (Citation2004) or Cerra and Woldemichael (Citation2017) for findings on export (acceleration) determinants for broader samples of countries. Jirasavetakul and Rahman (Citation2018) present a recent evidence on FDI determinants into transition countries (including Western Balkans but do not analyze the role of BITs), while Kaloyanchev, Kusen, and Mouzakitis (Citation2018) and Petreski (Citation2018) more recently looked into the broader determinants of intra-regional trade among Western Balkan countries.
6. In contrast, many transition countries in CEE had BITs in force (with other countries from the same group) already by 1995, many entered into force by 2000, and only a smaller fraction in early 2000s.
7. However, the method we apply (described below), using host country- and home country-time effects, should account for any country-year specific changes, including methodological ones.
8. Guerin et al. (Citation2010, p. 18, Table 2.9) present the large discrepancies in intra-regional bilateral trade values recorded by importer and exporter countries. While some differences are expected due to transport costs (which increase the value of imports relative to exports), for some country pairs the difference amounts to more than 50%. For other country pairs the value of exports exceeds the value of imports.
9. For the sake of simplicity, we drop the subscripts of the variables after the first mention.
11. While in our main approach the BIT variable assumes zero value for a pair of countries when they both become EU member states, we also ran the regressions and comment on the results with the alternative assumption that BITs still remain in force even after both countries have joined the EU. See the discussion in section 3, on the validity of the existing BITs after (both) signee countries become EU members.
12. While our data do not allow for such detailed analysis, we do check for other factors that may influence the relationship between BITs and FDI, such as institutional development.
13. It should be noted that the strategies of MNEs are often very complex, so that the traditional classification of FDI into strictly horizontal or vertical is not that meaningful in practice (Helpman Citation2006).
14. Bulgaria and Romania were also meant to become members of the “new” CEFTA, but they joined the EU in 2007. Other regional trade agreements include the Eurasian Economic Union, the Baltic Free Trade Area, or trade agreements with members of the European Free Trade Association.
15. While this was the case for the large majority of BITs, sometimes there was a lag of several years between the signing and the enforcement of a treaty, and there were also (rare) occasions when a BIT was signed but did not enter into force during the period under observation, although several years had passed since the signing.
16. We tried adding the institutional variables separately to each of the six specifications, which turned out to be infeasible, as this variable by itself appears collinear with the set of fixed effects. The same problem arises if we include only BIT and institutional distance variables without the interaction term, and if only institutional variables are included in the simplest model, only with the EU and SAA control variables.
17. We note that we also tried excluding single country pairs with Kosovo as importer and other Western Balkan countries as exporters; and with Serbia as exporter and other Western Balkan countries as importers. We found that no single country pair drives the above results to a large extent.
18. A similar development, though less pronounced in terms of the increase in overall exports as a share of GDP, can be observed for North Macedonia. Leaving North Macedonia out from our regressions in also increases the estimated coefficient for CEFTA, but much more mildly than for Serbia.
19. We note again that the sample is reduced when adding the tariff variables. Specifically, Kosovo is left out completely owing to a lack of tariff data. For this reason, our results in the fifth columns of Tables B6 and B7 produce results for the FTA_WB2 variable that are most similar to the result in the third column of , where Kosovo was also left out.