Complements or substitutes? New theoretical considerations and empirical evidence on the imports and FDI relationship in Central and Eastern European Countries

Abstract

This paper addresses the imperative need to understand the relationship between inward FDI and trade by developing a new conceptual approach and providing empirical evidence. We use an expanded time dataset, from 1992 to 2008 and an enriched dataset of countries, sectors and location factors. In regards to the inward FDI versus imports relationship, results comply with our theoretical formulation and strongly indicate an overall complementarity with each other. In the case of FDI we find strong locational characteristics such as the large market size, the gradual improvement of the macro-environment and finally the quality of labour force to play a positive role.

Keywords:

• Central and Eastern European Countries (CEEC)
• Investment Development Path (IDP)
• inward Foreign Direct Investment (FDI)
• imports

JEL Classifications:

• F210
• M110
• O520

Acknowledgements

The authors acknowledge constructive comments from an anonymous reviewer and S. Estrin, P.J. Buckley, J. H. Dunning, and R. Pearce. The usual disclaimer applies. This paper is part of work in progress under the title ‘An Investigation of the FDI and Trade Relationship: Evidence from Central and Eastern European Countries’.

Notes

1. United Nations Conference on Trade and Development (UNCTAD) changed the classification of countries in 2005; the eight former countries of CEE that first joined the EU are now classified into the developed ones along with Malta and Cyprus as the 10 new member states of the European Union (EU). The rest of the countries including Bulgaria and Romania have been classified in the region covering the South-East and the Commonwealth Independent States (CIS). Croatia belongs into this latter region. For analytical purposes we refer to these countries as commonly known, i.e. CEECs.

2. The new EU member states are Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Slovak Republic, Slovenia and the recently accessed Bulgaria and Romania.

3. Dunning, Kim, and Lin (2001) refer in general to FDI and trade; however, elaborating the argument, the relationship refers to either outward FDI and exports or to inward FDI and imports which is our case here.

4. We refer to the second stage as most of the countries under consideration were at this stage for the longest period of our analysis.

5. In the FDI equation we used alternatively the variables TECHNOLOGY and R&D/GDP but for simplicity we report here only the latter one.

6. It must be mentioned here that all of the studies carried out for the CEECs, are country-level, either dealing with total FDI inflows in the region or bilateral flows from some advanced origins to CEEC destinations, for there is lack of ‘consistent and detailed sectoral data’ (Resmini, 2000, 666). The only exception belongs to Resmini (2000) who analyses the determinants of FDI inflows in the manufacturing sector of 12 host CEECs, following the Pavitt (1984) taxonomy (the Pavitt taxonomy distinguishes among scale-intensive, high-tech and traditional sectors and specialised producers).

7. The entry modes are distinguished in green-field investment, mergers and acquisitions (through privatisation), nevertheless, all types of entry mode regard long-lasting interest in the respective host, and thus country location factors are significant prerequisites for all.

8. The use of a different measure of imports, normalised by the total trade, or the GDP of the country does not alter the results. In addition, the inclusion of a lagged FDI and Imports variable in the estimations does not alter the results, which are available upon request from the authors.

9. It is noteworthy that large markets hold also a particular role in new trade and new economic geography theories, as they reflect the potential of firms to capture economies of scale (Amiti 1998). Other studies that traditionally use real gross domestic product include Buckley and Casson (1981); Dunning (1993); Aristotelous and Fountas (1996); and Scott-Green and Clegg (1999).

10. The empirical evidence is vast; a few and influential studies belong to Dasgupta et al. (1996), Noorbakhsh and Paloni 2001).

11. Cost factors are at the heart of the Hecksher-Ohlin traditional trade theory. Foreign investments are considered to be motivated by production cost differentials, which investors exploit in order to increase their profits by reducing their cost of production. It is, however, beyond the scope of this paper to explain FDI determinants within this framework.

12. See also Lansbury, Pain, and Smidkova (1996), Meyer, (1998, 2001) and Hardy (1994).

13. Some of these variables have been use by other scholars, e.g. Daude and Stein 2007 (voice and accountability, political stability and lack of violence, government effectiveness, regulatory quality, rule of law and control of corruption) Bellos and Subasat 2012 (governance, corruption), Jadhav 2012 (control of corruption, government effectiveness, no violence) though not for Central and Eastern Europe.

14. A very interesting discussion from a policy implications’ perspective study, though indirectly linked to CEEC countries, can be found in Breuss, Egger, and Pfaffermayar (2001). They study the effect of Agenda 2000 reform of structural expenditures on outward FDI stocks from OECD countries in the EU.

15. http://www.prsgroup.com/ICRG_Methodology.aspx.

16. Data availability constraints exist after 2008 on institutional data (ICRG 2008).

17. The three time periods correspond to the early stages of transition (before 1997), the mature stage (between 1997 and 2002) and a stage where most of these countries are beginning their final preparation to the EU accession (after 2002). To choose among various breakpoints in our sample we also employed the Supremum F test, which involves estimating all Chow F statistics for each potential breakpoint in the sample and choosing the one where the F statistic was higher. When this test was implemented the breakpoints were determined at 1997 and 2002. The relationships before, between and after these two breakpoints are further explored in the following empirical part of the paper. The full correlation of our variables is presented in Appendix 2.

18. Stationarity tests (Im, Pesaran, and Shin 1997) showed no evidence on the existence of a unit root. These results were suppressed due to size constraints but are available upon request from the authors. The Im, Pesaran, Shin test were calculated using the IPSHIN routine in STATA v.10.

19. Results can be obtained from the authors upon request.

20. Results can be obtained from the authors upon request.