An Empirical Investigation of the Determinants of Foreign Direct Investment in the Central and Eastern European Countries Using Multi-level Data

In order to incorporate FDI factors for different MNEs, industries and countries, a unique dataset that includes country-, industryand firmlevel data is constructed. This approach avoids the aggregation or disaggregation problem by considering several levels of data simultaneously in to the analysis. The constructed database contains information on 1,223 foreign investment made by EU15 countries and other major investors in CEE (Japan, Norway, Russia, Switzerland and US) in 12 CEECs, namely 10 new EU member states (except for Malta and Cyprus) and Bulgaria, Croatia, Romania and Ukraine. This paper analyses two main questions. First of all, to what extent do different level determinants explain the size of investment at firm level? For this purpose OLS regression is used and it can be concluded that all firmand industry-level and some country-level factors can explain the size of investment at firm level. Hence, the insignificance of some countrylevel variables is not a sufficient argument to dismiss them from the analysis: those variables may not influence the size of the investment at the firm level, but they can affect the choice to invest or not in a particular country. As a result, the second question is raised, where do MNEs choose to locate in CEE? The results confirm that the responsiveness of FDI in CEECs to country-level variables differs across sectors and across firms of different sizes and profitability.