The investment location choices of multinational enterprises in Central and Eastern Europe: the multi-level data and discrete choice methodology approach

This thesis examines the principal economic factors explaining firms' foreign directinvestment (FDI) location decisions into 13 Central and Eastern European countries(CEECs) between 1997 and 2007 using discrete choice econometric methods.The first part employs Meta-analysis to systematically summarise, integrate andsynthesise the results of empirical studies that analyse two main reasons whymultinational enterprises (MNEs) locate their investment abroad: access to foreignmarkets and reducing productions costs. A large number of factors related to modelspecifications, dataset characteristics and methodologies in the primary studies explainthe variation in the estimates of the market size and labour costs effects on FDI acrossthe studies. Furthermore, the existing empirical literature on the market size effect onFDI is prone to publication bias more than the literature on the labour costs effect onFDI, as papers with statistically significant and larger market size effect on FDI are moreinclined to be published in international journals.The second part employs four alternative discrete choice methodologies, including theMixed logit (ML) model and the Latent Class (LC) model approaches to capture themain locational determinants of over a 1000 individual firm-level FDI location decisionsin 13 CEECs between 1997 and 2007. The results show that the choice where abroad toinvest does not only depend on the opportunities offered by foreign markets andindustries but also on investing firms' individual characteristics. These results supportthe presence of heterogeneity in the investment location decisions, which is not onlyrevealed by statistically significant interaction terms, but also by statistically significantstandard deviations of the random parameters in the ML model and statisticallysignificant class-specific explanatory variables in the LC model.