83rd International Atlantic Economic Conference

March 22 - 25, 2017 | Berlin, Germany

The case of missing foreign affiliates' investing behavior: Why it should matter for FDI policy and empirics

Friday, 24 March 2017: 15:30
Alexander Amati, Ph.D. , Finance and Economics, Rutgers Business School, Piscataway, NJ
To our knowledge, the issue of post-entry investing vs. non-investing multinationals takes up virtually no space in the literature on foreign direct investments (FDI), empirical or otherwise. This paper presents a theoretical discussion of why this issue matters not only for policy but also for FDI empirics, followed by an empirical investigation.

We model the foreignness of a direct investment from the perspective of a policy maker in a receiving small economy interested both in the initial FDI and in the recurring FDI, which derives from post-entry investing by the foreign affiliate. Foreignness of the investment refers, jointly, to foreign ownership of productive assets (giving rise to foreign rights to cash flow) and foreign control of post-entry investing decisions (foreign decision rights). The emphasis on the latter characteristic of these investments is the main originality of the paper.

Framing the issue in a principal-agent model, it is shown that, given any initial arrangement on the allocation of cash flow between the host economy and the multinational, the investment path of the latter is suboptimal in a Pareto sense. This result is interpreted as recognizing post-entry investing behavior of the foreign affiliate as a key variable in establishing FDI policy and expectations regarding, for example, subsidy, profit repatriation, or technology spillover.

Addressing the latter, we theorize, within the relevant theory (i.e. Hymer-Dunning ownership-location-internalization (OLI)), why investing foreign affiliates, but not non-investing ones, generate spillovers. Using a large plant-level panel data set on manufacturing plants in Chile, we implement quantile regression (Koenker and Bassett (1978)) estimation, controlling for plant fixed effects (a la Powell and Wagner (2010)). Our results corroborate our theoretical discussion of the relevance of foreign affiliates' investing behavior: gains in local productivity are associated with the presence of foreign plants engaging in recurrent investing, but not with the presence of non-investing foreign plants.