Date of Award

Spring 1-1-2012

Document Type


Degree Name

Doctor of Philosophy (PhD)



First Advisor

James Markusen

Second Advisor

Keith Maskus

Third Advisor

Yongmin Chen


Multinational firms play an important role in the world economy with both international trade and foreign direct investment being fast growing economic activities. Multinational firms need to make a series of decisions before and after they enter the world market. There are three major steps in the decision process relating to the productivity of multinational firms that have triggered a lot of research interest. The first step in the decision process is the relationship between a firm's export decision and its R&D investment (productivity) choice { what type of firm will self-select into the export market. The second step of the decision process is associated with the mode choice { exporting, licensing, or foreign direct investment. The third step of the decision process is how to choose a target firm to implement FDI or licensing in the host country if a parent firm has already made its mode choice. In order to set up a foreign direct investment, the parent firm can choose among greenfield investment, acquiring a more productive host-country firm and acquiring a less productive firm. As to the licensing target choice, the parent firm also can choose to license its technology to a more productive firm or a less productive firm in the host country.

My thesis aims to search for a general method to analyze these three major steps in the decision process and to relate these to the productivity decision at firm level in a single theoretical framework. I also find empirical support for some of the theoretical predictions from the Chilean plant-level data.

Following the first introduction chapter, in the second and third chapters, I set up an oligopolistic Cournot competition model in which heterogeneous firms can choose their productivity levels by adjusting their R&D investments before and after trade. I analyze the relationship between the productivity choice and the exporting decision in chapter 2, and discuss the interaction between the productivity choice and the mode choice in chapter 3. Similar to some recent literature, firms are heterogeneous in their ex-ante productivity. But I generalize this by allowing firms to make investment in R&D, with the marginal product of these investments increasing in their ex-ante cost functions. Productivity is no longer only a determinant of the export decision and mode choice by including this investment choice stage before and after trade. Instead the productivity choice interacts with the export decision and mode choice. After exposure to trade, letting firms adjust their productivity induces a further divergence in their ex-post productivity difference and leads to a stronger intra-industry reallocation effect on quantities, price and profits.

In the fourth chapter, I simplify the model in the previous chapters by _fixing firms' productivity levels to analyze the third step in the decision process. How a parent firm chooses among different alternative host-country targets to implement either FDI or licensing mode choice is a largely unexplored question in the trade literature. The optimal FDI or licensing implementation target in the host country depends on the level of heterogeneity among all firms (including the parent firm and host-country firms) and the profitability in the market. There is a trade-off between the FDI set-up costs and the parent firm's ex-post market share under different FDI implementation choices. Greenfield FDI is preferred when the market profitability is high, while cross-border acquisition of a more productive host-country firm is optimal when the market profitability is low. The parent firm will choose to acquire a less productive host-country firm with a medium-level market profitability. The choice of licensing implementation target is similar, with licensing to a less productive host-country firm if the market profitability is high and licensing to a more productive host-country firm if the market profitability is low.

In the fifth chapter, I use the Chilean plant-level data from 2001 to 2007 to check some of the theoretical predictions in chapter 3. The data shows that firms with access to foreign linkages (either FDI or licensing) are related to higher productivity and larger market share, with foreign subsidiaries even more productive and larger in size than domestic licensees. Empirical support is also found for the theoretical statement that FDI is associated with a larger productivity difference between less productive domestic firms and more productive foreign firms, while licensing is related to a smaller productivity difference.

Included in

Economics Commons