Profitability of foreign direct investment in global cities and CO-ethnic clusters

Dwarka Chakravarty, Paul W. Beamish

Research output: Contribution to journalArticleResearchpeer-review

4 Citations (Scopus)


This paper compares the profitability of foreign direct investment (FDI) in global cities (GCs), their metropolitan areas (metros), and other locations; and examines the impact of co-ethnic and co- industry FDI concentrations. GCs, metros, and clusters offer multinational enterprises (MNEs) a range of economic, institutional, and ecosystem advantages, but may also present substantial cost and competitive challenges. We use a sample comprising 1,832 unique Japanese subsidiaries in North America across 1,263 MNEs over the years 1990-2013. We apply a multi-level longitudinal analysis model and determine spatially significant clusters using geo-coding, proximal distance, and density analysis. We find that subsidiaries in GCs and metros are about twice as likely to be profitable relative to those in other locations. Services subsidiaries in GCs, and manufacturing subsidiaries in metros outperform peers elsewhere. Co-ethnic clusters improve subsidiary profitability in GCs and metros, but not in other locations. Our study responds to calls to examine the performance of FDI in global cities, and to bridge international business research with economic geography. It informs the subsidiary performance literature and the eclectic paradigm on fine-grained location specific advantages; and provides a large sample, longitudinal baseline to aid subsequent theoretical and empirical research.
Original languageEnglish
Journal Academy of Management Proceedings
Issue number1
Publication statusPublished - 2019
Externally publishedYes
Event79th Annual Meeting of the Academy of Management 2019: Understanding the Inclusive Organization - Boston, United States
Duration: 9 Aug 201913 Aug 2019 (Program)


Dive into the research topics of 'Profitability of foreign direct investment in global cities and CO-ethnic clusters'. Together they form a unique fingerprint.

Cite this