The papers already published in Teichova, Lévy-Leboyer and Nussbaum (eds.), Multinational enterprise in historical perspective (1986), and those now gathered in this volume, are too varied, wide-ranging and rich in fact and analysis to summarise. Perhaps I might, however, be allowed to comment on three issues that strike me as not without intellectual importance. I have in mind the role of technology in the multinational enterprise, the variety of forms of doing international business, and the fact that the choice among exporting, licensing technology, forming cartels, entering contracts, joint ventures, and forming a wholly-owned foreign subsidiary is often a very close one.

As Fieldhouse's useful summary reminds us (1986), we learn from Hymer that business operating away from its home base is always at a disadvantage, because of long lines of communication, operations in a culture which varies in degrees from that at home, at least slightly and perhaps very widely, and the agency problem that perhaps applies to all business and is usefully discussed by Nicholas (1986) – the difficulty at a distance of ensuring that the agent does what the principal has contracted with him to do, rather than operating in his own differing interest. To overcome these disadvantages, such a business must have a large advantage. Chandler (1986) finds this in large size, economies of scale, direct marketing rather than reliance on middlemen, and capital intensity, plus – a point bearing on technology – attention to research and development. But these advantages, it seems to me, start late rather than early in the course of foreign direct investment.