can account for the introduction of behavior constraints in transactions where output constraints predominate. The resulting institutions are contracts, and we can model their benefits and costs. The model is useful in explaining the strengths and limitations of alliances.
Lastly, the model has been described in terms of interdependencies. The reason is that some authors have erroneously thought that the use of the word 'transaction' by transaction cost theorists meant that the model was limited to the analysis of existing transactions, and could not handle the creation of new capabilities. This mistaken view may have arisen from an earlier emphasis on exploitation of advantages. Almost by definition, a firm can only exploit advantages it already owns.
In reality, the process by which MNEs get established is always one of merging complementary capabilities. A firm located in country X has some assets which have potential value in country Y if successfully combined with some country Y factors. These factors can be available on the market, or embedded in some local firm. This is what I mean by 'interdependencies'. Agents located in two different countries have the potential, if they combine their capabilities through international markets or within MNEs, of creating rents. MNEs arise when they offer the most efficient way to realize these potentialities, when they are the most efficient method of combining local and foreign assets. The process of establishing MNEs (i.e. market entry) can thus be seen as one of creating new combinations. Transaction costs theory tries to explain the institutional forms they will take.
Thinking in terms of 'interdependencies' also underlines that when the complementary assets are embedded in firms A and B located in two different countries, the concept of who takes the initiative to combine the assets is irrelevant to the major question of why MNEs exist. MNEs exist because the combination of the assets is more efficiently done within an MNE than through spot markets or contracts, but the initiative can come from either A or B. A firm with a mineral deposit in a given country but no mining expertise can set up a joint venture with a mining firm in another country which has the expertise it lacks, it can sell its deposit to that firm, or it can acquire the mining firm to obtain the expertise. In all three cases, an MNE will result. A theory of why MNEs exist should be able to handle all three cases with the same model, as we can if we think in terms of interdependencies.
Ackerlof, G. (1970). 'The Market for “Lemons”: Qualitative Uncertainty and the Market Mechanism', Quarterly Journal of Economics, 74: 448-500.