The above analysis, however, does allow scope for synthesis. Penrose supplements Hymer in explaining endogenous growth, the internal generation of advantages, and partially the direction of expansion. Hymer discusses pull and push factors for diversification. The various versions of transaction costs analysis explain the choice of mode. Dunning’s OLI addresses the issue of location. Throughout, the pursuit of long-term profit through innovation and/or monopoly restrictions and oligopolistic interaction motivates and shapes decisions and choices. It is worth pointing out that despite widespread recognition of difficulties associated with a ‘general theory’ in the 1990s, an increasing number of leading scholars have pointed out that the desirability and usefulness of synthesizing different pairs of the above pieces of the puzzle.8 This, we believe, supports our case for a synthesis. In addition to the aforementioned, this synthesis should consider labour. Both in terms of the Marxian notion of intra-firm competition between capital and labour,9 now widely acknowledged as the agency issue-see Alchian and Demsetz (1972), Jensen and Meckling (1976), Grossman and Hart (1986)—and in terms of the Marglin and Sugden focus on divide and rule strategies to reduce labour costs. Indeed, a synthesis along the above lines is not only possible but also suggested by the widely used assumption of ‘profit maximization’. Once this is accepted as an aim, it becomes difficult to explain why a firm should stop short of redeploying its whole arsenal to achieve it. This can include the reduction of market transaction costs and also of labour and other costs, and/or the increase of prices through, e.g., the monopolization of markets.10 The ability of a TNC to achieve all the above may, in fact, be viewed as a supply-side reason per se why firms choose to become TNCs (see also Dunning, 1992).