The missing factor: entrepreneurial networks, enterprises and economic growth in Ghana

In an earlier paper (Barr (1994)) I set out a theoretical model describing the role of entrepreneurial networks in both enterprises and the process of economic growth. In this paper I reformulate the model in such a way that it can be estimated using cross-sectional data from Ghana. The results of the analysis suggest that entrepreneurial networks are an input into the production process and possibly source of external economies of scale and economic growth. In the theoretical model entrepreneurial networks facilitate the process of economic growth by helping the entrepreneur capture of knowledge externalities. There is a pool of these externalities, which are productive, non-rival, free and potentially available to all firms. They may be used in conjunction with the rival factors, capital and labour, in the process of production. Where a factor of production is rival, the market determines how it is to be distributed between firms. Where a factor of production is non-rival and non- or only partially excludable, i.e. where the factor is an externality, the market cannot operate as the distributing mechanism. Faced with this market failure, economists interested in growth usually assume that the entire stock of knowledge is available to each firm. Here, as in my earlier paper, I propose that although the entire stock of knowledge is potentially available to each firm, the amount that is actually available is significantly less. In the theoretical model the extent to which an entrepreneur can access or harness the overall stock of knowledge for use in her firm’s production process is determined by the density, size and functionality of the network that connects all the entrepreneurs in the system. In the context of empirical work we must allow for asymmetry in the network, so the position of the agent within that network is also important. On a more general level the paper is one of few formal tests of the hypothesis that the structure of society is important in determining economic outcomes. Several researchers have suggested that the key to understanding the so called ‘East Asian miracle’ is to look at the structure of society and not only dynamic factors such as investment in physical and human capital. Indeed, the World Bank (1993) was able to explain only 36% of the difference in growth performance between Africa and the high performing Asian economies (HPAEs) with reference to these more traditional factors. The literature on this topic includes contributions from several disciplines. Economic sociologists, such as Biggart and Hamilton (1992), argue that the standard neoclassical model is not appropriate for the analysis of ‘Asia’s network capitalism’ (page 472) and that ‘persuasive explanations for the success of Asian business will ultimately come from an institutional analysis of Asian societies and the economies embedded in them.’ (page 488). Coming from a business orientation, Porter (1990) writes that ‘the principal function of the keiretsu (groups of companies affiliated by shareholdings) and the shita-uke structures is to facilitate interchange among related companies (the role of the keiretsu in strategy formulation and financing is greatly overstated in most Western accounts). Companies loosely linked in Japanese groups look to each other for guidance and input on new products, new processes, and new businesses. Japanese trade associations also promote the links between suppliers and buyers, by collecting information and sponsoring studies’ (page 408). Although this paper does not look directly at the HPAEs and the source of their success it does provide an analysis of the interplay between social structure and firm performance. The paper is arranged in five sections. Following the introduction, in Section 2 I develop the empirical formulation of the model. In Section 3 I explain how the entrepreneurial networks in Ghana were measured and provide a description of the structure of the Ghanaian manufacturing community based on the data collected. Section 4 contains the results of the regression analysis and a discussion of their significance for endogenous growth theory. I conclude in Section 5.