Complexity, Networks and Knowledge Flow

Abstract

The diffusion of knowledge plays an important role in both the creation and distribution of wealth. Attempting to construct a competitive advantage, managers of firms often seek to spread valuable knowledge widely across subunits of the organization, while simultaneously preventing its diffusion outside the firm to rivals. In the interest of stimulating local economies, regional planners and leaders of technological communities may also aim for an uneven flow of knowledge: rapid diffusion within geopolitical or community boundaries, but not beyond them. We argue that the degree of information inequality across social boundaries will reach its peak when moderately complex knowledge underlies these differences. Such knowledge generates the greatest disparities across boundaries because preferential access to private information - established through the relatively more dense social networks within these borders - provides the greatest advantage in assimilating and building on knowledge when that information involves an intermediate degree of interdependence. To test this proposition, we examine patent data, comparing citation rates across three types of social boundaries. Our findings establish that knowledge complexity moderates the proportion of future citations: (i) within versus outside the firm, (ii) geographically proximate to, versus distant from, the inventor, and (iii) internal versus external to the technological class - thereby supporting the idea that moderate knowledge complexity drives a wedge between the transfer of information internally and its diffusion across social boundaries.