Social networks are pervasive. Diverse examples range from friendship groups, neighbourhoods, and the decisions of companies over whom they conduct business with, to more formal networks such as workplaces, political groupings, local voluntary organisations, and international trade organisations. Analysing the formation and evolution of networks is complex and important.

* It is complex because there are two inter-related processes taking place. On one hand, the value any one individual or company gets from belonging to the network, their commitment to it, and the ethos and effectiveness of the network depend crucially on other members of the network. But, at the same time, these factors also influence how the membership changes over time. This complex relationship and feedback provides a rich diversity of potential outcomes and dynamics, not all of which are economically or socially beneficial.

* It is important because networks have important effects on educational outcomes, on employment levels, on social unrest, on how industries grow and decline, etc., and ultimately, through these effects, on the strength of the economy and society. Social networks have been widely studied in a sociological context, but it is only in recent years that the study of such networks has started to affect the way that we think of economic behaviour and the dynamic processes in society and economies. Both theoretical and empirical work in this area is difficult but recent developments in theory and newly available datasets make this possible. This is a new and exciting research agenda, which can provide insights into the importance of collective behaviour and collective norms for economic outcomes, and therefore on the growth of the economy. CMPO has already contributed to this agenda through its work in two distinct areas - corporate culture and neighbourhoods - and, building on this work, we are in a position to make a major contribution to this new field. It is clear from our work and our reading of the other research in this area that this emerging field of study could be pushed forward by a two pronged approach that undertakes (a) theoretical analysis that recognises that the existing

apparently diverse research is really one big issue, and (b) empirical analysis of the development, dynamics and consequences of critical social networks.

Our aim is to undertake a major study with two core, linked, components:

* Theoretical analysis of economic and social networks and communities, and of how different networks interact. The objective is to provide a global analysis that can be used to give deeper insight into specific types of networks. This will build on existing work that is currently based on three distinct but interrelated themes: (i) individual actions based on complementarities, (ii) group formation to achieve group objectives, and (iii) research that focuses on linkages between members.

* A series of econometric studies on large datasets and case studies to test the theories in specific contexts. The topics include (i) the long run implications of patent commons (open-sourced communities where firms share patents), (ii) the transition at retirement from work-based networks to social-based networks, (iii) how workplace networks affect career concerns and career paths, (iv) what is the impact of neighbourhood and peer groups in school on educational outcomes, (v) what is the impact of networks and neighbourhood on productivity and employment, (vi) when and with whom do lasting friendships form among teenagers and how these friendship networks influence educational attainment and criminal or anti-social behaviour.These projects will build on our experience of developing theory alongside empirical analysis, and exploit the skills and the large datasets we have built up over the past 7 years.