Markets and Regulation

Stream Director: In-Uck Park

Current Research

Patent Litigation and Takeovers

Funder: ESRC

Staff: In-Uck Park and Andreas Panagopoulos (University of Lancaster)

Commentators argue that a technological gap exists between the United States and the European Union, fuelled by the different levels of protection of intellectual property (IP) rights. In particular, the US courts are perceived to have changed their attitude consistently since the 1980s toward a stronger patent protection, a change that prompted the creation of large patent portfolios by dominant US firms. Given that new innovations increasingly tend to be sequential, ie they build upon previously-patented art; large patent portfolios can have an effect of discouraging future innovation activities by start-up firms, which are often considered as the core source of new ideas.

The objective of this project is to further our understanding of the impact of different levels of IP protection on the long-run innovation dynamics in such environments. In doing so, emphasis will be placed on the role of takeover deals (mutually advantageous out-of-court settlements) in cases of potential infringement, which are a common practise in the US but relatively rare in the EU. Specifically, the project aims (1) to formulate the argument through a mathematical model that can be computer-simulated and, (2) to calibrate the model using EU and US data, which hopefully can help form robust policy recommendations.

Adoption dynamics in network industries: A theoretical investigation under asymmetric information

Funder: ESRC
Staff: In-Uck Park

A good or service exhibits network effects if the benefit to the users increases as more people use/adopt it. In addition to traditional areas with network effects (such as telephone and fax), many additional markets with network effects have evolved in recent years owing to the advances in the digital/information technology. A well-known problem in network is coordination failure: the consumers may be trapped in an equilibrium in which no one adopts, even though there are better equilibria in which a group of consumers adopt (because the beliefs thereof are self-fulfilling). In many real-world situations, the adoption process is dynamic and the agents' preferences are unknown. This research aims to extend our understanding of this issue to such realistic situations, and tries to shed new lights on how dynamic adoption process might resolve the coordination problem via intertemporal coordination, and what kind of dynamic incentive schemes or policy measures might be useful to enhance the efficiency of network markets. Specifically, a game theoretical analysis will be conducted on how the consumers would behave in a model where they have private information on their own preferences and make adoption decisions based on other consumers’ behaviour up to then.

View research on ESRC website

Competition and competition policy - Complete

Funder: Leverhulme
Staff: Paul Grout, In-Uck Park, Sylvia Sonderegger

In a regulated network utilities with technology spillovers, an inherent dilemma of the regulator is that the greater the spillover the greater the desire to attract good entrants to promote competition, but also the harder it is to penalize poor quality entrants because they can free ride due to spillover. A theoretical research project analyzing the effect of incomplete information on the optimal access pricing policy to promote competition, was completed and published this year.

Market organisation - Complete

Funder: Leverhulme
Staff: In-Uck Park

A number of different strands of research have been under way depending on the market specifics. A prominent feature of professions with differentiated specialties is that the client often has to rely on experts' advice to identify the right service provider for each problem they face, as well as on their provision of service itself. Since such client-expert relationships are often recurrent, a question arises as to what kind of consultation behaviour would be optimal for the client to elicit reliable advice from the experts. A dynamic investigation addressing this issue was completed and published this year.

Network Formation and Evolution

(a) Theoretical modelling structure

Staff: In-Uck Park and Silvia Sonderegger

The quest for explaining the co-existence of multiple networks evolved to a little more general approach: what kind of effects does the payoff structure within a network have on the formation and stability of networks? Preliminary results have been obtained that suggest that a single network tends to form if there is increasing returns to scale in network size and all members of a network share the surplus equally. In addition, some groundwork has been done to examine the relationship between unequal payoff structure (according to some characteristics of the members) and the network formation.

On a different level, the sustainability of a network hinges on the members’ perception of other members’ future value to the network, and such perceptions are updated as the members accumulate experiences with one another. In collaboration with Bruno Jullien (University of Toulouse), a new project is under way that examines the evolution of the “reputation” of the members and seeks to characterize the stable dynamics of the evolution.

(b) Strategic Formation of Friendship Networks

Staff: Simon Burgess and Carol Propper

The questionnaire collecting the friendship data is now out in the field and we hope to receive the data in the coming year. Bruce Sacerdote of Dartmouth College, an expert in the economic analysis of friendship networks, is visiting in February.

(c) How network characteristics affect their development and their economic implications

Staff: Paul Grout, In-Uck Park and Sylvia Sonderegger

A Research Assistant has worked throughout the year to do broad literature review and example search in various aspects of network operation, including varying degrees of openness in membership and secretiveness of operation.  As a result, we recognize more clearly the importance of trying to understand different characteristics of networks in relation to the environments in which they arise. From economic perspectives, at least the following two aspects of the environment appear to be of significance: the source of the surplus of the network (e.g., self-generating from economies of scale vs. at the expense of non-members of the network) and the allocation structure of a network’s surplus among its members. A basic analytic framework is being formulated to address these issues in the coming years. Potential collaboration with Massimo Morelli (Columbia University, USA) is currently being explored. 

Interconnection, Competition and Overlap of Social Networks

(a) Interconnection and overlap of networks

Staff: Paul Grout, In-Uck Park and Sylvia Sonderegger

We have carried out preliminary work investigating data that might be used for an analysis of how networks affect board composition and behaviour, and subsequent effects on firm performance and executive remuneration.

We investigated two data sources. First, a database containing information on the schools attended and club memberships of board members of the top 500 companies in the UK stock market. While many executives were members of the same clubs and attended the same schools, we found limited evidence that these school or club networks existed within the boards of individual companies. We concluded that richer information on board members' university education and employment history would be required to provide a more complete picture of both members' expertise and the network connections between them. We therefore investigated a second data source BoardEx which is commercially available. While this provided a much better picture of board members' education and employment, together with information on remuneration and company accounts, at present the data did not cover a very long time span. This would make it difficult to capture changes in the composition of firms' boards over time. We therefore decided to postpone any purchase of the data until it became available over a longer period of time.

(b) Competing friendship networks in schools

Staff: Simon Burgess and In-Uck Park

Analysis of this project is scheduled for the second half of this research programme when the relevant data become available as previously explained in the project (b) under stream 1.

Neighbourhood Networks

(a) When and with whom do the retired form networks – and what are the implications

Staff: Sarah Smith

This project has been pushed back by 6 months and started in September 2008, partly to allow a full year with the same Research Assistant on the project (James Rossiter), and also to allow Ellen Greaves to complete a project on the ethnic composition of schools and the impact of these networks on educational outcomes.

(b) How do these friendship networks relate to educational attainment and criminal or anti-social behaviour?

Staff: Paul Gregg and Carol Propper

This project needs the results of the project (b) of stream 1, hence it will start when these results become available.

(c) Enclaves as Employment Networks

Staff: Simon Burgess, Fredrik Andersson (Census Bureau) and Julia Lane (NORC, Chicago)

We are investigating enclaves of immigrants from different countries in US cities. We ask whether these function as employment networks, and whether they help or hinder an individual’s transition into the US labour market. Either outcome seems possible: it may be that a network helps a new immigrant initially to find a job, but it also may be the case that this slows down the learning of the host language and slows down trading with the host group, reducing long term gains. This is an empirical question that forms the basis of our study. We also plan to study the nature and dynamics of the network itself.

The focus is on individuals immigrating into the US. Here, the networks are residential clusters of people of the same nation of birth. The broad question is about the incentives of individuals to join networks, or to remain separate, and make their own way in the new country and become part of the mainstream or host culture. The analysis is informed by a number of economics literatures, including for example ideas from Lazear on language, and language communities, and around thin or thick markets. We will quantify the relative returns of going in to the network, staying in, and moving out; that is, the incentives in terms of trading internally or moving out. We are also interested subsequently in the formation and evolution of enclaves or networks.

The initial stages of the project have concentrated on preparing the data. The data are extremely well suited to the task, but data preparation is very complex and large scale. The study is based on the Longitudinal Employer Household Dynamics (LEHD) programme ( ) that two of the authors were central in setting up (Lane and Andersson) and which all have great experience of working on.

Initial results so far suggest that enclaves as networks do indeed have implications for employment outcomes. These will be further explored during the second year of the project.

Note that this project refines and replaces two projects in the initial proposal, which are related to ethnic composition in neighbourhoods and workplaces (project (c) of stream 3 and project (b) of stream 4, respectively, in the initial proposal). We made this change as we found out that others were addressing precisely that topic using similar data. The project outlined above is in fact more closely connected to the theme of networks.

We are completing the analysis of employment networks. Our results suggest that there are indeed short run effects of being in an employment enclave.  We interpret these as network effects. The analysis uses a difference-in-difference design, and controls for neighbourhood and ethnic group effects. A discussion paper will be available shortly.

Workplace and Industry Networks

Research in this stream examines network interactions both within firms and between firms and other institutions.

(a) How workplace networks affect career concerns and career paths

Staff: Paul Grout, In-Uck Park and Silvia Sonderegger

The purpose of this project is to address the interaction of social and economic networks in the workplace. As indicated in the grant proposal, we begin by addressing gender issues, in particular, to what extent can network interactions explain ‘glass ceiling’ effects. At this stage a background paper has been written that explains how the glass ceiling phenomena survives competition, in spite of discriminatory treatment of women in managerial positions.  Based on game theoretic analysis under incomplete information, this paper suggests that outcomes where women have to work harder to achieve the same position as men may be ‘natural’ outcomes of competition in some markets.  The analysis shows that the market-place may naturally separate into ‘female-friendly’ and ‘female-unfriendly’ firms, with the former employing more women on comparatively favourable contracts relative to the female-unfriendly firms. This paper has been presented at several places around the world (e.g., Brown University, USA, and University of Bonn, Germany) and came out as a CMPO working paper in September, 2007.

We envisage that the existing model can be extended to include several types of workplace networks. For example, even if the ‘efficient’ equilibrium aims to exclude certain women then it may not be possible to implement this because of anti-discrimination rules. In this case, companies may gain by having leaders that favour networks that meet at inconvenient times and mix social activities with work. Even if the majority of male employees do not find this type of networking attractive it may still be the dominant organisational structure if the losses from exclusion are greater than the disutility suffered by the male employees. In such a case the social network becomes a powerful discriminatory tool in the workplace and may be very difficult to legislate against.

There has been a major revision of the background paper explaining how the glass ceiling phenomenon survives competition. The previous version (CMPO Working paper in September, 2007) contained technical analysis that was viewed as being too complex to convey the underlying insights with clarity. A new version adopts a standard, simultaneous-move model of Bertrand competition as opposed to a continuous-time model of the previous version. The new version, which also incorporates useful comments received from various presentations, is now complete and ready to be submitted to a top economics journal. In addition, Sonderegger has prepared, upon invitation, a summary version of the article for publication in the next issue of Research in Public Policy.

As hinted in the last year’s report, we expect that the analytic framework of the background model can be extended to include several types of workplace networks that may function as an indirect device to filter less career-oriented women in the workplace, e.g., to circumvent anti-discrimination rules. We anticipate that further research activities be extended in this general direction in the coming years.    

(b) Business-university networks

Staff: Laura Abramovsky (IFS) and Helen Simpson

The research has investigated evidence for spatially mediated knowledge transfer from university research to private sector innovative activity. Evidence suggests that academic knowledge benefits firms and that the mechanisms through which non-market knowledge spillovers occur may be localised. Geographic proximity may be crucial if the primary mechanism through which knowledge is transferred is direct personal interactions enabled by social networks. However, physical proximity may be less relevant if knowledge is codified or if tacit knowledge is transferred through well established links, such as formal collaboration agreements or alumni connections.

The research finds evidence that pharmaceutical firms locate their R&D facilities near to frontier chemistry research departments, consistent with accessing localised knowledge spillovers, but that location decisions are also linked to the presence of science parks. In industries such as chemicals and vehicles there is less evidence of immediate co-location with universities, but those innovative firms that do locate near to relevant research departments are more likely to engage in collaborative R&D with universities and to source information from universities to help them innovate. A research paper has now been completed and the research has been presented at an academic seminar and a policy-orientated conference. The project is jointly funded by a grant under the ESRC Impact of Higher Education Institutions on Regional Economies Initiative.

(c) Understanding co-operative innovative activity

Staff: Laura Abramovsky (IFS), Elisabeth Kremp (Banque de France), Alberto López(Universidad Carlos III de Madrid), Tobias Schmidt (Bundesbank), and Helen Simpson

This project investigates co-operative innovative activity in France, Germany, Spain and the UK, using internationally comparable firm-level data for manufacturing and service sectors. It examines the roles of knowledge flows, cost- and risk-sharing and public financial support in firms’ decisions to collaborate. The results suggest that firms which place greater value on external information flows are more likely to co-operate with universities than with other firms, and that firms facing appropriability problems are more likely to co-operate with universities and with upstream and downstream firms than with their direct competitors. Firms that receive public support are more likely to collaborate, and in line with the focus of policy this relationship is strongest for co-operation with universities. This research produced a manuscript entitled “Understanding co-operative innovative activity: evidence from four European countries.”  This project is jointly funded by the European Commission and the Gatsby Charitable Trust.

This project is now completed and the research paper has been accepted for publication in the journal The Economics of Innovation and New Technology. The research investigated co-operative innovative activity in four major European countries using internationally comparable firm-level data. The results suggest that firms which place greater value on external information flows are more likely to co-operate with universities than with other firms, and that firms facing appropriability problems are more likely to co-operate with universities and with upstream and downstream firms than with direct competitors. It also finds that receipt of public financial support is positively related to undertaking collaborative innovation, in particular with universities. This project is jointly funded by the European Commission and the Gatsby Charitable Trust.

(d) Industry localisation, cities and knowledge spillovers

Staff: Helen Simpson

This research will be continued in the latter part of 2008. The research examines the extent of geographic industrial clustering in the UK, and whether cities facilitate knowledge flows between firms. The research extends existing evidence on industry localisation to service sectors. It finds that many of the most geographically concentrated industries in the UK are in service sectors such as financial services and the film industry, and that firms in these industries tend to be located in major cities. Models of urban growth suggest that economic density may facilitate face-to-face networking, which may be beneficial in innovative and creative industries. The project examines the relationship between the degree to which firms benefit from knowledge flows, and location in urban and more educated areas.