Network Dynamics and Market Structure: The Case of Art Fairs

Abstract

Annual art fairs are a crucial element of the contemporary art market. Art galleries represent artists and exhibit their work at fairs all over the world. These fairs expose the gallery and its artists and create and maintain relationships with key actors in the global art market. Using data recording artists and galleries presenting at art fairs alongside qualitative interviews with different actors operating in the art market, we study the development of the art fair network over three years, 2005–2007. Our findings suggest significant network and homophily effects regarding the status and age of artists at fairs. However, the internationality of fairs and geographic distances between them do not seem to have any effect.

Acknowledgments

A previous version of this paper was presented in the Conference on Applications of Social Network Analysis (ASNA) in Zurich, Switzerland in 2008. For comments on earlier drafts of this paper we would like to thank Tom Snijders, Peter Hedström, Johan Koskinen, and seminar participants at the Nuffield-OII Networks Seminar.

Notes

¹Unfortunately, we do not have information about the number of applications and rejections of galleries wishing to participate in an art fair. Fair organizers were reluctant to publicly expose this aspect of fairs.

²It is noteworthy that this does not compromise the argument that the art market in general, and the art fairs system in particular, are western-dominated (Quemin 2006). The SIENA analysis looks at change of a network; it does not look at what is dominant in a network. More precisely, it looks at what is dominant in explaining the change that we observe. Thus, the internationality argument here means that the fact that two art fairs are similar in terms of how international they are at time t does not make them more likely to have galleries attending these two fairs at time t + 1.

³The distribution of the quality of galleries attending fairs is highly skewed. Therefore, using an upper percentile seemed most reasonable to assess the quality of a fair. We repeated our analyses using the top 10, 30, and 50 percent (median) of the artists' points. The results remained the same.

⁴We repeated our analysis with a link being formed when one or three galleries are shared by two fairs. The results were similar. We chose five galleries as a threshold to exclude more random connections.

⁵We use information from 2005 to 2007 to derive attributes for the fairs, yielding a more stable evaluation of fairs.

⁶Such a strategy, that calculates a status score for the galleries first and then uses these scores to assess the status of the fairs as a whole, seemed reasonable as essentially exposition space is given to the galleries (booths) and not to the artists per se.

⁷This measure is preferred over the simple number of different countries represented at a fair, because some fairs are larger than others. A Herfindahl-Hirschman concentration index takes this implicitly into account and then shows how much a fair is dominated by galleries from particular countries.

⁸There is no case in which all galleries at a fair are from the same country.

⁹This assumption can be relaxed with the appropriate specification of independent variables.

¹⁰For a full list of effects implemented in the SIENA software package, see Snijders et al. (2007).

¹¹For further details about the method we refer to Snijders (2001; 2005), Snijders and van Duijn (1997), van de Bunt et al. (1999). Non-technical introductions to the method are given by Steglich et al. (2006) and Snijders et al. (2010).

¹²Using the degree of a node to control for its size is common practice. However, we also used the pure number of galleries presenting at a fair as a control variable. The results remained the same.

¹³We used a simple Hirschman–Herfindahl index (see Herfindahl 1959 and Hirschman 1964).