Volume 19 · Number 3 · July 2018 economic econsoc.mpifg.de sociology _the european electronic newsletter 19.3 Note from the editor Content Networks for 1 Note from the editor Networks for economic sociology (and not the other way around) economic sociology by Olivier Godechot 4 Is social network analysis useful for (and not the other studying the family economy? by Céline Bessière and Sibylle Gollac way around) 11 Networks of Corporate Ancestry Olivier Godechot by Lasse Folke Henriksen et al. 19 Embeddedness and decoupling in innovation activities by Michel Grossetti ore than thirty years works are not only the warm social 25 A tale of two cities: the regional ago, Granovetter (1985) glue of kinship and friendship ties dimension of the Ecuadorian securities launched a research underlying social cohesion. They market M by Andrés Chiriboga-Tejada program for economic sociology also act as information processors, which rested to a large extent on and they enable not only the suc- 36 Neo-structural economic sociology networks as both conceptual and cess of some individuals in specific beyond embeddedness methodological tools. In order to network positions, but they also by ORIO Network understand concrete economic ex- improve global welfare beyond change, he not only argued in favor them. However, social networks’ 50 Book Reviews of a third way between under- and contribution to the economy is not over-socialized views of the econ- only positive. They also produce or Editor omy; he also proposed a path that fuel many inequality-generating Olivier Godechot, Sciences Po, differed from Williamson’s (1985) mechanisms. Laboratory experi- CNRS, and MaxPo, Paris articulation of market arm’s length ments have thus shown that actors Book Reviews Editor ties and hierarchical subordina- dependent on a limited set of con- Lisa Suckert, Max Planck Institute tion. Network embeddedness was tacts to access key resources tend for the Study of Societies thus the solution for understand- to accept unfavorable terms of ex- ing concrete patterns of both mar- change (Cook and Emerson, 1978), Editorial Board Patrik Aspers, Uppsala University; Jens ket and organizational life. thereby contributing to the power Beckert, Max Planck Institute for the Study Early work in economic so- of structural holes (Burt, 1992). of Societies, Cologne; Johan Heilbron, ciology insisted first on the infor- Opportunity hoarding (Tilly, 1998) Centre de sociologie européenne, Paris; mational dimension of networks is based not only on categorical as- Richard Swedberg, Cornell University, Ithaca (Granovetter, 1973). Social net- signation and identification, but economic sociology_the european electronic newsletter Volume 19 · Number 3 · July 2018 Note from the editor: Networks for economic sociology (and not the other way around) by Olivier Godechot 2 also owes much to homophilic relations, contributing times, comparatively little energy has been devoted to to discrimination and inner circles phenomena. Mech- analyzing the underlying social mechanisms that are anisms of asymmetric social comparisons following to be modeled. Some authors tend to apply the stan- the pattern of concrete social networks fuel relative dards of the network science field and forget that the frustration (Fligstein et al., 2017) and contribute in re- meaning of a given measure (centrality, transitivity, turn to hierarchization (De Vaan et al., 2018). etc.) in one social setting might have little to do with Networks therefore appear as a promise for a its meaning in another social setting. The network is deep and sound understanding of economic life, rang- reified, and we tend to forget that the coded network is ing from micro-interactions of actors in markets at best a very crude proxy of the underlying social re- (Baker, 1984) or organizations (Burt, 2004) to solidi- lations. fied chains of suppliers (Gereffi et al., 2005). As such, Conversely, many inspiring contributions in they became an important branch of economic sociol- economic sociology only use rough and simple net- ogy for the last forty years. However, for a long time, work measures. However, they innovate in forging re- scientific results did not always meet the expectations. lational mechanisms and finding simple network Informal networks underlying economic activity leave proxies for testing them. Hence, Granovetter’s approx- few traces, and social scientists lacked sufficient data imation and test of weak ties (1973) was very rudi- to make breakthrough contributions. Moreover, in ar- mentary: “Of those finding a job through contacts, eas where social networks were easier to collect, such 16.7% reported that they saw their contact often at the as US board interlocks, results have long been quite time, 55.6% said occasionally, and 27.8% rarely disappointing (Mizruchi, 1996). Economic sociolo- (N=54).” Padgett and Ansell (1992) proved that the Ol- gists might also have been discouraged by the high igarch-Medici divide was network-based rather than level of technicality in this subfield, with its large num- status-based using four pivot tables and one graph. ber of metrics (Wasserman and Faust, 1994) – be they The more complex block-modeling played little role in measures of centrality or techniques of clustering – or the paper. Uzzi (1996) implemented the concept of by the inherent complexity of its econometric models embeddedness with a simple “first order network cou- (Cranmer and Desmarais, 2011), including QAP re- pling” index that captures the concentration of trade gressions, ERGM, TERGM, or SAOM models. among business partners. More recently, Wilmers Indeed, the incursion of economic sociology (2018) gave empirical content to the notion of captive into the study of networks led this branch to integrate value chains (Gereffi et al., 2005) and showed how they the concepts, tools, and methods of network science – decreased workers’ power. To achieve this aim, he an interdisciplinary scientific field at the frontier of looked at how workers’ pay in supplier firms declined mathematics, physics, computer science, and social with the existence of dominant corporate buyers. sciences. While this encounter is very welcome, en- This reminder is not a rejection of sophisticated abling economic sociology to use more reliable and network measures and models. Moreover, simple robust tools and to avoid networks’ tricky artifacts, it crude measures are often quite complex to implement, also comes with some risks. Rather than using the net- simply because they are not provided in standard net- work techniques as a tool for proxying thoughtfully work software packages. This editorial tries to remind coined economic relations, economic sociology could the reader that concepts of economic activity should instead only provide empirical data for testing con- determine the choice of the network measure rather cepts coming from network science. This last objective than the reverse. is of course perfectly legitimate, but it does more to Following this line of thinking, the current issue help network scientists understand networks than it of economic sociology_the european electronic newslet- does to help economic sociologists understand eco- ter shows that networks are still a major tool for the nomic activities. understanding of economic activity, provided that Hence, those of us who sometimes venture into they are subordinated to economic sociology’s theo- network workshops – for instance, the excellent retical agenda. INSNA Sunbelt annual conference – might have been Céline Bessière and Sibylle Gollac open this is- struck by a sense of “déjà-vu” in many presentations. sue with a very inspiring reminder. Families are a) These often include the display of a spaghetti bowl economic units and b) a complex bundle of differenti- graph; a listing of the most central actors; the delimi- ated relations. The ethnographic analysis of family ex- tation of network clusters thanks to a given block- change networks therefore uncovers a householding modeling technique; and the use of an ERGM type of phenomenon which goes beyond the taken-for- regression for estimating many network effects pa- granted frontier of “households.” rameters, including transitivity, k-stars, popularity, as- Also inspired by the mechanisms of family rela- sortativity effects, etc. This is fun and fine. But some- tions, Lasse Folke Henriksen, Anton Grau Larsen, economic sociology_the european electronic newsletter Volume 19 · Number 3 · July 2018 Note from the editor: Networks for economic sociology (and not the other way around) by Olivier Godechot 3 Christoph Houman Ellersgaard, and Jacob Lunding Andrés Chiriboga studies the structure of the propose a very intriguing innovation for the study of exchange between brokers in the Ecuadorian stock ex- corporate networks. They analyze the appointment of change and suggests that the geographical split of the executives by chairmen as a form of genealogical suc- country around two centers, Quito and Guayaquil, is a cession. This enables them to establish a typology of major factor in the clustering of economic transac- “patrilineage” structures at the head of Danish firms. tions and could hamper the development of an inte- Michel Grossetti presents the notion of “decou- grated modern financial market. pling,” when a given network tie between actors ac- Finally, Emmanuel Lazega’s contribution with quires an existence beyond the two actors that initi- Julien Brailly, Catherine Comet, Sébastien Delarre, Fa- ated it, thereby becoming a frame of reference