I just finished reading Why the Garden Club Couldn’t Save Youngstown: The Transformation of the Rust Belt by Sean Safford. It’s a very good book. Safford compares the histories of Youngstown, Ohio and Allentown, Pennsylvania in an effort to discover why Allentown has been able to recover economically from the collapse of its primary industry (steel), while Youngstown hasn’t. He starts by tracking and comparing the social and economic histories of the two cities – which shows that differences in their social structures at the time of founding (200 years ago!) led to differences in social structures now, and that these differences explain a fair bit of their current situations. The main story is that the business elites in Youngstown formed a fairly closed circle, which was recreated across a number of social settings, while in Allentown there is much more of a history of cooperation across classes leading to effective collective action.

The story concludes with a very nice piece of network analysis, which supports this story. Safford looks at the interlocking board membership networks in both cities in 1950 and 1977 (both sample times precede instances of substantial economic change). He includes information about civic groups that are oriented around both business and community. What he finds is that in Youngstown the membership of business and community boards overlaps almost completely, while in Allentown these form two quite different networks. He concludes that this supports his earlier story – that in Youngstown the business elites faced a time of crisis and responded by mostly just talking to themselves, while in Allentown the business leaders encountered a broader range of opinion, which led to a better response to the economic changes.

This short summary obviously doesn’t do the book justice, and it’s definitely worth checking out yourself. It is a beautifully written book (even though it includes academic data, it reads like the sort of thing you could find in The New Yorker). It is also a very good example of how to put together a coherent research project, which PhD students can use as a model. The reviews of the book have been very positive. One thing that has struck me is that many people are pointing to the novelty of Safford’s use of network analysis.

While I think his combination of network analysis and more traditional sociological techniques is very well conceived and executed, I also think this is a bit of a strawman argument. There are obviously some prominent streams of network research that treat network structure as the only driver of action within the network (mostly coming from the physics-based literature). But I think that most people that are using network analysis in the social sciences (even some econophysicists!) acknowledge that agency is important, and try to take into account the impact that people and their relationships have on the formation and maintenance of networks. The fantastic work of Pip Pattison, Garry Robins and their research group is a good example of researchers that take these issues very seriously. Even in work that is more structurally based, agent-level action is taken pretty seriously – I was struck recently by parallel findings concerning the interplay between network structure and individual attributes in research by Andrew Stephen and Oliver Toubia (see their paper Explaining the Power-Law Degree Distribution in a Social Commerce Network) and some of my findings concerning the international trade network (while that particular piece of research doesn’t have much room for agency, the point is that the interplay between structure and actor is there, which is something I’m taking very seriously in my current work). And in our research group, Sam MacAulay, John Chen and Marco Fahmi are all doing studies that take the human side of networks very much into account.

While that angle isn’t unique, this doesn’t take away from the value of Safford’s book. It is a very well thought out piece of research, and he tells the story really nicely. It’s definitely worth a read.