Famous? Sure. Influential? You bet. But was Keynes a ‘Key Player’?

What did social networks look like 100 years ago?  In an era without Facebook, Twitter and other social media what did social networks look like and how much can this tell us about how networks operate today?  How can we work out what the effect of networks are on behavior and decisions that individuals make?

In a paper forthcoming in the Economic Record and to be presented this week at the Royal Economic Society Conference in Bristol I explain how a detailed study of the Bloomsbury Group network can tell us a lot about how networks form, change and operate.  Specifically it is possible to estimate what effect the structure of this network has on decisions that members of the Bloomsbury Group made.

Today social networks have a huge impact on our lives.  We influence our friends and they in turn have clear effects on us. This may have dramatic effects on our behavior and decisions. It has been shown that network effects – the possibility that who are friends are – may influence our outcomes – are ubiquitous.  For example, we know that if your friends are obese, then you are more likely to be obese.  The same is true of smoking.  Likewise our academic outcomes are likely to be influenced by the peers who we study with.

Modeling how social networks originate and how they operate may be tremendously important for all sorts of economic, educational, health and social outcomes.  The problem is that we can seldom observe how these networks formed and how they change over time.  One interesting exception – thanks to a large number of diaries, letters and memoirs – is the Bloomsbury Group – the group of London intellectuals which included, Virginia Woolf, John Maynard Keynes, Duncan Grant, Vanessa Bell, Lytton Strachey, E.M. Forster, Vita Sackville-West and many other writers and artists which formed before the First World War.  In the figure below we map the Bloomsbury Group network of connections in 1925.

This research goes on to construct a mathematical model of the Bloomsbury Group and this enables us to analyze the role that the network played in influencing the members decisions.  It also tells us a lot about how networks should be modeled.  The results could have major implications for how we think about all forms of social interactions and their effects.

A major finding of the research is that working out who is central to a network – the key player measure – is crucial to an understanding of how the network works. Specifically, the more central you are to the network in terms of connections the more likely you are to have your outcomes influenced.

The measure of ‘centrality’ in a network is important and determines who is the ‘key player’. This can lead to surprising discoveries. It turns out for instance that Keynes, despite the revolutionary role he was to go on to play in reshaping economics, was not the Key Player in this network in 1905, or in 1925.  This role belonged to Lytton Strachey, although Keynes was the person with the second highest centrality score by 1925.