The main idea of this book is to present contemporary understanding of human networks, how they are formed, behave, and facilitate relationships between people, including information dissipation, contagion, and power distribution between nodes. Lots of attention assigned to homophily – human tendency to attract to similar people as self and repulse dissimilar, and how it lead to polarization between groups of people. There is also discussion of “wisdom of crowds”, which depends on quality and diversity of the crowd as network, and intergenerational income mobility as result of maintenance of family networks across generations. Finally author presents his attitude to globalization as dramatic change in networks with some very significant consequences that may or may not be beneficial or dangerous.
- Introduction: Networks and Human Behavior
It starts with reference to the beginning of Arab Spring and the role of human networks in it. Author aims to discuss: “two different perspectives: one is how networks form and why they exhibit certain key patterns, and the other is how those patterns determine our power, opinions, opportunities, behaviors, and accomplishments.”
After that author provides examples of human networks in school, graphically demonstrating how networks split:
2. Power and Influence: Central Positions in Networks
In this chapter author discusses various position in networks, stressing importance of central position. Here is another graph demonstrating this point:
Author then discusses issue of centrality of individual in network and how it creates or undermines power. He provides a nice historical example when superior network led to the victory in power struggle:
3. Diffusion and Contagion
This chapter is about diffusion and contagion between different nodes in network. It starts with discussion of diffusion of plague and sexual diseases in networks of medieval Europe and contemporary school. Then author moves to vaccination and similar externalities that could impact the contagion process. He also discusses such measures as quarantine and their deficiencies.
4. Too Connected to Fail: Financial Networks
Here author expands these ideas to financial network discussing how failure of some financial institutions prompted failure of others during financial crisis of 2008. Here is graphic representation of the idea that distributed network is more stable than one that relies on a few core institutions:
Author then discusses regulation and externalities and poses the question whether crisis was like domino with pieces causing the fall of each other or like popcorn when conditions of market caused the individual pieces to pop up at approximately the same time, even if they were independent from each other.
5.Homophily: Houses Divided
Homophily here means love for people like self with rejection of people unlike self. Author uses Indian caste system to discuss separation of one village network sub-networks by caste. After that he discusses process of self-segregation and Shelling’s model of this process. At the end of chapter author discusses negative impact of segregation levels on overall productivity of society as expressed by GDP per capita,
6. Immobility and Inequality: Network Feedback and Poverty Traps
Here author discusses intergenerational income mobility in inequality using Gini and demonstrating how it changed over time:
He also looks at reasons for these changes that he finds in educational levels, changes in structure of economics when old manufacturing jobs disappeared, and role of job networks, which facilitate acquisition of a preferable place in network for next generation of its members. The chapter ends with call to fight homophily and inequality by increasing transparency of opportunities, providing education so people were qualified to move up and by removing barriers to such mobility.
7. The Wisdom and Folly of the Crowd
This is usual discussion of the wisdom of the crowd and need for diversity that is necessary for such wisdom to occur. Then author discusses polarization of news and political discourse in America, nicely supporting it by comparative diagram of voting patterns in US Senate in 1990 and 2015. The former shows relatively high number of votes cross party line, while the latter much lower number, making contrast very vivid:
8. The Influence of Our Friends and Our Local Network Structures
This chapter is about human behavior in crowds and the tendency of people to conform to majority. Author discusses different behavior of ants and lemmings, then providing examples of human behavior under carefully designed nudging, and that quite often achieves significant results comparatively with control groups. Author also discusses here clustering when connected nodes of network are linked to each other via multiple connection such as friends of a person are also friends between themselves. Here is a nice graphic presentation of this with and without clustering:
9. Globalization: Our Changing Networks
The final chapter is a look at contemporary world with extensive globalization when massive trade networks growth coincided with decreases in wars and increases in global GDP. To demonstrate increase of networking between countries author provides graph of military alliances from 1875 till now:
However the final thoughts are about disruptions, especially massive urbanization that dramatically increased density of networks, but also created conditions for growing impact of homophily and resulting polarization, which potentially could lead to explosion. Author ends by calling to develop better understanding of human networks and externalities so humanity could avoid such explosion and continue increase in productivity that linked to increase in networking.
MY TAKE ON IT:
It’s a nice book with decent set of facts and experiments description that provides more or less good picture of current understanding of human networks, their functionality, and impact on human relations and power distribution. Generally I think that presentation is correct, but it is somewhat minimalistic on the role of human individuals and their self-understanding and formation that has huge impact on functioning of networks. This leads to overuse of homophily as explanatory method for behavior of both networks overall and their human nodes. I believe that this is not exactly correct, and such thing as political polarization comes not from individual attraction / repulsion, but more for commonality of interest as they are perceived by individuals.