Reframing Bowling Alone
Reframing Bowling Alone
connecting the dots with network theory
Recently, my neighbor Daniel Golliher, author of Maximum New York, proposed a book club. I decided to participate, so I’ve written my reflections on the revised 2020 edition of Robert Putnam's Bowling Alone.
Bowling Alone presents a compelling narrative discussing the decline of social capital in America and supports its case with in-depth analysis of the factors that may have contributed to its decline while endeavoring to compensate for confounding factors and control for spurious correlations.
Social Capital and Its Decline
In Putnam’s framework, social capital refers to the networks of relationships among people who live and work in a society, enabling it to function effectively. Putnam posits that social capital can be broken down into two main categories:
Bonding capital refers to connections within homogeneous groups fostering a sense of community and shared identity.
Bridging capital refers to connections between heterogeneous groups. The theory proposes that bridging capital can link people to broader opportunities and resources outside their immediate social circles.
Proponents argue that social capital can reduce friction in social interactions and daily life and suggest that higher levels of social capital might correlate with more efficient societies, as trust and established networks could facilitate smoother transactions and collaborations.
As per the emblematic chart above, Bowling Alone argues that social capital in the United States has been declining since the 1950s by presenting decreases in a wide set of indicators of social capital, including civic engagement, political participation, religious involvement, workplace connections, informal social ties, philanthropy, and trust in institutions and other individuals. Putnum provides numerous charts depicting declining participation across a variety of institutions and attributes this to a combination of factors, including increased time spent watching television, two-career families, suburban sprawl, and generational change. He contends that this erosion of social capital has significant implications for community well-being, democratic participation, and overall societal health.
Putnam's approach seems robust. He takes a comprehensive look at the potential causes behind the decline of social capital and provides a well-reasoned rationalization for whether or not he thinks each meaningfully contributed. For example, in evaluating a factor, he might look at trends in comparable countries to determine if the same factor is present there and whether the magnitude of the effect was similar. While Putnam put range estimates on each of the factors he found likely to be responsible as shown in the image below, he left room for unexplained other factors.
Putnam's argument sparked considerable debate and inspired further research into the state of social connections. While his ideas are influential, the mechanisms and applicability of social capital is the subject of ongoing debate. As with many sociological concepts, social capital evades straightforward measurement and quantification.
As a professional forecaster, I found Putnam's approach well-considered and worthy of discussion, and, in my personal judgment, probably directionally accurate. However, with observational evidence, even with extensive controls, it’s hard to draw conclusions with high confidence.
While Putnam's work is thought-provoking, and there are many interesting arguments over the relative importance of the various factors he discussed, I don't want to focus on reanalyzing them since that has been done elsewhere. Instead, I want to look into the unexplained portion of the precipitous declines observed in many of the social and community organizations the book addresses. In the rest of this post, I’d like to explore how network effects might explain an important part of that mystery.
The Power of Connections: Network Effects
Network effects occur when the value or impact of a product or service increases as more people use it. Think of how much more useful a telephone is when the person you want to call also has a telephone. A phone wouldn’t be very useful if you were the only one that had one, or if the only other person who had one was someone you didn’t care much to talk to, but the more people that have them, the more valuable they become. In the context of social capital, we can think of community organizations, civic groups, or even informal social circles as networks whose value grows with increased participation.
One way to quantify network effects is through Metcalfe's Law, which states that the value of a network is proportional to the square of the number of connected users1. Even if the value of a social network doesn’t scale exponentially with the number of users, it seems reasonable to expect that many people would find larger organizations, at least up to a point, to be more valuable since they offer more opportunities to establish both bonding and bridging capital.
Presuming that, for most participants, participating in an organization depends on the cost-benefit balance, in the unfortunate event where a network loses a participant, perhaps due to moving, even if nothing else changes, it could lead to the next marginal member leaving as well as the decreasing value of the network moves their participation from net-positive to net-negative. This could lead to a snowball effect where more and more participants leave. Once enough people stop participating, there is no escaping the vicious cycle.
Combine this concept with rising opportunity-costs such as better at-home entertainment or longer commute times, and we can see a promising explanation for why the decline in organizational participation might have accelerated beyond what Putnam’s factors alone would predict.
This compounding effect could explain why Putnam observed such steep declines in participation across various social organizations. It's not just that people are choosing not to join these groups; it's that the groups themselves become less attractive as they shrink, creating a self-reinforcing cycle of decline.
To better demonstrate this idea, I created a simulation of a community center with a capacity of 200 members where every potential member is randomly assigned a benefit value that they get from the community along with an opportunity-cost. As long as the benefit is greater than the cost, and as long as there is room, the person becomes and stays a member (I also set some other parameters like having the value they get from membership increase slightly over time as they grow more deeply engaged, a random chance of quitting, a maximum monthly join and leave rate, and other variables, but those aren’t important to the overall narrative and you can play with the assumptions if you want to build on my simulation). However, if the cost exceeds the benefit, the member quits.
The simulation starts out with 150 members in 1940. Membership as a percentage of capacity is represented by the blue line, the calculated value of the community (as determined by the network effect, but maxing out in value after reaching saturation at 150 participants) is represented by the orange line. Finally, the baseline societal opportunity-cost is represented by the green line. In my simulation, I added two step-changes in opportunity cost, one after the wide-spread adoption of television and the second after the advent of the internet.
As the simulation starts out, new members rapidly join filling the community cent to its maximum capacity. This remains very stable until the scaled adoption of TV after which some members quit because they find staying home is a better alternative, but they are quickly replaced by other community members who put a higher value on what they get from the community. The network remains relatively stable until the arrival of the internet. For a few years, the value of the network remains high and new members continue to join as other members leave, but as more and more people stay home, the network quickly reaches a point of no return.
Understanding this dynamic is crucial for efforts to rebuild social capital. It suggests that simply addressing individual factors like time constraints or competing entertainment options may not be enough. Instead, strategies to revitalize community engagement might need to focus on reaching a critical mass of participation quickly, leveraging technology to reduce barriers to entry, and creating clear, immediate value propositions for potential members to overcome the inertia of non-participation.
In essence, the power of network effects works both ways – it can amplify growth in social capital when conditions are favorable, but it can also accelerate its decline when various factors align against it. Recognizing this can help us better understand the dramatic shifts in social engagement that Putnam documented and inform more effective approaches to fostering community connections in the future.
More precisely n(n - 1)/2 or about half of the square at the limit.