More connection, less community: network formation and local public goods provision
BibTeX
N/A. Working paper.
Abstract
This paper examines how social networks affect the provision of public goods within a com- munity. Here, networks spread information about whether people contribute to a public good,
which can create incentives for cooperative behaviour (in the form of contributions to a public good) without repeated interactions. I find a critical threshold in network connectedness at which the level of public good provision changes sharply. This is driven by a sudden change in how widely the network spreads information. The networks form endogenously, influenced by the amount of time people spend inside their community. And modest changes to the incentives to spend time inside the community can therefore lead to large falls in public good provision.
Notes and Excerpts
In a similar vein, Hillbilly Elegy (Vance, 2022) describes the decay of a community in Middletown, Ohio, and the problems that come with it.
Pretty funny to see that he cites the current vice president of the United States in the intro.
The key mechanism is that it is networks inside communities that can create accountability and incentives for public good provision. But new technologies that make connections easier overall can push people to spend time outside of their community—at the expense of time inside it. This weakens networks inside communities.
Basic idea for the model is that you reward in-community people who do pro-social things, but only if you hear about them. The in-community social network needs to be connected enough for that gossip to percolate.
changes that raise the benefits of interacting outside of one’s community can make everyone worse off, even though they would constitute a Pareto improvement were behaviour held constant.
Undoing a policy may not bring back public good provision. This is because a shock can knock the community out of the fragile equilibrium with public good provision into a stable equilibrium with no provision. To move back to the fragile equilibrium after the shock dissipates or reverses, a community would need to overcome a coordination problem—which may be beyond the ability of a policymaker.
This brings my paper closer to Elliott et al. (2023)—who consider how networks influence contributions to a firm’s overall corporate culture—although the setting is different.
Oh! Tha might be along the lines of what I was thinking Re: google overhiring!
Several assumptions about degree distribution. Odd one is $\mathbb E [d^2|t] - 2\mathbb E[d|t]$ strictly increasing in $t$ and equal to 0 for some $t\in(0,1)$. This is needed so that agents spending more time in their community can make the largest component smaller.
There are often a range of activities inside firms that cannot be contracted over, and which create significant spillover benefits for the firm as a whole. Informal training and mentoring (especially of junior staff) and informal knowledge sharing are two obvious examples. A careful application of my model would shed light on firms’ efforts to promote camaraderie and in-person attendance in offices, and help clarify the benefits of the proverbial water cooler.