Dilemmas of collective action
coordination games
2014-11-25 — 2025-06-13
Cooperation problems writ large. How do a thousand people work together? A million? This is a scrapbook. But I am very interested in coalition games at the moment and their relationship to various types of fairness.
See also democracy etc.
1 Coalition games
Much to learn here, e.g. Brandt and Bullinger (2022);Aziz and de Keijzer (2011);Aziz and Brandl (2012);Aziz, Brandt, and Harrenstein (2013) TBC
1.1 Rabbit and stag game
A game theoretical model of coordination. This is pertinent in problems of, e.g. having democracy or of getting peer networks off Facebook. As seen in Duncan Sabien’s Group Rationality musing, and Zvi Mowshowitz on the Facebook exodus problem. TBC
2 Fragmentation
TODO: write about the eternal problem of fragmentation in collective action, and standards
Case study: Which stack do you use for data science in 2024? See Full Steam Ahead: The 2024 MAD (Machine Learning, AI & Data) Landscape – Matt Turck
4 Coordination on conflict
5 Elite capture
Olúfẹ́mi O. Táíwò, On the Uses and Abuses of Identity Politics:
Elite capture, he explains, is a concept that emerged from the study of developing countries. It initially referred to the tendency of the upper class to gain control over foreign aid; in other words, the rich get richer. See more at institutions for angels.
6 Collective action dilemmas for oligarchs
A special case of collective action problem of interest to me. See e.g. Singh (2014).
In the state’s husbandry of the pigs of the economy, one might argue that the prime task is to ensure that said pigs “do not eat where they shit”.
However, each of the pigs has a self-interested view in being as lazy as possible in how it eats. This is the essence of the collective action for corporations; If every business wants highly paid customers and low-paid staff, whose job is it to provide customers? If every business wants state-provided infrastructure but must complain about their own personal share of the costs, which businesses are left holding the can? If every business knows that all their peers are lobbying to sell off the state assets for scrap, mightn’t they as well take a piece of the pie rather than be the only ones who were left out of the action?
Can you resolve this through a cartel? Our business won’t whine for special subsidy if yours doesn’t, we all win?
Eric Lonergan in a reply to Chris Dillow, argues
I should stress that it is incorrect to believe that vested interests do not act to protect their interests. But the specific interests of ‘capital’ at the current juncture of history are very unclear. Consider the lobbying of banks in the United States for further deregulation. Superficially, this presents itself as financial capital lobbying for its interests. But as an investor, I am totally unconvinced that this is in the interests of the shareholders of banks — regulations are barriers to entry (as Jamie Dimon has noted publicly). Deregulation usually results in credit booms, which are bad for the return on capital of the non-financial sector. As Mark and I argue in the book, the era of deregulation destroyed the bargaining power trade unions, but also set capital on a path of war with itself. Amazon, for example, is often presented as some evil capitalist semi-monopoly destroying our precious high streets (and who profited from these?). In practice, it operates as a marauding low-cost threat to all other capitalists with fat margins previously living in peace. Amazon’s valuation is treated as a scandal because it renders Jeff Bezos’s paper profits obscene. But what of the damage to other capitalists caused by a virtually zero cost of capital financing a maniacal blue sky investment philosophy? Warren Buffett correctly observed that the predictable winner from the internet was the consumer — that didn’t stop capital from throwing endless quantities of capital at it.
7 Switching games
Here is a feature matrix of coordination game models applied to the problem of switching social media platforms.
| Model Name | Summary | Link to Papers |
|---|---|---|
| Stag Hunt Game | Models the risk-reward tradeoff of switching. A high payoff requires coordinated action (everyone switches), but there’s a risk of being left alone on a new platform if others don’t follow. This captures the assurance problem of trust in others’ actions. Research suggests intuitive, fast thinking can lead to more cooperation (choosing “stag”). | Belloc et al. (2018), Aktipis, Whitaker, and Ayers (2020) |
| Dynamic Coordination with Switching Costs | Focuses on the costs (e.g., losing your network, content, or established presence) associated with leaving an established platform. Users will only switch if their belief in the new platform’s viability overcomes these costs. The model predicts that even with weak confidence in the old platform, switching remains improbable due to these costs.. | Ballard and Boosey (2025) |
| Coordination on Networks | Analyzes how the structure of a social network and the flow of information influence technology adoption. Decisions to switch are often based on local information within friend groups or cliques, which can create cascades. The network’s structure can determine if a switch will be adopted by everyone or only by specific, path-connected sets of users. | Leister, Zenou, and Zhou (2017),Arava (2017) |
| Pure Coordination Games | The most basic model where users receive a payoff simply for choosing the same platform as others, regardless of its features. This highlights the importance of creating a “focal point” (a mutually salient option) to rally users around a new platform, especially when choosing between two new, similar options without established network effects. | Perez-Zapata and Apperly (2022) |
8 Trust
TODO: Social capital. Development version, Bo Rothstein, How the Trust Trap Perpetuates Inequality.