Fresh SI: Collaboration leads to cooperation on sparse networks

It is traditional in game theory to model cooperation as the play of a given strategy in a social dilemma. This approach is subject to the criticism that cooperation has to be separately defined for each new situation in which it is considered.

Recently, collaboration — the ability to participate in collective decision making and optimization, has been proposed as an alternative approach to cooperative behavior.

Collaboration has the benefit that it can be defined independently of any game. In a paper published in PLOS – Computational Biology, Simon Angus and Jonathan Newton bring these two approaches together, showing that even relatively rare opportunities for collaboration can support robust levels of cooperation, especially when interaction networks are sparse.

This result is significant as human networks are often sparse and so our results support the wide distribution and persistence of cooperation across human populations.

Read the full paper here.

Fresh SI: Agency, potential and contagion

In a now published paper in Games and Economic Behavior, Jonathan Newton and Damian Sercombe consider the relationship between the aggregation of individual incentives into potential functions and the aggregation of individual agency into collective agency.

They consider two fundamental forces that can drive the diffusion of an innovation on a network. The first of these forces is potential maximization, a method of aggregating payoff incentives of players under individual agency. Potential maximization is related to the graph theoretic property of close-knittedness (Young, 2011). The second force is collective agency, under which sets of players decide together on whether to adjust their strategies. Collective agency is shown to be related to the graph theoretic property of cohesion (Morris, 2000). They compare the relative strengths of these forces under (i) different payoff specifications in coordination games and (ii) different network structures.

Read the full paper here.

Fresh SI: Watercooler chat, organizational structure and corporate culture

In a now published paper in Games and Economic Behavior, Jonathan Newton, Andrew Wait and Simon Angus show how collective choices made by employees around a watercooler can have profound effects on behavior within a firm.

Modeling firms as networks of employees, occasional collaborative decision making around the office watercooler changes long run employee behavior (corporate culture). The culture that emerges in a given team of employees depends on team size and on how the team is connected to the wider firm. The implications of the model for organizational structure are explored and related to trends in the design of hierarchies.

Read the full paper here.

Fresh SI: Agency Equilibrium

In a now published paper in Games, Jonathan Newton gives a fixed point solution concept that incorporates collective agency under incomplete information.

In keeping with the ethos of the SI project, agency may be exercised by different entities (e.g. individuals, firms, households). A given individual can form part of multiple agents (e.g. he may belong to a firm and a household).

However, the set of agents that act in a given situation might not be common knowledge. For example, Alice may not know whether Bob and Colm have formed a coalition and constitute a collective agent.

The paper models such considerations, inspired by the classic model of incomplete information over preferences of Harsanyi.

Read the full paper here.

Fresh SI: The evolution of collaboration in symmetric 2×2-games with imperfect recognition of types

In a recent paper published in Games and Economic Behavior, Hannes Rusch considers the evolution of collaboration. That is, he asks in what circumstances the ability to create shared intentions and undertake collaborative action choice can proliferate.

Previous work on this topic (see our discussion here) has considered the operation of group selection (Angus & Newton, 2015) or analyzed the complexities that arise from positive and negative externalities of collaboration on third parties (Newton, 2017).

The paper under discussion differs from prior work in that players with the ability to collaborate imperfectly recognize whether or not others have the ability to collaborate. For example, Alice may have the ability to collaborate, and may erroneously think that Bob also has the ability.

Alice may then see a large animal and say “Hey Bob, let’s hunt that animal!”, thinking that Bob will fulfill his role in a hunt. When Bob fails to adequately carry out this role, it may turn out that Alice would have been better off not hunting at all.

Read the full paper here.

Fresh SI: Stochastic stability under logit choice in coalitional bargaining problems

In a recent paper published in Games and Economic Behavior, Ryoji Sawa considers the emergence of distributive norms in situations in which surplus is produced by coalitions of players.

Previous work (Newton, 2012 – discussed here) found that if strategy updating takes place under a coalitional dynamic with uniform probabilities of a “mistake” (a non-best response), then in the long run there is a tendency towards Rawlsian social choice – the emergence of social norms that maximize the wealth of the poorest members of society.

In contrast, Sawa uses the coalitional logit choice rule (Sawa, 2014 – discussed here) and finds the emergence of social norms that minimize the wealth of the richest members of society. This is similar to the finding of Agastya (1999, see link together with a general discussion of the Evolutionary Nash Program here). However, Agastya’s model used uniform mistakes in an individualistic dynamic. Therefore, two models that differ in these two aspects (perturbation structure and presence or absence of SI) lead to similar outcomes.

Read the full paper here.

Fresh ENP: Conventional contracts, intentional behavior and logit choice: equality without symmetry

In a new paper in the Evolutionary Nash Program published in Games and Economic Behavior, Sung-Ha Hwang, Wooyoung Lim, Philip Neary and Jonathan Newton show how egalitarianism can emerge in asymmetric environments.

Specifically, when coordination games are played under the logit choice rule and there is intentional bias in agents’ non-best response behavior, the Egalitarian bargaining solution emerges as the long run social norm. Without intentional bias, a new solution, the Logit bargaining solution emerges as the long run norm.

These results contrast with results under non-payoff dependent deviations from best response behavior, where it has previously been shown that the Kalai–Smorodinsky and Nash bargaining solutions emerge as long run norms.

Experiments on human subjects, reported in the paper, suggest that non-best response play is payoff dependent and displays intentional bias. This suggests the Egalitarian solution as the most likely candidate for a long run bargaining norm.

Read the full paper here. Read more on the Evolutionary Nash Program on this site here.

Fresh SI: Agency Equilibrium

In a recent working paper, Jonathan Newton gives a fixed point solution concept that incorporates collective agency under incomplete information.

In keeping with the ethos of the SI project, agency may be exercised by different entities (e.g. individuals, firms, households). A given individual can form part of multiple agents (e.g. he may belong to a firm and a household).

However, the set of agents that act in a given situation might not be common knowledge. For example, Alice may not know whether Bob and Colm have formed a coalition and constitute a collective agent.

The paper models such considerations, inspired by the classic model of incomplete information over preferences of Harsanyi.

Read the full paper here.

Fresh SI: Newton & Sercombe on agency, potential and contagion

In a recent working paper, Jonathan Newton and Damian Sercombe consider the relationship between the aggregation of incentives at the individual level (using the mathematical tool of a potential function) and the aggregration of agency in decision making.

They find that the relationship between incentives and agency is related in striking ways to the payoffs of the underlying game they consider.

Results are proven by showing relationships between the pre-existing graph theoretic concepts of the close-knittedness and cohesion of sets of players whose interaction is mediated by a graph.

Read the full paper here.