Opening the Black Box of Preferences: A Framework for Modeling Choice (2026 Submission to The Journal of Economic Methodology)
Standard microeconomic models absorb into "preferences" a wide range of mechanisms. These include perceptions, beliefs about how the world works, social norms, emotions, and the rules agents use to evaluate options, that can each affect choice through distinct causal channels. This paper offers a framework that traces the path from an agent's preferences through perceptions, world models, and evaluation protocols to observed choice, and back again through learning. This integrates the relevant research scattered across economics, psychology, and sociology, into a single framework. By making explicit the assumptions a researcher adopts when choosing a model, the framework generates a fuller set of hypotheses at the pre-design stage, guides data collection and treatment design, and helps ensure unexpected results reflect a genuinely novel finding rather than failure to account for known mechanisms. More broadly, it suggests that often behavior we label irrational may be better understood as rational choice constrained by circumstances that the model used ignored.
Link to paper: Opening the Black Box of Preferences
Reputation, Punishment and the Enforcement of Social Norms (Published in The Journal of Economic Behaviour and Organization Oct 2026)
Reputation and social punishment are key mechanisms for norm enforcement, yet existing research often conflates or treats them in isolation. We distinguish reputational punishment as the automatic loss of trust-based interactions following defection, from social punishment, the voluntary avoidance in interactions maintained independently from reputation. Integrating both in a model of informal enforcement, we show that social punishment complements reputation when it is too weak to support cooperation alone, but as reputation strength increases, it substitutes for social punishment, resulting in an inverted-U-shaped relationship between punishment and reputational strength. The mechanism provides a novel motivation for third-party punishment driven by how reputation amplifies punishment threats across an agent's network. Payoffs depend on the distribution of interaction stakes, determining how many new transactions punishment facilitates. We consider social structure by defining ingroups by their capacity to independently enforce trust, implying incentives to renegotiate from universal to localised norms. However, when community enforcement is insufficient, ingroup members may impose additional punishment on defectors, mimicking collective reputation.
Extending the model to regulate multiple norms within a single domain has distinct implications for punishment intensity, and the model's equilibria provide a mechanism rationalizing the inconsistent cross-cultural relationship between cooperation and enforcement intensity. The framework contributes to the literature on informal enforcement, social capital, and social norms by explaining how the nature and intensity of interactions and the costs of complying with a community's norms influence patterns of social enforcement and norm compliance across different social structures.
Link to paper: Reputation, Punishment and the Enforcement of Social Norms
The three papers that comprise my PhD thesis focus on the informal enforcement of social norms.
Chapter 1: Reputation on Networks
This paper presents a network model of reputation. The model investigates how a reputation system can regulate opportunism and facilitate trust in low-frequency or one-off transactions beyond an agent's immediate network. The setup focuses on how the network facilitates access to transaction opportunities and determines the distribution of reputational information. When agents leverage trust over multiple homogeneously distributed transactions, the network structure determines a trust ranking of transactions between any two individuals. By incorporating how the salience of an agent's reputation affects who gets informed, the model identifies a unique set of gatekeeper agents that can regulate the flow of information. These gatekeepers may receive limited trust but can facilitate trust between their links. The model provides a flexible framework that is tractable on a general network. Extending this framework by relaxing various assumptions, the model can investigate several interesting questions. In particular understanding the connection between local and generalised trust, patterns of informal sanctioning and coordination on in-group vs out-group norms.
Link to paper: Reputation on Networks
Chapter 2: Reputation, Punishment and the Informal Enforcement of Norms
This paper introduces a model of informal enforcement that integrates both reputational punishment and costly social punishment, highlighting their complementary roles in sustaining cooperation. The model reveals that social punishment is often necessary to support a reputation system and that punishment without reputation is unlikely to yield positive payoffs to the punisher. Once cooperation is established, the two enforcement mechanisms act as substitutes, with the equilibrium level of social punishment decreasing as the power of reputation to enforce cooperation grows. The model is applied to norm enforcement by examining the regulation of two broad categories of behaviour, Leveraged behaviours, which are 'trust-like' in that cooperation yields a net benefit, and Costly behaviours, which are like public good contribution in that compliance results in a net cost. The findings suggest that understanding the nature and intensity of informal sanctions requires considering the array of enforced behaviours within a community rather than focusing solely on the enforcement of a single norm. By considering how other factors, such as culture, development and institutions, affect reputation systems by changing the nature and distribution of transactions, the paper contributes to understanding the variation in norm enforcement across societies and the impact of these factors on economic success.
Link to paper: Reputation Punishment and the Informal Enforcement of Norms
Chapter 3: Ingroup Norms and Relation Specific Punishment
This paper presents a model investigating ingroup favouritism and inter-group cooperation. While many existing models emphasise the role of information limitations, here, variation is driven by how patterns of interaction determine ingroup agents' incentives to punish other members for defecting outside the group. When agents are embedded in a reputation system, incentives for punishment arise indirectly. Committing to punish fellow ingroup members, means outsiders can trust them more, in effect increasing the value of their reputations. In turn, outsiders value ingroup relationships more, so they can also be trusted more by ingroup agents, including the punisher. Payoffs to punishment depend on the inter-group interaction frequency, transaction profitability, and the value of new transactions that increased trust facilitates with the outgroup. The model provides a mechanism which can be applied to understand how economic context impacts patterns of trust, cooperation and social punishment across different communities.
Link to paper: Ingroup Norms and Relation Specific Punishment
Working Papers:
Systems of Violence and Reputation
Many societies condone violence as a form of punishment or revenge that can be used to enforce informal contracts and agreements. We examine the relationship between this kind of enforcment and the structure of communities, and how this influences the scope and nature of social norms. We show that without formal enforcement, violence capacity is the main determinant of social structure and this has far reaching consequenses, undermining trust and the ability of collective action and public good provision. This is a potentially significant and understudied factor in limiting welfare in many developed contexts.
Community Enforcement and Public Good Provision
This paper extends a model of informal enforcement based on reputation and accounting for group structure, to analyse the drivers and frictions related to informal public good provision and collective action in developing contexts. The model identifies the conditions where cooperation is possible and the factors that undermine it. This provides applied practitioners a more detailed guide to when enlisting social enforcment can be used as an effective tool for change and when these attempts are likely to fail.
Truthful Reporting, Gatekeepers and the Limits of Reputation
This paper extends a model of reputation to look at when agents are willing to break with cooperative norms by renegotiating with defectors and how this depends on the communication and interaction network. We show that when an agent's network implies they are gatekeepers between many opportunities, in sparse networks this may be enough for the agent's links to abandon their community reputation and side with defectors. In more connected contexts, if agents can cooperate with defectors in secret, they do so while maintaining their reputations within the community. Here, an agent's complicity with a defector, is self enforcing because the threat of reputation loss. The results give underlying structure to groups that tollerate deviant norms, from group bullying to sexual harassment within firms.