Misspecified beliefs and the evolution of peer pressure

Paolo Pin, Roberto Rozzi · 2026

Read Misspecified beliefs and the evolution of peer pressure by Paolo Pin, Roberto Rozzi (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-02756.txt.

Introduction

Paolo Pin and Roberto Rozzi study how peer pressure can evolve when people's beliefs about their own returns are wrong. Agents choose effort under varied, possibly mistaken returns, and social contact does not change payoffs directly. Some agents trade off doing well against matching what they expect peers to do, which is where conformity enters.

For any set of beliefs, the authors characterize the best response and show a single optimal level of peer pressure exists. Within groups sharing the same mistaken belief, that level is evolutionarily stable. It is zero for agents with correct beliefs and can be positive for mistaken ones. In a key case, mistaken agents end up choosing the truly best effort, so their error confirms itself and persists. Pressure need not distort outcomes, but it gives social information a perceived value.

Catalog record

CREATORPaolo Pin, Roberto Rozzi
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE93 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.02756

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Misspecified beliefs and the evolution of peer pressure Thanks: We are grateful to Pierpaolo Battigalli, Alberto Dalmazzo, Sebastiano Della Lena, and Philip Neary for helpful discussions and suggestions. We also thank seminar participants at the Psychology Seminar Series at Royal Holloway, University of London, for useful comments. Financial support from Leverhulme International Professorship (grant number LIP-2022-001) is gratefully acknowledged. Paolo Pin, Roberto Rozzi Abstract We study the emergence of conformity preferences in an environment in which agents choose effort under heterogeneous, possibly misspecified returns, and social interactions do not directly affect material payoffs. Some agents choose effort by trading off performance and conformity to expected peer behavior. We characterize subjective best responses. For any given beliefs, an optimal and unique level of peer pressure exists and is evolutionarily stable within groups of agents sharing the same misspecification. Such a level is zero for correctly specified agents and may be positive for misspecified ones. When the efficient level of peer pressure is interior, misspecified agents choose effort equal to their true return, resulting in an equilibrium behavior that is both self-confirming and Nash, allowing the persistence of misspecifications. Peer pressure need not generate long-run allocative distortions, but it creates a perceived value of social information. In equilibrium, this value depends only on misspecification, generating scope for informational rents. Keywords: endogenous preferences; homophily; peer pressure JEL Codes: D83; D85; D91 1 Introduction Humans often feel the urge to conform with their peers in a wide range of contexts ( Mascagni, 2018 ; Farrow et al., 2017 ) . Susceptibility to peer influence appears to be particularly pronounced during adolescence ( Steinberg and Monahan, 2007 ; Laursen and Faur, 2022 ) . Despite the large empirical literature documenting conformity in behavior, much less is understood about the mechanisms through which conformity motives arise in the first place. We provide a simple explanation for this pattern by studying a model in which agents’ material utility does not directly depend on the actions of others, yet it may still be optimal for them to behave as if social pressure were payoff–relevant. We consider a population of agents partitioned into groups that differ in their returns to effort. Within each group, some agents know their own return to effort, while others do not. Agents’ material utility does not depend on what their peers do; however, some agents believe that it does. As a result, they choose their effort based on a perceived utility function that incorporates the expected effort of their peers. Specifically, for these agents, perceived utility is a weighted average of true material utility and a conformity component capturing the distance between their own effort and the expected effort of their peers. In this respect, misspecified agents in our model resemble users of a false but predictively successful theory: they may behave optimally even though their internal representation of the environment is incorrect. 1 1 1 A useful historical analogy is Ptolemaic astronomy. Although geocentric and ultimately false, Ptolemy’s system achieved substantial predictive accuracy in practice. At the same time, it is often contrasted with later frameworks as relying on a more elaborate corrective structure rather than on a correct account of the underlying system; see, e.g., Wray, 2018 . We first characterize agents’ subjective best response under this perceived utility. We then derive the efficient level of social pressure, defined as the level that maximizes material utility given the subjective best response, and assess the robustness of such a level to invasions by agents who share the same misspecification but differ in the intensity of social pressure they attach to peer behavior. We find that the subjective best response always exists and is unique, and that the efficient level of peer pressure is also uniquely determined. In particular, for correctly specified agents the efficient level is always zero, whereas for misspecified agents it may be interior, that is, strictly between 0 and 1 . We further show that, whenever the efficient level of social pressure is interior, the corresponding subjective best response is both a self-confirming equilibrium and a Nash equilibrium. It is self-confirming because, along the induced behavior, agents do not receive feedback that contradicts their beliefs; it is a Nash equilibrium because, at the efficient level of social pressure, the subjective best response also maximizes material utility. The implication is that agents may behave optimally without learning their true returns: misspecified beliefs can persist even when behavior is fully efficient. This also suggests a broader interpretation of the model: conformity concerns can survive not because they are directly payoff-relevant, but because they help agents act efficiently in environments with imperfect self-knowledge. Furthermore, the efficient level of peer pressure depends on three key features: the structure of peer composition, the fraction of misspecified agents, and the degree of misspecification within each group. In particular, greater homophily in the network—that is, a higher weight placed on peers with similar returns to effort—increases the efficient level of peer pressure when peers behave optimally. Intuitively, when peers are more informative about an agent’s own environment, for agents with incorrect beliefs, conforming to their behavior becomes a more effective way to compensate for misspecification. This effect may be attenuated or even reversed when peers themselves behave suboptimally, due to equilibrium feedback effects. The impact of the fraction of misspecified agents on peer pressure operates through the informativeness of peer behavior. When peers behave optimally, this fraction does not matter, since all peers exert the optimal level of effort. When they do not, its impact is conditional on whether changes in composition make peers more or less informative about the group’s true return to effort. The relationship between peer pressure and misspecification is instead conditional on the group’s relative position. The efficient level of peer pressure increases with misspecification if and only if the group’s return to effort exceeds the average effort in its peer network. In this case, placing more weight on peers helps correct individual misperceptions; otherwise, stronger misspecification reduces the value of conformity. Given the central role of peers’ behavior in agents’ decisions, we study the value that agents attribute to observing such behavior. Conditioning behavior on peer outcomes requires attention and computation, and may generate adjustment costs when beliefs are misspecified, even if behavior ultimately coincides with the efficient benchmark. More importantly, within the model agents perceive a direct benefit from observing the social statistic summarizing peers’ actions, because this allows them to condition their behavior on the prevailing environment. In equilibrium, however, the perceived value of monitoring depends only on the gap between agents’ misspecified beliefs about fundamentals and the true underlying returns. This implies that environments with larger misperceptions generate a higher perceived value of social information, creating scope for informational intermediaries—such as platforms providing access to social signals—to extract informational rents. The remainder of the paper is organized as follows. After a brief review of the related literature, Section 2 introduces the model. Section 3 presents the main results, while Section 4 studies the informational value and costs of observing peers’ behavior. Section 5 discusses the broader relevance of the mechanism and its connection to two strands of the literature. Section 6 concludes. Proofs are collected in Appendix A ,

Read the complete text