Perceived Social Norms and Information Disclosure
summary
The gist
Social-information messages are widely used to influence norm perceptions and normrelevant behavior, but their effects may depend on what they report and how they are disclosed This paper develops a…
In short
The study develops a belief-based framework linking personal values, perceived social norms, and empirical expectations through an inferred latent appropriateness standard. It shows that reported statistics encode private information differently, and public disclosure changes what recipients believe others know. This implies public disclosure has stronger effects than private disclosure for both message types.
Key concepts
- Latent Appropriateness Standard
- This is an underlying, unstated rule about what is considered appropriate in a given situation. Individuals don't always know this standard directly; instead, they form assessments based on imperfect private cues. This single standard connects personal values, perceived norms, and empirical expectations.
- Reported Statistics Encoding
- Different statistics summarize previous participants' private information in distinct ways. Because personal values and perceived norms encode private cues differently, the way these statistics are presented affects how a recipient infers the underlying appropriateness standard.
- Public Disclosure Effect
- When information is disclosed publicly rather than privately, it creates an additional belief channel. Recipients not only infer what the message says but also what they believe others know and infer from that same statistic, amplifying the message's influence.
Terminology used across episodes
This episode discusses
The paper
Perceived Social Norms and Information Disclosure · Read on arXiv
Southwestern University of Finance and Economics
Transcript
Introduction to the show: ident: Economics. The day in economics: theory, mechanisms and the evidence beneath them.
Ruth: We're looking at the paper: "Perceived Social Norms and Information Disclosure".
Warren: Here's the gist of the paper. Social-information messages are widely used to influence norm perceptions and normrelevant behavior, but their effects may depend on what they report and how they are disclosed This paper develops a belief-based framework connecting personal values, perceived social norms, and empirical expectations through inference about a latent appropriateness standard The framework identifies two mechanisms: reported statistics encode previous participants’ private information differently, and public disclosure changes what recipients believe others know and infer These mechanisms imply stronger effects for perceived-norm than personalvalue messages, and for public than private disclosure Pruning redundant reasoning tokens The central question is whether these different forms of information should have the same impact, or whether norm-relevant statistics affect perceived norms, and through them, norm-relevant behavior in systematically different ways This question is connected to a central distinction in the social-norms literature Norm-relevant beliefs are commonly grouped into personal values, perceived injunctive social norms, and empirical expectations A personal value captures what an individual herself regards as appropriate; a perceived injunctive social norm captures what she believes others regard as appropriate; and an empirical expectation captures what she believes others will do This distinction matters for social-information messages: because these beliefs are related but not identical, messages based on them can convey different information How the training signal is built The framework develops a belief-based mechanism where norm perception can be understood as an inference problem Individuals may not know the relevant appropriateness standard in a given situation; instead, they form assessments from imperfect private cues The same latent standard therefore connects personal values, perceived injunctive social norms, and empirical expectations This perspective implies that norm-relevant messages are not generally informationally equivalent Where the gains come from The framework clarifies two belief-based mechanisms through which norm-relevant messages affect perceived norms The first concerns what is disclosed: previous participants’ personal values or their perceived social norms Both concern appropriateness and can be reported on the same scale, but they encode private cues differently, implying they will shift current perceived norms to different extents The second concerns disclosure mode: holding fixed whether the message reports elicited personal values or perceived social norms, does it matter whether the information is disclosed privately to one individual or publicly to the current group Under public disclosure, this higher-order knowledge creates an additional channel through which public disclosure can shift perceived norms Message Forms The paper studies two message forms: previous participants’ elicited personal values and their elicited perceived social norms These forms are analyzed to show how different statistics summarize previous participants’ private cues and allow their effects on the recipient’s perceived social norm to be compared Disclosure Mode Public disclosure can amplify norm-relevant messages, although this effect is context-dependent The framework clarifies the belief mechanism: public disclosure changes not only what the recipient infers from the message, but also what she believes others know and infer The analysis yields two formal implications: first, holding the disclosure mode fixed, disclosed perceived social norms have a stronger marginal effect on current perceived norms than disclosed personal values, and second, holding the disclosed statistic fixed, public disclosure has a stronger effect than private disclosure The belief-based framework not only makes these comparisons possible, but also reveals the mechanisms behind them Different disclosed belief measures encode previous participants’ private cue information in different ways, generating different statistic-encoding effects across message forms Public disclosure, relative to private disclosure, adds a further belief channel: it changes what the recipient believes others know and infer from the same statistic, thereby amplifying the cue-decomposition effect The effect of a social-information intervention can depend on what statistic the message reports and how it is disclosed, not only on context or population Therefore, social-information messages should specify how the message is constructed, including what reports were elicited from previous individuals and how those reports were aggregated into the reported statistic Whether the message is received privately or observed publicly by peers is also not merely a matter of delivery: it determines whether the message changes only the recipient’s own inference, or also her beliefs about what others know and infer The effect of a norm-relevant message depends not only on the value it reports, but also on the belief process that produced it and on who is known to observe it References Arias, E.
Ruth: (2019).
Warren: How does media influence social norms?
Ruth: experimental evidence on the role of common knowledge Bicchieri, C.
Paper summary: Ruth: Alright folks, we’re looking at a paper called “Perceived Social Norms and Information Disclosure” by Senran Lin. The main idea here is that how social information messages actually work to change what people think about norms really depends on two things: what the message reports and how it gets shared.
Warren: It sets up this belief-based framework that connects personal values, perceived social norms, and empirical expectations all through an inference process about a hidden appropriateness standard.
Iris: What’s interesting is that the framework points to two specific mechanisms at play: first, reported statistics encode previous participants’ private information differently than just reporting what others think or do. Second, public disclosure changes what recipients believe others know and infer from that information.
Ruth: So these mechanisms lead to the implication that there are stronger effects for messages about perceived norms compared to messages about personal values, and for public sharing compared to private sharing.
Warren: It’s really focusing on the difference between those three types of beliefs—personal values, perceived injunctive social norms, and empirical expectations—and how they all come from that single underlying appropriateness standard.
Iris: This framework suggests that these norm-relevant messages aren't generally informationally equivalent; a message about what others personally think is appropriate doesn't necessarily convey the same information as a message about what others believe the group regards as appropriate.
Ruth: The paper builds this whole structure by defining this relevant appropriateness standard as some scalar S, which is specific to the situation, and it stresses that individuals don't directly observe this benchmark.
Warren: That means we're stepping away from assuming this benchmark is something everyone already knows, like Fehr and Schurtenberger did in their work on norms.
Iris: This implies that because the standard S isn't common knowledge, people might form different assessments of it, and they might hold different levels of belief about it.
Ruth: So we see this structure where the message reports things like previous participants’ personal values or their perceived social norms, but those things encode private cues in a way that shifts current perceived norms to different extents.
Warren: And that's not just about what’s reported, but also how it gets disclosed; public disclosure creates an extra channel where recipients believe others know and infer more about the situation.
Iris: The paper clarifies that holding the disclosure mode fixed, you see that disclosed perceived social norms have a stronger effect on current perceived norms than disclosed personal values.
Ruth: That comparison is important because it shows how different belief measures, even if they both relate to appropriateness, generate different statistic-encoding effects across message forms.
Warren: And when we hold the specific statistic being reported fixed, public disclosure has a stronger effect than private disclosure.
Iris: The paper also flags a limitation: this entire model relies on individuals forming assessments from imperfect private cues rather than directly observing the benchmark S itself.
Ruth: So for practical terms, if you’re designing a social information intervention, you have to specify exactly what statistics were reported from previous people and how those reports were aggregated into the statistic.
Warren: And whether it’s public or private delivery changes whether the message only shifts your own inference or also your beliefs about what others know and infer.
Iris: The final point is that the effect of a norm-relevant message depends not just on the value it reports, but on the belief process that produced it and who is known to be observing it.
Ruth: So looking at Senran Lin’s “Perceived Social Norms and Information Disclosure,” what it really means for us in markets or social settings is that we need to be extremely specific about how we build these messages.
Warren: It’s not just about sending out a message; it’s about specifying exactly what statistics were reported from previous people and how those reports were aggregated into the statistic.
Iris: And it also means whether the message is received privately or publicly by peers matters more than just the delivery method itself because of that extra belief channel created by public disclosure.
Ruth: For practical terms, if you're designing a social information intervention, you have to specify what you’re reporting and how that information was gathered from others.
Warren: And whether it’s public or private delivery changes whether the message only shifts your own inference or also your beliefs about what others know.
Iris: The final point is that the effect of a norm-relevant message depends not just on the value it reports, but on the belief process that produced it and who is known to be observing it.
Conclusion: Ruth: So "Perceived Social Norms and Information Disclosure" by Senran Lin is looking at how we use social information to get people to follow norms, and the big picture is that the effectiveness of those messages really hinges on two things: what kind of statistic you report and whether you’re telling it privately or publicly.
Warren: It sets up a framework where they connect personal values, how we see others' norms, and what we expect others to do all through this idea of a hidden appropriateness standard that nobody directly sees.
Iris: The authors find two ways this works: first, the way statistics are reported changes how private information from past participants is encoded, and second, if you make it public instead of private, it actually changes what recipients believe about what other people know and infer.
Ruth: This suggests that messages based on perceived norms tend to have stronger effects than messages based on personal values because of that way the data is reported and shared.
Warren: And public disclosure doesn't just deliver the information; it opens up a new belief channel where people think more about what others know and infer from the same statistic.
Iris: So when you look at this paper, it's showing that these different forms of belief aren't actually giving you the same piece of information, which is important for designing better interventions.
Ruth: It moves us away from just thinking about a simple message and forces us to think about the whole construction: what was reported and who is listening.
Warren: And when we look at the specific results, it shows that holding the type of statistic fixed, public sharing is stronger than private sharing.
Iris: It also shows that holding which belief measure you report fixed, perceived norms are more potent drivers than personal values in shifting current perceptions.
Ruth: So what this really means for us is that if you’re trying to change behavior using social information, the way you build the message—the statistics used and the disclosure mode—is just as important as the actual content of what you’re saying.
Warren: It's a lot to take in about how these different belief measures interact and amplify each other based on how they're presented.
Iris: This leads us into thinking about how this structure might apply when we look at other types of social information, maybe even political communication or things outside of just simple surveys.
More episodes
- Daily Summary for 2026-11-19
- 2608.21274-Recommendation Quality and the Concentration of Consumption: Experimental Evidence from Netflix
- 2408.02757-A nonparametric test for diurnal variation in spot correlation processes
- 2306.16393-High-Dimensional Canonical Correlation Analysis
- 2605.17299-Geometric Brownian motion with intermittent entries and exits
- 2607.04753-Fooling Yourself: how narratives shape beliefs
- 2408.14671-Inference on Treatment Effects with Repeatedly Mismeasured High-Dimensional Controls
- 2201.06898-Difference-in-Differences Estimators for Treatments Continuously Distributed at Every Period
- 2501.15307-Influence Functions: A Unified Differentiation-Based Framework
- The day in conversation for 2026-11-19