Fashionably Late: Differentially Costly Signaling of Sociometric Status Through a Subtle Act of Being Late

The paper investigates why arriving late to social gatherings—a deviation from norms of punctuality—can nonetheless yield positive social outcomes. Despite time being a valued resource and punctuality a societal norm, many people intentionally arrive late to social events. The authors propose that lateness functions as a costly and visible signal that communicates sociometric status (respect, admiration, and voluntary deference). They define “fashionably late” as the degree of tardiness constituting a separating equilibrium in a signaling game: high-status individuals can bear the opportunity costs of delayed arrival more easily than low-status individuals, making such delay a credible, differentially costly signal. The core research questions are whether late arrival elevates inferred sociometric status, whether it increases observers’ desire to affiliate with the latecomer, and whether it triggers consumption mimicry of the latecomer’s choices.

• Perceptions of time and punctuality vary by culture and context; punctuality at work is normative, but social events may tolerate or even favor lateness. Prior modeling work frames punctuality as equilibrium behavior contingent on expectations of others’ arrival patterns.
• Sociometric status (distinct from socioeconomic status) concerns the respect and deference accorded by others. Status motives are pervasive and guide behavior and impression management.
• Costly signaling theory explains how individuals credibly convey unobservable qualities through signals that are more costly for low-quality types to imitate. The authors model late arrival as differentially costly: high-status individuals incur lower opportunity costs of foregoing early socialization, permitting a separating equilibrium at a “fashionably late” point.
• Prior status-signaling research documents signals via conspicuous consumption, prosocial displays, subtle brand cues, product size, nonconformity (e.g., red sneakers), housing, and communicated busyness. This work extends that literature by proposing a real act (late arrival) as a credible, differentially costly signal of sociometric status. Supporting evidence suggests actions are perceived as more credible than mere bragging.
• Consumption mimicry literature shows that people imitate others, especially to affiliate and when status-seeking. The authors predict that inferred higher status from lateness increases affiliative intention and prompts imitation of the latecomer’s consumption choices, particularly among observers with higher chronic desire for status. Hypotheses H1–H4 specify effects on status inferences, mediation via perceived prior social engagement, affiliative intentions, and mimicry with moderation by desire for status.

The research comprises a pilot and five studies using field settings and online experiments.

• Pilot (N=190 MTurk, Mage=38.3): Participants viewed a party guest list with arrival times and rated fashionably late perceptions for an on-time, late (+45 min), and very late (+90 min) target. Arrival time strongly increased fashionably late perception (linear trend), validating the manipulation.
• Study 1A (field, European university mingling parties): Research assistants recorded students’ actual arrival times; attendees reported subjective sociometric status using a 10-rung ladder. Analysis focused on those not arriving early (effective n=41 after excluding early arrivals and outliers). Lateness (minutes) positively predicted self-reported sociometric status.
• Study 1B (N=197 MTurk, within-subjects): Participants compared an on-time (person 1) versus very late (person 31) target on number of friends, sociometric status (ladder), then fashionably late. Very late targets were judged higher in status and friends and more fashionably late.
• Study 2A (N=191 MTurk, between-subjects): Participants evaluated either an on-time or very late target on friends and sociometric status and rated mediators: perceived prior social engagement and prior academic/work engagement. Bootstrapped parallel mediation (5,000 resamples) tested indirect effects.
• Study 2B (N=294 MTurk, 2×2 between-subjects): Manipulated party size (small=10 vs large=60 new members) and arrival time (on-time vs very late). Measured friends, sociometric status, prior social engagement, prior academic/work engagement, perceived uniqueness, and perceived independence. Tested moderation by party size and parallel mediation including uniqueness/independence.
• Study 3 (N=81 undergraduates, within-subjects): Added affiliative intention measures (desire to be invited; desire to be in target’s social network) and perceived admiration. Constructed composite sociometric status (friends, status ladder, admiration) and composite affiliative intention (invitation, network). Tested indirect effect from fashionably late perception to affiliative intention via sociometric status using difference scores and bootstrapping.
• Study 4 (N=136 students, between-subjects): Scenario where target John arrived on-time (9:00 pm) or late (10:30 pm) to a party and brought Hoegaarden beer. Measured imitation intention (intent to try Hoegaarden) and chronic desire for status (7-item scale, α=.79). Regressed imitation on arrival time, desire for status, and their interaction.
• Study 5 (field, MBA cohort at nightclub): Two male confederates (matched appearance) wore different T-shirts of pretested equal baseline preference. One arrived at 7:30 pm (on-time), the other at 9:30 pm (late). After exposure, attendees selected which T-shirt they preferred in a raffle form. Analyzed proportion choosing the late-arriver’s T-shirt versus chance; controlled for image order and participant gender.

• Pilot: Arrival time strongly increased fashionably late perceptions (F(2,378)=131.95, p<.001, η²=.41); very late > late > on-time.
• Study 1A: Actual lateness predicted higher self-reported sociometric status (b=.05, β=.38, t(36)=2.47, p=.02; after excluding early arrivals and three Cook’s D outliers), supporting H1a.
• Study 1B: Very late target judged higher in sociometric status (M=5.87 vs 5.25; t(196)=2.47, p=.01, d=.30) and friends (M=4.22 vs 3.34; t(196)=5.21, p<.001, d=.59). Fashionably late manipulation check: very late > on-time (M=5.27 vs 1.98; t(196)=21.08, p<.001, d=2.27), supporting H1b.
• Study 2A: Very late > on-time on sociometric status (M=5.72 vs 4.82; t(189)=3.12, p<.01) and friends (M=3.98 vs 3.36; t(189)=3.03, p<.01). Perceived prior social engagement higher for very late (b=1.55, β=.51, t=8.05, p<.001), while prior academic/work not higher (p=.25). Prior social engagement predicted friends (b=.39, β=.42, p<.001) and status (b=.46, β=.34, p<.001); prior academic/work did not. Indirect effects via prior social engagement were significant for friends (b=.61; β=.21) and status (b=.71; β=.17), supporting H2a–b; indirects via academic/work were not.
• Study 2B: Main effect of arrival time: very late > on-time on friends (M=4.44 vs 3.73; F(1,290)=16.87, p<.001, partial-η²=.06) and sociometric status (M=5.75 vs 5.14; F(1,290)=9.96, p<.01, partial-η²=.03). Party size did not moderate effects. Mediation: Arrival time increased perceived prior social engagement (b=.86, β=.30, p<.001); no effect on prior academic/work (p=.18) or uniqueness (p=.83); small increase in independence (b=.30, β=.14, p=.02). Predictors of outcomes: prior social engagement predicted friends (b=.35, β=.34, p<.001) and status (b=.53, β=.46, p<.001); uniqueness positively predicted both; independence did not. Indirect effects of arrival time via prior social engagement were significant for friends (b=.30; β=.10) and status (b=.45; β=.14); indirects via academic/work, uniqueness, or independence were not, supporting H2b and ruling out alternatives.
• Study 3: Very late > on-time on fashionably late perception (M=4.73 vs 2.86; t(80)=6.73, p<.001), sociometric status composite (M=4.98 vs 3.88; t(80)=3.65, p<.001), and affiliative intention (M=4.36 vs 4.04; t(80)=2.41, p=.02). Mediation: fashionably late → higher sociometric status (b=.39, β=.36, p=.001); sociometric status → higher affiliative intention (b=.30, β=.71, p<.001). Indirect effect significant (b=.12; β=.25; 95% CI excludes 0), supporting H3a–b.
• Study 4: Arrival time × chronic desire for status interaction on imitation intention (b=.96, β=.22, t(132)=2.72, p<.01). Simple effects: effect of late (vs on-time) larger at +1 SD desire for status (b=1.70, β=.46, p<.001) and at mean (b=.88, β=.24, p<.01); non-significant at −1 SD (b=.06, p=.88), supporting H4a–b.
• Study 5 (field): 78% (18/23) chose the T-shirt worn by the late-arriving confederate; significantly above chance (z=2.71, p<.01). No effects of image order or participant gender. Confirms causal effect of arrival time on actual choice, supporting H4a.

Across controlled experiments and field studies, arriving very late to social gatherings operates as a credible, differentially costly signal that elevates inferred sociometric status. Observers infer that latecomers had prior social engagements, which mediates status attributions, rather than assuming work/academic busyness, independence, or uniqueness. Higher inferred sociometric status increases observers’ desire to affiliate with latecomers and leads them to imitate latecomers’ consumption choices, particularly among those with higher chronic desire for status. These findings extend conspicuous consumption research to the domain of time use, showing that a subtle nonverbal act (lateness) can function as a status signal with meaningful downstream social and consumer behavior consequences. The signaling-game perspective clarifies why late arrival can be a separating equilibrium: the opportunity costs of delaying are lower for high-status individuals, making the signal hard for low-status individuals to mimic credibly.

The paper introduces and formalizes “fashionably late” as a separating equilibrium in a costly signaling framework and empirically demonstrates that late arrival to social events increases perceived sociometric status, stimulates affiliative intentions, and drives consumption mimicry. Contributions include: (1) identifying a novel, product-free, time-based status signal; (2) isolating perceived prior social engagement as the key mediator of status inferences; (3) documenting downstream social (affiliation) and consumer (mimicry) effects with boundary conditions (moderation by chronic desire for status). Practical implications suggest leveraging arrival-time cues to identify influencers, craft advertising narratives, and design event policies and pricing that accommodate or capitalize on fashionably late behaviors. Future research could test generalizability across age groups, categories beyond beer and clothing, cultural contexts, and explore boundary conditions where lateness might backfire (e.g., formal or work settings).

• Sample and context skew young and social-party oriented; generalizability to older populations remains unknown.
• Downstream mimicry was tested in beer and clothing categories; applicability to other product domains is untested.
• Although multiple field components were included, some studies relied on scenarios and convenience samples, suggesting the need for broader populations and contexts in future work.