How close is close enough? When social closeness backfires on honesty[Jusqu’où peut aller la proximité sociale ? Quand la proximité sociale se retourne contre l’honnêteté]

The relationship between dishonesty and social closeness has garnered increasing attention from scholars. While the literature has long evidenced that social closeness increases cooperation, recent work suggests it may also enable cheating behaviour through in-group justification. We study this relationship in an online Die-under-the-cup task (DUTC), asking whether misreporting outcomes increases when participants are paired with socially close rather than socially distant counterparts. We recruited 288 participants and implemented two treatments that made social closeness salient along socioeconomic status (T1) and political alignment (T2). We modelled closeness objectively (living in localities with comparable socioeconomic levels and administered by the same political party), as well as subjectively (self-reported personal income and political preferences matching locality averages). Across pooled and treatment-specific analyses, we find little evidence that social closeness systematically increases misreporting in the DUTC, as differences in reported payoffs are small and sensitive to specification. While objective distance shows weak and non-robust associations with behaviour, subjective measures of closeness are consistently non-significant. Furthermore, we also examine whether being observed by a socially close counterpart amplifies misreports and do not detect a reliable effect, aside from isolated, non-generalisable patterns. Our results suggest that any relationship between social closeness and cheating behaviour in the DUTC is limited and contextdependent. Our findings underscore the importance of multi-method measurement when evaluating how social closeness relates to strategic decision-making.