Better late than never? Gift givers overestimate the relationship harm from giving late gifts

Abstract

Past work has found that there is often a mismatch between the types of gifts that individuals send and the types of gifts that recipients would prefer to receive. Moving beyond gift choice, the present work explores a novel type of giver–recipient mismatch—beliefs about the importance of sending an on-time gift. Specifically, the current work offers evidence that gift givers systematically overestimate the negative impact that a late occasion-based gift will have on their relationship with the recipient, which occurs because gift givers believe that sending a late gift will signal that they care about the recipient to a lesser extent than what the recipient perceives. As such, gift givers' overestimation of relationship harm from a late gift is attenuated when they signal care in some other way (e.g., with the amount of effort put into creating the gift). Finally, we explore the consequences of degree of gift lateness as well as the decision to not send an occasion-based gift at all on gift givers' overestimation of relationship harm.

INTRODUCTION

Consumer spending on gifts is substantial, with US holiday sales growing 3.8% to a record $964.4 billion in 2023 (Shearman, 2024). Given gift-giving's relevance to the economy and its pervasiveness across individuals, occasions, and cultures, it is no surprise that consumer researchers have long been interested in the psychology of gift-giving (Givi et al., 2023). Recent work in this area has focused on identifying both its antecedents (Givi & Galak, 2017; Tifferet et al., 2018) and outcomes (Chan & Mogilner, 2017), including giver–recipient mismatches, particularly as it relates to differences between the types of gifts that givers give and the types of gifts actually preferred by recipients (e.g., Givi & Das, 2023; Givi & Galak, 2017, 2020; Givi & Mu, 2023).

In this research, we explore a novel type of giver–recipient mismatch: a mismatch between gift givers' and gift recipients' beliefs about the importance of a gift being on time. Very little extant work on consumer gift-giving has explored the role of gift timing, with the notable exception of Givi and Galak (2022), who found that recipients anticipate significantly greater happiness from non-occasion-based gifts versus occasion-based gifts. In the current work, we focus exclusively on occasion-based gifts, that is, gifts given to celebrate special occasions that occur on a discrete date (e.g., a birthday, Christmas) and demonstrate that givers systematically over-estimate the harm that a late occasion-based gift will cause to the relationship compared to recipients. Just as mismatches between what givers give and what recipients want to receive occur due to givers paying more attention to social norms about what makes an appropriate gift (Givi et al., 2023), the mismatch we document stems from gift givers paying more attention to social norms about appropriate gift timing than recipients.

Beyond documenting the antecedent of gift givers' overestimation of anticipated relationship harm from late gifts (i.e., their greater attention to gifting norms, consistent with past work on giver–recipient mismatches), we additionally explore the underlying psychological mechanism driving gift givers' miscalibration. Specifically, we offer evidence of a perspective asymmetry in the extent to which a late gift is believed to signal care about the recipient, wherein gift givers believe that a late gift will signal relatively less care than what the recipient believes. In other words, givers are worried that a late gift will harm the relationship because they are mistakenly overconcerned that the lateness of the gift signals relatively less care for the recipient.

The present research therefore contributes to the growing body of literature in marketing and consumer behavior exploring gift-giving (Baskin et al., 2014; Belk & Coon, 1993; Givi & Galak, 2017; Ward & Broniarczyk, 2011). While a great deal of prior work has focused on gift giver decisions about which types of gifts to give (e.g., experiential vs. material in Goodman & Lim, 2018; new vs. used in Teigen et al., 2005) or even how gifts are presented (e.g., in extravagant vs. plain packaging in Yang & Urminsky, 2018; digital vs. physical gift cards in Reshadi et al., 2023), there is relatively little research on gift timing. As such, our work adds to the literature exploring perspective asymmetries between gift givers and recipients by documenting a novel asymmetry based on consumer perceptions of late versus on-time occasion-based gifts, in which givers fear more relationship harm from late gifts than receivers perceive. In doing so, we additionally add to the growing literature on the signaling value of time and time use (Bellezza et al., 2017; Donnelly et al., 2021; Keinan et al., 2019; Lee-Yoon et al., 2020; Whillans et al., 2021, 2022) and the literature documenting the role that perceptions of care play in driving consumers' choices for others (e.g., Garcia-Rada et al., 2022).

THEORETICAL DEVELOPMENT AND HYPOTHESES

Gift-giving norms and giver–recipient mismatches in type of gift

In a prototypical gifting context, gift giver choices highlight the recipient's gift preferences while focusing on creating, strengthening, and maintaining the giver–recipient relationship (Liu et al., 2019). Gift-giving, when it goes well, can strengthen cooperative bonds (Balconi & Fronda, 2020; Larsen & Watson, 2001; Sherry Jr., 1983) and can also be an effective signal of relationship value in that gifts are often used to signal thoughtfulness and care (Belk & Coon, 1993; Fuchs et al., 2015; Givi et al., 2021, 2023; Liu et al., 2019; Weinberger et al., 2024). One way in which givers may attempt to signal relationship value is to ensure that gift exchanges abide by appropriate social norms with respect to gifting, norms which have been argued to have originated and evolved as part of a “consumer gift system” dictated, in part, by rituals, symbolism, and anticipated reciprocity (Giesler, 2006). Just as social norms refer, generally, to the “socially accepted rules that people strive to abide by” (Kupor et al., 2017), gifting norms pertain, specifically, to the socially accepted rules people strive to abide by when exchanging gifts.

Seminal work on gift-giving has found that perceived norms strongly influence gifting intentions (Goodwin et al., 1990; Wolfinbarger, 1990) and has suggested that the violation of gifting norms can have negative social consequences (Sherry Jr., 1983). Due to the importance of giving the “right” gift to send the appropriate relational signal (Ward & Broniarczyk, 2016), gift givers (relative to recipients) tend to be particularly concerned with following social norms for what type of gift to give (Givi et al., 2023). Unfortunately, this focus on norms can lead to mismatches between the types of gifts givers give and the types of gifts recipients actually want. For example, Baskin et al. (2014) demonstrate that gift givers focus more on giving gifts that are desirable, unique, or special, placing less weight on the feasibility of being able to use the gift, whereas gift recipients care more about feasibility than specialness. Further, Kupor et al. (2017) find that gift givers avoid giving partial gifts (e.g., a $50 gift certificate to a restaurant where a typical meal costs $100) and instead give “complete” gifts (e.g., a $50 gift certificate to a restaurant where the typical meal costs $50) because of the belief that a partial gift would violate gift-giving norms, despite recipients preferring partial gifts. Along these same lines, givers often avoid giving used products as gifts (e.g., a used, 100,000-word dictionary), even when such products are superior to the new products that givers opt for instead (e.g., a new, 50,000-word dictionary). In contrast, recipients are quite open to receiving used products as gifts (Teigen et al., 2005).

Gift timing and the giver–recipient asymmetry in concern about lateness

While prior work has explored mismatches in gift giver versus recipient preferences for specific types of gifts, very little work has explored the role that timing plays in gift-giving. One stream of research that has explored gift timing has found that gift givers tend to focus on the moment a gift is exchanged and are primarily concerned with whether gift recipients will be happy when they open the gift versus the happiness the gift will bring the recipient over time through its ownership and use (Galak et al., 2016; Yang & Urminsky, 2018). More pertinent to the domain of the current work is the limited research on when gifts are given. In the only exploration of this question that we are aware of in the consumer behavior literature, Givi and Galak (2022) find that recipients anticipate significantly greater happiness from non-occasion-based gifts (e.g., a gift given on a random Tuesday) versus occasion-based gifts (e.g., a gift given on a birthday or on a specific holiday like Christmas). Interestingly, Givi and Galak (2022) find that even low-quality gifts increase happiness for non-occasion-gifts because the unexpected timing of the gift, decoupled from a special occasion, leads to lower expectations of gift quality (relative to occasion-based gifts, in which the recipient has a more defined baseline expectation of quality). As with prior work showing that givers focus more on perceived norms related to what types of gifts to give than do recipients (Givi & Galak, 2020, 2022), however, gift givers also fail to predict the positive reaction that recipients have to non-occasion-based gifts because these gifts violate the norm that gifts are typically given for a special occasion. Thus, this work suggests that givers are more sensitive to norms about gift timing than recipients, just as they are more sensitive to norms about what types of gifts to give.

Rather than focusing on timing of non-occasion-based gifts, in the current work we explore givers' and recipients' reactions to late occasion-based gifts, which we define as a gift given after a special occasion that occurs on a single discrete day (e.g., a birthday, Christmas). Such a gift would be counter to the norm of on-time gifts for special occasions, a norm held by the majority of participants in a pilot study that we conducted on Amazon's Mechanical Turk (MTurk; n = 203, M_age = 39.72, SD = 11.17; 41.87% female). Specifically, using a binary response scale, we found that 65.52% of participants agreed (vs. disagreed) that there is a social norm to give occasion-based gifts on time, rather than late (χ² = 19.55, p < 0.001; see Appendix S1: A for details). We propose that if the social norm of punctuality for occasion-based gifts is violated, both givers and recipients may perceive the gift as negatively impacting the relationship. That said, consistent with work showing that gift givers ascribe more importance to gift-giving norms than recipients, we predict that violating the norm of on-time gift-giving for occasion-based gifts will be perceived to result in more relationship harm (which we define as the subjective assessment of a negative impact on the relationship) by givers than recipients.

While we predict that the mismatch in anticipated relationship harm from a late gift will stem from gift givers' generally greater attention to gifting norms, we expect that it will be driven, more directly, by inferences related to care for the recipient. We make this prediction in light of prior research suggesting that signaling care for others is one of the primary roles of the custom of gift giving (Givi & Galak, 2022; Weinberger et al., 2024) and research that has documented the benefit that signaling care for a recipient through gift giving can have on giver–recipient relationships (and the negative impact that failing to signal care can have; Ruth et al., 2004). While past work has demonstrated the role that gift choice plays in the communication of care (e.g., by conveying knowledge about the recipient or demonstrating thoughtfulness via the appropriateness of the gift; Schwartz, 1967; Sherry Jr., 1983), we propose that gift timing, too, will be seen as a means of conveying care.

While giving the “right gift” involves some level of knowledge about the recipient and their identity, giving an on-time gift involves, more basically, an awareness of the occasion on which a gift is intended to be exchanged and enough forethought to ensure that a gift is purchased and received by that occasion. As such, in the same way that failing to give an appropriate gift can signal less care about the recipient due to a lack of thoughtfulness (Zhang & Epley, 2012), we predict that failure to send an on-time gift will lead to similar inferences, namely less care for the recipient. Importantly, because of gift givers' greater attention to the violation of gifting norms (Givi et al., 2023), we predict that they will overestimate the extent to which a late gift will signal a lack of care, which we propose will mediate their subsequent overestimation of relationship harm relative to recipients. We note that this prediction is consistent with work showing that givers focus more on the money spent on gifts than recipients (Flynn & Adams, 2009; Givi et al., 2023), since both time and money are resources gift givers can invest in gifts to indicate that they value the relationship. Formally, we therefore predict the following (summarized in Figure 1):

H1.Due to the greater importance gift givers ascribe to the norm of punctuality for occasion-based gifts, they will anticipate late (vs. on-time) gifts to cause more relationship harm than what is anticipated by gift recipients.

H2.Gift givers' (vs. gift recipients') overestimation of relationship harm stemming from a late (vs. on-time) gift will be mediated by a belief that the late gift will signal that the giver cares about the recipient to a lesser extent than what the recipient perceives.

FIGURE 1

Conceptual framework.

Attenuating the perspective gap

Given our theorizing, that gift givers' overestimation of the relationship harm from sending a late gift is driven by the belief that the gift will signal less care about the recipient than the recipient perceives, we additionally propose that if givers are able to signal care in an alternative way, the predicted perspective gap between givers and receivers of late gifts should be attenuated. One way in which gift givers may signal care is through the amount of effort that is put into the gift. We define effort in the context of gift-giving as the relative degree to which the giver's non-monetary resources, such as time and physical exertion, are dedicated to the gift's acquisition (e.g., driving a long distance to a particular store to acquire a gift), creation (e.g., creating a gift like a hand-knitted scarf or hand-carved wooden toy), or delivery (e.g., taking a flight to deliver a gift in person). Past work has shown that, in the case of wedding registries, for example, individuals who want to signal relationship closeness may deviate from the listed gifts to allow them to show that they put more effort into searching for an alternate gift that demonstrates how well they know the recipient (Ward & Broniarczyk, 2016). Additionally, other research has shown that when the creation of a gift involves more effort (e.g., when a gift is made by hand rather than by machine), it is believed to signal more care to the recipient via inferences about the product being imbued with “love” (even if the gift is not put together by the giver themselves; Fuchs et al., 2015).

Outside of the gift-giving context, a separate stream of work examining the role of care in consumer choices for others has found that individuals are averse to using effort-saving products in caregiving due to the belief that the decrease in effort put toward caregiving signals a lack of care for the other person (e.g., using effort-saving devices in one's role as a parent; Garcia-Rada et al., 2022). Taken together, this association between effort and perceptions of care suggests that, if gift givers' overestimation of relationship harm from late gifts is indeed driven by perceptions of a lack of care, a gift giver putting more effort into a gift (e.g., by creating the gift themselves or expending extra effort to obtain the gift) should attenuate the misprediction of relationship harm that we have hypothesized for gift givers. Importantly, this prediction additionally allows us to test our proposed process via moderation (Spencer et al., 2005). We therefore predict:

H3.The effect predicted in hypothesis 1 will be attenuated in instances where a gift giver devotes greater (vs. relatively less) effort into obtaining or creating a late gift.

How late is considered too late and is late truly better than never?

Thus far we have argued that gift givers' overestimation of relationship harm from sending a late gift stems from their greater attention to gifting norms (relative to recipients) and their concerns about the extent to which violating the norm of sending an on-time gift will signal relatively less care for the recipient. That said, the degree of gift lateness (i.e., how long after an occasion a gift is given) may affect the perceived severity of the norm violation, which should influence the size of the hypothesized perspective gap in anticipated relationship harm between givers and recipients.

In instances when a gift is sent only slightly late (e.g., 2 days after an occasion), it may be relatively ambiguous as to whether a gifting norm has truly been violated relative to instances in which a gift is sent considerably later (e.g., 2 weeks or 2 months after an occasion). In other words, a gift that is only a day or two late may not be considered sufficiently late to count as a norm violation (or may be seen as less severe of a violation), given that people often choose to celebrate an occasion a few days late (e.g., when a birthday falls during the week but is celebrated on the weekend). Further, gifts sent in the mail may arrive a day or two late and still be counted as more or less on time if they were sent before the occasion but arrived shortly after the occasion due to unavoidable shipping delays. Because of this, we anticipate that the hypothesized perspective gap will be larger for a moderately or severely late gift (which we operationalize as 2 weeks late and 2 months late, respectively) than one that is only mildly late (which we operationalize as 2 days late). The results of a pilot study (see Appendix S1: B for details) support our conceptualization of 2 days as mildly late, 2 weeks as moderately late, and 2 months as severely late for an occasion-based gift. That said, once a gift is more than a couple of days late, we expect that the perspective gap will become relatively insensitive to the degree of lateness. In other words, while the overall anticipated relationship harm from a gift that is severely late (e.g., 2 months) may be larger for both givers and recipients versus a gift that is moderately late (e.g., 2 weeks), we expect that the magnitude of gift givers' overestimation of relationship harm will not differ whether the gift is moderately or severely late since the norm of sending an on-time gift has been clearly violated in both cases We note that we only explore gifts that are late but still close enough to a special occasion to still be considered gifts meant for that occasion. We therefore formally predict the following:

H4.The magnitude of gift givers' overestimation of relationship harm from a late gift will be smaller for mildly late gifts versus those that are moderately or severely late but will not differ once a gift is moderately late and the norm of sending an on-time gift has been clearly violated.

Given gift givers' overestimation of relationship harm from sending late gifts, some givers may choose not to send a gift at all if they realize the gift will be late, particularly if the gift will be moderately or severely late. However, not sending a gift for an occasion when it is expected also constitutes a norm violation that could result in a similar giver–recipient mismatch in anticipated relationship harm as that caused by sending a late gift. This, therefore, begs the question: When it comes to gift giving, is late better than never from the perspective of the gift giver's concerns about anticipated relationship harm? Based on our theorizing, we expect that the answer to that question depends on which norm violation (sending a gift late or not sending a gift at all when one is expected) is perceived as more severe.

We propose that, while sending a late gift may be expected to signal relatively less care for the recipient because the sender failed to ensure that the gift would be delivered on time, not sending a gift at all may signal that the giver did not care enough about the recipient to even remember the occasion. Therefore, we predict that choosing not to send a gift for an occasion when it is expected will lead to an even greater overestimation of relationship harm by gift givers (relative to recipients). Thus, we predict that both the timing of a gift and the choice of whether to send a gift at all can create an asymmetry in predicted relationship harm between givers and receivers. As a result, from the giver's perspective, it is better to be late than to never send a gift at all since not sending a gift when one is expected will lead givers to predict relatively greater relationship harm than receivers perceive compared to sending a gift late. More formally:

H5.Gift givers will overestimate the relationship harm that will result from not sending a gift to an even greater extent than sending a gift that is late.

STUDY OVERVIEW

Across six studies, we examine gift givers' (relative to gift recipients') overestimation of the relationship harm that results from a late gift and the psychological mechanisms underlying this perspective gap. In study 1, we demonstrate that gift givers anticipate more relationship harm from late gifts than gift recipients due to gift givers' greater ascribed importance of adhering to gift-giving norms, consistent with prior research. We then test whether givers' overestimation of relationship harm is due, more specifically, to gift givers' overestimation of the extent to which sending a late gift will signal that they care about the recipient to a lesser extent, using two different occasion-based gifting scenarios (Christmas in study 2a; a birthday in study 2b). Study 3 then offers evidence of process by moderation by demonstrating that the perspective gap between givers and recipients is attenuated when gift givers are able to signal care by putting more effort into acquiring the gift. Finally, in studies 4 and 5 we explore whether the degree of gift lateness impacts the magnitude of gift givers' overestimation of relationship harm from late gifts (study 4) and whether the magnitude of anticipated relationship harm differs when a gift giver sends a clearly late gift, compared to when the giver does not send a gift at all (study 5). For all studies, we report all manipulations, measures, and data exclusions. Sample sizes were determined prior to data collection, and all studies were pre-registered, with the exception of study 5. Data, syntax, and pre-registrations are all available via the Open Science Framework (https://osf.io/aurz8/?view_only=fe210b8902574cffa6a775aabc681e70).

STUDY 1

In study 1, we test whether gift givers' overestimation of the negative impact that a late gift will have on their relationship with the recipient (relative to the negative impact that the recipient anticipates) stems from gift givers' greater ascribed importance of adhering to gifting norms.

Participants and procedure

A total of 344 undergraduate students (M_age = 20.01, SD = 1.13, 46.22% female) were recruited to participate in this pre-registered study (https://aspredicted.org/YQ7_H2F) in exchange for course credit. This study employed a 2 (role: giver vs. recipient) × 2 (gift timing: on-time vs. late) between-subjects design. Participants in the giver [recipient] condition first imagined that their friend [they] really liked ice cream. We then manipulated when a birthday gift was sent, such that participants in the on-time condition imagined that they had gifted their friend Alex [their friend Alex had gifted them] a pint of ice cream that would be delivered today, on their birthday, while those in the late condition imagined that they had gifted their friend Alex [their friend Alex had gifted them] a pint of ice cream that would be delivered today, for their birthday that was 2 weeks ago (see Appendix S1: C for full details of manipulations used in all studies). We then collected a measure of anticipated relationship harm, for which participants in the on-time condition were asked to indicate, “How likely do you think this on-time gift from Alex [to Alex] would be to have a negative impact on your relationship?” while those in the late condition were asked to indicate, “How likely do you think this gift that is two weeks late from Alex [to Alex] would be to have a negative impact on your relationship?” measured on seven-point scales (1—Not at all likely to have a negative impact, 7—Extremely likely to have a negative impact). All participants were then told that we were interested in their thoughts about the social norms of gift-giving (which we described as “the socially accepted rules that people strive to abide by”; Kupor et al., 2017) and responded to a scale intended to measure beliefs about the importance of gifting norms, which included three items presented in randomized order: “As a gift giver [recipient], I think it's important that a gift giver follow social norms when giving a gift,” “As a gift giver [recipient], I always pay attention to whether a gift follows the social norms for gift-giving in my social circle,” and “As a gift giver [recipient], I find that an important part of giving a good gift is whether the giver follows social norms about gift-giving.” All items were measured on a seven-point scale (1—Strongly disagree, 7—Strongly agree; α = 0.92) and were combined into a composite measure of beliefs about the importance of adhering to gifting norms. Next, all participants responded to a manipulation check in which they indicated whether the gift described in the scenario would be received “on the birthday” or “two weeks after the birthday” before finally indicating their age and gender. We note that, because this study was run as part of a lab session that included studies from multiple researchers, additional demographic information (e.g., ethnicity and political orientation) were also collected. We did not conduct any analyses with these additional measures, and they are not discussed further.

Results and discussion

In line with our pre-registration, we excluded those who incorrectly answered the manipulation check from our analyses, resulting in a final sample of 312 participants (M_age = 20.03, SD = 1.16, 47.76% female). We first conducted a two-way ANOVA predicting anticipated relationship harm from role, gift timing, and their interaction. This analysis yielded a significant effect of role (F(1, 310) = 41.59, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.12), such that those in the giver condition anticipated a greater negative impact on the relationship than those in the recipient condition (M_Giver = 2.83, SD = 1.87 vs. M_Recipient = 1.74, SD = 1.31), a significant effect of gift timing (F(1, 310) = 98.00, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.24), such that those in the late condition anticipated a greater negative impact on the relationship than those in the on-time condition (M_Late = 3.14, SD = 1.81 vs. M_OnTime = 1.51, SD = 1.13), and, importantly, a significant interaction (F(1, 310) = 21.93, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.07). An analysis of simple effects revealed that, in the late condition, gift givers anticipated that the gift would have a negative impact on the relationship to a greater extent than gift recipients (M_Giver = 3.93, SD = 1.69 vs. M_Recipient = 2.19, SD = 1.48; F(1, 310) = 60.21, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.16). In the on-time condition, however, we observed no difference in anticipated relationship harm between gift givers and gift recipients (M_Giver = 1.65, SD = 1.23 vs. M_Recipient = 1.37, SD = 1.02; F(1, 310) = 1.61, p = 0.21, $$$ {\eta}_p^2 $$$ = 0.01; see Figure 2), consistent with Hypothesis 1.

FIGURE 2

Study 1: Anticipated negative impact on the relationship based on role and gift timing (error bars indicate ±1 SEs. ***p < 0.001).

We next conducted a two-way ANOVA predicting beliefs about the importance of gifting norms from role, gift timing, and their interaction. This analysis yielded a significant effect of role (F(1, 310) = 22.42, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.07), such that those in the giver condition indicated that gifting norms were significantly more important than those in the recipient condition (M_Giver = 3.95, SD = 1.57 vs. M_Recipient = 3.09, SD = 1.56), a significant effect of gift timing (F(1, 310) = 4.79, p = 0.03, $$$ {\eta}_p^2 $$$ = 0.02), such that those in the late condition indicated that gifting norms were significantly more important relative to those in the on-time condition (M_Late = 3.76, SD = 1.67 vs. M_OnTime = 3.31, SD = 1.55), and a non-significant interaction (F(1, 310) = 1.04, p = 0.31, $$$ {\eta}_p^2 $$$ = 0.00).

Finally, we conducted a PROCESS Model 14 (Hayes, 2017) moderated mediation analysis, treating role as the independent variable (giver = −1, recipient = 1), anticipated negative impact on the relationship as the dependent variable, rated importance of gifting norms as the mediator, and gift timing as the moderator (on-time = −1, late = 1). In other words, this analysis allows us to test whether the relationship between the rated importance of gifting norms on anticipated negative impact on the relationship was dependent on whether the gift was late or on-time (i.e., whether moderation occurs on the b-path). This analysis revealed a significant index of moderated mediation (b = −0.14, SE = 0.06, 95% CI: [−0.2705 to −0.0424]). We observed a significant indirect effect of participants' rated importance of gifting norms on anticipated negative impact on the relationship in the late gift condition, indicating that, when the gift was late, gift givers' increased rated importance of gifting norms led to an increase in anticipated negative impact on the relationship (b = −0.19, SE = 0.06, 95% CI: [−0.3195 to −0.0907]). Further, we observed a non-significant indirect effect in the on-time gift condition (b = −0.05, SE = 0.03, 95% CI: [−0.1172 to 0.0050]; see Figure 3).

FIGURE 3

Study 1: PROCESS Model 14 moderated mediation analysis (***p < 0.001).

Study 1 therefore provides evidence consistent with Hypothesis 1, that gift givers expect a late gift to have a significantly greater negative impact on their relationship with the recipient than the recipient anticipates. Additionally, we demonstrate that this effect is due, in part, to gift givers' (vs. recipients') relatively greater belief in the importance of following gifting norms. Specifically, as shown in Figure 3, givers place more emphasis on gift-giving norms than recipients, but it is only when these norms are violated that this greater ascription of importance of following gifting norms results in a difference in anticipated relationship harm between givers and receivers. Finally, while we anticipated (and observed) a significant main effect of role on the rated importance of gifting norms, we also observed a significant effect of gift timing, which we did not anticipate. We note that the effect size for gift timing was markedly smaller than that of role and believe this result was due to those in the late gift condition having just failed to adhere to gifting norms, thus leading to a greater focus on the importance of gifting norms.

Having offered evidence that the giver–recipient perspective gap in anticipated relationship harm from giving a late gift stems from differences in the importance placed on gifting norms by givers versus receivers, in study 2 we begin to explore the more specific underlying psychological mechanism.

STUDIES 2A-2B

The goal of study 2 was to test Hypothesis 2, that the giver–recipient mismatch we have documented is due, more specifically, to givers' belief that a late gift will signal relatively less care for the recipient. Further, we test whether the demonstrated effect is generalizable to additional gifting occasions by testing the proposed mismatch in the context of both Christmas gifts (study 2a) and birthday gifts (study 2b).

Study 2a

Participants and procedure

A total of 403 participants (M_age = 42.87, SD = 12.68, 44.64% female) were recruited from MTurk to participate in this pre-registered study (https://aspredicted.org/193_89X) in exchange for monetary compensation and were randomly assigned to one of two conditions (role: gift giver vs. gift recipient). Participants in the giver [recipient] condition were first told to imagine that their friend [they] really like gadgets. Participants were then told to picture that Christmas was 2 weeks ago and they had sent their friend [their friend had sent them] a smart mug that was delivered late, 2 weeks after Christmas. All participants were then asked to indicate how likely the late gift would be to have a negative impact on their relationship, using the same measure from study 1, before responding to a measure of perceived care for the recipient. Specifically, participants were asked to indicate the extent to which they agreed, “Based on this gift to Alex [from Alex] that is two weeks late, Alex will believe I care about them [I believe Alex cares about me].” Finally, all participants responded to a manipulation check in which they indicated whether the gift described in the scenario was received on time or 2 weeks late before finally indicating their age and gender. We note that the gift was always 2 weeks late in this study.

Results and discussion

In line with our pre-registration, we excluded those who incorrectly answered the manipulation check from our analyses, resulting in a final sample of 396 participants (M_age = 42.98, SD = 12.70, 44.95% female). We first conducted an ANOVA predicting anticipated relationship harm from role, which yielded a significant effect (F(1, 394) = 18.10, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.04), such that those in the giver condition anticipated a greater negative impact on the relationship than those in the recipient condition (M_Giver = 2.76, SD = 1.58 vs. M_Recipient = 2.12, SD = 1.41). We then conducted an ANOVA predicting perceived care from role, which also yielded a significant effect (F(1, 394) = 15.32, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.04), such that those in the giver condition anticipated that the late gift would signal care to a lesser extent than those in the recipient condition (M_Giver = 4.68, SD = 1.55 vs. M_Recipient = 5.25, SD = 1.35). Finally, we conducted a PROCESS Model 4 mediation analysis treating role as the independent variable (giver = −1, recipient = 1), anticipated negative impact on the relationship as the dependent variable, and perceived care for the recipient as the mediator. This analysis yielded a significant indirect effect (b = −0.18, SE = 0.05, 95% CI: [−0.2792 to −0.0865]) indicating that gift givers' overestimation of relationship harm from a late gift was driven by the belief that the late gift would signal that they care about the recipient to a lesser extent than recipients anticipated.

This study therefore provides initial evidence in support of Hypothesis 2, that gift giver (vs. recipient) overestimation of relationship harm from a late gift is mediated by the belief that sending a late gift will signal less care for the recipient. In addition, this study also demonstrates that the observed giver–recipient asymmetry is robust to a different gifting occasion than that tested in study 1. We note, however, that this study did not include an on-time gift condition, which would allow us to determine whether this mismatch in perceived care is specific to late gifts. We address this in study 2b.

Study 2b

Participants and procedure

A total of 801 MTurk participants (M_age = 39.93, SD = 11.93, 49.81% female) were recruited to participate in this pre-registered study (https://aspredicted.org/W5Z_9FG) in exchange for monetary compensation. This study employed a 2 (role: giver vs. recipient) × 2 (gift timing: on-time vs. late) between-subjects design. Participants in the giver [recipient] condition first imagined that their friend [they] really liked board games. Those in the on-time condition were then told that they had gifted their friend Alex [their friend Alex had gifted them] a board game that would be delivered today, on their birthday, while those in the late condition imagined that they had gifted their friend Alex [their friend Alex had gifted them] a board game that would be delivered today, for their birthday that was 2 months ago. Participants then indicated how likely they believed the gift would be to have a negative impact on the relationship, using the same measure as in prior studies, before responding to the same measure of perceived care as in study 2a. Finally, all participants responded to a manipulation check in which they indicated whether the gift described in the scenario would be received “on the birthday” or “two months after the birthday” before indicating their age and gender.

Results and discussion

In line with our pre-registration, we excluded those who incorrectly answered the manipulation check from our analyses, resulting in a final sample of 786 participants (M_age = 39.96, SD = 11.94, 50.51% female). We first conducted a two-way ANOVA predicting anticipated negative impact on the relationship from role, gift timing, and their interaction. This analysis yielded a significant effect of role (F(1, 784) = 20.61, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.03), such that those in the giver condition anticipated a greater negative impact on the relationship than those in the recipient condition (M_Giver = 2.24, SD = 1.65 vs. M_Recipient = 1.72, SD = 1.36), a significant effect of gift timing (F(1, 784) = 188.57, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.19), such that those in the late condition anticipated a greater negative impact on the relationship than those in the on-time condition (M_Late = 2.66, SD = 1.70 vs. M_OnTime = 1.30, SD = 0.93), and, importantly, a significant interaction (F(1, 784) = 6.12, p = 0.014, $$$ {\eta}_p^2 $$$ = 0.01). An analysis of simple effects revealed that, in the late condition, gift givers anticipated that the gift would lead to significantly greater relationship harm than gift recipients (M_Giver = 2.98, SD = 1.70 vs. M_Recipient = 2.30, SD = 1.65; F(1, 784) = 24.72, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.03). In the on-time condition, however, we observed no difference in anticipated relationship harm between gift givers and gift recipients (M_Giver = 1.41, SD = 1.10 vs. M_Recipient = 1.21, SD = 0.74; F(1, 784) = 2.12, p = 0.15, $$$ {\eta}_p^2 $$$ = 0.00; see Figure 4).

FIGURE 4

Study 2: Anticipated negative impact on the relationship based on role and gift timing (error bars indicate ±1 SEs. ***p < 0.001).

We next conducted a two-way ANOVA predicting perceived care from role, gift timing, and their interaction. This analysis yielded a significant effect of role (F(1, 784) = 9.92, p = 0.002, $$$ {\eta}_p^2 $$$ = 0.01), such that those in the giver condition believed the gift would signal less care relative to those in the recipient condition (M_Giver = 5.44, SD = 1.62 vs. M_Recipient = 5.82, SD = 1.36), a significant effect of gift timing (F(1, 784) = 332.25, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.30), such that those in the late condition believed the gift would signal less care relative to those in the on-time condition (M_Late = 4.81, SD = 1.58 vs. M_OnTime = 6.46, SD = 0.82), and, importantly, a significant interaction (F(1, 784) = 7.72, p = 0.006, $$$ {\eta}_p^2 $$$ = 0.01). An analysis of simple effects revealed that in the late condition, gift givers anticipated that the gift would signal significantly less care than gift recipients (M_Giver = 4.56, SD = 1.65 vs. M_Recipient = 5.09, SD = 1.45; F(1, 784) = 17.65, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.02). In the on-time condition, however, we observed no difference in perceived care between gift givers and gift recipients (M_Giver = 6.44, SD = 0.80 vs. M_Recipient = 6.47, SD = 0.85; F(1, 784) = 0.07, p = 0.79, $$$ {\eta}_p^2 $$$ = 0.00).

Finally, we conducted a PROCESS Model 7 moderated mediation analysis, treating role as the independent variable (giver = −1, recipient = 1), anticipated negative impact on the relationship as the dependent variable, perceived care for the recipient as the mediator, and gift timing as the moderator (on-time = −1, late = 1). In other words, this analysis allows us to test whether the relationship between role and perceived care for the recipient was dependent on whether the gift was late or on-time (i.e., whether moderation occurs on the a-path). We observed a significant index of moderated mediation (b = −0.16, SE = 0.06, 95% CI: [−0.2740 to −0.0487]), such that there was a significant indirect effect of role on anticipated negative impact on the relationship when the gift was late, indicating that a decrease in the extent to which givers believed the gift would signal care for the recipient led to an increase in anticipated negative impact on the relationship (b = −0.17, SE = 0.05, 95% CI: [−0.2732 to −0.0727]). In contrast, we observed a non-significant indirect effect when the gift was on time (b = −0.01, SE = 0.03, 95% CI: [−0.0637 to 0.0425]; see Figure 5).

FIGURE 5

Study 2: PROCESS Model 7 moderated mediation analysis (***p < 0.001, **p < 0.01).

Taken together, the results of studies 2a and 2b offer evidence that gift givers' (vs. recipients') overestimation of relationship harm from a late gift is driven by the belief that sending a late gift will signal relatively less care about the recipient, consistent with our predictions in Hypothesis 2. Having offered evidence for the mediating role of care in the giver–recipient asymmetry in predicted relationship harm from late gifts, the goal of study 3 was to provide additional evidence of process via moderation by exploring a context in which the gift giver can communicate care even when a gift is late.

STUDY 3

In study 3 we test our prediction in Hypothesis 3 that, if gift givers' overestimation of relationship harm is driven by inferences that a late gift will signal less care for the recipient, the perspective gap will be attenuated if the giver is able to signal care in some other way (which we operationalize as acquiring and creating items to fill a gift basket).

Participants and procedure

A total of 805 participants (M_age = 43.99, SD = 12.86, 53.29% female) were recruited from MTurk to participate in this pre-registered study (https://aspredicted.org/WCW_BB9) in exchange for monetary compensation. This study employed a 2 (role: giver vs. recipient) × 2 (effort expended by giver in acquiring and creating gift: high vs. low) between-subjects design. Participants in both the giver and recipient condition imagined that they sent their friend Alex [their friend Alex sent them] a gift basket that was delivered 2 weeks after Alex's [their] birthday. Hence, the gift was 2 weeks late in all conditions. We then manipulated the amount of effort put into the gift by the giver, such that, in the high effort by giver condition, participants were told that their friend Alex notices [they notice] that the gift basket was put together by hand by the giver and involved driving to multiple local businesses to purchase items from different locations, including coffee from a local roaster, chocolate from a local candy store, coasters from a local bar, and homemade cookies. Those in the low effort by giver condition were told that the gift basket included the same items, but that the gift basket was pre-made by a company. In a separate manipulation check study, we confirmed that the gift basket that was put together by hand (i.e., the one in the high effort condition) was seen as requiring significantly more effort than the one that was put together by a company (M_HighEffort = 6.27, SD = 1.06 vs. M_LowEffort = 3.55, SD = 1.81; see Appendix S1: D). All participants then indicated how likely they believed the late gift would be to have a negative impact on the giver–recipient relationship, consistent with prior studies, before responding to a manipulation check in which they indicated whether the gift described in the scenario was put together by hand or pre-made by a company, before finally indicating their age and gender.

Results

In line with our pre-registration, we excluded those who incorrectly answered the manipulation check from our analyses, resulting in a final sample of 768 participants (M_age = 44.28, SD = 12.91, 53.26% female). We first conducted a two-way ANOVA predicting anticipated negative impact on the relationship from role, giver effort, and their interaction. This analysis yielded a significant effect of role (F(1, 766) = 34.94, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.04), such that those in the giver condition anticipated a greater negative impact on the relationship from the late gift than those in the recipient condition (M_Giver = 2.35, SD = 1.43 vs. M_Recipient = 1.82, SD = 1.28), a significant effect of giver effort (F(1, 766) = 47.11, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.06), such that those in the high giver effort condition anticipated less of a negative impact on the relationship than those in the low giver effort condition (M_HighEffort = 1.77, SD = 1.20 vs. M_LowEffort = 2.40, SD = 1.48) and a significant interaction (F(1, 766) = 4.01, p = 0.046, $$$ {\eta}_p^2 $$$ = 0.01). An analysis of simple effects revealed that gift givers anticipated a greater negative impact on the relationship in both the high effort condition (M_Giver = 1.95, SD = 1.16 vs. M_Recipient = 1.58, SD = 1.21; F(1, 766) = 7.76, p = 0.006, $$$ {\eta}_p^2 $$$ = 0.01) and the low effort condition (M_Giver = 2.79, SD = 1.56 vs. M_Recipient = 2.04, SD = 1.30; F(1, 766) = 30.83, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.04). That said, the interaction indicates that there was a significant difference in the size of these effects, such that the effect was smaller in the high giver effort condition. In other words, consistent with our predictions, gift givers' overestimation of relationship harm was attenuated in the high (vs. low) effort condition, as shown in Figure 6.

FIGURE 6

Study 3: Anticipated negative impact on the relationship based on role and effort (error bars indicate ±1 SEs. **p < 0.01, ***p < 0.001).

Thus, the results of study 3 show that, in line with Hypothesis 3, when a gift giver is able to signal care through the amount of effort they put into acquiring and creating a gift, it lessens the extent to which they anticipate the late gift will have a negative impact on their relationship with the recipient. Having offered evidence for the proposed giver–recipient perspective gap that emerges from a late gift, as well as the underlying psychological driver of this mismatch, in the remaining studies we turn our attention to whether this perspective gap varies based on degree of gift lateness, as well as whether a greater perspective gap emerges in scenarios in which a giver fails to send a gift altogether when a gift is expected (i.e., whether late truly is better than never from the gift giver's perspective).

STUDY 4

In study 4 we test whether the degree of lateness of a gift affects the magnitude of gift givers' overestimation of relationship harm relative to what recipients anticipate. As outlined in Hypothesis 4, we expect the magnitude of gift givers' overestimation to be larger when a gift is moderately or severely late (i.e., 2 weeks or 2 months) versus when it is mildly late (i.e., 2 days) but not to differ across moderately and severely late gifts given that the norm of giving occasion-based gifts on-time has been clearly and unambiguously violated in both of these cases.

Participants and procedure

A total of 802 participants (M_age = 38.65, SD = 12.13, 45.76% female) were recruited from MTurk to participate in this pre-registered study (https://aspredicted.org/Z4J_Q5H) in exchange for monetary compensation. This study employed a 2 (role: giver vs. recipient) × 4 (gift timing: on-time, 2 days late, 2 weeks late, 2 months late) between-subjects design. Participants in the giver [recipient] condition first imagined that their friend [they] really liked plants. We then manipulated when a birthday gift was given, such that participants imagined that their friend Alex's birthday [their birthday] “is today,” “was two days ago,” “was two weeks ago,” or “was two months ago” (depending on condition) and that they sent Alex [Alex sent them] a potted plant for their birthday today. Participants then indicated how likely they believed the gift would be to have a negative impact on their relationship, using the same measure as prior studies. Finally, all participants responded to a manipulation check in which they indicated whether the gift described in the scenario would be received “on the birthday” or “after the birthday” before indicating their age and gender.

Results and discussion

Consistent with our pre-registered analysis plan, we excluded those who incorrectly answered the manipulation check from our analyses, resulting in a final sample of 754 participants (M_age = 38.71, SD = 12.03, 46.55% female). We first conducted a two-way ANOVA predicting anticipated negative impact on the relationship from role, gift timing, and their interaction. This analysis yielded a significant effect of role (F(1, 752) = 62.08, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.08), such that those in the gift giver condition anticipated a greater negative impact on the relationship than those in the gift recipient condition (M_Giver = 2.42, SD = 1.53 vs. M_Recipient = 1.72, SD = 1.22), a significant effect of gift timing (which we outline in further detail below; F(3, 750) = 39.32, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.14), and a significant interaction (F(3, 750) = 13.24, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.05).

In line with our pre-registration, we conducted an analysis of simple effects to interpret the significant interaction between role and gift timing, which revealed that there was no difference in anticipated negative impact on the relationship between gift givers and gift recipients in the on-time condition (M_Giver = 1.21, SD = 0.64 vs. M_Recipient = 1.39, SD = 1.08; F(1, 752) = 0.93, p = 0.34, $$$ {\eta}_p^2 $$$ = 0.00), but that there was a significant difference in anticipated relationship harm in the 2 days late (M_Giver = 2.29, SD = 1.23 vs. M_Recipient = 1.70, SD = 1.23; F(1, 752) = 10.81, p = 0.001, $$$ {\eta}_p^2 $$$ = 0.01), 2 weeks late (M_Giver = 2.94, SD = 1.49 vs. M_Recipient = 1.73, SD = 1.19; F(1, 752) = 44.78, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.06), and 2 months late (M_Giver = 3.30, SD = 1.69 vs. M_Recipient = 2.03, SD = 1.29; F(1, 752) = 46.51, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.06) conditions, such that gift givers believed there would be a significantly larger negative impact on the relationship than gift recipients in all three late gift conditions (see Figure 7).

FIGURE 7

Study 4: Anticipated negative impact on the relationship based on role and gift timing (error bars indicate ±1 SEs. **p < 0.01, ***p < 0.001).

To understand whether there was a difference in the magnitude of the perspective gap based on when the gift was given, we next constructed a set of orthogonal contrast codes (in line with our pre-registered analysis plan) to allow us to compare (1) whether there was a difference in the size of the perspective gap for the on-time condition compared to the three late conditions, (2) whether there was a difference in the size of the perspective gap for the 2-day late condition compared to the 2 weeks and 2 months late conditions, and (3) whether there was a difference in the size of the perspective gap for the 2 weeks late condition compared to the 2 months late condition. These contrast codes partition the ANOVA sums of squares into interpretable subsets (Rosenthal et al., 2000) and, because the codes are orthogonal, there is no need for special alpha levels (see Appendix S1: E for coding schemes). We then conducted an ANOVA predicting anticipated negative impact on the relationship from role, each of the three orthogonal contrast codes, and the interaction between role and each of the three orthogonal contrasts. This analysis revealed a significant main effect of role (F(1, 752) = 68.02, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.08), a significant main effect of the first contrast code comparing the on-time condition to the three late conditions (F(1, 752) = 92.11, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.11), a significant main effect of the second contrast code comparing the 2 days late condition to the 2 weeks and 2 months late conditions (F(1, 752) = 20.82, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.03), and a significant main effect of the third contrast code comparing the 2 weeks late condition to the 2 months late condition (F(1, 752) = 6.38, p = 0.012, $$$ {\eta}_p^2 $$$ = 0.01). In other words, collapsing across gift givers and gift recipients, we see that a late gift is believed to cause more relationship harm than an on-time gift, a gift that is 2 weeks or 2 months late is believed to cause more relationship harm than a gift that is 2 days late, and a gift that is 2 months late is believed to cause more relationship harm than a gift that is 2 weeks late.

We additionally observed a significant interaction between role and the first contrast code comparing the on-time versus three late conditions (F(1, 752) = 31.47, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.04), a significant interaction between role and the contrast code comparing the 2 days late versus 2 weeks and 2 months late conditions (F(1, 752) = 8.53, p = 0.004, $$$ {\eta}_p^2 $$$ = 0.01), and a non-significant interaction between role and the contrast code comparing the 2 weeks versus 2 months late condition (F(1, 752) = 0.03, p = 0.86, $$$ {\eta}_p^2 $$$ = 0.00). In other words, there was a significant difference in the magnitude of the perspective gap that resulted from a gift being on-time versus late in general (i.e., for all three of the late conditions compared to the on-time condition) and between a gift being 2 days late versus later (such that the perspective gap was significantly larger when the gift was 2 weeks or 2 months late than when it was 2 days late). For 2 weeks compared to 2 months late, on the other hand, there was not a significant difference in the size of the perspective gap.

Taken together, the results of this study demonstrate that the degree of lateness of an occasion-based gift does affect the magnitude of the giver–receiver perspective gap in anticipated relationship harm but that, in line with Hypothesis 4, the size of the perspective gap depends on whether the gift is clearly seen as violating the norm of sending an on-time gift (i.e., whether the gift is more than mildly late). Further, while a gift that is 2 months late is seen as worse overall than a gift that is 2 weeks late (i.e., collapsing across givers and recipients), we see that, in line with our predictions, the size of the perspective gap between these conditions does not differ. In the next study, we turn our attention to whether a larger perspective gap emerges for what should be seen as a relatively more severe violation of a gifting norm: not sending an occasion-based gift at all when one is expected.

STUDY 5

The goal of study 5 was to test Hypothesis 5 by comparing gift givers' overestimation of relationship harm from giving a late gift to the violation of a different gifting norm, in which a giver chooses not to send a gift at all. If the mismatch we have documented does stem from inferences about care, then the more severe norm violation of not sending a gift at all when one is expected should produce an even larger perspective gap than sending a gift late.

Participants and procedure

A total of 903 participants (M_age = 39.18, SD = 12.13, 49.94% female) were recruited from MTurk to participate in the study in exchange for monetary compensation. This study employed a 2 (role: giver vs. recipient) × 3 (gift received: on-time, 2 months late, never sent) between-subjects design. This study used the same initial setup as the prior study, with those in the on-time and 2 months late condition being told that their friend [they] really liked plants, after which participants in all conditions were told, “You and your friend Alex typically exchange gifts for your birthdays.” Participants then imagined that they had gifted their friend Alex [their friend Alex had gifted them] a potted plant that was either delivered on the birthday, 2 months after the birthday, or that a gift was never sent. As in prior studies, participants again indicated how likely they believed the gift (or lack thereof) would be to have a negative impact on the giver–receiver relationship. Finally, all participants responded to a manipulation check in which they indicated whether the gift described in the scenario was received on the birthday, after the birthday, or was never sent, before indicating their age and gender.

Results and discussion

Consistent with prior studies, we excluded those who incorrectly answered the manipulation check from our analyses, resulting in a final sample of 871 participants (M_age = 39.25, SD = 12.11, 50.40% female). We first conducted a two-way ANOVA predicting anticipated negative impact on the relationship from role, condition of when (or if) the gift was given, and their interaction. This analysis yielded a significant effect of role (F(1, 869) = 43.19, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.05), such that those in the giver condition anticipated greater negative impact on the relationship than those in the recipient condition (M_Giver = 2.64, SD = 1.76 vs. M_Recipient = 2.11, SD = 1.46), a significant effect of condition of when (or if) the gift was given (which we outline in further detail below; F(2, 868) = 182.39, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.30), and a significant interaction (F(2, 868) = 12.53, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.03). An analysis of simple effects revealed that in the on-time condition there was no difference in anticipated negative impact on the relationship between givers and recipients (M_Giver = 1.29, SD = 0.87 vs. M_Recipient = 1.28, SD = 0.82; F(1, 869) = 0.00, p = 0.95, $$$ {\eta}_p^2 $$$ = 0.00), but that in both the 2 months late condition (M_Giver = 2.87, SD = 1.56 vs. M_Recipient = 2.21, SD = 1.44; F(1, 869) = 17.87, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.02) and the never sent condition (M_Giver = 3.97, SD = 1.59 vs. M_Recipient = 2.84, SD = 1.58; F(1, 869) = 48.54, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.05), there was a significant difference in givers' versus receivers' anticipated negative impact on the relationship, such that gift givers anticipated significantly greater relationship harm than gift recipients (see Figure 8).

FIGURE 8

Study 5: Anticipated negative impact on the relationship based on role and gift timing (error bars indicate ±1 SEs. ***p < 0.001).

To understand whether there was a difference in the magnitude of the perspective gap between these conditions, we again constructed a set of orthogonal contrast codes to allow us to compare (1) whether there was a difference in the size of the perspective gap for the on-time condition compared to the 2 months late and never sent conditions, and (2) whether there was a difference in the size of the perspective gap for the 2 months late condition compared to the never sent condition (see Appendix S1: E for coding schemes). We conducted an ANOVA predicting anticipated negative impact on the relationship from role, both of the orthogonal contrast codes, and the interaction between role and both of the orthogonal contrast codes. This analysis revealed a significant main effect of role (F(1, 869) = 43.19, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.05), a significant main effect of the contrast code comparing the on-time condition to the late and never sent conditions (F(1, 869) = 310.62, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.26), and a significant main effect of the contrast code comparing the 2 months late condition to the never sent condition (F(1, 869) = 59.01, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.06). In other words, collapsing across givers and recipients, both sending a late gift and not sending a gift at all were believed to result in greater harm to the relationship than sending an on-time gift, and not sending a gift at all was believed to result in greater harm to the relationship than sending a late gift.

We additionally observed a significant interaction between role and the contrast code comparing the on-time condition to the late and never sent conditions (F(1, 869) = 21.16, p < 0.001, $$$ {\eta}_p^2 $$$ = 0.02), as well as a significant interaction between role and the contrast code comparing the 2 months late condition to the never sent condition (F(1, 869) = 4.25, p = 0.04, $$$ {\eta}_p^2 $$$ = 0.01). These results indicate that the size of the perspective gap between givers and recipients was larger for both sending a late gift and not sending a gift at all than for sending an on-time gift and, importantly, was larger for not sending a gift at all compared to sending a late gift.

Taken together, these results offer evidence in support of Hypothesis 5. While we observe a perspective gap in both the late and never sent conditions, we see that the gap is larger when an individual fails to send a gift at all when it is expected than when they send a severely late gift. This is consistent with our theorizing that both late gifts and not sending a gift at all (when a gift is expected) are violations of gifting norms, but that not sending an expected gift is perceived as a more severe norm violation. Thus, late is, in fact, better than never when it comes to gift giving from the giver's perspective.

GENERAL DISCUSSION

In this research, we have demonstrated that gift givers systematically overestimate the negative impact that a late occasion-based gift (i.e., a gift given after a special event's discrete date) will have on their relationship with the recipient and have provided evidence of the psychological mechanism that underlies this perspective gap. In study 1, we demonstrated that gift givers' (vs. recipients') overestimation of relationship harm derives from the relatively greater importance that gift givers (vs. recipients) ascribe to gifting norms, consistent with prior research. Studies 2a and 2b then show that this mismatch is driven, more precisely, by gift givers' inference that sending a late gift will signal that they care about the recipient to a relatively lesser extent than what the recipient perceives. Study 3 then provided process evidence via moderation (Spencer et al., 2005) by showing that this overestimation of relationship harm is attenuated if the giver is able to signal care in some other way (i.e., with the amount of effort put into acquiring and creating the gift). Finally, in studies 4 and 5 we demonstrated that the magnitude of the giver–receiver perspective gap for late gifts is larger for moderately and severely late gifts than for gifts that are only mildly late (study 4) and that not sending a gift at all leads to an even larger giver–receiver perspective gap with respect to relationship harm compared to sending a gift late (study 5).

Theoretical contributions

Our work contributes to the gift-giving literature by furthering our understanding of the role of social norms in gift-giving (Givi et al., 2023; Givi & Das, 2023; Goodwin et al., 1990; Kupor et al., 2017; Teigen et al., 2005; Weinberger et al., 2024) by exploring the role of gift timing. With the exception of recent work by Givi and Galak (2022), prior work has largely overlooked the relevance and impact of when a gift is given. While Givi and Galak (2022) evaluated the benefits of giving non-occasion-based gifts at unexpected times, we show that timing also influences perceptions of gifts exchanged for occasion-based holidays (e.g., a birthday, Christmas). Specifically, we show that while late gifts are associated with relationship harm for both givers and recipients, givers systematically overestimate the harm a late gift will have on a relationship compared to what recipients anticipate. Thus, while prior work has shown that givers are often miscalibrated in terms of what gifts recipients would most like to receive (Baskin et al., 2014; Kupor et al., 2017; Lee-Yoon et al., 2020; Teigen et al., 2005), we show a new type of giver–recipient mismatch: Givers are more concerned about potential relationship harm from late occasion-based gifts than recipients.

This work also contributes to the growing literature on the signaling value of time (Bellezza et al., 2017; Donnelly et al., 2021; Keinan et al., 2019; Lee-Yoon et al., 2020; Whillans et al., 2021, 2022) by focusing on the interpersonal inferences consumers make from timing of gift-giving. Prior work has demonstrated how time use can signal human capital characteristics (Bellezza et al., 2017), specifically that lacking time (i.e., busyness) can signal status and competence. Relatedly, recent work on the signaling value of time in the workplace shows that workers are hesitant to request a deadline extension on a work assignment out of concern that doing so would burden their co-workers, leading others to see them as incompetent and inconsiderate (Whillans et al., 2021, 2022). The current work demonstrates a different way in which time may have perceived interpersonal consequences in the form of an anticipated signal of care (or lack thereof) for others. Further, similar to the way in which individuals' fear that time use can send a negative signal to others in the workplace, which drives workers to avoid asking for deadline extensions (Whillans et al., 2021), additional empirical work we have conducted demonstrates downstream consequences of givers' concern that a late gift can signal a lack of care and harm the relationship. Specifically, in two consequential choice studies preceding real holidays (Mother's Day and Father's Day), we show that gift givers who were given a choice between a free gift for their mother or father (respectively) chose a lower-value gift basket that would arrive on-time for the holiday over a higher-value gift basket that would arrive late (see Appendix S1: F and G). These studies suggest that fear of relationship harm may result in behaviors to actively avoid gift lateness, including potentially choosing worse (e.g., lower value) gifts if they are guaranteed to arrive on time.

Finally, our work also provides additional insight into the role that effort can play in signaling care for others. Just as caregivers eschew effort-saving products in contexts where it is important to them to show care for others (e.g., caring for an infant; Garcia-Rada et al., 2022), we demonstrate in study 3 that late gift givers may be able to mitigate their fear that a late gift will harm the relationship if they put extra care into creating and/or assembling the gift, thus using their own effort as a way to signal their care to the recipient even though the gift itself is late.

Practical implications

Our findings suggest a number of practical insights for marketers. First, given the role that we have identified that perceptions of care play in gift givers' overestimation of relationship harm from late gifts, it seems likely that marketers may benefit from emphasizing how much care a gift will signal after an occasion has passed (e.g., “Forgot Mother's Day? Show mom you really care with...”). Further, while we demonstrate that gift givers can signal care by putting more effort into the gift itself (i.e., through effortful acquisition and creation; study 3), our theorizing (along with prior work) would suggest that this could also be achieved through other means, for example, with how much money is spent on the gift (Flynn & Adams, 2009). Thus, it is plausible that marketers may benefit by offering higher-cost, premium options after an occasion-based gifting holiday (in contrast to the usual strategy of offering steep discounts) as a way to encourage late gift givers to signal care via how much money they spend on the gift. Our findings additionally suggest that, because the documented perspective gap is driven by beliefs that failing to ensure that the gift was delivered on time will signal less care for the recipient, marketers may also benefit from allowing consumers to offload the blame to the firm for a gift being late. For instance, a firm could potentially provide an option to include a note with the gift indicating that it would have been delivered earlier, but that there was a supply chain issue or unavoidable shipping delay that was outside of the sender's control.

Aside from identifying tactics that could potentially limit gift givers' overestimation of relationship harm from late gifts, our work also offers practical implications with regards to encouraging consumers to send a gift when they may have otherwise foregone the decision to purchase one. Specifically, in study 5 we demonstrate that an even larger perspective gap in anticipated relationship harm occurs when consumers do not give a gift at all, compared to when they send a gift that is severely late. Marketers may benefit from emphasizing that consumers should still send a gift after an occasion has passed, even if the gift is (very) late as our studies suggest that consumers will be less concerned about potential relationship harm if they do decide to send a late gift rather than if they opt to not send a gift at all. Our work therefore also has implications for consumer welfare, as it suggests that sending a gift late will result in fewer giver worries about the state of the relationship than not sending a gift simply because it is going to be late.

Limitations and future research

The present research is subject to a number of limitations. First, while we examined gift givers' overestimation of the relationship harm that occurs from giving a late gift using two different occasion-based gifting contexts (birthdays and Christmas) and tested potential downstream consequences using other occasions (e.g., Mother's Day and Father's Day), we did not compare the magnitude of the documented perspective gap between different types of holidays. It is possible that gift givers' overestimation of relationship harm may vary, for example, between one-way gifting contexts (e.g., a birthday, where only the giver is giving a gift) and reciprocal gifting occasions (e.g., Christmas, where both the giver and recipient are exchanging gifts), especially if the other person does send an on-time gift in the latter context. Further, given that gifting norms for some occasions are relatively more ambiguous (e.g., weddings, where the acceptable timeframe to send a gift has evolved in recent years; Martin, 2018), it seems likely that gift givers' overestimation of relationship harm may differ for gift-giving occasions where the norms around gift timing are less well-defined, such that it may not be clear that a norm has been violated (as we observed to be the case when a gift was only mildly late in study 4).

Beyond comparisons between different types of gifting occasions, the present work also does not distinguish between different types of giver–recipient relationships. In other words, we focused on those giving (and receiving) gifts from friends, but it seems likely that the effects we have demonstrated might differ if the relationship were closer (e.g., gifting to a family member) or more distant (e.g., gifting to a coworker). We also explored only one type of effort (effort in creating a gift) as a moderator of the asymmetry we explore. Future work could explore effort in delivering a gift as a moderator as well as whether spending more money on a gift actually does compensate for a late gift in the giver's mind.

Finally, another area for potential future research is with regard to individual differences between gift givers. The current work was limited in that we used samples of participants in the United States, but, given that social norms in general (and gifting norms in particular) vary throughout the world (e.g., Shen et al., 2011), it seems likely that the effects we observed may differ with cross-cultural samples. Aside from cultural differences, we expect that there are also a number of other individual differences that likely affect the perspective gap we have identified. For example, gift givers higher in self-monitoring (Lennox & Wolfe, 1984) or those who are more relationally oriented (Whillans et al., 2022) may be likelier to overestimate the relational harm of late gifts to an even greater extent. We leave an exploration of these and other questions for future research.

CONFLICT OF INTEREST STATEMENT

We have no conflicts of interest to disclose.

Open Research

DATA AVAILABILITY STATEMENT

Data for all studies, as well as pre-registrations and syntax, are available on the Open Science Framework (https://osf.io/aurz8/?view_only=fe210b8902574cffa6a775aabc681e70).