Self-Gifts Through the Lens of Attribution Theory

Corinne Faure, University of Florida
David Glen Mick, University of Florida
ABSTRACT - This paper examines some of the antecedents of self-gift likelihood. Specifically, we explore how the three dimensions of causal attributions developed by Weiner (1986)Clocus, controllability, and stabilityCcan serve to predict self-gift likelihood in achievement contexts. Using previous research on emotions and on deservingness, we propose that causal attributions can lead to self-gift behavior through two routes, one affective and the other cognitive. Nine hypotheses are developed. We conclude by examining some ways by which our hypotheses can be tested through correlational and causal approaches.
[ to cite ]:
Corinne Faure and David Glen Mick (1993) ,"Self-Gifts Through the Lens of Attribution Theory", in NA - Advances in Consumer Research Volume 20, eds. Leigh McAlister and Michael L. Rothschild, Provo, UT : Association for Consumer Research, Pages: 553-556.

Advances in Consumer Research Volume 20, 1993      Pages 553-556


Corinne Faure, University of Florida

David Glen Mick, University of Florida


This paper examines some of the antecedents of self-gift likelihood. Specifically, we explore how the three dimensions of causal attributions developed by Weiner (1986)Clocus, controllability, and stabilityCcan serve to predict self-gift likelihood in achievement contexts. Using previous research on emotions and on deservingness, we propose that causal attributions can lead to self-gift behavior through two routes, one affective and the other cognitive. Nine hypotheses are developed. We conclude by examining some ways by which our hypotheses can be tested through correlational and causal approaches.


In the consumer behavior literature, there has been new and increasing research attention to the concept of self-gifts. Empirical work focused on self-gifts has so far been mostly exploratory (e.g., Mick and DeMoss 1990a) or descriptive (e.g., Mick and DeMoss 1990b, Mick, DeMoss and Faber 1992), which is not surprising during the earliest theory-building stages of a novel construct. We believe it is time for some self-gift research to shift to a hypothetico-deductive arena wherein past insights from self-gifts research are combined with significant theoretical developments in the social sciences to produce empirically testable claims about self-gift behavior (see also Mick 1991). The main purpose of this paper is to identify certain systematic factors that increase the likelihood that consumers will give gifts to themselves and to generate hypotheses about these factors for future investigation.

According to Mick and DeMoss (1990b), two of the most prevalent contexts of self-gifts (and thereby two types of self-gifts) are rewards for accomplishments and therapeutics for disappointments, associated with positive and negative life situations respectively. This insight points to achievement tasks and outcomes as possible triggers for giving gifts to oneself.

Not only may self-gifts originate from achievement situations, but more importantly from the interpretations consumers make about the outcomes. Relative to this point, Mick and DeMoss (1990b) intimated that attribution theory might be useful in understanding self-gifts. Attribution theories concern the way by which individuals try to make sense of the world in which they live and of the actions of the people around them, including themselves. Therefore, just as people make causal attributions about the behavior of others, they also make similar attributions for their own actions. In achievement situations they will infer different causes for the outcomes of the tasks they engage in. In this paper, we review Weiner's (1986) attribution theory of motivation and emotion, and we suggest ways by which achievement task outcomes, causal attribution dimensions, deservingness, and emotions influence the likelihood of a self-gift. On the basis of these insights, we generate specific hypotheses for future research and discuss potential methodological approaches to test these hypotheses.


Gift-giving and Self-Gifts

One of the characteristics of gifts in general, and of self-gifts in particular, is that they are typically highly motivated and, thereby, perceived as justifiable. For example, gifts are offered on culturally normative occasions determined by calendar (e.g., Christmas, birthdays) or determined by emergent life situations (e.g., to celebrate a job promotion, to cheer up someone who is ill). Whatever the occasion, it is a common trait of gifts that they are offered under some pretext, and that this pretext helps to make the gift acceptable to the receiver (Tournier 1963). The fact that individuals readily identify some self-directed purchases as self-gifts and distinguish them from other self-directed purchases (see Mick and DeMoss 1990a) points to the existence of normative criteria for distinguishing between these different types of purchases. This fact also suggests the existence of a partial rationalization process for self-gifts, similar to those identified for gifts to others. Indeed, Mick and DeMoss (1990a) found that personal accomplishments, disappointments, and holidays were the three prevalent circumstances under which people give gifts to themselves. In further research, Mick and DeMoss (1990b) identified three essential aspects of self-gifts. First, self-gifts are a means to communicate with one's self, and in particular to influence one's self-definition and self-esteem. This function of communication appears especially relevant to self-gifts following accomplishments or disappointments. Second, self-gifts are a special type of exchange, acting as a kind of self-contract. Mick and DeMoss (1990b) pointed out that this notion of exchange is particularly linked to reward self-gifts and illustrated by the theme of deservingness. As we will discuss, deservingness and exchange can also be associated with therapeutic self-gifts. Third, there is an aspect of specialness that refers to the idea that self-gifts often have special meanings for consumers as compared to common, everyday, self-directed purchases. This aspect also accrues from the highly context-bound character of self-gifts. In sum, the insights gained from prior self-gift research point to the potential importance of achievement contexts as common precursors of self-gift behavior.

Bernard Weiner's (1986) attributional theory of motivation and emotion seems particularly well-suited for elucidating certain self-gifts. This theory focuses on achievement contexts and identifies different dimensions along which people make causal ascriptions and the consequences different ascriptions have on specific emotions. Moreover, this theory offers two main characteristics that make it particularly attractive for understanding self-gift behavior. First, Weiner (1986) is explicitly concerned with issues concerning the self. Second, the theory is based on laws, rather than individual differences, and stresses the situational specificity of behavior. Given the highly context-bound character of the situations in which people indulge in self-gifts identified in previous research, Weiner's theory seems appropriate for self-gift research.

Achievement Outcomes

According to Weiner (1986), achievement outcomes first generate attribution-independent emotions. Failure produces a general feeling of sadness, while success leads to happiness, both feelings being independent of the specific cause ascribed to the success or failure. Moreover, to the extent that the outcome is either particularly important or unexpected, individuals will engage in a process of causal ascriptions in order to explain the outcome.

Previous research in social psychology has addressed the effects of success and failure on concepts such as self-gratification (e.g., Mischel, Coates and Raskoff 1968) and deservingness (e.g., Feather 1992), which are closely linked to self-gift issues. Mischel et al. (1968) found that children who experienced success in an achievement task were more likely to indulge in self-gratification than non-successful children. This finding might also apply to Western adults and self-gift behavior by virtue of the Protestant work ethic and its underlying achievement values (McClelland 1961). That is, as Mick (1991) has discussed, success in everyday Western life often leads to an "I earned it" attitude that translates into consumption indulgences. This is not to say, however, that self-gifts do not follow from failed achievement contexts. Further research by Underwood, Moore, and Rosenhan (1972) refined Mischel et al's (1968) finding by showing that while children tend to be more generous toward others and themselves after a success, they still indulge in self-gratification after a failure, but share less with others. In general though, we posit:

HI: In achievement contexts, successes will lead to greater self-gift likelihood than failures.

Causal Attributions and Emotions

Weiner (1986) proposes three major dimensions that underly causal inferences: locus (internal/external), controllability (low to high volitional control), and stability (low to high). According to his theory, these dimensions are orthogonal and they generate different attribution-dependent emotions. The first dimension, locus, refers to the perceived origin of the cause for the outcome. This dimension appears as a continuum from internal (outcome due to something within the individual him/herself) to external (outcome due to something outside or beyond the individual). The second dimension, controllability, concerns the degree of volitional control ascribed to the cause. This dimension is orthogonal to the first in the sense that control can be exercised by oneself (internally) or by someone else (externally). For instance, success on an exam can be attributed to an internal controllable cause (hard work), an internal uncontrollable one (a general positive mood), an external controllable one (the teacher likes you), or an external uncontrollable one (luck). Finally, the last dimension identified by Weiner (1986), stability, refers to the perceived persistence of the cause. Ability, for instance, is perceived as a more stable cause of success than effort.

On the basis of these causal dimensions, Weiner (1986) suggests a number of attribution-dependent emotions. For successful outcomes, attributions to internal controllable causes (e.g., strong effort) tend to generate high self-esteem (a feeling of pride). As pointed out by Weiner (1986), attributions for success are also characterized by a self-serving bias, which means that people will tend to attribute most of their successes to internal controllable causes. An external controllable success generates feelings of gratitude toward the external agent, while in general uncontrollable successes (both internal and external) tend to elicit more general feelings of luckiness and happiness. Compared to successful outcomes, failures result in a greater variety of outcome-related affect (Weiner 1986). For example, a failure attributed to an internal controlled cause (e.g., weak effort) elicits guilt, whereas a failure attributed to an external controlled cause (e.g., a devious person) elicits anger. Feelings of shame are likely to be experienced for failures attributed to an internal uncontrolled cause (e.g., low intelligence), whereas an individual failing because of an external uncontrolled cause (e.g., bad luck) is likely to feel victimized while also experiencing feelings of self-pity and helplessness. These various emotions and the intensities with which they are experienced may have different implications for self-gift likelihood. Thus, Weiner's (1986) focus on the emotions generated by causal attributions suggests the existence of an affective route to self-gift behavior.

A complementary cognitive route to self-gift behavior is implicated in Feather's (1992) work. Feather (1992) has explored the effects of success and failure attributions on perceptions of outcome deservingness. His results show that an outcome itself (success or failure) is perceived as deserved when the actor is responsible for it, and when there is a fit between the value of the action and the value of the outcome. For instance, someone who passes a test because of hard work will deserve this outcome more than someone who passes it through cheating; conversely, the cheater will be perceived as deserving to fail the test, while the hard worker will not. It is important to note that Feather (1992) manipulates the controllability of the cause and keeps locus constant (internal). Also of interest is the fact that his dependent variable is outcome deservingness, that is to say, the extent to which the outcome obtained is perceived as deserved. This should not be confused with self-gift deservingness as identified by Mick and DeMoss (1990b). Indeed, while one can expect a match between these two types of deservingness for success situations (the individual who deserves to win also deserves a reward), a mismatch can be expected for failure situations (the individual who deserves to fail is not entitled to get a gift, whereas the one who does not deserve the failure might deserve a gift as a compensation for an unfair outcome). Nonetheless, Feather's (1992) research provides additional credence to Mick and DeMoss's (1990b) theme of deservingness as an important cognitive antecedent of self-gift behavior based on achievement outcomes and attributions.

To recapitulate, the preceding discussions on Weiner (1986) and Feather (1992) suggest two possible routes for the effects of causal attributions on self-gift likelihood. That is, there appears to be two parallel mediating routes for self-gift likelihood, an affective route (emotions experienced) and a cognitive route (sense of deservingness).

Specific hypotheses on the effects of the causal dimensions on self-gift likelihood can also be delineated. Self-efficacy theory (Bandura 1982) suggests that people monitor their own behavior as a consequence of the attributions they make for their own actions and outcomes. Specifically, Bandura (1982) posits that people motivate themselves in achievement contexts by making reward-incentives contingent on the attainment of certain performance goals. Then, people give themselves rewards when they attain their goals and are proud of it. On the other hand, they engage in self-criticism when they fail to attain goals, especially if they feel responsible for failing the goals. Previous research suggests that, following a successful achievement outcome, people feel particularly proud of their success when they can attribute it to an internal controllable cause such as effort (Bandura 1982). The effect of locus has been well-documented in the attribution literature, and there is some evidence that internal attributions after successful outcomes lead to higher self-esteem than external ones (Bandura 1982, Weiner 1986). Thus, one can expect self-gift likelihood to be highest following success when causal attributions are internal rather than external. The effects of controllability are less straightforward, but there is nevertheless a tendency to consider outcomes due to controllable causes as more valuable than those due to uncontrollable ones. Indeed, as Weiner (1986) points out, successes attributed to external controllable causes (e.g., a friend's help) tend to generate higher self-esteem than successes attributed to external uncontrollable causes (e.g., luck), because people tend to believe that the help from the external controllable cause was triggered by their own likable nature. Hence, self-gift likelihood following success should be greater when the cause is considered controllable rather than uncontrollable.

In failure situations, a reverse pattern is expected. Research on excuse-giving (Weiner, Amirkhan, Folkes and Verette 1987) has shown that people tend to avoid attributing their failures to internal controllable causes when making an excuse. This is due to the fact that this type of attribution generates guilt (Weiner 1986) and the negative outcome is perceived as deserved (Feather 1992). Bandura (1982) stresses that in this type of situation people do not engage in self-reward because that behavior would not be socially acceptable. By contrast, failures due to external causes generate emotions such as anger or self-pity, and generally minimize the degree to which the individual feels that he/she deserves the outcome realized. Similarly, failures due to uncontrollable causes also minimize the feeling that the outcome was deserved. Given the specific types of negative feelings experienced in those situations (mentioned earlier), the likelihood that consumers will engage in self-gift behavior to cheer themselves up should be elevated. That is, following a failed achievement outcome, self-gift likelihood should be higher when the cause is considered external rather than internal, and when it is considered uncontrollable rather than controllable.

The third dimension identified by Weiner (1986), stability, may come into play in a different manner. Weiner (1986) notes that stability affects expectancy of performance on future tasks. When the individual has reason to think that the causes for a successful outcome are internal and stable (e.g., giving a good speech as a result of intrinsic personality qualities), he/she should anticipate similar results in the future. To this extent, a successful achievement on any particular speech-giving occasion may not be sufficiently special to trigger a self-gift. On the other hand, an attribution to an internal unstable cause for the same outcome (e.g., giving a good speech as a result of a detailed research effort regarding the specific topic) should make the outcome more special since it was not as predictable or expected due to the instability of the cause. Indeed, an attribution to an unstable internal cause as opposed to a stable internal cause should heighten the specialness of the situation, which then should increase the likelihood of a reward self-gift. Similarly, in a failed achievement context, an attribution to an unstable external cause (e.g., doing poorly in an annual race due to bad weather) should also be perceived as more special than an attribution to a stable external cause (e.g., doing poorly in an annual race due to the hilly nature of the course which doesn't suit your abilities). The instability of the cause lends more specialness to the situation, thereby potentially increasing the likelihood of a therapeutic self-gift.

In summary, the preceding discussion has articulated these formalized hypotheses.

H2: Following successful achievement outcomes, attributions to internal causes lead to greater self-gift likelihood than attributions to external causes.

H3: Following successful achievement outcomes, attributions to controllable causes lead to greater self-gift likelihood than attributions to uncontrollable causes.

H4: Following failed achievement outcomes, attributions to external causes lead to greater self-gift likelihood than attributions to internal causes.

H5: Following failed achievement outcomes, attributions to uncontrollable causes lead to greater self-gift likelihood than attributions to controllable causes.

H6a: Following successful achievement outcomes, attributions to internal unstable causes lead to greater self-gift likelihood than attributions to internal stable causes.

H6b: Following failed achievement outcomes, attributions to external unstable causes lead to greater self-gift likelihood than attributions to external stable causes.

H7: Following an achievement outcome, the effects of causal attributions on self-gift likelihood are mediated by emotional responses (affective route).

H8: Following an achievement outcome, the effects of causal attributions on self-gift likelihood are mediated by the extent to which the gift is perceived as deserved (cognitive route).


We now turn to a brief discussion on how the preceding hypotheses may be tested. One possible correlational approach would require that the subjects themselves classify causes of achievement outcomes. For instance, using the method of critical incident, as used by Mick and DeMoss (1990b), one could ask subjects to describe achievement occasions where they gave themselves a gift after a failure or a success. Then, using a causal attribution instrument developed by Russell (1982), subjects could be asked to rate the perceived causes for these outcomes in the described circumstances. Russell's scale enables the classification of causes along Weiner's causal dimensions, which would then facilitate insights about the most common dimensions of causal attributions preceding self-gift behavior. The main problem with this approach is that subjects self-select the situations, and most likely would describe situations where, according to our hypotheses, self-gift propensity is highest (internal, controlled attributions for successes and external, uncontrolled attributions for failures). Such a selection would not enable a full test of our hypotheses.

A more promising causal approach would involve manipulating the dimensions experimentally. One variation of this approach would be to create situations where subjects are actually asked to perform a task, and then receive feedback on how they performed and the reasons for their performance. Another variation would be to use imagined achievement scenarios. While the use of direct manipulations appears more appealing, the problems of credibility encompassed in dealing with feedback are quite big. Indeed, as noted by Weiner (1983), there are some difficulties in manipulating causes experimentally. Weiner (1983) identifies three sources of possible discrepancies between the instructions and feedback given to the subjects: the objective characteristics of the task, the life experience of the subject, and the experience of the subject during the experiment. Thus, in an experimental setting, one has to worry whether the subjects will really perceive the causes and their dimensions as they were intended to be manipulated. Manipulating the dimensions by imagined scenarios has risks as well, since it would require from the subjects that they project themselves into the situations, which presents the disadvantage of lower involvement and less concreteness than real-life situations. On the other hand, attribution dimensions can be manipulated with more control, and this approach has been used in the past by other attribution researchers (e.g., Feather 1992). Thus, experimental research with the use of scenarios would enable researchers to manipulate the achievement outcomes and to manipulate the dimensions of causal attributions. In particular, based on Weiner's (1986) theory, it appears quite possible to manipulate orthogonally the dimensions of locus and controllability in an unstable context. A 2 by 2 by 2 experimental design (outcome - success/failure, locus - internal/external, controllability - controlled/uncontrolled) appears to be a reasonable way to test most of the hypotheses developed here.

The third dimension identified by Weiner, stability, appears more difficult to manipulate. Indeed, it may be impossible to have experimental subjects really believe that they are always bad or good at a particular achievement task (e.g., tests, sports). Thus, while it appears possible to manipulate experimentally some unstable causes, stable causes appear much more difficult given the contrived setting of an experiment. One option, however, would be to measure beforehand the stable attributions individuals typically make of their achievements in certain contexts. On the basis of these measures individuals could then be assigned to experimental groups where the attributions would be manipulated in an unstable way. Thus, stable attributions would be used as a blocking factor. It should be the case that when stable and unstable attributions are congruent, individuals tend to react less than when they are at odds with one another. Imagine for instance the case of a subject who tends to perform poorly in a specific sport and thinks that he/she does not have the abilities required to be a good performer at this sport. Now, put him/her in a situation where he/she wins a tough competition because of hard training. The likelihood that he/she will give himself/herself a gift as a reward for his/her performance should be higher than for another student who tends to perform well at this sport and believes that he/she has the required abilities to do so. Assigning subjects to experimental groups on the basis of the most important dimension of causal attributions (locus) seems quite promising.


The objectives of this paper were to develop theoretically-derived hypotheses about the antecedents of self-gift likelihood and to propose ways to test these hypotheses. Through the effects that causal dimensions have on both cognitive and affective processesCas described or implicated in Weiner's (1986) attribution theoryCwe believe it is possible to identify specific achievement situations in which people give themselves gifts, thus extending our understanding of self-gift behavior beyond a description of the phenomenon. We believe that this approach offers two important contributions. First, it expands our view of the substantive topic of self-gifts, which appears to be a fairly common and important phenomenon in Western consumer behavior. Second, it also extends the usefulness of attribution theories as they have been used so far in the consumer behavior literature.

Although researchers have long ago perceived the potential of attribution theories for explaining consumer behavior (e.g., Folkes 1984, 1988), attribution theories have not been used to their full potential. These theories have been used to predict the reaction of individuals toward causal agents (e.g., consumer reaction toward the company who is considered responsible for a product failure, Folkes 1988) or to study the motivational effect that attributions have on future behavior in the same activity (e.g., the tendency to exert more or less effort at work depending on the attributions made about one's prior performance, Sujan 1985). By contrast, this paper considers attribution theories for the consequences attributions have on consumer buying behavior, particularly self-gifts. Hence, this paper extends the application of attribution theories in consumer research.

Finally, as discussed in the beginning of this paper, Mick and DeMoss (1990b) identified an exchange dimension relevant to self-gift behavior. Research on resource exchange theory (Brinberg and Wood 1983) and on mental accounting (Thaler 1985) suggests that people give special meanings to the things they exchange or buy, and tend to match what they buy with the reasons why they engage in buying or exchanging in the first place. Thus, of interest in the framework we have developed is to test whether there are any differences between the types of self-gifts one buys after a success or after a failure (e.g., purchasing a new sweater vs. going to an expensive restaurant), and whether one can identify a mental match between the type of self-gift, the achievement outcome (success or failure), and the causal attributional dimensions. Such insights would extend our framework further and yield a more complete understanding of the self-gift phenomenon.


Bandura, Albert (1982), "The Self and Mechanisms of Agency", in Psychological Perspectives on the Self, Vol. 1, ed. Jerry Suls, Hillsdale, NJ: Erlbaum.

Brinberg, David and Ronald Wood (1983), "A Resource Exchange Theory Analysis of Consumer Behavior", Journal of Consumer Research, 10 (December), 330-338.

Feather, N. T. (1992), "An Attributional and Value Analysis of Deservingness in Success and Failure Situations", British Journal of Social Psychology, 31 (2), 125-145.

Folkes, Valerie S. (1984), "Consumer Reactions to Product Failure: An Attributional Approach", Journal of Consumer Research, 10 (March), 398-409.

Folkes, Valerie S. (1988), "Recent Attribution Research in Consumer Behavior: A Review and New Directions", Journal of Consumer Research, 14 (March), 548-565.

McClelland, David C. (1961), The Achieving Society, Princeton, NJ: Van Nostrand.

Mick, David Glen (1991), "Giving Gifts to Ourselves: A Greimassian Analysis Leading to Testable Propositions", in Marketing and Semiotics: Selected Papers from the Copenhagen Symposium, eds. Hanne Hartvig Larsen, David Glen Mick, and Christian Alsted, Copenhagen: Handelshojskolens Forlag, 142-159.

Mick, David Glen and Michelle DeMoss (1990a), "To Me From Me: A Descriptive Phenomenology of Self-Gifts", Vol. 17, Advances in Consumer Research, eds. Marvin E. Goldberg, Gerald Gorn and Richard W. Pollay, Provo, UT: Association for Consumer Research, 677-682.

Mick, David Glen and Michelle DeMoss (1990b), "Self-Gifts: Phenomenological Insights from Four Contexts", Journal of Consumer Research, 17 (December), 322-332.

Mick, David Glen, Michelle DeMoss and Ronald J. Faber (1992), "A Projective Study of Motivation and Meanings of Self-Gifts: Implications for Retail Management", Journal of Retailing, 68 (2), 122-144.

Mischel, Walter, Brian Coates and Antonette Raskoff (1968), "Effects of Success and Failure on Self-Gratification", Journal of Personality and Social Psychology, 10 (4), 381-390.

Russell, Dan (1982), "The Causal Dimension Scale: A Measure of How Individuals Perceive Causes", Journal of Personality and Social Psychology, 42 (6), 1137-1145.

Sujan, Harish (1986), "Smarter Versus Harder: An Exploratory Attributional Analysis of Salespeople's Motivation", Journal of Marketing Research, 23 (February), 41-49.

Thaler, Richard (1985), "Mental Accounting and Consumer Choice", Marketing Science, 4 (3), 199-214.

Tournier, Paul (1963), The Meaning of Gifts, Richmond, Virginia: John Knox Press.

Underwood, Bill, Bert S. Moore and D.L. Rosenhan (1973), "Affect and Self-Gratification", Experimental Psychology, 8 (2), 209-214.

Weiner, Bernard (1983), "Some Methodological Pitfalls in Attributional Research", Journal of Educational Psychology, 75 (4), 530-543.

Weiner, Bernard (1986), An Attributional Theory of Motivation and Emotion, New York: Springer-Verlag.

Weiner, Bernard, James Amirkhan, Valerie S. Folkes and Julie A. Verette (1987), "An Attributional Analysis of Excuse Giving: Studies of a Naive Theory of Emotion", Journal of Personality and Social Psychology, 52 (2), 316-324.