The Role of Labeling Processes in Elderly Consumers' Responses to Age Segmentation Cues

This investigation explores labeling processes underlying age segmentation cue effects on discount usage intentions. Depth interviews regarding participants' experiences using senior-citizen-type discounts reveal three levels of responsiveness to consumer offerings promoted with age segmentation cues: rejecting senior citizen discounts to avoid self-devaluation, rejecting senior citizen discounts to avoid stigmatization, and assigning positive meanings to the status that promotes senior citizen discount usage. An experimental investigation, undertaken to assess the sequential ordering of these levels of responsiveness, reveals that self-devaluation and perceived stigma mediate age segmentation cue effects on discount usage intention only for younger-aged elderly. Results lend support for a stage model of consumers' progression through phases of responsiveness to "senior citizen" labeling. Copyright 1994 by the University of Chicago.


marketers' attempts to target them explicitly on the basis of age.
This investigation offers labeling-theory-based explanations for consumers' resistance to discount offerings promoted with age segmentation cues. Labeling theory, which originated in social psychological studies of deviance to explain causes and outcomes of labeling individuals with stigmatized statuses (see Lemert 1951), provides two explanations for low participation rates in senior citizen discount programs. Given that an elderly status may be viewed as a stigmatized attribute that evokes negative connotations (Ward 1977), elderly consumers may reject senior-citizen-type discounts to avoid self-devaluation that might occur with personal acceptance of the status (resisting private self-labeling) and/or to avoid stigmatization from others who become aware of the status (resisting social labeling).

LABELING THEORY
Early labeling theory conceptions (Lemert 1951;Schur 1971) and applications (Scheff 1966) emphasized societal reaction as the force responsible for defining acts and actors as deviant. For example, Scheff (1966) described a social labeling process in which labeling patients "mentally ill" would negatively affect the patient's social interactions via devaluation and discrimination by others responding to the label. The labelee's internalization of the "mentally ill" label was posited to maintain and potentially promote deviant behavior.
More recent conceptions of labeling theory recognize individual motivations to resist personal acceptance of labels imputed by social audiences. Esteem motivations direct individuals to resist acceptance of or association with negative labels because such labels pose a threat to their self-concept (Allen 1982). Consistency motives, which prompt individuals to reject negative labels that are incongruent with their self-image, accentuate the esteem motive in directing behavior, because the selfimage of some individuals is subject to greater "spoilation" as a result of violating a specified social norm. For example, Rotenberg (1974) notes that public labeling as a "thief' will "spoil" the self-concept of the business professional to a greater extent than that of the wayward pauper. Sirgy's (1982Sirgy's ( , 1986) research on self-concept in consumer behavior similarly suggests that consistency motivations combine with esteem motivations in directing the purchase of labeled or branded products that become associated with undesirable user images.
Recent research further recognizes that personal acceptance of a negative label may occur through a private self-labeling process (Thoits 1985). Individuals can label themselves a "failure" in reaction to actual or imagined norm violations, even though the violation is undetected by others (Rotenberg 1974). Private self-labeling can occur if (1) the actor shares the normative expectations of larger society, (2) there is some degree of cul-tural agreement as to the rule-breaking behaviors that result in attachment of stigmatized labels, and (3) the actor is motivated to conform to social expectations (Thoits 1985).
Researchers have applied Goffman's (1963) notion of stigma (i.e., an attribute that deviates from expectations is discrediting and results in disgrace or shame) to labels resulting from a variety of "deviant" statuses, including (1) "physical stigmas" or deviations from physical norms, (2) "tribal stigmas" of race, nation, religion, and social status, and (3) "blemishes of individual character" resulting from violations of cultural social norms. Findings from these studies suggest that the labelee's interpersonal interactions are inhibited as a result of the label (e.g., Gray, Russell, and Blockley 1991; Harris et al. 1990; Ray and Lee 1989). Kuypers and Bengtson's (1973) social breakdown syndrome theory describes public labeling effects on the elderly's social interactions and self-image. According to the theory, Western society, which values productivity and achievement, conveys negative expectancies of the elderly, who eventually adopt a self-image as "incompetent." However, social breakdown syndrome theory possesses several shortcomings. First, previous research provides mixed support for the theory's assumption that society holds negative stereotypes of the elderly (see Brubaker and Powers 1976). Second, other stigmatizing attributes may be associated with an "old" age status (e.g., loss of physical attractiveness; Ward 1977). Finally, social-breakdown syndrome theory does not allow for individual differences in the elderly's personal acceptance of socially applied labels or recognize the potential for private self-labeling. Given these limitations, study 1 was designed to assess the experiential relevance of core theoretical constructs suggested by current labeling theory to the elderly's use of senior citizen discounts.

STUDY 1
Study 1 identified characterizations of senior citizen discount users that are potentially stigmatizing and determined the relevance of private self-labeling and social-labeling processes to elderly consumers' responses to discounts promoted with age segmentation cues.

Method
A snowballing technique generated a purposive sample of 38 elderly consumers, age 50 or older, who agreed to participate in personal, depth interviews. Because labeling theory suggests that labeling processes are initiated by institutional or public attempts to label individuals, setting the demarcation age at 50 was desirable, as it represents the earliest age at which marketers label consumers with age-based labels. (American Association of Retired Persons [AARP] membership cards, which can be obtained at age 50, are accepted by a va-riety of retailers, entitling the holder to senior citizen discounts; see Heilman 1990.) Participants were 20 females and 18 males, all caucasians from a middle-class socioeconomic bracket. Twenty-five (65 percent) were high school graduates and 10 (27 percent) possessed college or professional degrees. Eleven participants (29 percent) were age 50-54, 21 (55 percent) were age 55-64, and six (16 percent) were 65 years of age or older. Thirty three (87 percent) were married, and 19 (50 percent) were either retired or previously retired but now working.
Participants were informed of the general purpose of the study, informed that the interview would be audiotaped, and assured anonymity. The interview format had five parts: (1) a request for an example of an occasion on which the participant had used a senior citizen discount, (2) the participant's description of the occasion, (3) probing questions designed to elicit details of each usage episode, (4) follow-up questions to obtain an exhaustive list of each participant's usage occasions and settings, and (5) closing questions to assess the participant's familiarity with senior citizen offerings, regardless of previous usage.
Participants' first description of a usage occasion was elicited with the general statement, "Tell me about any experiences you've had using senior-citizen-type discounts." Follow-up and probing questions were developed in early interviews. Responses in these interviews also resulted in additions to the interview questionnaire. For example, when one woman answered the introductory question by tracing the changes in her experiences through time, the "working" questionnaire was expanded to include questions such as "How did you first become aware that you were eligible for seniorcitizen-type discounts?" "Did you begin using senior citizen discounts as soon as you became eligible?" and "Do you recall the experience that led you to use senior citizen discounts for the first time?" The semistructured interviews were conducted by the author in the respondents' homes and lasted about 20-45 minutes.
Transcriptions of the audiotaped interviews were examined for the presence of content suggested by labeling theory. The emergence of unanticipated categories and themes rendered a partially inductive process reflective of the constant comparative method (categories and themes were revised through successive comparisons with new data; Glaser and Strauss 1967). The coding of the data involved three separate coding iterations, with categories and themes at each stage agreed upon by a second researcher who had not participated in the fieldwork and could therefore bring a different eye to the analysis (Eisenhardt 1989). The initial coding involved searching described encounters across all interviews for categories that captured similar experiences. Identified categories, which were not mutually exclusive, made up broader themes including (1) characterizations of senior citizen discount users, (2) perceived or experienced evaluations and reactions from sales-clerks, other customers, and/or other people in the shopping party, (3) participants' affect resulting from the usage experience, and (4) recognition of personal eligibility for senior citizen discounts.
The second stage of analysis involved a part-towhole assessment of each participant's transcript. Movement through the data in the first and second coding iterations provided an overall view that highlighted acceptance of a senior citizen status and perceived reactions and evaluations associated with senior citizen discount usage as the most prominent differences among participants.
In the third stage, participants similar on these dimensions were grouped together. Interviews were then compared and contrasted within and between groups. This process led to the refinement of themes that better discriminated among the individuals with respect to their avoidance of negative self-concept and social outcomes via consumer responses to senior citizen discounts. This final stage of analysis also illuminated possible interrelationships among the themes.

STUDY 1: IDENTIFIED THEMES
As illustrated in Table 1, participants' comments revealed that the senior citizen status encountered in consumer settings conflicts with a more youthful selfconception and raises concerns over others' evaluations and reactions. Participants indicated that they were uneasy about "approaching that age" or "getting old," and expressed concerns that others associated senior citizen discount usage with several stigmatizing characteristics related to an old age status (e.g., appearing physically older and financially limited).
The stigmatizing nature of financially limited characterizations (i.e., senior citizen discount users "don't have extra money" and "have very little funds") was reflected in participants' concerns that others would view senior citizen discount users as financially dependent, unable to afford the full price, and lower-status customers or "second-class citizens" who should not be entitled to the same services (see Table 1).
The senior citizen status and associated characterizations were also perceived to lead to negative social reactions. As illustrated in Table 1, participants relayed experiences involving embarrassment, ridicule, and the denial of equal consumer rights with respect to accessing or receiving desired service outcomes, including expected conduct during the interaction. Thus, participants' concerns paralleled the conceptualization of perceived stigma as the belief that others will devalue (negatively evaluate) or discriminate (negatively react) against individuals who have acquired a labeled status (Link et al. 1989).
These stigmatizing aspects of a senior citizen status reflect the motivations and/or meanings that are associated with two dominant themes identified from the interview data: (1) information management strategies

SELF-DEVALUATION AND PERCEIVED SOCIAL EVALUATIONS AND REACTIONS RESULTING FROM SENIOR CITIZEN DISCOUNT USAGE
Examples from participants' comments Self-devaluation: Loss of youthfulness "We're so prone to want to be young in America especially. It hurts when you first hit 50 and you realize that you're eligible for this discount-it's a psychological thing" (f, 59).
"Many of us refuse to accept the fact that we've approached that age. Pride might enter into it but I think that would be chief of my objections. The fact that you go through life refusing to accept certain things and that would be one of mine. I've noticed people using them but it seems like it's something that's around the corner and not imminent" (m, 60). Social evaluations: Older physical appearance "At 55 you can't accept it but say at 57 or 58 you begin to. They offered it to me for the first time in a bookstore when I was 55 and I was really insulted. I felt like they were sort of looking at me, like they could tell, you know" (f, 58). Financially limited: Financially dependent "The first thing people think is it's a charity program with a senior" (f, 50). Unable to afford full price "They might feel self conscious, especially if they're not used to using it and it's a first-time purchase, that type thing. And they're sort of feeling their way: 'Is this right?' or 'Am I doing the right thing?' 'What will people think?' That type thing." "So you think some people might be self-conscious about what other customers think?" "I do. Other customers may think the person just can't afford more than that so they're getting the product at that price" (m, 55). Lower status or class "I think a person might tend to feel like they (the waiter or maitre d') would think if you're in a high-class restaurant like this, why do you need a discount. You know: 'You shouldn't be here if you need the discount' " (f, 54). Undeserving of equal rights "They may think that other people will think they couldn't afford (the price). They might think, what is this person behind me thinking. . . . The person paying full price may think they should have more privilege or maybe go first or pay first or something like that" (m, 50). Social reactions: Embarrassment or ridicule "Recently, three couples of us were going to Florida and we stopped at a Residence Inn.. . . I just simply asked would my AARP card help reduce the price of the room.. . . But I felt like two of the men were extremely embarrassed that I had pulled my AARP card out and one of them even had an AARP card in his wallet and had not pulled it out, he admitted later." "What led you to believe they were embarrassed by it?" "Well, they just ragged me and kidded me about pulling the AARP card there in the lobby" (f, 54). Denial of equal rights with respect to desired service outcomes "I've found that you don't get the same treatment if you go in and ask for that [AARP] discount that you do if you don't ask for it." "How do they treat you if you ask for it?" "Well, you're liable to get one of their lesser rooms and that type thing. And they might even tell you that they don't have a room." "Do you take advantage of any discounts available with this card?" "When I'm traveling I do get discounts from AAA rather than AARP, for some reason." "Are the discounts higher?" "No, usually they are about the same but they accept that better than they do the AARP, they seem to. I don't know if it's because of a relationship that's established between older people and them, with more of them being difficult to deal with or what it is, but there seems to be a difference whether you go in and say 'Hey I want a AAA discount' or going in and saying 'Do you have an AARP discount?' Or you can offer them both and they will take the AAA. You can't have both. But you can say 'Hey, do you take an AARP or AAA' and they'll say 'Yeah, we'll take a AAA discount.' I don't know why.... If I'm in a crowded situation and I know the room is tight, I go with the AAA.. . . I walked in a motel one time and this couple was older than I was but . . . he just told them he didn't have a room. And I walked up and didn't ask for any discount and I got a room" (m, 57). NOTE.-Concerns regarding others' evaluations were often expressed as rationales for someone else's unwillingness to use senior citizen discounts. These comments are assumed to be self-referential projections for participants who later in the interview acknowledged that they avoided usage for similar reasons or reported usage patterns consistent with the expressed concerns for others' evaluations. that consumers employ to avoid the negative effects of senior citizen labeling (i.e., self-devaluation and stigmatization) and (2) levels of responsiveness to age segmentation cues.

Information Management Strategies
The information management strategies employed by participants, in order of increasing disclosure of the Resisting private self-labeling as a senior citizen: Status denial "I guess I've noticed them but I haven't looked Age 50-54: perceive self-devaluation, at the age where I could use them so I haven't perceive stigma, resist usage thought about it. In shopping with my mother though or for her, if I had seen they offered it, I would ask" (f, 54). Status rejection "Shoney's for one, gives a senior citizens discount.... I'm 56 and I believe they give it at 55, but I never even think about it. Ruth is in her 60s and she never asks for it. But yet, I'm aware that it's there if you want it. I think that's just getting old. I don't think I've reached that age where I felt like I really need a discount, like at where you eat" (f, 56).

Resisting social labeling as a senior citizen:
Secrecy "I take advantage of them [senior citizen Age 55-64: perceive self-devaluation, discounts] in my banking . . . and don't pay for perceive stigma, do not resist checks and get a discount on my safety deposit usage box . . . motels that have offered it and restaurants that have offered it." "Which restaurants?" "The only one I recall is called Perkins." "Is that here in town?" "[No,] Jackson, Mississippi" (m, 57). Selective concealment "People usually have friends with like circumstances and if they were with their friends who had a card they'd use it, I'm sure. If they were with friends or acquaintances that did not have the card, they might not use it" (m, 56). Redefining the senior citizen status: Unqualified disclosure "Well, I use them to get my hair done all the Age 65 plus: do not perceive selftime over at First Place and save close to $2. devaluation, do not perceive Quincy's, Hardees, Burger King. I go over here stigma, do not resist usage to Burger King sometimes and get a sandwich. I always ask for it. I always say 'I'm a senior citizen'.. . . He's usually the one to rent the motel.room, so he uses his AARP card. But if I was not with him and it said that they had the discount, then I would use it [AARP card] anytime. If I was going to a motel and they didn't have a sign up, I would ask them if they had a discount for senior citizens" (f, 68). senior citizen identity, are (1) status denial, (2) status rejection, (3) secrecy, (4) selective concealment, and (5) unqualified disclosure (see Table 2). Status Denial. Participants practicing status denial refused to acknowledge eligibility for senior citizen discounts, stated that they had never used senior citizen discounts, and disavowed the personal applicability of the interview topic. Two women indicated that they may be selectively screening out promotional content informing them of their own eligibility while securing such discounts for their mothers. One of them stated, "I'm aware of it [senior citizen discounts] probably because mother's, you know, at the age and there are several places that if I go to the drugstore to pick up her medicine, she gets the senior citizen discount there. And then there's several restaurants occasionally that she and I will be. . . . I'm not looking forward to getting 60" (f, 57). Following direct questioning both participants reported awareness of AARP's "50 and older" eligibility criteria. Nevertheless, status deniers viewed themselves as different from senior citizens, whom they characterized as "old." The previously quoted woman continued: "I don't come into much contact with things that it makes a difference that I'm 57, you know. We go to Walkers, and Cobbs Corner, Nickies and John's downtown, O'Charlie's and Olive Garden." "I'm surprised Walker's doesn't." "There are so many old people in there. . . . I haven't been in there in a year and a half. But always on Friday night, that place is loaded with senior citizens, so they might and I don't know it." Status Rejection. Individuals practicing status rejection recognized their eligibility but resisted senior citizen discount usage because they did not want to accept a senior citizen status. Status rejecters associated the senior citizen status with "being old'' and admitted sacrificing desirable seniors' benefits to prevent the "spoiling" of their more youthful identity. The dialogue with one man (age 55, who, just prior to the interview, shrugged off congratulations on becoming a first-time grandfather) revealed this esteem motive: "I do relate to a senior citizens' program that was mailed to me that I've thrown away and it was from one of the hospitals where you get cheaper parking and these kinds of things. And I haven't followed through on it. I think agewise I'm eligible but I haven't followed through on it. I don't know if I'm following the question or not." "Yes, just any experiences you've had. I'm wondering why you threw this away or why weren't you interested in it?" "Because I don't want to admit that I'm old enough to use it." "Is it something that you had a need for?" "Yeah. I think it would probably be nice to have but I just don't want to take advantage of it." Secrecy. Similar to Becker's (1963, p. 11) "secret deviants," several participants revealed a willingness to use senior citizen discounts on occasions that enabled them to conceal their usage from others. The comment of one woman (age 60) that there are "definitely" people who "might not want to get the card out in front of people because they might be a little embarrassed" appeared to be self-referential projections of her own motivations for secrecy via mail transactions.
"Are you a member of AARP?" "Yes and Silver Service . . . checking is free." "Have you ever used any seniors' discount cards at restaurants?" "No." "Have you requested or received one anywhere before?" "No, I'm just beginning to get there." While these secret users avoid detection by any audience, operative motivations are provided by the potential for salesclerks to devalue senior citizen discount users as "old," "unattractive," or "having money problems," as well as react negatively toward them (e.g., being "impatient" or "rude," and denying service; see Table 1). However, attributing participants' use of secrecy exclusively to store employees' reactions is difficult, because store employees' reactions can also make the senior citizen discount transaction more visible to others. As one participant recognized, the clerk could call out " 'Hey, I've got a senior citizen discount card here for Ms. so and so. Do we honor these?' and then everybody in the store knows immediately that you're using a discount card" (f, 54).
Another form of secrecy involved efforts to avoid stigmatization from familiar others by using senior citizen discounts only in out-of-town transactions, often for hotel accommodations and restaurants. One man, age 61, noted that "a lot of people do things in secret that they don't do among their friends." In addition to expressing concerns that friends might think that they are "old" and/or "can't afford to pay full price," participants described incidences in which familiar onlookers "acted embarrassed," "about crawled under the table," or "got a charge out of it" when they or other shopping companions either requested a senior citizen discount or were offered one by a salesclerk.
Selective Concealment. Whereas secretive users eschewed all local face-to-face transactions, participants practicing selective concealment avoided usage of senior citizen discounts only in the presence of dissimilar others with whom they were either familiar or newly acquainted. Participants' frequent remarks that they hesitated to use senior citizen discounts in the presence of others who were younger, financially better-off, or both (as indicated by reference to "yuppies") reflected concerns for others' evaluations of them as "old" and/or "financially limited" and the reactions stemming from these attributions.
There are people who will not take advantage of them. I guess for two or three reasons. One would be that they don't want to admit they're old enough to be in that age group. And another, people may think they're cheap or can't afford to pay the full price if they do take advantage of it. It depends on who they're with at the time whether they can use that discount. . . . If I'm among people that I know real well, then I wouldn't hesitate to use the discount and ask for it. Whereas if I'm with people that I don't know their reaction to it, and whether I would continue to fit in with that group, for whatever reason, say a seminar where everybody goes out... until I get to know them. Maybe after we had been through a familiarity period it wouldn't make a difference. [m, 59] Unqualified Disclosure. Several participants reported frequent usage of senior citizen discounts, with disclosure of the status being unaffected by the store setting or the presence of others. Participants practicing unqualified disclosure perceived positive evaluations and experienced positive affect as a result of senior citizen discount usage (e.g., one man recounted his eagerness to display his senior citizen discount cards in front of ineligible others, announcing, "When you get my age, you can get one of these too"). These frequent users viewed senior citizen discounts as "rewards" offered for past inputs of "valued" and "deserving" elderly, as opposed to need-based incentives offered to assist financially limited elderly.

Levels of Responsiveness to Age Segmentation Cues
Synthesis of the preceding information management strategies in terms of whether concerns for self-devaluation or stigmatization served as the primary motivation for rejecting or accepting senior citizen discounts revealed three levels of responsiveness to consumer offerings promoted with age segmentation cues (see Table  2). Resisting private self-labeling as a senior citizen (i.e., avoiding self-devaluation via the rejection of senior citizen discounts) is reflected in status-denial and statusrejection strategies. Resisting social labeling as a senior citizen (i.e., avoiding stigmatization via the rejection of senior citizen discounts) is evidenced in secrecy and selective concealment strategies. Redefining the senior citizen status (i.e., assigning positive meanings to the status that promotes senior citizen discount usage) is evidenced in unqualified disclosure.
Resisting private self-labeling was manifested by individuals who were primarily concerned that their youthful self-image would be blemished by association with the senior citizen status. These individuals expressed concerns about becoming associated with the senior citizen status whether or not the transaction was visible to salient others. Because concerns for self-devaluation motivated unconditional rejection of senior citizen discounts, resisting private self-labeling represents the lowest level of responsiveness to age segmentation cues.
Resisting social labeling was manifested by individuals who rejected the senior citizen status as a social image. Many of these individuals seemed to accept the senior citizen status as a self-image, as evidenced by their lack of concern for self-devaluation and their willingness to use seniors' discounts in transactions that were less visible to salient others. Because concerns for stigmatization curtail but do not completely preclude senior citizen discount usage, resisting social labeling represents a higher level of responsiveness to age segmentation cues than resisting private self-labeling.
Redefining the senior citizen status to connote positive meanings was exhibited by individuals who demonstrated not only a lack of awareness or concern for the negative effects of labeling (i.e., self-devaluation and stigmatization) but who attached desirable self-references (e.g., pride) and social references (e.g., respect) to the senior citizen status. Individuals manifesting positive redefinition conveyed that, during early periods of eligibility, they had viewed the senior citizen status as undesirable. Their comments revealed that redefinition of the senior citizen status occurred through (1) identifying and recognizing potential contributions of senior citizens and (2) reversing stigmatization, whereby rejection of senior citizen discount offerings was labeled "irrational" or "vain." These individuals reported the highest level of responsiveness to consumer offerings promoted with age segmentation cues. A woman, age 59, stated "I think I had not accepted the fact that I was over 50 and each time you present that card you're saying 'I'm over 50.' And it took me a couple to three years before I would accept the fact that I had reached that age in life. It took me that long, I couldn't accept that I was that old. And once I accepted it I thought 'Hey, why not?' because you can get some nice discounts, especially in motels. And now with my company, I kid them about the fact that I am very, that I won't cost them as much money as the young people will because of the senior citizen discounts. Sometimes I even get 20 percent." "What happened that helped you decide to use these discounts?" "Other friends would say 'Hey, we can get a discount at so and so with AARP.' So when I realized my other friends were using it, I thought 'My goodness, I'm really silly not to use it.' " Remarks of other current users suggested that resistance to private self-labeling may be an initial reaction to senior citizen discounts that dissipates over time. Hence, it may be that the three levels of responsiveness form sequentially related phases of a stage model characterizing consumers' progressive acceptance of consumer offerings promoted with age segmentation cues. Transition into the senior citizen status may begin with the initial exposure to explicit public labeling as a senior citizen, which might constitute a sudden and involuntary loss of the elderly consumer's self-perceived youthfulness (a condition that produces negative affect; cf. Belk 1988). Participants relayed experiences in which they or their friends felt "killed," "devastated," "insulted," "offended," or "miffed" by a salesclerk's unelicited offer of a senior citizen discount. However, labeling may become internalized when consumers experience other age-related transitions. Participants commonly relayed that their initial requests for senior citizen discounts followed retirement and/or increasing health problems, conditions associated with the acceptance of "older" self-perceived age (see, e.g., Ward 1977).
Following self-acceptance of a senior citizen status, resisting social labeling may postpone complete responsiveness to senior citizen discounts. Several participants' accounts of their initial usage reflected con-cerns for the evaluations or reactions of others. Other accounts revealed that, after postponing usage for more than several years, new users continued to monitor reactions to their behavior, concealing usage from select friends and acquaintances.
Consumers may eventually seek congruity between a private self-image as a senior citizen and a more "youthful" social self-image, given the motivation to resolve discrepant selves (see Goodwin 1992). Congruity may be achieved through redefining the senior citizen status to connote positive meanings. Redefinition is consistent with Brubaker and Powers' (1976) suggestion that upon acceptance of an "older" identity the elderly adopt a positive stereotype of old.
Although plausible, the preceding stage model conceptualization of the relationship among the three levels of responsiveness to age segmentation cues warrants more systematic investigation. A second study was conducted to validate the roles of self-devaluation and perceived stigma in explaining consumers' responses to age segmentation cues and to test the proposed stage model.

STUDY 2
Study 2 experimentally tested a labeling theory explanation of how age segmentation cues influence elderly consumers' discount usage intentions. To assess the levels of responsiveness identified in study 1, two variables, self-devaluation and perceived stigma associated with discount usage, were tested as mediators of the effect of age segmentation cues on discount usage intention. Study 2 also investigated the moderating effects of two variables identified in study 1 as augmenting the influence of age segmentation cues: self-perceived age and the social visibility of the transaction. The following section outlines hypothesized mediation and mediated-moderation relationships and describes how the proposed stage model was tested by introducing chronological age as a blocking variable.

Research Hypotheses
Study 1 interviews revealed that "old" characterizations associated with senior citizen status hold negative connotations for some consumers, who resist using discounts promoted with age segmentation cues to avoid self-devaluation. Several participants reported that they rejected senior citizen discounts because acknowledging the personal applicability of the status through usage was a distressful self-admission. Thus, reflective of consumers' resistance to private self-labeling, responsiveness to age segmentation cues may engender self-devaluation that inhibits consumers' willingness to use a discount. Hla: Self-devaluation will mediate the relationship between age segmentation cue and discount usage intention. (Marrs 1984). Study 1 depth interviews suggested that consumers resist usage of senior citizen discounts because it links the user to an old age status that results in perceived social devaluation and discrimination. In addition, as a "master stigma," a "senior citizen" label may accentuate other forms of negative labeling connoted by an older age, such as "financially limited." Thus, reflective of consumers' resistance to social labeling, age segmentation cues used to promote discounts may result in perceived stigma associated with discount usage that inhibits consumers' willingness to use a discount.

Hlb: Perceived stigma will mediate the relationship between age segmentation cue and discount usage intention.
Consistent with labeling theory, the depth interviews suggested that self-devaluation resulting from senior citizen discount usage was accentuated by inconsistency between consumers' more youthful self-image and their image of senior citizen discount users as "old" (cf. Rotenberg 1974). Usage of a discount in which an age segmentation cue identifies the user as older should "spoil" the self-image of those who view themselves as young or middle-aged (i.e., those with a younger age identity) to a greater extent than those who view themselves as older or elderly (i.e., those with an older age identity; see Cutler 1982). Therefore, consumers with younger age identities should be more likely to reject discounts promoted with age segmentation cues because incongruity between the consumer's youthful self-image and "old" images evoked by senior citizen labels accentuate concerns for self-devaluation. H2: Self-devaluation will mediate the effect of the interaction between age segmentation cue and age identity on discount usage intention.
Both younger age identity and high social visibility transactions are expected to augment age segmentation cue effects on perceived stigma. Gerontologists have found that the elderly attempt to avoid negative societal labeling by holding fast to views of themselves as young or middle-aged (Bultena and Powers 1978). Hence, age segmentation cue effects on discount usage will be greater when age identity is "younger" versus "older" because younger age identity consumers will perceive greater stigma associated with discount usage.
Although social visibility of the discount transaction should not moderate the effect of age segmentation cue on self-devaluation (given that self-devaluation can occur even in the absence of real or perceived negative social reactions or evaluations via the private self-labeling process), high social visibility transactions should evoke greater perceptions of stigma associated with discount usage than low social visibility transactions. Concerns for stigmatization motivated several study 1 participants to restrict their usage of senior citizen discounts to transactions that concealed their senior citizen identity. Hence, high social visibility transactions should augment perceptions of stigma associated with using senior citizen discounts (thereby reducing willingness to use a discount), particularly among elderly consumers possessing a younger age identity.
H3: Perceived stigma associated with discount usage will mediate the effect of the interaction between age segmentation cue, age identity, and social visibility on discount usage intention.

Figure 1 graphically depicts hypothesized mediation and "mediated moderation" relations as the indirect main and interaction effects, respectively, on discount usage intention that occur through self-devaluation and perceived stigma (see Baron and Kenny 1986).
Chronological age is introduced as a moderator of these mediation and mediated moderation relations to assess whether consumers' responsiveness to age segmentation cues reflects a stage model that culminates in redefinition of the senior citizen status. Specifically, resistance to senior citizen labeling is expected to decrease and responsiveness to consumer offerings promoted with age segmentation cues is expected to increase with chronological age. Findings that the presence of an age segmentation cue does not affect self-devaluation, perceived stigma, and discount usage intention for the oldest-aged elderly, despite the presence of these effects for younger-aged groups, would be consistent with "redefinition" as the proposed final stage.

STUDY 2: DESIGN AND METHOD An Overview of the Experimental Design
The experiment was a two (age segmentation cue: present or absent) by two (social visibility: high or low) by two (age identity: younger or older) between-subjects experimental design. Each of the four age segmentation cue by social visibility treatments were operationalized as a department store's discount program offered through the mail. The discount allowed consumers to save 10 percent off their total bill on all future purchases and could be received at the time of purchase by presenting the card issued by the store (labeled "XYZ Store"). An illustration of the purchase area and the card were juxtaposed with the description to enhance the strength of the manipulations (see App.). Following presentation of the stimulus material, subjects completed items measuring perceived stigma, self-devaluation, discount usage intention, manipulation checks, perceived credibility of the discount, and age identity, in sequence. Age Segmentation Cue. Presence of an age segmentation cue was operationalized by referring to the discount as a "senior citizen discount" offered to customers "over a certain age."' Absence of an age segmentation cue was operationalized by referring to the discount as a "privileged customer discount" offered to "special" customers.
Social Visibility. The high social visibility condition was operationalized by requiring the prospective discount recipient to present a specially designed discount card at a register designated for the discount card holders only. In the corresponding illustration, the discount name appeared prominently on a sigii hanging above the register and on the discount card. The low social visibility condition was operationalized by requiring the prospective discount recipient to present a discount card, which looked like and replaced the department store's credit card, at any register. The card was coded so that the clerk would be aware that the customer receives the discount but would not be aware of the reason for the discount. In the corresponding illustrations, the discount name did not appear on the sign or the card.
Age Identity. Age identity, as a labeled self-perception of age, was viewed as more appropriate than the cognitive age scale that yields perceived age in years (e.g., a response of "60 years of age" would not clearly indicate whether subjects viewed themselves as "young" or "older"; Wilkes 1992). The age identity response categories ("young," "middle-aged," "older," and "el-'Two alternative operationalizations of a present age segmentation cue condition were designed to validate this manipulation: (1) a "senior citizen discount" offered to customers "over 50 years of age," and (2) a "privileged customer discount" offered to customers "over a certain age." These alternative operationalizations were administered to two groups serving as control groups on the present age segmentation cue/high social visibility treatment group. Findings regarding the demographic profile, potential nonresponse bias, and effectiveness of the manipulations paralleled findings for the present age segmentation cue/high social visibility treatment group. The MANCOVA and ANCOVA conducted for the overall sample and age group subsamples indicated that, generally, findings presented here would be replicated by replacing the present age segmentation cue/high social visibility treatment with either of the alternative operationalizations. Similar to results produced by the chosen present age segmentation cue operationalization, discount usage in the control group employing the "privileged customer" label in addition to an age segmentation cue, as compared with the absent age segmentation cue/high social visibility treatment group was associated with greater self-devaluation and perceived stigma. (Although, in contrast, the presence of an age-segmentation cue in conjunction with a privileged customer label negatively affected discount usage intention for the overall sample.) However, because a privileged customer discount offered to consumers "over a certain age" may not be the affective equivalent of a privileged customer discount offered to "special" customers, the employed operationalization of the absent age segmentation cue condition may be a limitation of the present research. Future research should consider alternative operationalizations that are less affectively charged (i.e., input-based incentives that would not engender age segregation of the store environment in the present age segmentation cue condition).

Levels of Responsiveness Requisite R&elafionships Moderators
Resisting Private Self-labeling ASC + Self-devaluation -e Discount Usage Because "younger" as compared (Hla) Intention with "older" age identity consumers will associate greater selfdevaluation with the usage of discounts accompanied by age segmentation cues, ASC effects on discount usage intention through self-devaluation will be greater for consumers with "younger" age identity. (H2)

Resisting Social Labeling ASC -> Perceived Stigma)'Discount Usage Because high social visibility and
(Hlb) Intention "younger" age identity heighten perceived stigma associated with age segmentation cues, ASC effects on discount usage through perceived stigma will be greater in high social visibility transactions when age identity is "younger' versus "older." (H3) Redefining the ASC has no effect on self-devaluation, Senior Citizen Status perceived stigma or discount usage intention.
derly") were adapted from the most frequently employed age identity measure in gerontology studies (Bultena and Powers 1978). The "older" label was employed to replace the "old" response option in previous measures because it provided (1) a less negatively charged option, (2) a more viable response alternative to "elderly" for subjects who have relinquished a middle-aged self-conception but do not view themselves as very old or aged, and (3) clarity in the implicit ordering of the labels in terms of a young-old gradient (see Cutler 1982). In the analyses, "young" and "middle-aged" responses were collapsed to form a "younger" age identity subsample; "older" and "elderly" formed an "older" subsample.

Pretest of Stimulus Materials. During a pretest, 75
subjects age 50 years and over, randomly assigned to one of the four main treatments, provided responses to an open-ended questionnaire administered in personal interviews and then to a fixed-alternative questionnaire, which they returned by mail. Content analysis of their reactions and opinions, using two judges to assess reliability, indicated that there were no significant differences across treatments in perceived difficulty of using the discount or the effort required to receive the discount. However, discounts described in the absent age segmentation cue condition were perceived as less credible than those in the present age segmentation cue condition (16 percent as compared with 54 percent; X2 (1,75) = 11.89, p < .00 1). Participants' comments suggested that the lack of eligibility criteria for the absent age segmentation cue operationalization largely contributed to this difference. Thus, all scenarios were revised to state that the discounts were mailed to "customers who have been coming to the store for a long time." Remaining concerns for credibility in both age segmentation cue conditions reflected general attitudes toward discount offers (e.g., "you can't get something for nothing" or "the percentage off is going to be included in the price anyway"). Finally, checks for demand characteristics included in the mail survey revealed that no one correctly guessed the purposes of the study.

Measurement of Constructs
Discount Usage Intention. Respondents indicated their intention to use a 10 percent discount for a purchase amount of $10. The three-item, summed scale was made up of two five-point items ("extremely interested" to "not at all interested" and "extremely likely to use" to "not at all likely to use") and one 11 -point probability scale. The alpha coefficient was .82 (range = .78-.85 for treatment groups). The validity of this measure was assessed by submitting the usage intention scale as a dependent variable and a behavioral measure (operationalized as the subject's return or failure to return a request form for additional information on the described discount) as an independent variable in an ANOVA. Subjects who returned the request card indicated a greater intention to use the discount described in the scenario than subjects who did not return the request cards (F(1,514) = 26.18, p < .0001; Xdid return = 17.56, Xdid not return = 15.5 7). Further analyses revealed that the predictive validity of the usage intention measure held across scenarios; neither behavior by age segmentation cue nor behavior by treatment interaction effects were significant.

Self-Devaluation Associated with Discount
Usage. An item representing loss of self-perceived youthfulness and two global items (e.g., "Some people may not use this discount to maintain a positive image of themselves") made up the self-devaluation scale. The three-item scale (five-point, "'strongly agree" to "strongly disagree") yielded an alpha coefficient of .79 (range = .74 to .81 for treatment groups).
Perceived Stigma Associated with Discount Usage. The perceived stigma items examined perceived devaluation and perceived discrimination for each of three social audiences: store employees, other customers in the store, and others in the user's shopping party. Perceived devaluation was operationalized with three items for each social audience representing others' views of the user as "old," "physically unattractive," "financially limited," plus a global measure. Sample items include the following: "Store employees may think that users of this discount have less money to spend than the average customer," "Other customers may view the user of this card as a 'second class citizen,' " and "If someone uses this discount, the people they are shopping with will be less likely to view them as youthful." The construct also captured the perceived discrimination imparted by these social audiences through global items as well as items representing a range of actions: disrespect, amusement, embarrassment, impatience, pity, condescension, and reduction of desired service outcomes including assistance with locating outof-stock merchandise, responsiveness to complaints, and service quality. Sample items include the following: "The complaints and comments of the customer using this card may be taken less seriously by store employees than those of other customers," "Other customers may find it humorous that another customer would take advantage of this discount," and "People may act uncomfortable or embarrassed when someone they are shopping with requests this discount." The 23-item scale (five-point, "strongly agree" to "strongly disagree") produced an alpha coefficient of .94 (range = .91-.94 for treatment groups). All items were worded consistent with a projective technique and were interwoven with an additional 16 items (measuring positive reactions) to reduce acquiescence bias.
Perceived Credibility of the Discount. Theory and research on reference pricing suggests that consumers doubt the credibility of advertised savings (often communicated in the form of "X percent off'; Urbany, Bearden, and Weilbaker 1988) and "discount discounts" (Gupta and Cooper 1992). Perceived credibility of the discount may be compromised by frequent exposure to reference price advertising claims and habitual puffing, resulting in lower intention to take advantage of the discount offer (see Biswas and Blair 1991). Pretest interviews provided evidence that participants' general attitudes about the credibility of discounts affected their responsiveness to both senior citizen and privileged customer discount scenarios. Thus, study 2 treated perceived credibility of the discount (defined as perceptions of the discount as a bona fide offering) as a covariate. The five-item scale (five-point, "strongly agree" to "strongly disagree"; alpha = .91 for the full sample, range = .85-.94 for treatment groups) included items such as "I would think this was a very credible discount program." Subjects A mail survey was distributed to consumers who were identified through a snowball sampling technique. This method was selected for its ability to yield: (1) a high overall rate of response, (2) an approximately equal number of responses in all treatment cells, and (3) an unbiased sample in terms of subjects' orientations toward age segmentation cues. Sources assisted in the generation of a mailing list of 1,206 noninstitutionalized persons residing in two large southeastern cities. Subjects were assigned to treatment cells on the basis of age and sex in an effort to match cells on most demographic variables. The matching procedure was deemed desirable given that age, education, income, sex, and marital status have been found to influence age identity and/ or discount usage (see Bawa and Shoemaker 1987;Ward 1977).
Of the 1,206 questionnaires distributed, 784 (65 percent) useable questionnaires were returned. Total responses were distributed almost equally across treatment groups (range = 121-133). Significant differences emerged across treatments for chronological age (X2 = 25.72, df = 15, p < .05). For the 50-54 age group, lower rates of response were obtained in both treatment groups receiving the present age segmentation cue condition as compared with the absent age segmentation cue condition. There were no significant differences across treatment groups in education, income, sex, and marital status.
Potential nonresponse bias was assessed by comparing two successive waves of the questionnaire representing subjects responding to the first mailing and subjects responding after the added stimulus of a postcard reminder. Significant differences in chronological age emerged only for the two treatments in which an age segmentation cue was present; subjects responding to the second wave were more likely to be chronologically younger than subjects responding to the first wave. Age identity significantly differed across waves for the treatment group receiving a present age segmentation cue under a high social visibility condition (X2 = 4.1, df = 1, p < .05). Subjects indicating a young age identity were more likely to respond in the second wave (72.22 percent).

Manipulation Checks
The scale for the age segmentation cue manipulation check included three Likert-type items such as "Only people over a certain age are eligible for this discount" and yielded an alpha of .76 (range = .67-.82 for treatment groups). Strength of the social visibility manipulation was assessed by a five-item scale (five-point, "strongly agree" to "strongly disagree;" alpha = .76, range = .71-.78 for treatment groups) that included statements such as "Other customers may observe the customer taking advantage of the discount." Verification of the manipulations entailed submitting the manipulation check scales as dependent variables in two separate full-factorial ANOVAs. The age segmentation cue manipulation check yielded a main effects for age segmentation cue (F(1,775) = 549.92, p < .0001, w 2 =.41; Xpresent =I10.53, Xabsent = 6.55). Although a main effect for social visibility also emerged (F(1,775) = 6.44, p < .05, w 2 < .0 1), there was no evidence to render the discriminant validity suspect (w 2 was small relative to that for the main effect for age segmentation cue). As expected, only a main effect of social visibility influenced the social visibility manipulation check scale

Mediation Effects of Self-Devaluation and Perceived Stigma
Initial tests of the hypothesized mediation and mediated moderation relations involved assessment of main and interaction effects on self-devaluation and perceived stigma (see Baron and Kenny 1986) using MANCOVA, followed by univariate analyses. Age segmentation cue and social visibility treatment manipulations, age identity, and all two-way and three-way interactions were entered as independent variables, and perceived credibility of the discount was entered as a covariate in each model. Subsequent analyses involved tests of treatment effects on discount usage intention and, where treatments obtained parallel effects on the mediators and criterion, more rigorous assessments of mediation (i.e., partial correlation analysis of the mediators with the criterion, and covariance analyses of treatment effects on the criterion after removing variance attributable to the mediator). Assessment of mediation effects was undertaken for the full sample and for each of the four age groups (50-54, 55-59, 60-64, and 65 and older). Table 3 presents treatment effects for the multivariate and univariate analyses by age group; Table 4 presents the means corresponding to treatment effects on discount usage intention.
Although Hypotheses la and lb were not supported for other age groups, age segmentation cue did have the expected effect on the proposed mediators for the two middle chronological age groups. Self-devaluation was greater when an age segmentation cue was present versus absent: for the 5 5-59 age group, F(l, 126) = 19.04, C2 = .1 l, p = .0001; and for the 60-64 age group, F(l ,96) siveness to age segmentation cues. Whereas subjects in the youngest age group (i.e., age 50-54) were less likely to use a discount promoted with an age segmentation cue (to avoid private self-labeling and social labeling), chronologically older elderly (i.e., 65 and older) were not affected by the presence of a senior citizen label. (Results suggested that age segmentation cues did not threaten their self-image or perceived social acceptance and did not inhibit their responsiveness to a discount offer.) Responses of the middle-aged groups (age 55-64) suggested an intermediate degree of acceptance of the senior citizen label, although the "moderate" level of responsiveness exhibited by this age group departed from expectations in two ways. First, although these middle-aged elderly perceived that others do not accord senior citizen discount users the same regard and respect that more positively labeled discount users receive, age segmentation cues did not negatively impact their willingness to use a discount. Second, consumers in the middle-aged groups acknowledged that senior citizen discount usage was damaging to their self-image. Hence, middle-aged elderly demonstrated concerns for negative labeling but, unlike their younger-aged counterparts, they did not resist private self-labeling nor did they resist social labeling.
One explanation for this pattern of results is that, relative to their younger counterparts, middle-aged elderly have had greater exposure to senior citizen labeling, unelicited awards of senior citizen discounts (from salesclerks), and age peers who model discount usage behaviors. Consequently, middle-aged elderly become accustomed to using senior citizen discounts, even though their behavior engenders self-devaluation and stigmatization. The comments of one participant in study 1 support the argument that elderly consumers who have yet to accept the status may reluctantly use senior citizen discounts. "Those just barely 50, they won't quite admit it yet. But I think when you get 55 or 60 everybody knows you're over 50 and it really doesn't make any difference (f, 59)." The dissonance between discount usage behavior and self-perceptions is resolved when elderly consumers redefine the senior citizen status as a new and proud identity. Older elderly, who did not attach negative meaning to the senior citizen status, appeared to have achieved this level of responsiveness. Hence, the negative effects of senior citizen labeling prompted the development and resolution of dissonance between self-perceptions and behavior.
Although younger age identity and high social visibility did not augment the mediating effects of selfdevaluation and perceived stigma, the influence of self-perceived age and social visibility should not be dismissed, given limitations of the operationalizations employed. The phrasing of the age identity measure ("How do you view yourself?") may have evoked the cultural term most commonly attached to subjects' chronological age as opposed to a more personalized assessment. In addition, the "young" and "elderly" response options may possess value-charged connotations in Western society, which glorifies and sanctions youthfulness. The high social visibility condition, operationalized by making transactions observable by any social audience, may not have produced consistent conceptions of the makeup and characteristics of present others. For example, perceived stigma may be greater for elderly who envision other customers as younger as opposed to older (given that stigmatization is more pronounced in the presence of individuals who do not share the status; Goffman 1963). The absence of specific targets may have introduced variability that masked the effects of social visibility. Hence, future research should explore alternative operationalizations of age identity and social visibility.
The present investigation may be regarded as an important preliminary step in an area that warrants further research. Confidence in the stage model would be bolstered by longitudinal research that (1) documents elderly consumers' invariant progression through the levels of responsiveness uncovered in study 2, (2) investigates responsiveness to alternative operationalizations of age segmentation cue (e.g., other labels connoting age segmentation, elderly models, and products developed specifically for the elderly), and (3) compares the rate of progression through levels of responsiveness across communities that vary in terms of age segregation. Such research would extend the present findings by identifying the triggering forces that drive individuals from one stage to the next and, possibly, lay the foundation for building theoretical frameworks that describe individuals' acceptance of other stigmatizing statuses that are used as a basis for segmentation (e.g., overweight, disabled, low income, or homosexual).
Replication or extension efforts that validate the present findings would lend credence to marketing strategies that target the elderly on the basis of their stage of responsiveness to age segmentation cues. To enhance responsiveness from consumers resisting private self-labeling, marketers could develop discount programs that target the elderly without the explicit use of age segmentation cues (e.g., frequent-buyer discounts at pharmacies) or that disassociate age segmentation cues from stigmatizing attributes while reinforcing positive attributes of elderly (e.g., "seniors' " banking discounts offered for large balances). Given that age segmentation cues may elicit perceived stigma, reducing the social visibility of the transaction (e.g., separating the transaction area from the waiting queue, electronically tagging credit cards for senior citizens to ensure that they receive discounts automatically) may make eligible consumers more comfortable with using senior citizen discounts. Finally, to encourage responsiveness to age-related offerings, marketers may assist the elderly in redefining the senior citizen status by providing employee training on how to reinforce the senior citizen label as a desirable status during the interaction.

Sample Stimulus Material: Present Age Segmentation Cue/High Social Visibility Treatment Group
A department store sends the following announcement in the mail to customers who have been coming to the store for a long time. This announcement describes a discount offer as follows: Because you are over a certain age, our store is making a discount available to you, a Senior Citizen, to save 10% off your total bill on all future purchases.
In order to receive your 10 percent discount, you need only begin using the enclosed special membership card. There is no monthly or annual fee. This new department store credit card has been especially designed to identify you as a customer entitled to a discount. Use it in place of your regular department store credit card or as an identification card when making purchases by cash or check. You must present this card to your salesclerk at the special register designated for Senior Citizen card holders only. The clerk will enter the Senior Citizen Discount code number into the register which will automatically deduct the discount from your purchase amount. The special register and card are needed to make sure our staff identify you as a member of this program. We hope you will take advantage of this Senior Citizen Discount. After all, this is an offer we are making specifically to people your age! The illustration below (Fig. Al)