“I” Give, But “We” Give More: the Impact of Identity and the Mere Social Information Effect on Donation Behavior

Abstract

Three field studies and one laboratory experiment explore the”mere social information” effect: When donation contributions are affected by information about another’s level of giving. Experiments 1a and 1b demonstrate the effect. Experiment 2 shows it can occur when identity congruence is high. Experiment 3 shows that the effect is a joint function of identity congruence, the strength of identification, and the collective nature of underlying focal thoughts at the time of deciding the amount to give. Theoretical and substantive implications of these findings are discussed. Running Head: Identity, Social Information And Donation Behavior 3 ”I” Give, but “We” Give More: The Impact of Identity and the Mere Social Information Effect on Donation Behavior The impact of social information (e.g., information about what others have done in the past) is important in a variety of marketing environments (e.g. Reingen, 1982). In the absence of other information, consumers often rely on social information as a basis to make their own consumption decisions (Amaldoss & Jain, 2005). In this research, we examine a particular kind of social information: another person’s contribution to a nonprofit entity—and how that affects contribution behavior. We try to delve into the identity implications of this “mere social information” effect. This package of studies makes several contributions to the field of marketing and to the knowledge of psychological processes in consumption and donation decision making. First, it is the first study in marketing, to our knowledge, to demonstrate in a field study setting that social information significantly influences the level of contributions. There has been a call for consumer research to focus on more “messier” environments in order to buttress the mundane realism of the body of knowledge in the field (Cohen, 2005). Moreover, we examine a fundraising environment that is different from those used in previous research in nonprofit marketing, where donors were solicited by individual fundraisers before they had made a decision about whether they wanted to contribute. Instead, our research examines the influence of social information on consumers who have already crossed the motivational threshold to give or not (i.e., donors who have called in to a public radio station in order to contribute to the fundraising campaign). In this environment we can observe the effect of social information and identity on the amount of contribution that is cond itional on a consistent 100% response rate. Secondly, there is a growing body of work in consumer behavior that illuminates the importance Running Head: Identity, Social Information And Donation Behavior 4 of activated identities on consumption decision making (Reed 2004; Forehand, Reed and Deshpande 2002) and how identities shift as a function of contextual and situational cues (cf. DeMarree, Wheeler and Petty 2005; Mandel 2003). We demonstrate and build on this research in the domain of donation behavior. Third, an examination of key identity based constructs in the substantive domain of nonprofit giving should be informative for nonprofit marketing and consumption behavior research. These theoretical and practical contributions should also be informative for practitioners in charge of figuring out how to maximize charit able contributions. OVERVIEW OF EMPIRICAL WORK In the first two field experiments we show a very subtle pattern of results we refer to as the “mere social information” effect: Awareness of even a single other’s contribution amount can affect the target’s contribution level (Experiments 1a and 1b). In Field Experiment 2, we show that the effect can occur due to the congruence between the source of the social information and the target donor along the gender identity dimension. This result compellingly argues that the congruence of the target’s identity to the other is one possible precursor for the results (see Sirgy 1982). Finally, in order to examine this effect more closely, we conducted a laboratory study to test at least one possible mechanism through which the kind of social information and congruent identity used in the previous field studies influence behavior. Our results suggest that the kind of effects found may be identification based (as opposed to compliance based). More specifically, consistent with prior theorizing on identity activation (Reed 2004, Forehand et. al 2002), the impact of social information as a reference point for one’s own contribution decision is most potent when an identity is activated through congruence with a contributing other, when that identity is chronically important to the giver, and when focal thoughts about how much to give are relatively more collective and interdependent in nature (cf. Mandel 2003). Our results Running Head: Identity, Social Information And Donation Behavior 5 suggest that social information influences behavio r when the social information connects with the giver. Identity driven processes impact donors to give more, because “we” give more than “I” do. We begin by reviewing the relevant marketing literature on acquiescence to requests for charitable giving (i.e., compliance based techniques). We discuss how the environment that we use and the conceptual nuances of the story our data tell are quite different from those in previous research. We then present three field experiments that show the influence of socia l information and its identity based implications. Possible psychological mechanisms through which the effect may occur are then discussed within the findings of the laboratory experiment. We then discuss the contributions and limitations of this research, and its implication for future research and for practitioners. THEORETICAL BACKGROUND Compliance and Normative Influence. A large literature examines how to increase participation rates of contribution by using social influence techniques. For example the foot-inthe-door technique (Freedman and Fraser, 1978) involves making an initial request that is small enough that virtually everybody will comply. The second (actual) request is then made. Participants who are exposed to both requests are significantly more likely to comply with the second than those who are exposed to only the second request (see Burger 1999 for a review and meta-analysis). Evidence suggests that individuals’ desire to be (or appear) consistent causes the increased compliance rate (Guadagno, Asher, Demaine & Cialdini, 2001). The low-ball technique (Cialdini et. al., 1978) also triggers consistency pressure. This technique involves securing agreement with a request that is only partially specified, then naming more demanding or onerous details for its fulfillment. This is a technique typically used by car salesman. After Running Head: Identity, Social Information And Donation Behavior 6 they made customers agreed to a certain model, they kept adding on more and more features that are worth at least a few hundred dollars more (Cialdini, 1993). This effect has been replicated in getting compliance to answer highly personal interview questions (Hornik, Zaig, & Shadmon 1991), and in nonprofit marketing, where Brownstein and Katzev (1985) showed that this technique increased participation rate for an additional contribution request after asking for an initial donation in a museum. Doorin-the-face (Cialdini, Vincent, Lewis, Catalan, Wheetler and Darby, 1975) relies on consistency and it appeals to guilt and the dissonance mechanisms of guilt alleviation. This technique is also sequential requests. Here, the first request is large enough that virtually everybody will deny. The second (actual) request is then made. Participants who are exposed to both requests are significantly more likely to comply with the second than those who are exposed only to the second request (see Fern, Monroe & Avila 1986 a review and metaanalysis). More recent meta-analysis suggests that an individual’s desire to alleviate guilt from rejecting the first request causes the increased compliance rate with the second (O’Keefe & Hale, 1997). Finally, legitimization-of-small-donations also relies on guilt alleviation by highlighting the lowest-cost option available. The most commonly tested phrase for this technique is “Even a penny will help.” Past research shows that adding this phrase increases the participation rate without changing the contribution amount (Brockner, Guzzi, Kane, Levin, & Shaplen, 1984). These aforementioned techniques are overt and compliance based, that is—they all rely on an aggressive interaction between the fundraiser and the donor, and work because they trigger some kind of normative “pressure” to respond to internal or external social sanctions and rewards (Fishbein and Ajzen 1974). Conceptually such influence is most likely temporary, because normative influence affects behavior, or expressed attitudes through a compliance based process (Kelman 1961) without affecting true, underlying attitudes. When internal or external pressures Running Head: Identity, Social Information And Donation Behavior 7 are removed or attenuated (e.g., when decisions are made in private vs. public, or once guilt has been alleviated), then influence should be correspondingly mitigated (Burger, 1999). Moreover, consumers are savvy and may--over time learn to be more defensive, in that a "fool me once-shame on you--fool me twice, shame on me" sentiment may take over in the long run. This kind of "schemer schema" may render normative influence strategies ineffectual at motivating long term behavioral change (Wright, 1986). Informational and Identification based influence. We examine a subtle form of influence in which the mere presence of social information may affect behavior through informational and identification based processes (Kelman 1961). For example, researchers have found that individual’s product evaluations can be significantly affected by quality ratings ostensibly provi