Judgment and Decision Making, vol. 3, no. 5, June 2008, pp. 404-410.

In the “I” of the storm: Shared initials increase disaster donations

Jesse Chandler* , Tiffany M. Griffin, and Nicholas Sorensen
Department of Psychology
University of Michigan

People prefer their own initials to other letters, influencing preferences in many domains. The “name letter effect” (Nuttin, 1987) may not apply to negatively valenced targets if people are motivated to downplay or distance themselves from negative targets associated with the self, as previous research has shown (e.g., Finch & Cialdini, 1989). In the current research we examine the relationship between same initial preferences and negatively valenced stimuli. Specifically, we examined donations to disaster relief after seven major hurricanes to test the influence of the name letter effect with negatively valenced targets. Individuals who shared an initial with the hurricane name were overrepresented among hurricane relief donors relative to the baseline distribution of initials in the donor population. This finding suggests that people may seek to ameliorate the negative effects of a disaster when there are shared characteristics between the disaster and the self.

Keywords: name letter effect, donations, charity, hurricane, Katrina.

1  Introduction

Katrina Petrillo raised $1000 for Hurricane Katrina victims by selling lemonade (Salkin, 2005). When asked why, she said “I realized my name is going to go down in history as one of the biggest storms ever” (Salkin, 2005). The current research explores whether Katrina Petrillo’s response to the association between Hurricane Katrina and her name might be more likely than what would be expected by chance. Specifically, we predict that donors who share an initial with a disaster will be more likely to donate to disaster relief efforts for that particular disaster than would individuals who do not share an initial with a disaster.

The prediction that individuals who share an initial with a disaster will be more likely to donate to fundraising efforts targeted at that disaster is consistent with research that has demonstrated that people prefer the initial of their first name over other letters. This preference, called the name letter effect (Nuttin, 1987), is reflected in both subjective ratings of initials (Nuttin, 1987) and in choices of same-initial targets over comparable targets with different initials (Brendl, Chattopadhyay, Pelham, & Carvallo, 2005). Shared initials have been shown to influence choices in numerous domains including product preferences, donations to political candidates, spouses and occupations (Pelham, Mirenberg, & Jones, 2002; Jones, Pelham & Carvallo, 2004). For example, one study found that people with names beginning with the letter “D” are overrepresented in the profession of dentists and people with names beginning with the letter “L” are overrepresented in the population of lawyers (Pelham et al., 2002). Similarly, people named “George” are more likely to live in Georgia than expected by chance and people named “Louise” are more likely to live in Louisiana (Pelham et al., 2002). Together, these findings indicate that similarities affect judgment and choice, presumably by creating a “unit-relation” (Heider, 1958) that facilitates the assimilation of positive feelings of the self into the representation of the target (implicit egotism; Pelham et al. 2002).

Psychologists have theoretically assumed that people should avoid associations with negative stimuli (Jones et al., 2004). This claim is based on several studies that have shown that people “cut off reflected failure” (Snyder, Lassegard, & Ford, 1986) of close others in order to protect the ego. For example, people are less likely to wear clothing displaying the logos of sports teams that have recently lost a game as compared to teams that have recently won a game (Cialdini et al., 1976). Similarly, Snyder and colleagues (1986) randomly assigned groups of college students to receive positive or negative false feedback on a group task. They found that compared to a no-feedback control, the successful group demonstrated more affiliative tendencies, while the failure group demonstrated less afilliative tendencies.

Under other circumstances, individuals may downplay negative features of a target with which they share a unit relationship. Finch and Cialdini (1989) found that subjects who believed that they shared a birthday with Rasputin rated him less negatively after reading a description of his misdeeds than those who did not. In a recent study, Nelson and Simmons (2007) extended this work by demonstrating that pairing an initial with a negative outcome undermines task performance. Specifically, they found that people whose first names begin with C and D initials have lower GPAs than people with other initials and baseball players with K initials (the letter used to represent strikes in baseball) are more likely to strike out than players with other initials, presumably because people tend to downplay the severity of the negative outcome, much as they downplay Rasputin’s deeds when they think he shares their birthday. In other words, people with first names that begin with C and D may implicitly downplay the severity of getting these lower grades, and consequently dampen efforts to perform well.

Extending research on implicit egotism and negative stimuli, we contend that people may not always minimize their associations with negative stimuli (Snyder et. al, 1986) or downplay their negativity (Finch & Cialdini, 1989; Nelson & Simmons, 2007). Clearly, Katrina Petrillo understood the negative impact of Hurricane Katrina. She claimed that sharing her name with a severe storm motivated her to help its victims. If she minimized her association with the storm, it is unlikely that she would attend to the common initial and consequently would feel no more motivated to donate money than anyone else. Similarly, if she downplayed the severity of the storm, it is unlikely that she would mention that it is the “worst storm in history” or invest so much time raising money. In this case, it is again unlikely that she would be so compelled to donate given previous research suggesting that people are less likely to donate to less severe disasters (Strömberg, 2007).

Katrina Petrillo’s efforts to raise money for the victims of Hurricane Katrina is contrary to both of the processes outline above, as well as to self-enhancement literature in general. (For a review see Leary, 2007.) However, there are circumstances when people may not avoid their association with negative targets, and may even make reparations for the actions of others with whom they share an association. For example, people will endorse compensatory behavior to right the wrongs committed by their ingroup against an outgroup (e.g., Brown et al., 2008; Doosje, Branscombe, Spears & Manstead, 1996) even if this compensatory behavior comes at a personal cost to the self. Doosje and colleagues (1996) found that this phenomena was mediated by feelings of collective guilt, suggesting that responsibility, like positive affect can transfer between members of a unit-relation. If group membership increases the likelihood of compensatory behavior for the actions of other group members, then perhaps other kinds of unit-relations can do the same.

Taken together, previous literatures on self-target associations allow us to speculate that Katrina Petrillo’s sentiment towards the victims of Hurricane Katrina may be shared by others. Shared-initials are capable of producing a unit relation and there is at least some evidence that a unit-relation with a target that inflicts harm can lead people to help their victims. This led us to expect that people would: i) be more likely to donate money to same-initial disasters; and possibly ii) donate more money to same initial disasters.

Table 1: K-initial donations by funding stream.

 Prior to KatrinaAfter Katrina
International & general funds
National disaster services
Hurricane season 2005
Hurricane Katrina relief
Other donors19272861568349592
K-initial847 8021210
Percentage of K-initial donors4.20%a2.39%a4.85%ab5.72%b9.80%b
Note. Proportions with differing subscripts are significantly different, p < .05. Donations prior to Hurricane Katrina occurred in the six months prior to August 29, 2005 (landfall in Louisiana). Donations after Hurricane Katrina occurred in the two months after August 29, 2005.

2  Method

2.1  Red Cross Donor Records

To test our predictions we examined archival donation records of a Midwestern chapter of the Red Cross. The database consisted of sixty-six thousand donations from 1930 to early 2006. The vast majority of donations were made by current residents of the county the Red Cross chapter was located in. Due to privacy concerns, the Red Cross released only the first name of the donor, the amount donated, the date of the donation and the donor’s wishes as to how the donation should be used (international or domestic relief, with the addition of disaster specific funds in 2005). The structure of the database also made it impossible to determine the number of unique donors who contributed the donations.

Although the most recent records are comprehensive, records of earlier years were fragmented and did not specify how donations were allocated. For example, there were only eleven entries prior to 1983 and less than 1000 donations recorded per year prior to 1989. More than one-sixth of the donations came from 2005, reflecting a mixture of the public’s response to Hurricane Katrina, better record keeping and the growing population of the county served by this particular chapter of the Red Cross. Contributors’ intentions for their donations began to be recorded in 2005. Prior to 2005, the Red Cross only recorded whether donations were intended for international or domestic disaster relief purposes.

After examining the data we were able to identify several hurricanes that seemed destructive enough to influence donor behavior. Hurricanes Katrina (August, 2005) and Rita (September, 2005) both had specific funds dedicated to them, making them ideal. Hurricane Mitch produced a substantial increase in donations and caused extensive damage in the Caribbean, but not the United States, allowing us to examine its influence on M-initial donations by focusing on contributions to international disaster relief. Finally, to increase the generalizability of our findings, we then extended our analysis to four additional hurricanes that were not associated with a noticeable increase in Red Cross donations but were particularly destructive.

2.2  Hurricane Katrina

2.2.1  Creation of comparison groups

Donations to the Red Cross increased dramatically following Hurricane Katrina. There were 240 donations recorded in August 2005 (prior to Katrina), compared to an average of 289 over the previous five Augusts (1999–2004). Donations to the chapter increased to 3959 in September, 2005, significantly higher than the average of the previous five Septembers (M = 216). There were 968 donations in October 2005, which was above average in comparison to the donations in October for the previous five years (M = 237). Donations returned to their year-over-year average by November (647 compared to an average of 664 donations, SD = 92.39, over the previous five Novembers). Thus, the two months following a hurricane seemed to be a period in which people were actively donating to hurricane relief. We compared donations that occurred during the two month period to donations in the six months prior to Katrina’s landfall (N = 2012).1

2.2.2  Evidence of implicit egotism

Of interest was whether sharing an initial with a hurricane increases the likelihood that one would donate to relief efforts for that specific hurricane. We examined both the effect of time (before vs. after Katrina) on disaster donations and where people allocated their money after Katrina. The analyses showed a pronounced name letter effect. First, 4.2% of the donors in the six months prior to Katrina had a name with the initial K; this proportion increased to over 5% (a 31% increase over baseline) after Katrina (χ2 (1, N = 7,469) = 5.49, p < .02, ϕc = .027).

Second, the more directly a disaster relief fund targeted Katrina relief efforts, the greater the proportion of K-initial donors. Specifically, the greatest proportion of K-initial donors was to Katrina specific relief funds (money designated for either Hurricane Katrina relief and Louisiana Disaster Relief; 9.8%), followed by donations to hurricane relief in the United States (5.72%), followed by domestic disaster relief funds more generally (4.85%), followed by the baseline measure of pre-Katrina disaster relief (4.2%), as indicated by a planned linear contrast on proportions (Furr & Rosenthal, 2003) (z = 2.47, p < .02, Table 1). Finally, we compared the proportion of K-initial donations during the six months prior to Katrina to the K-initial donations in the two months following Katrina that were not allocated to domestic disaster or hurricane relief (2.39%). The observed increase in the number of contributions of K-initial donors to Hurricane Katrina relief was not associated with a significant decrease in the number of their contributions to other funds (2.39%, χ2 (1, N = 2,304) = 2.16, p > .14, ϕc = .03).

We also examined whether the increased number of K-initial donors was driven primarily by people with names closely related to Katrina (specifically beginning with “Kat”). The most sensitive test of this hypothesis is to compare the proportion of Kat-onset donors that make up the K-initial donor population before Hurricane Katrina to the proportion of Kat-onset donors who made up the K-initial donors who contributed money to domestic disaster relief following Katrina. If Kat-onset donors are driving the effect, they should make up a greater proportion of donors after Hurricane Katrina. The proportion of Kat donors in the K-initial donor population was 30% prior to Katrina and only 27% after Katrina, a non-significant decrease (χ2 (1, N = 296) < 1).

Finally, we examined the average amount of K-initial donations. The data were non-normally distributed (Skewness = 80.64) so amounts were rank-transformed and subjected to a 2 (money donated to Katrina compared to money donated the six months prior) X 2 (K-initial or other initial) ANOVA. This procedure produces equivalent results to the Kruskall-Wallace test and is appropriate for nonparametric tests of two-way interactions (Seaman, Wall, Wise & Jaeger, 1994). Donations were significantly greater after Katrina (Mrank = 3 587, SE = 61.02) than before Katrina (Mrank = 5, 152, SE = 122.62, F(1, 7818) = 135.56, p < .001. However, there was no interaction with name initial (F < 1, p > .35, ηp2 < .001).

Since we lack information about the proportion of K-initial and non-K-initial donors after Katrina who would have donated nothing had Katrina not occurred, the mean rank of donors prior to Katrina is inflated because the people who donate nothing are not measured.2 Since there is a greater proportion of K-initial non-donors who are not measured before Katrina, this may obscure any differences between the two groups. To estimate the total number of donors who would have contributed had Katrina not occurred, we calculated the average number of K-initial donations for the same time period in the previous five years (M = 16) and the average number of non-K-initial donors (M = 438). We then selected the top K-initial and non-K initial donors based on these averages. We submitted the rank-transformed contribution of these donors and the rank-transformed contribution of donors from the six-months prior to a 2 (money donated by the top donors following Katrina compared to money donated the six months prior) X 2 (K-initial or other initial) ANOVA. Again, donations were significantly greater after Katrina, reflecting both increased donations discussed above and our selection criteria. However, we found no interaction with initial (F < 1, p > .45, ηp2 < .001). Taken together, these findings indicate that people with K-initials were more likely to donate after Katrina. The more relevant the fund to Katrina, the greater the proportion of K-initial donors, but Katrina donations did not come at the expense of non-Katrina donations.

2.3  Hurricane Rita

Hurricane Rita also had a specific funding account, thus enabling us to analyze the name letter effect with donations to these relief efforts as well. Given that Hurricane Rita made landfall less than one month after Katrina, donor response to Rita-specific funds was weaker than responses to the Hurricane Katrina fund: only fourteen people donated specifically to the Hurricane Rita relief fund. However, of the small number of people who donated specifically to Rita, nearly a third (four out of fourteen) had names that begin with R, representing a substantial increase over the six months prior (N = 5969, χ2 (1, N = 5,973) = 8.51, p <.01 ϕc = .038). The proportion of R-initial donations to Hurricane and National Disaster relief in the two months following Hurricane Rita (7.20%) did not differ from the proportion of R-initial donations in the six months before (7.71%, χ2 (1, N = 7, 416) < 1).

We also examined the amount donated to Hurricane Rita relief efforts by R-initial and non R-initial donors before and after the hurricane. There was no main effect of time or initial. As with Katrina, the rank order of R-initial donations did not increase more than other donations, F < 1 for the interaction indicating that the average donation of R-initial donors did not increase after Hurricane Rita.

2.4  Hurricane Mitch

As a further test of our hypotheses, we examined international disaster relief following Hurricane Mitch. Hurricane Mitch was the deadliest Atlantic Hurricane since 1780. It killed between 11,000 and 18,000 people, almost all of them in Nicaragua and Honduras and caused more than $6 billion in damage (NCDC: Hurricane Mitch, n.d). Although this Red Cross chapter did not maintain records of where funds were allocated prior to 2005, they did maintain separate records of international and domestic donations. We assumed that most of the international donations made immediately after Hurricane Mitch were probably intended for the countries affected by it. This assumption was based on the increase in donations in the wake of Hurricane Mitch.

2.4.1  Creation of comparison groups

Hurricane Mitch made landfall in Central America in October 29, 1998. There were 113 donations recorded between October 1st and October 29th, 1998, compared to an average of 140 donations for the previous and following five Octobers (1993–1997; 1999–2004). Donations to the Red Cross increased dramatically following this hurricane. In November, 1199 donations were made to this chapter of the Red Cross, significantly higher than the previous and following five years (M = 483). Donations remained above average in December (N = 1129 compared to M = 690 for the previous and following five Decembers) and returned to baseline by January (N = 256 compared to an average of 360 donations for the previous and following five Januarys). Thus we compared donors in the six months prior to Mitch’s landfall (N = 2402) to the two months following Mitch (N = 2109).

2.4.2  Evidence of implicit egotism

Although the increase in Red Cross Donations intended for international use is an imprecise measure of funds intended for Hurricane Mitch, donations to other disasters would only work against our hypothesis. Consistent with our hypothesis however, the proportion of M-initial names donating to international disaster relief in the two months after Mitch (N = 410) was higher (12.9%) than the proportion of M-initial donations to all funds for the six months prior (10.05%), a 30% increase, χ2 (1, N = 2,810) = 3.97, p < .05, ϕc = .038. M-initial donations to international disaster relief were also greater than M-initial donations to domestic disaster relief and undesignated donations (9.13%), χ2 (1, N = 2, 106) = 5.32, p < .02, ϕc = .05. As with Katrina, M-initial donations to international disaster relief did not appear to reduce M-initial donations to domestic disaster services or unspecified gifts below the average of the previous six months, χ2 (1, N = 3, 708) < 1.

Again, using an ANOVA on rank-transformed donation amounts, we found that donations to Hurricane Mitch (Mrank = 1124) were smaller than the six months prior (Mrank = 1491), F(1, 2846) = 33.22, p < .001. The average rank of M-initial donations did not differ from other donations and there was no interaction between time and initial, Fs < 1.

2.5  Other hurricanes and meta-analysis

2.5.1  Selection of additional hurricanes

We identified additional hurricanes to test the generalizability of the influence of shared initials on disaster donations. There were no other hurricanes in our data set that had specific accounts dedicated to them, or that produced a noticeable increase in donations, thus we used different criteria to identify storms that were likely salient for potential donors. We examined the ten most destructive hurricanes from the list of retired Atlantic Hurricane names. Two of them (Katrina and Rita) are discussed above. Of the remaining eight, four had occurred before the Red Cross database records appeared to be comprehensive (Agnes, 1972; Andrew, 1992; Hugo, 1989 and Betsy, 1965), leaving four hurricanes that seemed suitable for analysis: Charlie (August, 2004), Francis (September, 2004), Ivan (September, 2004), and Wilma (October, 2005). Unlike Katrina, Rita and Mitch, these hurricanes affected both the United States and other Caribbean countries so we examined their influence on donations in general.

2.5.2  Evidence for Implicit Egotism with Additional Hurricanes

Across all four hurricanes we found that donation patterns were in the predicted direction, with the proportion of same-initial donations increasing after each hurricane made landfall. The effect sizes however were statistically non-significant, most likely due to small sample sizes (see Table 2). In order to evaluate the robustness of the effect, we conducted a meta-analysis of these storms as well as the storms discussed above. The effects were not heterogeneous, Q (6) = 6.17, p = .40 indicating that fixed effects analysis was appropriate. Fixed-effects analysis revealed that the effect was robust, OR = 1.17, (95% CI 1.06, 1.28), z = 3.18, p < .002. Differences in the amount donated to hurricane relief were not examined.

Table 2: Same-initial donations to other hurricanes.

Hurricane Different-initial donorsSame-initial donorsProportion of same-initial donors
 χ2 (1, N = 1755) = 2.31, p = .065
 χ2 (1, N = 1960) = 1.81, p = .086
 χ2 (1, N = 2025) = 1.85, p = .086
 χ2 (1, N = 8051) = 1.93, p = .082
Note. All tests one-tailed.

3  Discussion

Sharing an initial with a natural disaster increases the likelihood that people will donate to relief efforts for that disaster, but does not increase the average size of the donation. The only appropriately powered comparison that did not detect results in the predicted direction was for R-initial donations to domestic disaster and hurricane relief after Hurricane Rita. One reason we may not have found an increase of R-initial donors is that the two month time period following Hurricane Rita overlapped substantially with the two month time period following Hurricane Katrina. Katrina specific donations were seven times greater than Rita-specific donations. Based on these numbers, it seems safe to assume that the majority of the funds donated to domestic disaster relief and hurricane relief after Rita were actually intended for Hurricane Katrina. Regardless of why this effect did not occur with Hurricane Rita, a meta-analysis of the most severe hurricanes between 1998 and 2005 revealed that the effect of shared initials on donations is small but robust.

This finding replicates and extends earlier work done by Nelson and Simmons (2007) that suggests that people can be attracted to negative targets with which they share an initial. However, unlike previous studies, our findings cannot easily be accounted for by the assimilation of positive feelings about the self into the target of evaluation. Current theories on implicit egotism predict that people should either avoid association with a negative stimulus (in a manner similar to Snyder et al. 1986) or downplay the negativity of same initial stimuli (Finch & Cialdini, 1989).

As a field study, these data offer little insight into underlying psychological processes, aside from demonstrating that self-image management techniques and egotistic motives alone do not appear to be sufficient to produce this phenomenon. There are a number of alternative processes that alone or together could account for this phenomenon. First, people may process information about same-initial hurricanes differently than information about hurricanes with different initials. People process self-relevant information more carefully and are more likely to retain self-relevant information. This effect also generalizes to close others (for a meta-analytic review see Symons & Johnson, 1997). The unit-relation formed between the self and a hurricane may similarly facilitate in-depth processing and retention of hurricane relevant information. This may make people particularly likely to remember concrete information about the victims of a disaster as opposed to the abstract details conveyed by a headline. People are more likely to help identifiable victims (Small & Lowenstein, 2003), so exposure to more information about suffering individuals should make them more likely to donate.

Second, the increase in donations may be a result of more intense negative feelings experienced by same initial donors while processing information about the disaster. According to the “feelings as information” model (Schwarz & Clore, 2007) people’s feelings while they are exposed to the storm may be seen as diagnostic of the suffering it caused. Assuming that people feel that their affective reaction to the storm is diagnostic of the suffering of those who are victims, the negative affect experienced by those who share an initial with the storm should make them more likely to donate. A related explanation is that charity donation can function as a kind of mood repair. According to this account, people should make efforts to repair their mood after experiencing negative affect. Charity is one method by which people may do this (Harbaugh, 1998). Exposure to a same-initial hurricane makes people feel worse, and the most salient way to repair this feeling is the opportunity to donate money to Katrina.

Third, people may differ in how responsible they feel for the consequences of same-initial and different-initial hurricanes. Shared initials may lead to the assimilation of responsibility for a target’s actions into the self, much as they lead to the assimilation of positive feelings about the self into judgments of the target. This explanation is compatible with previous research on intergroup relations that has found that people’s willingness to compensate outgroup members for the actions of other ingroup members is mediated by feelings of collective guilt (Brown et al., 2008; Doosje, et al., 1998). Although our data do not make it possible to test this hypothesis directly, it remains an intriguing possibility for future research.

Finally, people may construe disaster relief as “contributing to a noble cause” thus freeing disasters from their negative connotations. We cannot rule this possibility out, but it seems unlikely, as a change in the overall assessment of the worthiness of the cause would have to occur without revising beliefs about the magnitude of the disaster (which would have negative implications for the self) or despite attempts to downplay the severity of the hurricane. The former is improbable given the importance of maintaining a positive view of the self (Sedikides, 1993) while the latter seems unlikely given the relatively small effect sizes produced by the name-letter effect. Despite these counterarguments, the possibility that K-initial donors construe Katrina-relief as a noble cause cannot be conclusively ruled out with the current data. More generally, very little is known about which charities people choose to donate to (Bekkers & Wiepking, 2007), suggesting that this may be an issue ripe for studying in the future.

While in the present study we consistently find that people are more likely to donate to same-initial disasters, in none of the cases did same-initial donors donate more money. This may be because donating any amount of money may make people feel like they have “done their part.” Alternatively, the relationship between donation amount and initial may be obscured by demographic factors such as income that have a substantial effect on the amount of money donated to disasters (Todd & Lawson, 1999). Name-letter effects, although statistically robust and of theoretical interest are quite small in comparison to other determinants of donor behavior.

Despite this limitation, our data do support our primary hypothesis that people are more likely to donate to the relief of disasters that they share an initial with. This contributes to the existing literature by showing that not only are people likely to associate themselves with same-initial negative targets (Nelson & Simmons, 2007) but that this phenomena cannot entirely be accounted for by a failure to appreciate how negative the target really is. The processes that lead people to associate with same-initial negative stimuli require further exploration but could involve a number of interrelated changes to information encoding, the construal or management of negative emotions evoked by exposure to information about the hurricane, or the reconstrual of negative targets in positive terms.


Bekkers, R. & Wiepking, P. (2007). Generosity and philanthropy: A literature review. SSRN. Retrieved June 6, 2008, from http://ssrn.com/abstract=1015507.

Brendl, C. M., Chattopadhyay, A., Pelham, B. W., & Carvallo, M. (2005). Name letter branding: Valence transfers when product specific needs are active. Journal of Consumer Research, 32, 405–416.

Brown, R., González, R., Zagefka, H., Manzi, J., Cehajic, S. (2008). Nuestra Culpa: Collective guilt and shame as predictors of reparation for historical wrongdoing. Journal of Personality and Social Psychology, 94, 75–90.

Cialdini, R. B., Borden, R. J., Thorne, A., Walker, M. R., Freeman, S. & Sloan, L. R. (1976). Basking in reflected failure: Three (football) field studies. Journal of Personality and Social Psychology, 39, 406–415.

Doosje, B., Branscombe, N., Spears, R., & Manstead, A. S. R. (1998). Guilty by association: When one’s group has a negative history. Journal of Personality and Social Psychology, 75, 872–886.

Finch, J. F., & Cialdini, R. B. (1989). Another indirect tactic of (self-) image management: Boosting. Personality and Social Psychology Bulletin, 15, 222–232.

Furr, R. M. & Rosenthal, R. (2003). Evaluating theories efficiently: The nuts and bolts of contrast analysis. Understanding Statistics, 2, 45–67.

Harbaugh. W. T. (1998). What do Donations Buy? A Model of Philanthropy Based on Prestige and Warm Glow. Journal of Public Economics, 67, 169–284

Heider, F. (1958). The psychology of interpersonal relations. Hoboken, NJ: John Wiley & Sons, Inc.

Jones, J. T., Pelham, B. W., & Carvallo, M. (2004). How Do I Love Thee? Let Me Count the Js: Implicit Egotism and Interpersonal Attraction. Journal of Personality and Social Psychology, 87, 665–683.

Leary, M. R. (2007). Motivational and emotional aspects of the self. Annual Review of Psychology 58, 317–44.

NCDC: Hurricane Mitch (n.d.) Retrieved December 22, 2006, from http://www.ncdc.noaa.gov/oa/reports/mitch/mitch.html.

Nelson, L. D., & Simmons, J. P. (2007). Moniker Maladies: When names sabotage success. Psychological Science, 18, 1106–1112.

Nuttin, J. M. (1987). Affective consequences of mere ownership: The name letter effect in twelve European languages. European Journal of Social Psychology, 17, 381–402.

Pelham, B. W., Mirenberg, M. C., & Jones, J. T. (2002). Why Susie sells seashells by the seashore: Implicit egotism and major life decisions. Journal of Personality and Social Psychology, 82, 469–487.

Salkin, A. (2005, September 11). What’s in a Name, Katrinas? New York Times, p. 9–1.

Schwarz, N., & Clore, G. L. (2007). Feelings and phenomenal experiences. In E. T. Higgins & A. Kruglanski (eds.), Social psychology. Handbook of basic principles (2nd ed.; pp. 385–407). New York: Guilford.

Seaman, J. W., Walls, S. C., Wise, S. E. Jaeger, R. G. (1994). Caveat emptor: Rank transform methods and interaction. Trends in Ecology & Evolution, 9, 261–263.

Sedikides, C. (1993). Assessment, enhancement, and verification determinants of the self-evaluation process. Journal of Personality and Social Psychology, 65, 317–338.

Small, D. A. & Loewenstein, G. (2003). Helping a victim or helping the victim: Altruism and identifiability. Journal of Risk and Uncertainty, 26, 5–16.

Snyder, C. R., Lassegard, M., & Ford, C. E. (1986). Distancing after group success and failure: Basking in reflected glory and cutting off reflected failure. Journal of Personality and Social Psychology, 51, 382–388.

Strömberg, D. (2007). Natural Disasters, Economic Development, and Humanitarian Aid. Journal of Economic Perspectives, 21, 199–222.

Symons, C. S., & Johnson, B. T. (1997). The self-reference effect in memory: a meta-analysis. Psychological Bulletin, 121, 371–394.

Todd, S. J., & Lawson, R. W. (1999). Towards a better understanding of the financial donor: An examination of donor behavior in terms of value structure and demographics. International Journal of Nonprofit and Voluntary Sector Marketing 4, 235–244

We would like to thank Norbert Schwarz and two anonymous reviewers for comments on earlier versions of this paper. This research was supported by SSHRC doctoral fellowship awarded to the first author, a National Science Foundation Graduate Research Fellowship awarded to the second author and the Michigan Prevention Research Training Grant (NIH grant number T32 MH63057). Correspondence concerning this article should be addressed to Jesse Chandler, Department of Psychology, University of Michigan, 530 Church Street, Ann Arbor, MI 48109, U.S.A. Email: jjchandl@umich.edu
As one reviewer pointed out, the six month time period prior to Katrina is arbitrary. We repeated these analyses using different cut-off intervals ranging from six months to a year. Results remained substantially the same.
Thanks to an anonymous reviewer for this comment.

This document was translated from LATEX by HEVEA.