Abstract
Research suggests that people are typically more generous in later life. Terror management theory offers one explanation for this pattern. The theory suggests that humans’ innate desire to survive results in strong reactions to increased awareness of mortality, thereby affecting behaviors. Older adults’ increasing proximity to life’s end has been associated with greater generative concern, as a means of caring for future generations and ensuring symbolic immortality. This experimental study evaluated the effects of age and mortality salience on charitable giving. Age and mortality salience interacted to affect donations overall and specifically for in-group and out-group donations. Compared with respective age-based control groups, findings indicate that mortality salience primes led young adults to donate less overall to the in-group, and older adults to donate more overall and more to the out-group. Middle-aged adults’ donations did not vary according to priming condition. Results suggest that contemplation of mortality differentially affects charitable donations of young and older adults.
Giving generously to meet the needs of others is a valued social and cultural norm. Though the rate of American giving has remained stable at approximately 2% of the gross domestic product (Philanthropy Roundtable, 2018a), the amount given to charity in 2017 was $410 billion, with 70% of donations made by individuals (Lilly Family School of Philanthropy, 2018). Evidence suggests charitable giving behavior, a form of generativity, varies over the adult life course, wherein older adults typically exhibit greater generosity than young or middle-aged adults (77% of donations made by older adults compared with 60% made by young adults; Philanthropy Roundtable, 2018b). The reasons this age difference exists are not fully understood and little is known about whether psychosocial benefits to giving change with age. One possibility for the age-related increase in generosity may be that older adults’ awareness of nearing death prompts generative concern, or interest in the well-being of future generations (Maxfield et al., 2014), and generative behavior (Erikson & Erikson, 1997) as a means of ensuring symbolic immortality (Pyszczynski, Solomon, & Greenberg, 2015). This experimental study expands on prior research by examining the effects of death reminders on charitable giving, operationalized as the amount of money donated to charity, across the adult life span. Constructs from terror management theory are used as a theoretical frame for understanding age differences in giving.
Charitable Giving Over the Life Span
Charitable giving varies with age, with the majority of charitable donations being made by people older than 65 years of age, though a decline in giving has been shown to occur among the oldest old (i.e., 85+ years of age; Wiepking & James, 2013). Variability in charitable giving behavior across the adult life span likely occurs for a variety of reasons. From a developmental perspective, young adults are more likely to focus on autonomy-based goals, such as accumulating wealth for their own well-being and benefit (see Kotre, 1984). Because many middle-aged and older adults may have already accumulated social, psychological, and financial resources, they may be more likely to focus on generativity-based goals, such as working to improve one’s community or spending for the well-being of future generations (see Erikson’s psychosocial theory of development in Erikson & Erikson, 1997).
Data from the World Values Study (WVS; World Values Survey Association, 2016) provide an example of how generative attitudes vary by age. In wave six of the WVS, participants in the United States were asked to rate how similar they thought they were to a person for whom it is important to do good for society. Overall, 91.4% of respondents identified themselves as similar to this individual. However, age differences were evident: respondents younger than 29 years reported 85.6% similarity, respondents aged 30 to 49 years reported 92.1% similarity, and those older than 50 years reported 93.9% similarity. It appears that engaging in activities good for society is important at all ages; however, the importance of generativity, or behaving in ways that benefit future generations (Erikson & Erikson, 1997), seems to be even greater for adults in middle age and beyond.
Providing additional support for the positive correlation between charitable giving and age, Freund and Blanchard-Fields (2014) conducted studies examining age-related differences in charitable giving values and behavior. Participants (18-85 years old) chose between receiving an Amazon voucher in exchange for their participation or donating the equivalent amount to Doctors Without Borders. Older age significantly predicted charitable behavior, even after controlling for income. Not all research, however, supports the idea that giving is associated with older age. Landry, Lange, List, Price, and Rupp (2006) reported demographic data suggesting that adults older than 65 years give less to charity when approached at their home than those younger than 30 years; perhaps because the authors’ focus was not age-related differences, they did not elaborate on or interpret this outcome. Furthermore, Wiepking and James (2013) assert that the oldest old (those older than 85 years of age) give the least to charity, which the authors interpreted as the potential result of declining health and/or declining cognitive function (e.g., decreased involvement in social activities).
In light of the conflicting data concerning charitable giving over the adult life span, further research examining potential factors was considered worthwhile. It is likely that middle-aged and older adults have access to resources that allow them to be more generative or charitable than young adults. For example, after retirement, older adults may have more time and motivation to volunteer (Barlow & Hainsworth, 2001) compared with those at younger ages. It is also possible that increasing nearness to and salience of death affects older adults’ generative concern and willingness to give. Increased understanding of the relationships between age, death awareness, and giving can assist in clarifying the factors that influence generous giving in adulthood.
Terror Management Theory
Humans’ expansive cognitive capacity includes the ability to think abstractly about the world around them, and these thoughts relate to what has been, what is, and what will be. The ability to project into the future leads to contemplation of diverse abstract possibilities and ideas, including one’s eventual death. Terror management theory (Greenberg, Pyszczynski, & Solomon, 1986; see also Pyszczynski et al., 2015, for a review) proposes that humans’ awareness that death is unavoidable, unpredictable, and the end of one’s physical existence, leads to the potential for anxiety, referred to as terror in terror management theory literature. According to the theory, humans attempt to manage the potential anxiety that death awareness brings by accepting and living up to cultural worldviews. Living up to societal standards strengthens self-esteem and enables pursuit of immortality, in either symbolic or literal forms. Furthermore, these efforts are particularly important following events that increase awareness of mortality (Solomon, Greenberg, & Pyszczynski, 1991).
Psychological protection against the awareness of death’s inevitability is partly accomplished by the provision of either literal or symbolic immortality (see Pyszczynski et al., 2015, for a review). Cultures define what a person must do to be a “good” and “valuable” member of society, and provide some form of immortality to those who meet cultural expectations. Literal immortality is the idea that some part of a person (e.g., a noncorporeal spirit) will exist beyond death (Pyszczynski et al., 2015) and is usually provided via religion (see Vail et al., 2010, for an overview). Symbolic immortality involves being a valuable member of an enduring and meaningful culture and leaving something of one’s personhood behind that continues to exist beyond death (Pyszczynski et al., 2015). Symbolic immortality can be achieved through memorable accomplishments (e.g., receiving a notable award), or through financial legacies (e.g., donating charitable endowments to lasting organizations). Such achievements create opportunities to be remembered and renowned beyond the scope of a single person’s lifetime.
According to terror management theory, cultural worldviews are created and supported through socialization into the groups to which a person belongs (Greenberg & Kosloff, 2008). Furthermore, previous research has found that being part of a chosen culture (e.g., religious groups, political affiliations, communities) influences the desire to contribute to the group, as well as share in the benefits of the group (Bennett, 2003). It has been suggested that prosocial behavior fulfills cultural standards, especially when bestowed on people who are judged as “worthy” within a particular culture. Thus, prosocial behavior has the potential to strengthen a person’s personal value, thereby providing protection against thoughts of death (see Hirschberger, 2010, for a review).
When reminded of mortality, people are more likely to support groups that share or validate their cultural worldview (the in-group; Castano, Yzerbyt, Paladino, & Sacchi, 2002; Jonas, Schimel, Greenberg, & Pyszczynski, 2002) and to derogate or respond more aggressively toward groups that do not validate their cultural worldview (the out-group; Greenberg et al., 1995). For a review of the literature concerning reactions to in-groups and out-groups following mortality salience (events that provoke awareness of mortality), see Pyszczynski et al. (2015). Taken together, results of these studies indicate that prosocial behavior should increase in reaction to mortality salience, particularly in response to in-group members, though responses were expected to vary with age. For the purposes of this study, prosocial behavior was defined as the amount of money donated to charity; additionally, in-group versus out-group support was examined based on the amount of money donated to either a local, in-group charity (Care and Share Food Bank of Southern Colorado) or a global, out-group charity (The United Nations World Food Programme).
Age Differences in Terror Management Processes
The majority of terror management research has been conducted with college-aged students (Burke, Martens, & Faucher, 2010), with only a few studies including middle-aged adults (e.g., Rosenblatt, Greenberg, Solomon, Pyszczynski, & Lyon, 1989) and older adults (e.g., Maxfield et al., 2007; Maxfield et al., 2014). Early terror management research indicated that young adults (19-29 years old; Florian & Mikulincer, 1997) and middle-aged adults (Rosenblatt et al., 1989) typically respond with harsher judgments to moral transgressors (the out-group) after mortality salience primes than after control primes. As death becomes more personally salient over the course of the life span, people’s responses to mortality reminders may differ.
In an early study of age differences in terror management, Maxfield et al. (2007; Study 2) found that older adults (59-92 years old) who received a subtle mortality salience prime responded with less punitive evaluations of moral transgressors (the out-group) than did older adults in the control condition. As seen in previous studies, young adults (17-34 years old) in Maxfield and colleagues’ study endorsed harsher judgments of moral transgressors in the mortality salience condition, compared with young adults in the control condition. Stated differently, older adults rated moral transgressions more harshly than young adults in the control condition, but this age difference was eliminated in the mortality salience condition.
Terror Management and Prosocial Behavior
Terror management theory asserts that people are motivated to adhere to cultural norms and standards of value, particularly following reminders of death (see Solomon et al., 1991). If prosocial behavior (e.g., generosity or charitable giving) is valued in a person’s culture, mortality salience would be expected to increase the importance of behaving prosocially (e.g., giving more to a charitable organization; Jonas et al., 2002). In contrast, if consumerism and wealth are valued in a person’s culture, reminders of mortality would be expected to increase the importance of consuming and garnering wealth (Jonas, Sullivan, & Greenberg, 2013; Kasser & Sheldon, 2000; Zaleskiewicz, Gasiorowska, Kesebir, Luszczynska, & Pyszczynski, 2013). According to life course theories of development (see Erikson & Erikson, 1997, for a description of psychosocial developmental stages), an individual’s value of and ability to engage in prosocial behaviors is likely to vary in meaningful ways across the life span. Thus, aging may influence how and when people conform to cultural standards, such as emphasizing others over one’s self via the prosocial behavior of giving.
Indeed, age-related differences in response to mortality reminders have been observed in studies examining generative concern. Maxfield et al. (2014, Study 1) examined the effects of mortality salience on generative concern among young (17-22 years old) and older adults (62-92 years old). No mortality salience effects were observed on generative concern for the young adult group. However, older adults reported higher generative concern after receiving a mortality salience prime, compared with older adults in the control condition. In a conceptual replication of this finding, Maxfield et al. (2014, Study 2) examined the effects of mortality salience on a more specific form of generative concern, operationalized as interest in leaving a legacy for the benefit of others (prosocial preference) compared with interest in leaving a legacy of fame and recognition for the benefit of self (proself preference). A main effect of mortality salience revealed that participants in the mortality salience condition reported greater levels of prosocial preference compared with participants in the control condition. Additionally, a main effect of age indicated that older adults (60-90 years old) reported higher levels of prosocial preference than young adults (18-33 years old). These main effects were qualified by a significant interaction between age and mortality salience, such that older adults in the mortality salience condition reported higher levels of prosocial preference compared with older adults in the control condition. Young adults’ prosocial preference did not differ as a result of the priming condition.
Specific to the effects of mortality salience on charitable behaviors, Jonas et al. (2002) found that young adults donated more money to an American charity than an international charity after being primed with mortality salience compared with young adults primed with an aversive topic, unrelated to mortality (dental pain). More recently, Zaleskiewicz, Gasiorowska, and Kesebir (2015) reported similar results, wherein Polish university students exhibited greater prosocial attitudes and behavior after a mortality salience prime, compared with same aged participants in the control condition. Neither of these studies examined the potential role of age.
In summary, the reported studies examining the effects of mortality reminders on self-reported attitudes of young and older adults exhibit consistent patterns. For young adults, death reminders generally result in negative evaluations or neutral prosocial attitudes (e.g., harsher judgments of moral transgressors, no differences in generative concern) compared with a control prime. One exception to this pattern is the research conducted by Jonas et al. (2002), who found that mortality salience was associated with greater generosity toward in-group organizations. Older adults are generally more prosocial after being primed with mortality (e.g., more lenient judgments of moral transgressors and higher levels of generative concern), compared with a control prime. Because mortality salience differentially affects older and young adults’ attitudes and intentions, it is plausible that mortality salience would also differentially affect prosocial behaviors (e.g., charitable giving).
Study Purpose and Hypotheses
This study was modeled after previous research (Jonas et al., 2002; Maxfield et al., 2014) and used quasi-experimental methods to examine the effect that age (young adult: 18-39 years of age; middle-aged adult: 40-64 years of age; older adult: 65+ years of age) and death-related awareness (mortality salience prime vs. control prime) had on generativity, herein operationalized as the amount of money given to preselected charities. Young, middle-aged, and older adult participants were randomly assigned to priming condition. This study was conducted to determine whether (a) age and reminders of mortality interact to affect total donations to charity and (b) whether age and reminders of mortality interact to differentially affect donations to in-group and out-group organizations. A priori hypotheses were set forth as follows:
Method
Materials
Study materials are described below in order of their presentation.
Generativity
The Social Generativity Scale (SGS; Morselli & Passini, 2015) is a six-item scale measuring social generativity specific to concern for future generations in adults older than 18 years of age. Participants indicate their agreement with each of the statements on a 7-point Likert-type scale (1 = strongly disagree to 7 = strongly agree), with higher mean scores indicating greater levels of social generativity. “I give up part of my daily comforts to foster the development of next generations” is an example statement from the SGS. Because any mention of death prior to the priming conditions was undesirable, one of the questions on the SGS was modified from “I commit myself to do things that will survive even after I die” to “I commit myself to do things that will last.” The SGS had acceptable internal reliability (Cronbach’s α = .77) for this sample.
Social Desirability
The Marlow–Crowne Social Desirability Scale (SDS; Crowne & Marlowe, 1960) is a 33-item measure of social desirability with statements such as “I have never intensely disliked anyone.” Participants respond by indicating whether they believe the statements to be “true” or “false” for them personally. One point is scored for each socially desirable response, with higher sum of scores indicating higher social desirability bias. The SDS had acceptable internal consistency (Cronbach’s α = .75) for this sample.
Priming Manipulation
Participants completed primes commonly used in terror management research, wherein participants are asked to write about their own death or an aversive control topic unrelated to death (Burke et al., 2010). Participants in the mortality salience condition were asked to write responses to two questions specifically related to death: “Please briefly describe the emotions that the thought of your own death arouses in you” and “Jot down, as specifically as you can, what you think will happen to
Affect
The Positive and Negative Affect Schedule (PANAS; Watson, Clark, & Tellegen, 1988) is a 20-item PANAS measure of emotional state. It was used as a distractor task in this study. Distractor tasks are included in terror management research to allow thoughts of death to leave conscious awareness (see Steinman & Updegraff, 2015, for a related meta-analysis). The PANAS includes 10 positive (e.g., “content”) and 10 negative (e.g., “anxious”) words related to affect. Participants indicate the extent to which they feel each emotion “right now” on a 5-point scale (1 = not at all to 5 = extremely) with mean higher scores indicating higher levels of positive and negative emotion. The internal reliability of the PANAS positive subscale (Cronbach’s α = .88) and the PANAS negative subscale (Cronbach’s α = .90) were high for this sample.
Charitable Giving Materials
According to the World Values Survey (2016), 53.1% of people report that the most serious world problem is meeting the needs of people living in poverty. As such, two charities providing food for those in need, one local (the “in-group charity;” Care and Share Food Bank for Southern Colorado) and one global (the “out-group charity;” The United Nations World Food Programme) were selected for use in this study. To increase experimental control, these charities were selected to be as similar as possible in terms of their mission to feed impoverished individuals. Participants received a packet containing an instruction sheet, a brochure describing the charities, and two pre-labeled donation envelopes. The introduction sheet informed participants that the researchers had agreed to share information about a charitable giving drive, and asked participants to consider donating to one or both of the charities indicated on the attached brochure. Two versions of the brochure were created and the order was counter-balanced for the charity presented at the top of the page to minimize recency effects (Jonas et al., 2002). The donation envelopes were pre-printed with instructions to place the desired donation amount, if any, into the designated envelope, and place both envelopes, regardless of whether they chose to donate, into a locked metal donation box.
Demographic Questionnaire
Participants completed a demographic questionnaire with questions relating to age, gender, education, ethnicity, race, relationship status, approximate annual income, religion, language, health status, 2016 presidential vote, and political leaning.
Cognitive Screen
To rule out the effect of possible cognitive impairment, which may have affected participants’ ability to respond to the study materials and instructions, all participants received a brief cognitive screen. The Short Orientation-Memory-Concentration Test (OMCT; Katzman et al., 1983) is a six-item measure for use in detecting cognitive impairment. Participants receive a point for each incorrect response (up to 12) and errors are weighted to calculate a total score (maximum of 24). Scores less than or equal to eight generally indicate normal cognitive functioning, thus data from participants who score a nine or greater were excluded from analysis (n = 9).
Procedure
This study was approved by an institutional review board and met all requirements for studies involving human participants. Participants were told they were completing a study related to personality and aging. Each participant was paid $15 at the onset of the study; they were given one $5 bill and ten $1 bills to allow for flexibility with amount of donation. To minimize suspicion at the payment combination, participants were told that the researcher was out of $10 bills, and only $1 bills were available. After providing informed consent, participants completed measures of generativity and social desirability followed by random assignment to one of the priming conditions (either mortality salience: MS or dental pain: DP). Random assignment was achieved through the use of Qualtrics software. Following the priming manipulation, participants completed the distractor task and a demographic questionnaire.
The charitable giving request was patterned after that of Jonas et al. (2002). The researcher informed participants that students in the psychology department were participating in a charitable giving drive. All participants were asked to read a brochure with information about the two charities and decide whether to donate. The researcher then left the room to retrieve additional measures in order to reduce potential bias and to keep the researcher blind to participant donations. When the researcher returned to the room, participants completed additional filler measures and were given a cognitive screen. Finally, all participants were debriefed about the study purpose and hypotheses and given the opportunity to ask questions.
Participants
Participants aged 18 years and older were recruited from a university psychology department, a research participant database, and the general community. Because many psychology students are familiar with terror management research, those who had recently participated in a terror management study were excluded from participation. A total of 139 participants (young adults = 49; middle-aged adults = 43; older adults = 47) completed this study. Three participants donated more than the initial $15 payment (two donated $16 and one donated $35). To include these participants in the analyses, the amounts donated were capped at the maximum of $15. Based on a priori exclusion criteria (pilot participants, possible cognitive impairment, and nonstandard testing conditions), 15 participants were excluded from analysis. The 124 remaining participants constitute the study sample (young adults = 41, middle-aged adults = 39, older adults = 44). Participants ranged in age from 18 to 86 years old (M = 51.85, SD = 20.70). Participant demographic data are presented in Tables 1 and 2.
Descriptive Statistics by Age Group.
Note. SGS = Social Generativity Scale; SDS = Marlow–Crowne Social Desirability Scale; PANAS = Positive and Negative Affect Schedule; OMCT = Short Orientation-Memory-Concentration Test; CS = Care and Share Food Bank for Southern Colorado; WFP = The United Nations World Food Programme.
Frequencies by Age Group.
Christian included responses for Catholic, Christian, and Protestant. bOther included responses for agnostic, atheist, and not practicing.
Results
Statistical Assumptions and Group Equivalency
The skewness statistic was used to evaluate the assumption of normality. All variables of interest were normally distributed with the exception of the dependent variable (amount of money donated to charity). Many participants donated $0, which resulted in a positive skew in donations. A square root transformation, which is frequently used with positively skewed data, was applied to the nonnormal variables, which resulted in approximately normal distributions: donations to Care and Share Food Bank for Southern Colorado (skewness = 1.67), donations to the United Nations World Food Programme (skewness = 1.75), and total donations (skewness = 1.24). The transformed means were used in all data analyses; however, the raw means and standard deviations are reported throughout the text and tables for ease of interpretation.
Equivalency of groups across experimental conditions was evaluated using t tests and chi-square tests. Aside from age, the only significant difference occurred within the relationship variable (χ2 = 9.42, p = .05), see Table 2. After the relationship variable was recoded into a binary factor (married/partnered vs. not married: single, divorced, widowed, or other), no significant difference was found between priming conditions (p = .37). After modification of the relationship variable, no significant differences were found on any of the measures or demographic variables between priming condition (ps > .05).
Hypothesis 1: Total Donations to Charity
Pearson correlations were performed to assess the relationships between potential covariates (generativity, social desirability, gender, education level, relationship status, health, and political leaning) with the total amount of money donated to charity. No significant relationships were found between the covariates and total giving. A significant relationship was found between total amount of money donated to charity and income (r = .20, p = .03); however, because annual income was reported within categories (e.g., 1 = $0-10,999, 2 = $11,000-29,999, etc.), correlation is not the ideal analysis. Examining the impact of age group and income level on total donations indicated a lack of main effects and interaction, suggesting that higher mean income within the middle-aged and older adult participants does not explain age differences in total donation amounts.
A 2 (priming condition: MS, DP) × 3 (age: young, middle-aged, older) analysis of variance was used to evaluate the mean differences of total amount of money donated to charity. No main effects were found for age group (p = .82) or for the effect of priming condition (p = .71) on the total amount of money donated to charity.
A significant age × priming condition interaction was found, F(2, 118) = 4.20, p = .02, η p 2 = .07. A least significant difference comparison was used to evaluate the pattern of mean differences. Young adults in the mortality salience condition donated significantly less total money to charity (M = 1.05, SD = 1.75) compared with young adults in the dental pain condition (M = 4.05, SD = 5.62), p = .05. Among middle-aged adults, no difference was observed in total overall donations between the mortality salience (M = 2.19, SD = 4.79) and dental pain (M = 2.89, SD = 5.17) conditions, p = .55. The amount of money donated by older adults in the mortality salience condition (M = 3.52, SD = 5.06) was significantly more compared with older adults in the dental pain condition (M = 1.09, SD = 3.25), p = .04. See Figure 1.
Age × priming condition interaction on total amount donated.
Hypothesis 2: In-Group Versus Out-Group Giving
A two-way multivariate analysis of variance was calculated with two independent variables (age group and priming condition) and two dependent variables (in-group support and out-group support). Because assumptions of homogeneity of variance and covariance were violated, reported results are based on Pillai’s trace, which is a more conservative approach because it is robust to assumption violations.
No significant main effect was found for age on overall group-specific support, F(4, 236) = 1.72, p = .15, Pillai’s trace = .06, η p 2 = .03. Furthermore, no significant main effect was found for priming condition on overall group-specific support, F(2, 117) = 0.67, p = .51, Pillai’s trace = .01, η p 2 = .01. A significant interaction was found between age group and priming condition on group-specific support, F(4, 236) = 3.40, p = .01, Pillai’s trace = .11, η p 2 = .05. The effects for in-group and out-group donations are explored separately.
Age and Priming Condition Effects on In-Group Support
A significant interaction effect was found between age group and priming condition for in-group donations, F(2, 118) = 5.38, p = .006, η p 2 = .08. Least significant difference comparisons were used to evaluate the pattern of mean differences in donations made to the in-group (Care and Share Food Bank for Southern Colorado). For young adults, the mean amount of money donated to the in-group in the mortality salience condition (M = 0.24, SD = 0.54) was significantly less than the mean amount given in the dental pain condition (M = 2.62, SD = 4.12), p = .006. For middle-aged adults, the mean amount of money donated to the in-group in the mortality salience condition (M = 0.76, SD = 2.39) was not different than the amount given in the dental pain condition (M = 0.92, SD = 2.12), p = .69. For older adults, the mean amount of money donated to the in-group in the mortality salience condition (M = 1.97, SD = 2.63) was not different than the amount given in the dental pain condition (M = 1.04, SD = 3.24), p = .08.
Specific to the mortality salience prime, the mean amount of money donated to the in-group by older adults (M = 1.97, SD = 2.63) was significantly more than the mean amount donated by young adults (M = 0.24, SD = 0.54; p = .02) and middle-aged adults (M = 0.76, SD = 2.39; p = .04). No significant difference was observed between the mean amount donated by young adults and middle-aged adults (p = .75).
Within the dental pain condition, the mean amount of money donated to the in-group by younger adults (M = 2.61, SD = 4.12) was significantly greater than the mean amount donated by middle-aged adults (M = 0.92, SD = 2.11, p = .05) and older adults (M = 1.04, SD = 3.24, p = .03); middle-aged and older adults’ in-group donations did not differ, p = .96. See Figure 2.
Age × priming condition interaction on in-group specific support.
Age and Priming Condition Effects on Out-Group Support
A statistically significant interaction effect was found between age group and priming condition for out-group donations, F(2, 118) = 3.72, p = .03, η p 2 = .06. Least significant difference comparisons were used to evaluate the pattern of mean differences in donations made to the out-group. For young adults, the mean amount of money donated to the out-group in the mortality salience condition (M = 0.81, SD = 1.60) was not different than the amount donated in the dental pain condition (M = 1.43, SD = 2.27), p = .33. For middle-aged adults, the mean amount of money donated to the out-group in the mortality salience condition (M = 1.43, SD = 3.80) did not differ from the amount donated in the dental pain condition (M = 1.97, SD = 3.96), p = .37. For older adults, the mean amount of money donated to the out-group was significantly greater in the mortality salience condition (M = 1.55, SD = 2.62) compared with the dental pain condition (M = 0.04, SD = 0.21), p = .02.
Within the mortality salience condition, the age groups did not differ in mean amount given to the out-group, ps > .05. However, within the dental pain condition, both young adults (M = 1.43, SD = 2.27; p = .02) and middle-aged adults (M = 1.97, SD = 3.96; p = .02) donated significantly more to the out-group than older adults (M = 0.04, SD = 0.21). The mean amount donated to the out-group by young adults was not significantly different than the mean amount donated by middle-aged adults (p = .99). See Figure 3.
Age × priming condition interaction on out-group support.
Discussion
One purpose of this study was to identify factors that may facilitate charitable giving, as measured by financial donations made after an experimental prime, over the adult life span (e.g., mortality reminders, age, in-group vs. out-group status of the charity). Though the hypotheses were not entirely supported, reminders of death resulted in differences in charitable giving behaviors. Young adults responded to reminders of death by giving less overall to charity, and by being less generous toward the in-group, though this appears to be consistent with developmental expectations. Middle-aged adults’ charitable donations did not differ as a result of the priming condition, whereas older adults responded to death reminders by being more generous in overall charitable donations and in donations specifically to the out-group.
Prior researchers have proposed relationships between social and demographic factors and charitable giving (gender: Eagle, Keister, & Read, 2018; political leaning: Brooks, 2006; education: Oesterle, Johnson, & Mortimer, 2004). However, in this study, giving patterns were not associated with gender, education level, relationship status, income, health, and/or political leaning. Thus, they were not discussed in-depth.
Young Adults
Young adults were not expected to differ in the total amount of money given to charity as a result of the priming condition. However, they were expected to give more to the in-group and less to the out-group after being reminded of death. After a mortality salience prime (compared with a dental pain prime), young adults donated less total money to charity overall, and less to the in-group. Conflicting research exists relating to the frequency of young adult prosocial behaviors after a death reminder. Jonas et al. (2002) report that young adults engage in more prosocial behaviors toward an in-group after a death reminder (compared with a dental pain prime). In contrast, Jonas et al. (2013) found that reminders of death resulted in lower levels of charitable giving to an out-group (i.e., larger sums of money kept for oneself) when an in-group option for giving was not present. In support of the idea that mortality salience results in greater proself bias, Kasser and Sheldon (2000) and Zaleskiewicz et al. (2013) report that young adults display higher levels of greed following a reminder of mortality, compared with a control prime. Regarding the total amount of money donated to charity by young adults, it appears that their behavior was consistent with research suggesting that thoughts of death result in higher levels of greed, or self-bias. Because young adults are often focused on creating personal security for the future (see Kotre, 1984), self-oriented bias, particularly after being primed with thoughts of death, may be developmentally appropriate.
Research has shown that for young adults, the more similar others are to an individual, the more likely that individual is to provide support after a death reminder (Bennett, 2003; Castano et al., 2002; Jonas et al., 2002). The young adult sample in this study primarily consisted of university students. It is possible that this group identifies as part of global culture rather than with local culture, which may have affected giving behavior toward the local charity. If young adults in this sample view themselves as members of a global culture, they may have viewed the global (out-group) charity as part of their in-group. In support of this idea, Killick (2012) suggests that university students who are exposed to opportunities to develop global citizenship, report stronger global identities. If students, who made up the majority of the young adult sample, have participated in educational courses intended to increase their global awareness, they may have developed identities as members of a global community, which would make the out-group charity in this study more aligned with their worldviews than the in-group charity. However, group identification was not assessed in this study, presenting a limitation to further understanding the young adult findings presented here. Future research would benefit from assessing participant cultural worldviews and global versus national group identification.
Middle-Aged Adults
Though no prior research has evaluated the effects of mortality salience on charitable giving behaviors in middle-aged adults, developmental theorists suggest that members of this age group are likely to have developed the psychosocial and financial ability to care for others in prosocial ways (see Erikson & Erikson, 1997). As such, it was expected that middle-aged adults would give more in total donations to charity after a mortality salience prime than after a control prime. Additionally, based on prior research suggesting that middle-aged adults are more punitive to the out-group after thoughts of death (Rosenblatt et al., 1989; see also Greenberg & Kosloff, 2008, for a review), they were expected to give more to the in-group after a mortality salience prime than after a control prime. However, middle-aged adults giving patterns did not differ between mortality salience and dental pain conditions.
Prior researchers have reported age differences related to death anxiety. Young and older adults report lower levels of death anxiety compared with middle-aged adults’ higher levels (Neimeyer, Wittkowski, & Moser, 2004). Though middle-aged adults report high levels of explicit death anxiety, they do not appear to react to reminders of death, such as mortality salience. It is possible that, because middle-aged adults acknowledge their fear of death, reminders of their mortality do not affect them as strongly as other age groups. However, death anxiety was not measured in this study. Future research would benefit from including evaluation of middle-aged adults’ self-reported death anxiety and their responses to mortality salience.
Older Adults
Older adults were expected to donate more to charity after a mortality prime compared with a dental pain prime. It was also expected that older adults would give more to the out-group after a mortality prime, compared with a dental pain prime. In the present study, older adults reminded of death displayed generative behavior, giving significantly more in total donations to charity after a mortality prime, compared with a dental pain prime, suggesting a conceptual replication of prior terror management findings with older adults (e.g., Maxfield et al., 2014). Additionally, older adults displayed greater levels of charitable behavior toward a charity benefiting an out-group after mortality prime, compared with the control group, replicating older adults’ tendency toward greater leniency toward individuals outside their cultural worldview (Maxfield et al., 2007).
Study Limitations
This study is limited by its evaluation of a narrow scope of charitable giving (i.e., donations given after an experimental priming prompt); thus, the present findings may not generalize to all forms of generativity and charitable giving. However, based on the present results, future studies may benefit from assessing a variety of charitable giving behaviors (e.g., regular payroll donations, annual donations to organizations of personal significance, spontaneous giving, such as placing money in a jar at the store, and volunteering one’s time, etc.) after mortality salience primes.
In a literature review examining research from a variety of fields (e.g., economics, psychology, sociology, marketing), Bekkers and Wiepking (2011) reported eight mechanisms that play a role in charitable giving behaviors, thus providing context, beyond terror management theory, for why people may be motivated to donate to charity. Bekkers and Wiepking (2011) suggest that “awareness of need, solicitation, costs and benefits, altruism, reputation, psychological benefits, values, and efficacy” are all factors that influence giving (p. 927). From a terror management perspective, mortality salience motivates behavior that affirms one’s values; in the present case, mortality salience motivated older adults’ generative behavior. Given this positive outcome, the present study may fall within Bekkers and Wiepking’s domain of psychological benefits. However, factors from Bekkers and Wiepking’s other seven mechanisms were not assessed. It is possible that other motivational factors of giving were present in participants’ donation behavior, which may have affected the study outcomes. Future research would benefit from measuring and evaluating other factors that may influence patterns of giving beyond age and mortality salience.
Another limitation of this study is the manner in which the charitable giving request was presented. Despite the cover story that the charitable giving request was unrelated to the study, presenting the request within the study likely raised participant suspicion. In fact, at the conclusion of the study, it was noted whether participants indicated suspicion that the charitable giving drive and study purpose were connected (suspected n = 22; 17.6%). Additionally, during the debriefing portion of the study, some participants told the researcher they felt it was inappropriate to ask for donations during a research study, though this data were not explicitly tracked. Future studies would benefit from increasing the external validity of the charitable giving measure by making the donation request seem less contrived.
Though using predetermined charities increased experimental control, it is possible that the charities selected for this study did not adequately represent areas of generative concern for participants. Food banks benefit children as well as adults, but they do not explicitly focus on care of future generations. Future research would benefit from pilot surveys allowing participants to list in-group and out-group charities they believe best represent generativity, or care for future generations. Researchers could then select participant generated charities, which may better capture generativity and cultural worldviews in a more meaningful manner.
Conclusion
The present study adds to the existing giving literature by increasing understanding of the combined effects of age and awareness of one’s mortality on adult life span generativity, operationalized here as the amount of money donated to charity. It also adds to the existing terror management theory literature by reporting prosocial behaviors rather than self-reported attitudes alone, in response to a death reminder. Due to social desirability bias, self-reported attitudes may not always be congruent with actual behavior. Though the studies conducted by Maxfield et al. (2007; Maxfield et al., 2014) examined age-related differences to mortality salience, the dependent variables were based on self-reported attitudes. The current study examined behavioral, age-related differences in adult life span generativity in reaction to mortality salience, albeit in a contrived laboratory setting, which was still likely influenced by social desirability.
This study also adds to the current literature base by examining the effects of mortality salience for middle-aged adults, who are underrepresented in generativity and in terror management studies. Middle-aged adults differ from young and older adults in self-reported levels of death anxiety; middle-aged adults report high death anxiety, whereas young and older adults report low (Neimeyer et al., 2004). Additionally, middle-aged adults’ behavioral responses to mortality salience differ from young and older adults; young and older adults are more reactive to mortality salience, whereas middle-aged adults are not. This pattern of attitudes and behavior may highlight that open acknowledgement of a person’s fear of death may minimize the effects of mortality salience. It is yet to be determined whether the relationship between self-reported death-related anxiety and responses to mortality salience affects long-term behavior. However, based on the outcomes of this study, it appears that being less reactive to mortality salience makes participants more neutral in the realm of prosocial behaviors. Additional research evaluating the effects of mortality salience in middle-aged adults is recommended to better understand how death anxiety and responses to mortality reminders interact for this age group.
Though the results of this study did not fully support the hypotheses, the outcomes provided information regarding the effects of mortality salience on charitable giving behaviors over the adult life span. A primary theme identified for future research relates to the importance of evaluating personal identification with charities. It would be worthwhile to conduct a follow-up study addressing the limitations previously noted to obtain a broader picture of how mortality salience affects charitable giving behaviors.
When considering ways to apply the results of this study to charitable giving drives, it is recommended that organizers consider the age of their target audience. When young adults are reminded of their mortality, they tend to save more money for themselves, whereas older adults tend to give more to others. In designing charitable giving drives intended for young adults, it would be prudent to avoid language related to mortality (e.g., leaving a legacy for the future, leaving something of oneself behind). In contrast, reminding older adults of their mortality may result in more generous giving behaviors and attention to future generations. In summary, the results of this study were not completely consistent with prior giving research, nor with terror management research, and may reflect changing worldviews. Nevertheless, given the positive impact generativity and charitable giving have on psychological health (e.g., Dunn, Aknin, & Norton, 2008), further evaluation of the reported outcomes would be warranted.
Footnotes
Acknowledgements
The authors would like to thank Tom Pyszczynski, PhD, and Katie Tomlin, PhD, for their helpful feedback on the design and results of this project. They also thank Evan Furr, BA, who assisted in participant recruitment and scheduling.
Author Contributions
Jennifer R. Roberts primarily authored this article. She made substantial contributions, including modifying study design, applying for institutional review board approval, digital programming of study materials, recruiting participants, and collecting and analyzing data. She drafted the original paper, participated in revisions, and approved the final paper draft. Molly Maxfield developed the original study idea and hypotheses; she supervised data analysis and interpretation and provided substantial revisions of the article.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This project was funded by the Lilly Family School of Philanthropy at IUPUI.