Abstract
Keywords
The recent upsurge of research on prosocial behavior has led to a large number of emotions and cognitions being proposed as explanations for
This typology contributes to a better and more nuanced understanding of the motivations underlying prosocial behavior in several ways. First, a common strategy for investigating psychological mechanisms underlying prosocial decisions is to test a specific emotion or cognition as a mediator of an observed effect (e.g., Brunel & Nelson, 2000; Ma et al., 2023; Zagefka, 2018), but this approach can easily lead to erroneous conclusions as there are almost always unmeasured variables that could explain the effect better (Fiedler et al., 2018; Pieters, 2017). A central tenet in this article is thus that prosocial behavior can be motivated by several qualitatively different psychological mechanisms, and our typology classifies psychological mechanisms that (de)motivate prosocial behavior into four overarching, interrelated categories: (a) emotions; (b) moral principles; (c) anticipated impact; and (d) anticipated personal consequences (see Figure 1).
The Four Overarching Categories and Their Respective Subcategories of Helping Motives and How They Differ on Two Dimensions.
Second, although there already exist several models and frameworks of prosocial motives (see Appendix), we argue that each of these fails to cover at least some psychological mechanisms underlying human prosociality. Therefore, this typology was developed by merging valuable and complementary insights from existing frameworks, while organizing and reconceptualizing these insights into a comprehensive and pedagogical whole.
Third, while earlier frameworks have typically focused on prosocial motives of a single form of prosociality such as charitable giving (Bekkers & Wiepking, 2011a; Konrath & Handy, 2018), blood donations (Ferguson et al., 2020), or volunteering (Clary et al., 1998), our integrative typology aims to encompass motives for all types of prosocial behavior.
Our typology should be valuable to scholars on human prosociality as it provides a tool for organizing research on psychological mechanisms underlying helping so that it extends the generation of testable hypotheses. It should also be useful for practitioners by providing a conceptual overview of different types of motives which solicitors can “appeal to” when asking for help.
We first set the stage by clarifying what this article is, and what it is not, and explain how key terms should and should not be understood (Pfattheicher et al., 2022). We next introduce the typology, describe how it was derived, and distinguish the inherently psychological “why-question” (which emotions and cognitions motivate helping?), from the “when-question” (which situations increase helping?), and the “who-question” (which people are more likely to help?). In the following four sections, we describe each of the four overarching categories and their respective subcategories. We then showcase similarities and differences between our typology and four other influential frameworks before concluding the article.
Setting the Stage
This article proposes a novel and alternative way to organize and classify
Our contribution should be understood as a theoretical article which includes a critical review. Critical reviews emphasize conceptual innovation and typically present, analyze, and synthesize material from diverse sources and often attempt to integrate different bodies of work, resulting in a synthesis of existing models or a new interpretation of existing data (M. J. Grant & Booth, 2009). The aim of our critical review is to present a framework that covers the various types of psychological mechanisms that have been suggested to motivate or demotivate prosocial behavior.
This article should not be understood as a systematic literature review. Despite citing several hundreds of articles, no formal literature search determined which empirical studies to include. Cited articles instead represent research the two authors have come upon “naturally” while reading, reviewing, and publishing scientific articles on human prosociality for over a decade. Thus, the cited empirical research is used to illustrate and exemplify the various proposed psychological mechanisms but may not be a complete depiction of the extensive literature on prosocial behavior.
Defining Key Terms
We use
It should be clarified that underlying mechanisms of human behavior can be investigated on different levels (Preston & De Waal, 2002). Evolutionary psychologists typically focus on the “ultimate causes” of human prosociality and explain prosocial behavior in terms of natural selection (Burum et al., 2020; van Vugt & van Lange, 2006), whereas medical scientists focus on biological or chemical mechanisms (e.g., oxytocin; de Dreu et al., 2010; Marsh et al., 2020), and neuroscientists investigate activation in certain brain areas when people respond to prosocial decisions (Cutler & Campbell-Meiklejohn, 2019; Harbaugh et al., 2007). The different levels of analysis should be seen as complementary accounts. We are not denying the evolutionary origin of prosocial emotions, that increased activity in a specific brain area makes us experience these emotions, or that thoughts and beliefs are represented in the brain. Still, in this article, we understand psychological mechanisms as emotions, thoughts, or beliefs that,
Examples of Psychological Mechanisms From Each Subcategory.
Four Overarching Categories
We propose a typology with four overarching categories: (a) emotions; (b) moral principles; (c) anticipated impact; and (d) anticipated personal consequences. Each overarching category includes psychological mechanisms that can increase as well as decrease helping, and there are two or three subcategories within each category (see Table 1). We emphasize that the mechanisms are interrelated and rarely experienced in isolation, but we argue that they can be distinguished theoretically as well as empirically and that all can increase or decrease helping motivation independent of each other.
From a requester perspective, the overarching categories can be understood as four psychological strings that one can play on to increase helping. We borrow terminology from Weber and Lindemann (2007) to illustrate how the different psychological strings relate to helping. Appealing to the
The four overarching categories can be understood as separate cells in a 2 × 2 matrix (see Figure 1) with two underlying dimensions. The first dimension differentiates psychological mechanisms that are strongly linked to the helper (internal focus) from mechanisms that are primarily linked to something outside the helper such as the beneficiaries or to universal principles (external focus). The second dimension differentiates psychological mechanisms that are inherently forward-looking and dependent on what the helper anticipates will happen if he or she helps or does not help (future-oriented), from mechanisms that are determined more by what is currently experienced or what has happened previously (past or present-oriented).
Developing the Categories
The four overarching categories (and their respective subcategories) were derived primarily by deduction. Specifically, we conceptually analyzed and merged insights from several influential theories and frameworks (see Appendix). Four influential frameworks that particularly inspired the development of our typology are described and compared against our typology in the end of the article.
The complementary inductive approach implied integrating empirical articles which tested or made claims about the underlying emotions or cognitions of any form of prosocial behavior, while classifying their proposed psychological mechanisms according to the existing typology. In the beginning of this process, new subcategories were sometimes created by the authors to accommodate all encountered psychological mechanisms until we reached a “saturated typology.” We tentatively suggest that the current categories and subcategories—sometimes individually but more often in combination—can explain all specific emotions, thoughts, and beliefs that have been proposed in the covered literature to motivate or demotivate human prosocial behavior.
The Who, When, and Why Questions
This article distinguishes between three questions that are central in research on prosociality; “the Who?,” “the When?,” and “the Why?” (Bhati & Hansen, 2020; Erlandsson, 2014). The
The
The
The Who, the When, and the Why questions of helping are not always clearly separated in the literature. In a seminal review, Bekkers and Wiepking (2011a) suggested a theoretical framework including eight mechanisms:
Emotions
Emotions, feelings, and affect have been intimately linked to moral attitudes and behavior in general (Greene et al., 2001; Haidt, 2001), and even stronger so to attitudes and behavior related to helping (Loewenstein & Small, 2007; van Kleef & Lelieveld, 2022). Emotions can refer to many things in relation to helping. A person faced with a need situation can help because (a) she feels sad for the person in need (emotional reaction elicited by the need situation); (b) she has affectionate emotions toward the beneficiary or the requester (emotions linked to an entity); (c) she is happy because of an unrelated event (incidental mood); (d) she desires the positive emotions that she expects as a result of helping (anticipated warm glow due to helping); (e) she wants to avoid the negative emotions that she expects as a result of not helping (anticipated guilt due to not helping). Moreover, other emotions involved in helping decisions are (f) emotions, feelings, or attitudes toward the very act of helping and (g) emotions that are experienced as a result of helping or non-helping.
An important feature of this typology is that it treats helping driven by
Type (f) (emotions toward helping) is almost indistinguishable from helping motivation, which is the proposed dependent variable, and therefore outside the scope of this article. 2 Finally, Type (g) (emotions experienced after a helping decision) is a consequence rather than an antecedent of helping—even if emotions felt as a response of an earlier helping decision can reinforce future helping by influencing one’s mood (Aknin et al., 2012). Next, we explain and exemplify the three subcategories of emotions (Types 1–3 above).
Emotional Reactions Felt as a Response to a Need Situation
Emotional reactions are immediately experienced as a response to being exposed to a need situation, such as hearing a story about a child who is about to die from malnourishment. Labels used for describing these emotional reactions include empathic concern, empathic arousal, empathy, sympathy, pity, distress, tenderness, compassion, and sadness. Pinpointing the nuances between these is outside the scope of this article (see Cuff et al., 2016; Decety & Cowell, 2014), but one distinction should be clarified.
Both empathic concern and distress are emotions with negative valence that are typically experienced when learning about a need situation (Batson, 2011; Dickert & Slovic, 2009). However, while distress is directed inwards (e.g., “I feel bad”), sympathy is directed outwards (e.g., “I feel bad for the person in need”). Sometimes distress and empathic concern can motivate identical behaviors, but if there are alternative ways to feel better, such as switching the channel to a comedy rather than watching the news, the person feeling distress is likely to do so, whereas the person feeling empathic concern is not (Cialdini & Kenrick, 1976).
Empathic concern clearly fits in the emotional reaction subcategory. Distress, on the other hand, can be categorized as either an emotional reaction or as an anticipated personal consequence. This depends on whether the person believes that engaging in helping will reduce the personal distress or not. Helping because of feeling sad from hearing about a need situation (an emotional reaction) is different from helping because one anticipates feeling better if one helps (a type of emotion regulation), even if these motives often co-occur. Helping to avoid feeling bad is included in the anticipated personal consequences category.
Many studies comprise one or several types of emotional reactions as a possible psychological mechanism underlying helping (e.g., Bagozzi & Moore, 1994; Graziano et al., 2007; Kogut & Ritov, 2005b, 2007; Pagano & Huo, 2007; Pfattheicher et al., 2020; Saito et al., 2019; Västfjäll et al., 2014). Empathic concern has been found to predict helping rather well in many situations but less well in other situations. For example, in one study, native Germans and Turkish immigrants read either about Markus (a German) or Mohammed (a Turk) experiencing problems. When hearing about an ingroup victim, but not when hearing about an outgroup victim, empathic concern predicted helping motivation (Stürmer et al., 2006; see also Maner & Gailliot, 2007).
Empathic concern and similar emotional reactions have been suggested as the mediating psychological mechanism of both individual differences (the who-question; e.g., women are more prone than men to feel and to express emotions when observing a need situation, Rueckert et al., 2011; Sisco & Weber, 2019) and situational differences (the when-question; e.g., people help identified beneficiaries more than statistical because the face and name of a single person in need increase empathic concern; Erlandsson et al., 2015, 2017).
Emotions Linked to a Specific Entity
Not all emotions that influence helping behavior are elicited by the need situation. Some emotions such as admiration, attraction, and “aww” (the feeling of cuteness; Buckley, 2016) are linked to a specific entity such as a person, an animal, or an organization. Unlike empathic concern and distress, these attitude-like emotions are present also in the absence of a need situation, but if an individual that we admire or find unusually cute is in need, the emotions elicited by the need situation are likely boosted to higher levels. In line with this, Batson (2011) argues that a perception of need together with valuing the beneficiary’s welfare are necessary antecedents for empathic concern. Entity-linked emotions can be positive (e.g., respect, awe; Keltner & Haidt, 2003), but also negative (e.g., anger or irritation; Fong, 2007; van Diepen et al., 2009), and thus both motivate and demotivate helping.
Liking of the beneficiary predicts specifically some forms of helping. Cryder et al. (2017) suggest that people have an intuitive push toward helping more attractive beneficiaries. Relatedly, the research by Stürmer et al. (2005, 2006; Stürmer & Snyder, 2010) found that while empathic concern was a stronger predictor of helping one’s ingroup, liking and perceived attractiveness were stronger predictors of helping people from one’s outgroup. Also, small details about beneficiaries’ worldviews or prior behavior can increase as well as decrease their perceived attractiveness and consequently influence helping toward them (Gershon & Fridman, 2020; Sole et al., 1975).
Entity-related emotions can be directed toward the beneficiary but also toward the requester or fundraiser (Chapman et al., 2022). Sometimes the beneficiary and the requester are the same such as panhandling, or when an injured person screams for help; but often, the requester and the beneficiary are different such as when giving to a charitable organization. Physically attractive requesters made people donate more in a door-to-door fundraiser (C. E. Landry et al., 2006), and requesters who show gratitude with kind words or smiles will likely successfully solicit help also in the future, as emotions toward the requester become more positive (Merchant et al., 2010). A solicitation for help can also induce negative emotions toward the requester, if the solicitation is aggressive, annoying, or perceived as manipulative (Cotte et al., 2005; Hibbert et al., 2007; Kang et al., 2022). Requests such as “you simply have no choice—from a moral perspective you must help” can induce psychological reactance meaning that people feel that their free will is threatened and stop helping to reinstate their sense of autonomy (Berkowitz, 1973; Quick, 2012; Reinhart et al., 2007).
Incidental Mood
Emotions completely unrelated to the helping situation can also influence decisions to help (Anik et al., 2011; Small & Lerner, 2008; Västfjall et al., 2016). A positive mood makes it more likely that we notice need situations, whereas a negative mood makes us more self-focused. A happy mood makes us more energetic, approach-oriented, and interested in others’ well-being (M. Carlson et al., 1988; Fredrickson, 2001; van Kleef & Lelieveld, 2022), and when in a positive mood (e.g., because one passed an exam), helping can be done to commemorate one’s personal success. Participants who are smiled at by an unrelated person tend to help more when subsequently observing a need situation, likely because of an improved mood (Guéguen & De Gail, 2003), and an induced positive mood in combination with a high self-awareness has been suggested to increase helping (Berkowitz, 1987). In contrast, experiencing unrelated negative emotions (e.g., anger; Gummerum et al., 2016) or negative visceral states (e.g., hunger; Harel & Kogut, 2015) makes people focus on their own suffering more than others’ suffering.
Still, the finding that positive mood increases and negative mood decreases helping is disputed. If a happy person needs to be exposed to a situation that will make her distressed, or if helping is costly, it is possible that she will refrain from helping to maintain her positive mood (Isen & Simmonds, 1978; Sabato & Kogut, 2021). Also, if a person in a negative mood believes that helping will make her feel better due to an increased sense of meaningfulness or because of anticipated social or emotional benefits, she is more likely to help (Cunningham et al., 1990; Dovidio et al., 2006). These findings suggest that people experiencing incidental negative mood, as well as people experiencing personal distress from observing a need situation, sometimes help because helping can be a way to regulate their emotions and improve their mood (c.f. the negative-state-relief model, Cialdini & Fultz, 1990; Cialdini et al., 1987). As helping to improve one’s mood is future-focused, it is included in the anticipated personal consequences category in our typology.
To say that emotions motivate helping is not the same as to say that feeling stronger prosocial emotions necessarily make us help more. Also, it is not the same as to say that a change in helping is always a result of a change in emotions. Emotions can increase even without a subsequent increase in helping, and helping can increase even without a preceding increase in emotions. We now turn to other psychological mechanisms that have been shown to motivate or demotivate helping.
Moral Principles
The second overarching category is based on perceived rules, obligations, or internalized norms and is here referred to as moral principles (Bekkers & Ottoni-Wilhelm, 2016; Gross & Vostroknutov, 2022; Tomasello, 2020). Whereas the emotion category includes mechanisms about “
Several types of moral principles are involved when making helping decisions. For instance, religious people might follow the decrees stated in their holy scriptures because they believe that they ought to. Likewise, many non-religious people are primarily motivated to abide the law, act justly, do their duty, not violate others’ rights, or simply to “do the right thing” (Capraro & Rand, 2018; Tappin & Capraro, 2018). Loosely inspired by the moral foundation theory (Haidt, 2012), the three subcategories of moral principles are: (a) responsibility, (b) fairness, (c) not causing harm.
Responsibility
Helping others is sometimes perceived as mandatory, so not helping implies that one has failed to fulfill one’s personal responsibility. At other times, helping is over and above one’s responsibility, and therefore, optional (Leliveld et al., 2008; McManus et al., 2020). Responsibility in helping situations can refer to a causal responsibility (e.g., “I caused the need situation”) or to a moral responsibility (e.g., “I am personally responsible to help”). We are focusing on the latter, but causal responsibility is typically strongly related to moral responsibility. If someone is suffering because of my mistake, I will likely believe that I have a responsibility to rectify my mistake by helping, but if the same person is suffering because of her own mistake or because of someone else’s mistake, my perceived personal responsibility to help will be weaker, and helping is less certain (Clayton et al., 2013; Henry et al., 2004). In one study where different situational factors were tested as predictors of helping motivation, having caused the situation was the best predictor (Fritzsche et al., 2000).
There are also other types of contextual factors that can result in a high perceived moral responsibility (Erlandsson et al., 2016). For example, some occupations (police officer or nurse) or social positions (being an authority figure) come with an increased role-based responsibility (Haidt & Baron, 1996; Jeske, 2021). Similarly, people feel a greater responsibility to help children, elderly, and baby animals because these beneficiaries are perceived as vulnerable and defenseless, whereas non-disabled adults are typically perceived to have a greater responsibility to take care of themselves (e.g., Chasteen & Madey, 2003; Moche & Västfjäll, 2021). Another factor that can increase our responsibility to help is promise-making in that we perceive ourselves to be more obliged to act when we have made a promise to do so (Kerr et al., 1997; Vanberg, 2008). To a lesser extent, just the mere existence of an expectation to receive support could increase the perceived personal responsibility to help (Goldstein et al., 2011; Nagatsu et al., 2018). One can also experience a personal responsibility to reciprocate for privileges previously enjoyed, for instance, alumni donating money to their alma mater (McDearmon, 2013).
Perceived moral responsibility to help can be diluted by the presence of other potential helpers (i.e., the bystander effect; Cryder & Loewenstein, 2012; Darley & Latane, 1968). In such a situation, the responsibility to help is perceived to be split between all bystanders meaning that the personal responsibility to help is smaller than when one is the only potential helper. One study found that people (no matter their own financial situation) believed that those richer than themselves had a responsibility to help more (Berman et al., 2020). Moreover, people in countries with a sophisticated welfare system tend to believe that the governments are responsible for providing health care and social services to those worse off and consequently perceive a weaker personal responsibility to help (Nelson et al., 2006; Vamstad & von Essen, 2013). Simply put, the amount of perceived responsibility can be understood as a zero-sum game, where an increased responsibility of others (the beneficiary or other potential helpers) decreases one’s own responsibility to help.
Individual and cultural differences in perceived personal responsibility to help have been suggested as variables that determine people’s motivation to help (Baron & Miller, 2000; Barrett et al., 2004; Schwartz & Howard, 1980). To illustrate, Ottoni-Wilhelm and Bekkers (2010) suggest that the predictive power of dispositional empathy drops in magnitude and often loses significance after internalized moral principles related to helping are controlled for (see also van der Linden, 2011), and Winterich and Zhang (2014) found that the main reason for why people high on power-distance are helping less is that their perceived personal responsibility to help is lower.
Perceived responsibility is frequently included in studies on helping, sometimes as a proxy for helping motivation (e.g., Jenni & Loewenstein, 1997; Nagel & Waldmann, 2013), but more often as a possible underlying psychological mechanism of helping (e.g., Basil et al., 2006; Duval et al., 1979; Kleber et al., 2013; Lee et al., 2014). Perceived personal responsibility has been argued to be the best psychological mechanism for understanding the ingroup effect (Erlandsson et al., 2015, 2017), meaning that we help ingroup beneficiaries such as relatives and fellow countrymen more than outgroup beneficiaries such as non-relatives and foreigners, primarily because we believe that we have a personal responsibility to do so (see also Baron et al., 2013).
Fairness
Fairness is another moral principle that people strongly care about and want to uphold, both on a procedural level and on an outcome level (Ajzen et al., 2000; Tyler, 2000). Most people value equality (equal outcomes) and, to an even greater extent, equity (equal outcomes to equal inputs), and both these concepts are intimately linked to perceived fairness (Gordon-Hecker, Choshen-Hillel, et al., 2017; A. Shaw, 2013).
In prosocial situations, fairness concerns can increase helping, for instance, when helping improves the lives of those who are worse off, reduces inequalities, or when helping is provided to those who deserve it the most (A. Shaw & Olson, 2012). Fairness concerns can also decrease helping, when the beneficiaries one can help are extremely privileged (Bradley et al., 2019; Fehr & Schmidt, 1999), or when helping implies that one must give unequal treatment to two equally deserving beneficiaries. People are even prepared to sacrifice efficiency by discarding resources when they cannot help everyone equally much (Choshen-Hillel et al., 2015; Gordon-Hecker, Rosensaft-Eshel, et al., 2017).
Another way fairness concerns influence helping is that people can feel obliged to help after learning that others have helped, to avoid unfair free-riding (Frey & Meier, 2004; Goldstein et al., 2011). To illustrate, an intervention intended to recruit blood donors with the help of making reciprocity norms salient involved first asking “Would you be willing to accept a transfusion if you needed one in the future” before asking “Are you willing to donate blood?” (D. W. Landry, 2006).
Not Causing Harm
People are aversive toward unfair outcomes, but even more aversive toward the notion of personally implementing unfairness (Beattie et al., 1994; A. Shaw & Olson, 2014). For instance, people are more willing to create inequitable outcomes when they can let a random device determine which of two possible beneficiaries will be helped, than if they must make the decision themselves (Gordon-Hecker, Rosensaft-Eshel, et al., 2017).
Relatedly, people normally think that an omission to help is more acceptable than a helpful action that causes harm to innocent bystanders. This is clearly illustrated in the sacrificial dilemmas commonly used in moral psychology (Bauman et al., 2014; Kahane et al., 2018). Many people are reluctant to help five (or more) people about to be run over by an out-of-control trolley, if helping implies killing one person, especially if the killing implies physical force (Greene, 2008, 2013). This tendency can be explained by an internalized aversion toward causing harm that is stronger than the aversion toward allowing harm to happen (omission bias; Baron, 2008; Ritov & Baron, 1990, 1999). Expressed differently, actively harming someone is equivalent to breaking a moral rule or violating a protected value, whereas omitting to help is normally more excusable (Baron & Spranca, 1997; Krettenauer & Johnston, 2011). This type of thinking sometimes makes people refrain from helping, not only when they know that helping will harm others but also when there is uncertainty about possible side-effects of prosocial behavior (C. J. Anderson, 2003; Everett et al., 2015; Sunstein, 2005).
Anticipated Impact
While emotions and moral principles focus primarily on the past and on the present, the remaining two categories focus on the anticipated future consequences of helping decisions. We first discuss anticipated consequences for the beneficiaries and then move to anticipated consequences for the helper.
Anticipated impact represents a more calculating psychological mechanism than emotions or moral principles, and concepts such as perceived impact, utility, outcome, effectiveness, efficiency, and “making a difference” are all included in this category (Bodem-Schrötgens & Becker, 2020). People value effectiveness when making decisions about themselves, but also when making decisions about others (Duncan, 2004), and help more when they believe that their help will significantly improve the beneficiary’s condition (Smith et al., 1989). On the other hand, people are demotivated when they expect that their helping will not do any difference, or be a “drop in the bucket,” and even more demotivated if they expect that helping will aggravate the need situation. Anticipated impact is often measured next to some type of emotional reaction when trying to determine if a helping effect is driven by quick intuitive System 1–like processes or by slow deliberate System 2–like processes (Camilleri & Larrick, 2019; Cryder et al., 2013a, 2013b; Dickert, Kleber, et al., 2011; Friedrich & McGuire, 2010; A. M. Grant et al., 2007).
We distinguish two types of anticipated impact—self-efficacy and response efficacy (e.g., Bandura, 1978; Basil et al., 2008; Sharma & Morwitz, 2016). In a helping context, self-efficacy refers to the belief that “I am able to make a difference,” whereas response efficacy refers to a belief that “a difference can be made,” or that a specific project or helping enterprise is effective and worthwhile, regardless of who is doing the helping.
Self-Efficacy
Helpers who have a desire to personally make a difference are referred to as “impact philanthropists” (Duncan, 2004). This desire can give rise to a preference to give directly to a homeless person rather than to an organization or to a preference to devote one’s helping effort to a small, clearly defined, cause (Caviola et al., 2020). The prospect of “fixing a small problem once and for all” is more appealing than the prospect of contributing to mitigate a much larger problem, despite the fact that the actual amount of good one can do is typically greater when focusing on large-scale problems (Li & Chapman, 2009; Zhang & Slovic, 2018). Likewise, people are most likely to provide support when a fundraiser is close to its goal, and this “goal gradient effect” has been linked to perceived self-efficacy (Cryder et al., 2013b; Kuppuswamy & Bayus, 2017). Similarly, self-efficacy mediates the proportion dominance effect, meaning that people are more motivated to rescue 10 out of 12 otters for $200, than they are to rescue 10 out of 1000 otters for the same amount. The reason for this is that the anticipated personal impact is perceived to be higher when helping 10 out of 12 than if helping 10 out of 1000 (Erlandsson et al., 2014, 2015, 2017).
A field study by Gneezy et al. (2014) showed that if a large sum of money is used to cover all administrative costs of a charity organization (implying that 100% of the subsequently donated money will reach the beneficiaries), donations from the public will increase more than if the large sum of money is used as seed money or as matching money (List, 2011). The authors suggested that this is because people perceive that the impact of their own donation is greater when someone else has paid for all overhead costs.
Response Efficacy
Whereas an impact philanthropist is primarily interested in self-efficacy and in “personally making a difference,” other impact-motivated helpers focus on response efficacy, which means that they want to help in ways that are as effective as possible, generally speaking. Perceived response efficacy (or outcome efficacy) can differ as a function of the charitable cause, as some problems in the world are perceived as more alterable than others, or as a function of the charitable organization, as some organizations are perceived to be more competent than others. A higher perceived organizational effectiveness has been shown to increase helping motivation, and non-profit organizations perceived as professional, credible, transparent, and trustworthy typically elicit more support (Jones, 2017; Sargeant & Woodliffe, 2007; Xiao et al., 2022).
On the contrary, organizations that have a relatively high overhead cost or reside in a fancy office space are perceived as wasteful and inefficient, whereas low overhead costs or modest salaries to employees are typically (often mistakenly) seen as evidence for an effective charity (Caviola et al., 2014; Newman et al., 2019; Portillo & Stinn, 2018; Qu & Daniel, 2021). Sending out a charity appeal in a gloss envelope with a printed color picture to sporadic givers rendered fewer donations than an identical charity appeal in a blank simple envelope (Bekkers & Crutzen, 2007). In another study, a 4-star rating rather than a 2-star rating on a homepage indicating how efficiently charities use their money increased the proportion of people who donated money by forfeiting a personal discount (Winterich & Barone, 2011). Low-impact information decreased while high-impact information increased donations among people who requested information about a charitable project, but not among those who did not, suggesting that the importance of impact varies across helpers (Metzger & Günther, 2019).
Perceived response efficacy is not the same as actual response efficacy. Due to framing effects and heuristics, we sometimes perceive effective helping as ineffective and vice versa (Slovic, 2007). The well-known effective altruism movement argues that people should maximize response efficacy when making a prosocial decision (MacAskill, 2015; Singer, 2016). However, rather than using crude heuristics such as minimizing overhead costs, or as always, giving to those worse off (e.g., Baron & Szymanska, 2011; Sunstein, 2005), effective altruists are explicitly motivated to help in ways that do the most good from an impartial and universal point of view and devoted to using scientific methods to determine the response efficacy of different helping behaviors in an objective way.
Both types of efficacies can be linked to individual differences in trust in the charitable sector and more generally to social capital (Konrath & Handy, 2018; Sargeant & Lee, 2004). Some people have much confidence in charitable organizations, whereas others are inherently skeptical. Also, some people believe that they have the resources and the ability to help in efficient ways, whereas others do not. One study investigating gender differences in volunteering found that women help primarily through emotional reactions and moral principles, whereas men’s helping is motivated more through trust and social networks (i.e., anticipated impact motives; Einolf, 2011). Other studies found that impact information increases helping only among the relatively educated (Verkaik, 2016) and among major prior donors (Karlan & Wood, 2017).
Anticipated Personal Consequences
The fourth and final overarching category includes different types of anticipated consequences for the person faced with the helping decision. Anticipated consequences for oneself can be classified as positive (personal
Anticipated Personal Consequences That Motivate Helping
Anticipated Positive Material Consequences of Helping
People sometimes see helping as an investment in that a smaller personal cost now will pay back in the future because of direct reciprocity norms (Leimgruber, 2018; Trivers, 1971). Prosocial activities can also be done to gain material benefits, for instance, by boosting one’s chances of being hired by adding volunteering activities to their resume (Handy et al., 2010). People might also help to be eligible for a tax deduction (Duquette, 2016) or to obtain a symbolic gift or a free health check when donating blood (Goette & Stutzer, 2020).
Anticipated Negative Material Consequences of Not Helping
It is possible to help primarily to avoid a greater material or physical loss. To exemplify, a person who encounters a stranger who behaves erratically in a deserted place might lend his or her cell phone or give $20 to “tend and befriend” and reduce the risk of being the target of aggression (Taylor, 2006). There is also some support for “protective donations” meaning that people anticipate that “bad things happen to bad people” and help because they believe that a decision to not help might affect their “karma” and physical well-being negatively (Converse et al., 2012; Kogut & Ritov, 2011; Lerner, 1980).
Anticipated positive social consequences of helping
Even in the absence of any tangible benefits, people often help to expand their social network (meet new friends) or to receive praise and improve their reputation among others (R. A. Anderson et al., 2020; Exley, 2017; Milinski et al., 2002; Raihani & Smith, 2015). Helping is usually seen as a social norm, and as norms get more important in the presence of others, people help more when they believe they are being observed (Caviola & Faulmüller, 2014; Rayniers & Bhalla, 2013). For instance, people offer to volunteer more if they believe their decision is made in public rather than in private, and offering to volunteer (vs. not doing it) typically improves one’s social reputation (Bereczkei et al., 2007; Fisher & Ackerman, 1998). 4
Individual and situational factors influence anticipated social benefits of helping. Men help more when they believe they are being observed by an attractive female than if they believe they are being observed by a male or not observed by anyone (C. E. Landry et al., 2006; van Vugt & Iredale, 2013). Similarly, helping can be used as a humble way to signal one’s prosperity (Hardy & Van Vugt, 2006; Sisco & Weber, 2019). A person who gives away tens of thousands of dollars to help sick children communicates both “I am a nice person” and “I am a rich person” to observers in general and to potential romantic partners in particular.
Anticipated Negative Social Consequences of Not Helping
Whereas non-helping is rarely associated with material personal costs, it is often associated with social personal costs. A person who decides not to help when aware of a need situation runs the risk of being perceived as a callous and unsympathetic person by observers (Vonasch et al., 2018). People who find themselves in a public helping situation are often aware of this and might therefore help to avoid the social condemnation or loss in reputation they expect will follow if they do not help (Bradley et al., 2018). This type of helping has been referred to as “giving in,” meaning that the helper prefers not to be asked but still helps to avoid the social costs associated with saying no (Cain et al., 2014; see also Andreoni et al., 2017; DellaVigna et al., 2012; Flynn & Lake, 2008; Lin et al., 2016).
Helping to evade negative social consequences (avoiding blame) is not the same as helping to obtain positive social consequences (approaching praise; Wiltermuth et al., 2010). Helping to avoid blame is typically done in situations where perceived external expectations are high (e.g., when the descriptive norm is to help; Cialdini et al., 1990; van Teunenbroek et al., 2020), whereas helping to obtain praise is typically done in situations where helping is costly, unexpected, and perceived as extraordinary (Berman & Silver, 2022; Wiepking & Heijnen, 2011).
Anticipated Positive Psychological Consequences of Helping
Most people can benefit personally from helping, also in the absence of any material or social benefits. Even helping that is performed in private, such as anonymously donating money to a charity fundraiser, can reward the helper by making her feel meaningful or self-satisfied because of the prosocial act (Baumann et al., 1981; Cialdini & Kenrick, 1976; Duclos & Barasch, 2014). The positive emotion that is felt after a moral decision is often referred to as “warm glow” or “emphatic joy,” and anticipating these positive emotions can motivate helping (Andreoni, 1990; Batson et al., 1991; Västfjäll et al., 2015).
A relevant finding is that material or social rewards associated with helping tend to reduce the anticipated psychological rewards of helping (Andersson et al., 2022; Mellström & Johannesson, 2008; X. Wang & Tong, 2015). Consequently, helpers who are primarily motivated by anticipated psychological rewards prefer non-incentivized helping done in private over incentivized helping done in public. Relatedly, people help more when helping behavior is framed as a “charitable purchase” rather than as a “donation rewarded with a gift.” The reason for this is that a “charitable purchase” is internally compared against other (non-charitable) purchases, which provides the helper with warm glow for doing something relatively moral. “Donation rewarded with a gift” is instead internally compared against other non-rewarded donations, thus not eliciting as much warm glow (Chao, 2017; Krishna, 2011; Savary et al., 2020; Zlatev & Miller, 2016).
Mechanisms in the anticipated impact category are often correlated with anticipated warm glow. Studies have shown that when people believe their helping is directed to credible and worthwhile causes or organizations (i.e., high response efficacy), and when they believe that their personal effort is making a noticeable difference (i.e., high self-efficacy), they also anticipate and actually experience more warm glow (Aknin et al., 2013; Cryder et al., 2013b).
Anticipated Negative Psychological Consequences of Not Helping
Just as anticipated warm glow as a result of helping can motivate prosocial behavior, so can anticipated guilt (or other anticipated negative emotions or cognitions) in the case of non-helping motivate us to help even in the absence of any observers (Ahn et al., 2014; Basil et al., 2008; Elgaaied, 2012; Lindsey, 2005; X. Wang, 2011). 5 People try to forecast their own emotional responses, and as we are more motivated to avoid the negative than to approach the positive (e.g., Rozin & Royzman, 2001), we tend to adjust our behavior so that we avoid negative emotions as much as possible.
As with praise and blame, the helping situations that we anticipate will make us feel warm glow are different from the situations that we anticipate will make us feel guilt (Erlandsson et al., 2016; Krettenauer & Johnston, 2011). We anticipate feeling guilty after not helping primarily in situations where we believe we are personally responsible to help but anticipate feeling warm glow after helping primarily when there is no external or internal pressure to help (Dunn et al., 2014).
Anticipated Personal Consequences That Demotivate Helping
Anticipated Negative Material Consequences of Helping
The most obvious reason for deciding not to help is that helping comes with a cost (Ferguson et al., 2019; Kelting et al., 2019; Rubaltelli et al., 2020). The material cost can be monetary, for example, “giving away money will reduce my disposable income,” temporal, for example, “volunteering will take time from more personally beneficial activities,” or physical, for example, “donating blood is painful and implies some risk.” Some digital helping behaviors might also increase the risk of being scammed (Whitty, 2020). All these costs can be avoided simply by deciding not to help.
Anticipated Positive Material Consequences of Not Helping
This is a different framing of the abovementioned mechanism. If I opt out from my monthly $500 donation to a charitable cause, I will have additional money to spend on myself or my family. Illustrated differently, people tend to donate less when reminded of alternative, more lucrative, ways to spend their money (i.e., opportunity costs; Moche et al., 2020).
Anticipated Negative Social Consequences of Helping
Research on moral impression formation clearly shows that helping is not unequivocally perceived in a positive light (Berman & Silver, 2022; Cramwinckel et al., 2015; Critcher & Dunning, 2011). At times, helping is perceived as a sign that the helper is either motivated for “the wrong reasons” or helps to put others in a bad light (Erlandsson et al., 2020; Newman & Cain, 2014; Raihani & Power, 2021). Derogatory terms such as “do-gooder” or “goodness-junkie” suggest that helping, at times, does not only “not improve” but even worsen the helper’s reputation (Minson & Monin, 2012; Monin et al., 2008; Silver et al., 2021). For example, although people like helpers and promise-keepers, they do not like those who help too much or those who exceed their promises (Gneezy & Epley, 2014; Klein & Epley, 2014; Pleasant & Barclay, 2018).
To the extent people are aware of this, they might refrain from helping if the helping is inherently public and if they believe that non-helping is the local norm (Jones & Linardi, 2014). Similarly, when helping publicly, people are careful not to appear motivated by ulterior motives (Lin-Healy & Small, 2012; Scopelliti et al., 2015), and this could result in less helping in some situations. In one study where participants could help a well-liked charity organization by pressing a button, performance was lower if participants also had publicly known egoistic incentives for pressing (Ariely et al., 2009; White & Peloza, 2009). On the other hand, helping that comes with a high physical pain (e.g., pouring an ice bucket over oneself to raise awareness of a disease) are more socially and psychologically rewarding than pleasurable helping (e.g., a charity picnic; Gneezy et al., 2012; Olivola & Shafir, 2013, 2018).
Anticipated Positive Social Consequences of Not Helping
In circumstances involving intergroup animosity or beneficiaries from stigmatized groups, a potential helper seeing an outgroup member in need might help if the helping is made in private, for example, because of empathic concern or moral principles, but refrain from helping when in public (Snyder et al., 1999). Both helping and non-helping can signal one’s loyalty and group-belonging, and if I anticipate that non-helping will improve my status within the group, that could result in less helping. To exemplify, one study found that people dislike coworkers who act prosocially toward superiors (Vonk, 1998), and another found that people want to interact with people who “do their part” but not with those who behave extraordinary unselfish (Parks & Stone, 2010) because extreme helpfulness can be perceived negatively as it sets undesired norms and makes others look bad in comparison. To the extent that people are aware of this, they might consciously reduce their helping in public situations.
Anticipated Negative Psychological Consequences of Helping
Sometimes helping is psychologically difficult as it exposes oneself to needs and sufferings one was previously blissfully unaware of (Cameron & Payne, 2011; Kirby et al., 2019; L. L. Shaw et al., 1994). Also, agreeing to help once tends to increase the number of requests one will get in the future and intensify expectations that one will continue to help henceforward. By committing to help someone in acute need, one also faces the risk of failure such as a person dying while you are performing CPR on her, and this can be traumatic as well as reduce your self-confidence. Again, people can predict these consequences, and anticipation of an increased exposure of suffering with subsequent personal distress, of more requests and greater expectations to help, or a fear of not being able to help successfully can demotivate people from helping in the first place (Cameron et al., 2015). Providing indirect support for this mechanisms, one field experiment gave donors the option to donate once and never be contacted again, and this increased subsequent donations not only in the short but also in the long term (Kamdar et al., 2015).
Anticipated Positive Psychological Consequences of Not Helping
Although not a widespread mechanism, it is possible to imagine people who are in a position to help but refrain from doing so because they anticipate pleasure from non-helping (c.f. schadenfreude; Leach et al., 2003).
Discussion in the Context of Other Theoretical Frameworks
The current typology was developed partly by integrating and re-organizing insights from earlier theories and frameworks. In this section, we discuss the similarities and differences of our typology to four seminal frameworks for our theorizing (see also Appendix).
First, the
Despite several similarities, the revised theory of planned behavior lacks specific prosocial motives, such as emotional reactions elicited by the need situation and some types of anticipated personal consequences. Moreover, we argue that “attitude toward a helping behavior” is a fuzzy concept typically assessed by asking people how “negative-positive” or “bad-good” they think helping is (White et al., 2023). Someone stating that they helped because they think helping is good is not necessarily communicating any meaningful information about their specific prosocial motives.
Second,
In Weber’s framework, the calculation-based decision mode focuses exclusively on consequences for the helper, which makes it identical to our anticipated personal consequences category. However, a helper could also calculate consequences for the beneficiaries, and then we are very close to response efficacy, which is a part of the anticipated impact category. Weber’s affect-based decision mode superficially resembles our emotion category but does not clearly spell out how emotional reactions can motivate helping when observing others in need.
Third, Batson’s (2022)
Batson’s collectivism motivation can, in our typology, be understood both as “positive emotions toward the beneficiary,” as we value the welfare of some groups more than the welfare of other groups, and as “responsibility,” as we see helping the ingroup as a moral obligation but helping the outgroup as supererogatory. Nonetheless, Batson’s framework lacks goals related to anticipated impact and welfare-maximization.
Bekkers and Wiepkings Mechanisms Finally, Bekkers and Wiepking’s (2011a) seminal and influential article suggests eight mechanisms (psychological and non-psychological) that drive charitable giving, including five that are represented in our typology—
Awareness of Need
We see awareness of need as a requirement for helping rather than a psychological mechanism that influences helping. Without noticing a situation and interpreting it as a need for help, it is impossible to experience the emotions, thoughts, or beliefs that are included in the typology. 6
However, one can argue that awareness can come in different degrees and that a greater perceived need elicited, for example, by increased media coverage, increases helping. We suggest that perceived need can motivate helping both through an anticipated impact mechanism such as a belief that one can make a greater difference by focusing on the worst-off beneficiaries due to a higher marginal utility and through moral principles such as fairness concerns or a perceived obligation to prioritize those in the greatest need.
Moreover, Bekkers and Wiepking exemplify awareness of need by suggesting that the more people know about a specific cause, or about a specific type of plight, the more likely they are to support the charitable cause that focuses on the underlying problem. For instance, people donate more to cancer research if they know people who suffered from cancer (Bennett, 2003; Small & Simonsohn, 2008; Zagefka et al., 2013). We suggest that both emotional reactions and moral principles might underlie this “personal-experience effect.”
Emotional reactions could explain this effect because being reminded of, and learning more about a specific cause, or about the fate of a specified individual elicits more emotions toward that cause and makes emotional reactions more persistent. Likewise, studies show that emotional reactions toward beneficiaries are weakened by unrelated pictures or by knowledge about beneficiaries that cannot be helped, as these distractors divert attention away from the need situation and from the beneficiaries one can help (Dickert & Slovic, 2009; Västfjäll et al., 2015). The personal-experience effect could also be explained by moral principles. When someone you know is suffering from a specific illness, fighting that illness becomes a mission that one must engage in personally, whereas helping is optional for others, for example, “It’s my obligation to donate to Alzheimer research because my dad suffered from it, but as your brother died from cancer, you should focus on that.”
Solicitation
Bekkers and Wiepking’s “
With that said, several of the psychological mechanisms in our typology can be used to explain why some types of solicitations elicit more helping than others. Anticipated social and psychological personal consequences play a major role as it is more costly—both reputationally and psychologically—to decline to help when solicited face to face as opposed to indirectly. Also, some charity requests (e.g., those presenting one identified victim; Kogut & Ritov, 2015; Lee & Feeley, 2016) attempt to increase helping by eliciting compassionate emotions, whereas other requests attempt to elicit a high anticipated impact and increase helping that way (e.g., those emphasizing low overhead costs, e.g., Gneezy et al., 2014). As people differ in their primary motives to help, it is possible to purposely “appeal to” different psychological mechanisms when advertising for different segments of helpers.
Altruism
There exist many definitions of altruism (Pfattheicher et al., 2022), and Bekkers and Wiepking seem to use altruism primarily in its “economic meaning,” not in its “utilitarian meaning,” nor in its “social psychological meaning.” Altruism in economy is understood as something very close to self-efficacy, meaning that people motivated by altruism want to make a substantial difference and that others’ helping may “crowd out” personal helping because the personal impact of helping is then reduced, relatively speaking (Duncan, 2004).
An alternative way to understand altruism is to shift from self-efficacy (how much good can I do?) to response efficacy (how much good can be done in total? Genç et al., 2020; Singer, 2016). This type of altruism is proposed by the effective altruism movement that base their suggestions on what the anticipated consequences of different decisions are in terms of utilitarian cost-benefit calculations. Both the economic (self-efficacy) and the utilitarian understanding of altruism (response efficacy) are included in the anticipated impact category in our typology.
To complicate things further, altruism is used in a third way in social psychology. Specifically, Batson (2011) argues that the only altruistic motive for helping is empathic concern which is a type of emotional reaction that people often feel when observing a need situation. Altruism, in this sense, is then related to emotions rather than to anticipated impact. Because of the multiple ways the term “altruism” is used in the literature, we prefer not to include it at all in our typology. 7
Conclusion
We have proposed a new comprehensive typology that we hope will help other researchers synthesize and organize the numerous emotions, thoughts, and beliefs that have been shown to increase or decrease different helping behaviors. Our typology consists of four overarching categories (emotions, moral principles, anticipated impact, and anticipated personal consequences) and has been inspired by several other influential models and frameworks (see Appendix).
We acknowledge that our contribution is somehow unconventional. For instance, we opted against performing a systematic literature search, not only because it would be a massive task as we do not limit ourselves to one form of prosocial behavior but also because it would give readers a false sense of comprehensiveness because of severe jingle and jangle problems in the field (Hall & Schwartz, 2022; Pekrun, 2023). The jingle fallacy occurs when the same word is used to describe several different phenomena, whereas the jangle fallacy occurs when different words are used to describe the same phenomenon, and both fallacies imply substantive problems when determining search words. Compared to a systematic literature review approach, our critical review approach is less vulnerable to jingle-jangle problems, as it relies less on the names and labels of constructs, and more on how the constructs are operationalized.
It is possible that our typology might need refinement in the future (e.g., added subcategories), but we believe that the idea to conceptually disentangle the “why-question” of helping from the “when-question” and from the “who-question” will make it easier to evaluate and understand existing and future studies related to the psychological underpinnings of human prosociality. All these questions are valuable in their own right (and even more valuable in combination), but it is problematic to draw inferences about underlying psychological mechanisms based only on observed individual differences or situational factors alone. We argue that it is fruitful to recognize that these factors can influence helping behavior through several different types of emotions, thoughts, and beliefs.
Footnotes
Data Availability Statement
Declaration of Conflicting Interests
Funding
ORCID iD
