Navigating emotional depletion: the role of emotional intelligence and donor status in shaping donation intentions Open Access

Recent crises, like the COVID-19 pandemic, wars in Ukraine and Gaza, and disasters such as the Los Angeles fires, can inspire sympathy and support for non-profits. However, the size and ongoing nature of these events can leave people feeling drained and overwhelmed, known as emotional exhaustion (Nong and Mei, 2024). As a result, individuals are experiencing a growing desire to “simply switch off” (Ariza, 2023). This is concerning, given non-profits’ reliance on community empathy for donations, as emotional exhaustion can lead donors and potential donors to become less likely to engage in donation behaviour. This presents a significant challenge for non-profits to continue their service provision and engage in effective marketing for donations.

Existing literature identifies the importance of understanding the distinctions in the thoughts, emotions and behaviours of non-donors and donors (Previte et al., 2019), and how this may guide marketing strategies. Current knowledge mainly focuses on demographic and geographic factors, such as gender (Wymer et al., 2021) and age (Benish-Weishman et al., 2019), as well as attitudinal and cognitive factors like attitudes towards non-profits (Yang et al., 2024). The current research aims to add new insight to non-profit marketing studies by using COR to explore how emotional factors, like emotional exhaustion, compassion fatigue and emotional intelligence (EI), affect donors’ and non-donors’ intentions to donate.

Marketing literature highlights the critical role of emotions in motivating charitable giving, with reviews by Bekkers and Wiepking (2011) and Chapman et al. (2022) underscoring the psychological benefits and emotional drivers of giving, including guilt and empathy. While this established body of work identifies the importance of emotional elicitation for donation behaviours, a critical gap in understanding involves when donors and potential donors may undertake psychological processes that supress emotional elicitation. Considering COR theory, it’s also important to understand how donors manage emotions, which may explain why some experience stronger or weaker emotional responses (Bekkers and Wiepking, 2011; Chapman et al., 2022). From a COR perspective, the literature does not yet explain how individuals may use existing abilities to manage their emotions. One such ability is EI, the capacity to recognise and manage emotions (Chaouali et al., 2021), which may help protect against emotional exhaustion and compassion fatigue. As such, this research explores EI as a mitigation factor against emotional resource depletion while fostering donation intentions.

COR theory further suggests that as emotional resources are depleted, individuals may enter a loss spiral, focusing on self-protection by conserving their remaining resources (Ford et al., 2023), such as money, which may, in turn, reduce their likelihood of donating. As such, this research examines key aspects of how emotional resource depletion impacts financial donation likelihood, emotional exhaustion and compassion fatigue. Specifically, it explores how both donors and non-donors experience emotional exhaustion, including the tendency to “switch off” emotional responses (D’Souza et al., 2023). This is suggested to lead to compassion fatigue, a stress response that reduces the ability to empathise or feel compassion for others (Gonzalez-Mendez and Diaz, 2021) and may ultimately reduce monetary donations. Against this backdrop, the second aim of the current research is to examine the mediating roles of emotional exhaustion and compassion fatigue in the relationship between EI and intentions to donate.

Building on the previous research aims, this study also examines whether these processes differ between donors and non-donors. Prior non-profit and marketing research has explored how donation motivations and attitudes vary based on factors such as gender, age (Wymer et al., 2021; Lee and Park, 2025) and culture (Wang et al., 2015), often analysing first-time donors (non-donors) and recurring donors separately. However, this study takes a different approach by investigating how emotional resource management and depletion processes vary between donors and non-donors. In doing so, it seeks to enrich both academic scholarship and practical applications by identifying where these processes exert the greatest (or least) influence on donation behaviour. In turn, the third aim of the study is to determine whether the interrelationships among EI, emotional exhaustion, compassion fatigue and donation intentions differ between donors and non-donors.

To test these aims, the current research, guided by COR theory, develops a series of hypotheses empirically tested using a sample of donors and non-donors from an Australian-based charity. The findings offer new insights into the role of emotions in non-profit marketing research, underscoring the importance of considering how donors manage or struggle to manage their emotional resource depletion and the implications for donation likelihood. It also provides valuable insights into how these effects vary not only by donor type (donor vs. non-donor) but also by an individual’s pre-existing ability to manage emotional resources, such as EI. The following sections introduce the theoretical foundation of the research, COR theory, before presenting the development of hypotheses related to donors, non-donors, donation behaviour, emotional exhaustion, compassion fatigue and EI.

Donation research draws on multiple theories to explain giving behaviour. A systematic review by Kumar and Chakrabarti (2023) identifies key approaches, including the Theory of Planned Behaviour (TPB) (White et al., 2023) and Andreoni’s Warm Glow model (Ranga et al., 2025). A well-established critique of TPB is its narrow focus on attitudinal and cognitive predictors, with limited attention to emotions (Parkinson et al., 2018), despite strong evidence that emotions play a central role in donation behaviour (Chapman et al., 2022; Kemp, 2025). Although theories such as Warm Glow acknowledge emotions, they primarily address contexts of emotional arousal and elicitation (Ranga et al., 2025). For example, Liang et al. (2016) examine how strengthening emotional responses can encourage donation. However, these approaches overlook opposite marketplace conditions in which donors or non-donors suppress emotions or experience emotional resource depletion. As a result, prior theorising has focused on varying levels of emotional arousal while neglecting how and when emotions are effectively “switched off.” This represents a critical gap, as TPB and Warm Glow do not explain how diminished emotional resources may suppress donation behaviour. The present research addresses this gap by applying COR (Hobfoll, 1989) to explain how emotional resource loss constrains donation motivation, leading to emotional exhaustion and compassion fatigue.

COR theory helps explain individual motivation to maintain existing resources or acquire new ones (Hobfoll, 1989). COR is also recognised as an important theory for understanding negative emotional responses such as burnout, exhaustion, and fatigue (Sheng et al., 2023), all of which are central to the current research and considering where emotional resource depletion may occur. According to Halbesleben et al. (2014), the basic tenets of COR are as follows: (1) resource loss is more salient than resource gain; (2) people must invest resources to gain resources, protect themselves from losing resources, or recover from resource loss; (3) individuals with more resources are better positioned to achieve resource gains, whereas those with fewer resources are more likely to experience resource losses; (4) initial resource losses tend to lead to further losses; (5) initial resource gains lead to future gains and (6) a lack of resources triggers defensive attempts to conserve the remaining resources. COR is particularly important in understanding resource losses and defensive efforts to conserve remaining resources, which aligns with the current study examining how and why intentions to donate may be lower (or higher) under particular conditions.

Literature on COR theory further suggests that resources can be both emotional and financial (Yu et al., 2023). The current study focuses specifically on emotional resources through the examination of emotional exhaustion and compassion fatigue, and financial resources through intention to donate to charity. It also incorporates the concept of loss spirals (Ford et al., 2023), where resource depletion, such as emotional exhaustion, heightens vulnerability to further losses, triggering self-protective behaviours like emotional withdrawal, exemplified in compassion fatigue. The overarching conceptual framework, guided by COR, is presented in Figure 1.

In the current research, EI is defined as an individual’s capacity to reason about emotions and to reflectively regulate emotions to generate, recognise, understand and evaluate their own and other’ emotions to guide their thinking and action to cope with environmental demands and pressures (Chaouali et al., 2021; Mayer et al., 2004). When considering this through the lens of COR, EI is considered as a pre-existing emotional resource that donors and non-donors can draw upon to guide their considerations and actions of donating to non-profit organisations. In this way, the current research examines one of the key tenants of COR, which is “lack of resources leads to defensive attempts to conserve remaining resources (Hallbesleben et al., 2014, p. 1337)”.

When investigating EI, COR and why even under high stress situations, donors may continue to donate and non-donors may consider beginning to donate, the following is considered. First, research shows that individuals with high EI can mitigate emotional resource depletion, indicated by aspects such as stress (Singh and Sharma, 2012), and are more likely to protect themselves against emotional resource depletion via coping strategies. If such thinking were to hold and generalise to the current research, higher EI should be associated with stronger intentions to (continue or begin to) donate, due to an individuals’ ability to utilise this resource as a mechanism to guard against negative and stressful situations and their associated responses, enabling them to continue or begin to help others via donating.

Studies provide evidence of EI predicting prosocial behaviour, further supporting the consideration that it may be associated with higher intentions to donate. Martin-Raugh et al. (2016) show that EI indirectly influences prosocial behaviour, including demonstrations of care and compassion, via prosocial knowledge. Thus, in line with this thinking guided by COR, and support of prior literature suggesting EI is associated with a higher likelihood of prosocial behaviours, including donating, the following is hypothesised:

To date, much of the non-profit and marketing literature has explored factors that moderate (differ in the likelihood or strength of) donation behaviours, including gender (Wymer et al., 2021), age (Lee and Park, 2025) and religious beliefs (Teah et al., 2014). Research considering how donors and non-donors’ perceptions and behaviours differ in conjunction with the psychological factors that affect their intention to donate, however, is less well known. This could be attributed in some part to much of the research often considering these segments, donor and non-donor, separately rather than together, within a single study.

Of the limited studies that show differences between donors and non-donors on factors beyond demographics, the majority are situated within blood donation. For instance, Griffith et al. (2014) show social responsibility to be a stronger driver of involvement with blood donation in donors than non-donors. Of particular relevance to the current research is that emotional factors are evidenced to differ between donors and non-donors, and in particular, non-donors being more highly sensitive and have heightened negative emotions, reducing prosocial behaviours, indicating they are less likely to regulate their emotions (O’Carroll et al., 2011).

Thus, to extend upon this limited body of evidence, the current research again draws upon COR to consider why donors may use EI more effectively when considering donations, in comparison to their non-donor counterparts. According to COR, people who have already invested resources (e.g. time, money and emotional energy) are more motivated to continue investing to protect their contributions and reduce the likelihood of loss. In this way, donors with higher EI can consider the need to continue donating even in stressful periods, as they wish to protect their already invested resources in the cause. Alternatively, high EI non-donors will consider the opposite, whereby investing time, money and emotional effort in donating the first time will be a potential loss, given the perceived sacrifices needed to make this first-time action. In this way, non-donors see donating as an extra burden during highly stressful situations, even with high levels of EI. In contrast, donors with high EI see donating as resource generating or resource protecting, given their prior donations and emotional investments with the cause. Thus, the following is hypothesised:

H3 proposes that the relationships between EI, donor status and donation intention are mediated by emotional depletion, specifically through emotional exhaustion and compassion fatigue. Prior research shows these constructs act as mediators in roles requiring high emotional resources, such as nursing (Ding and Wu, 2023).

With respect to the first mediator, emotional exhaustion, marketing research has largely focused on its experience and mitigation among salespeople and frontline employees (Sharma et al., 2025; Lewin and Sager, 2009; Edmondson et al., 2019). Far fewer studies have examined emotional exhaustion among other marketplace actors, such as consumers and, more specifically, donors and non-donors. However, prior work identifies emotional exhaustion as central to understanding coping resources (Lewin and Sager, 2009), supporting the present study’s theoretical lens and relevance.

The second mediator, compassion fatigue, should be distinguished from compassion fade. Compassion fade refers to the decline in helping behaviour as the number of people in need increases (Meier, 2025) and is primarily cognitive, reflecting perceptions of limited individual impact at scale. Compassion fatigue, by contrast, is emotionally driven and often described as the “cost of caring” (Noor et al., 2025). It arises from sustained exposure to others’ suffering rather than from the size of the need. Research has begun to examine compassion fatigue’s triggers (Patel and Weberling McKeever, 2014). For instance, Patel and Weberling McKeever (2014) found that health nonprofit websites used stewardship strategies and positive framing to mitigate compassion fatigue, though greater emphasis on long-term stakeholder relationships is needed.

Given the lack of empirical evidence on emotional exhaustion and compassion fatigue as serial mediators, we theorise their connection using COR theory, which explains resource depletion and loss spirals (Ford et al., 2023). COR theory suggests that once emotional resource depletion is triggered by emotional exhaustion, individuals enter a resource loss spiral. Specifically, as individuals lose emotional resources, further investment becomes increasingly difficult. Consequently, higher levels of emotional exhaustion may reduce individuals’ capacity to care for and support others, thereby increasing their susceptibility to compassion fatigue.

To prevent further depletion, individuals may enter a loss spiral (Halbesleben et al., 2014) and withdraw from prosocial behaviour, reducing willingness to help and intensifying compassion fatigue. COR theory suggests that resource scarcity prompts defensive strategies to conserve remaining resources (Elahi et al., 2024). Accordingly, when emotional resources are depleted, as reflected in emotional exhaustion, individuals are more likely to adopt self-protective responses such as compassion fatigue rather than empathy, which signals emotional resource availability (Pohl et al., 2015). This depletion may also extend to financial self-preservation, resulting in lower donation intentions.

Building on previous hypotheses, COR theory suggests that individuals with limited resources are more vulnerable to loss spirals, where constrained resource reserves contribute to further depletion (Halbesleben et al., 2014). We propose that loss spirals are particularly strong in two cases: (1) among individuals with lower EI, who have fewer emotional resources to draw upon and (2) among non-donors, who may perceive donating as a loss rather than an investment, unlike existing donors. These patterns align with COR theory, which highlights that individuals with low resources or prior losses are more susceptible to further depletion.

Accordingly, we expect donation intentions to be mediated by emotional exhaustion and compassion fatigue, particularly for individuals with lower EI. With fewer emotional resources, they are more likely to prioritise self-protection over altruism, reducing their intention to donate. By contrast, individuals with higher EI can draw on this resource to buffer against emotional exhaustion, suppress compassion fatigue and maintain stronger donation intentions. They may even bypass emotional exhaustion and compassion fatigue altogether, continuing to donate despite emotional strain. Furthermore, the impact of emotional exhaustion and compassion fatigue should be more pronounced for non-donors with low EI, as they may perceive donations as both an emotional and financial loss. Thus, based on the prior discussion, the following is hypothesised:

The data collection for the study was in Australia in collaboration with a regional-based charity that provides support for purchasing vital medical equipment, funding support services and hospital accommodation, among other initiatives. Data were collected using an online survey administered by Qualtrics with ethical approval from the researcher’s university (ethics approval: A221845). The use of a cross-sectional design with an online survey was deemed appropriate to test the hypotheses as a first step to obtain preliminary results with large samples from both donors and non-donors and identify significant relationship patterns. While it should be noted that causality cannot be inferred from the results obtained from this research design, this approach does provide a necessary foundation for identifying relationships that should be tested in causal designs such as experiments and longitudinal surveys.

Participants were recruited via two approaches to ensure both donors and non-donors were sampled. For donors, the charity that partnered with the researchers sent an email invitation via their donor database, inviting participation in the research. For the non-donor sample, Qualtrics was used to source participants who identified as non-donors of the charity who lived in the geographical area in which the charity operated. This was done to ensure that geographical and potential cultural variations were minimal and would not confound the results. This is particularly important given that the work of the charity partner is based upon supporting a regional area, as opposed to an urban area, and that this reflects its market segments of focus. In total, 749 completed donor surveys were recorded and 301 completed non-donor surveys with Table 1 providing an overview of their characteristics.

The data were analysed utilising Partial Least Squares Structural Equation Modelling (PLS-SEM) and the PROCESS MACRO Extension of SPSS. Both analysis techniques were utilised to leverage their advantages in line with other studies (e.g. Tryapkin et al., 2025). Specifically, PLS-SEM was able to examine the entirety of the measurement and structural model and allow for probing of the hypotheses. PROCESS MACRO was used to compliment the analysis of PLS-SEM by identifying, particularly when considering non-donors and donors, whether direct or indirect effects of EI were significant or non-significant. By using both analysis techniques, we can gain more comprehensive insights into the moderated mediation hypothesis (Tryapkin et al., 2025).

All constructs measured in this research utilised pre-existing scales within the literature. For EI, we adapted the seven items of Chaouli et al. (2021). For emotional exhaustion, seven items from the Weisberg and Sagie (1999) scale were utilised. Compassion fatigue was measured using the five-item scale of Eng et al. (2021). Intentions to donate was measured using four items of Eng et al.’s (2021) scale.

The study took several steps to mitigate endogeneity and common method bias, strengthening the rigor of the analysis and findings. Endogeneity was addressed in two ways. First, the analysis controlled for key demographic variables, gender, age, income and education, which prior research shows influence donation and related prosocial behaviours. Second, a competing model reversing the roles of emotional exhaustion and compassion fatigue was tested to assess the robustness of the hypothesised mediation, demonstrating that the proposed specification was the only viable model. Common method variance (CMV) was further assessed using variance inflation factor (VIF) tests, appropriate for PLS-SEM. All VIF values were below 5, indicating that CMV was not a significant concern.

Prior to assessing the structural model and hypothesis tests, the measurement model was assessed for reliability and validity. As evidenced in Table 2, all constructs had satisfactory levels of convergent validity with factor loadings varying between 0.710 and 0.928. Reliability was also established with all constructs having Cronbach Alpha and Composite Reliability scores above 0.856. AVE scores were also above the recommended threshold of 0.500.

A critical aspect of cross-sectional survey research is also establishing discriminant validity to ensure that the constructs measured are not conflated and are indeed distinct. As per the recommendations in the literature, we sought to establish discriminant validity by using both the HTMT and Fornell and Larcker (1981) tests. As evidenced in Tables 3 and 4, discriminant validity was confirmed via both tests, further supporting the rigour of the data and measurement of the constructs. After establishing reliability and validity, the structural model and hypothesis tests were examined.

H1. To assess H1, the direct effect of EI on intentions to donate was examined. The results showed a significant positive association between EI and intention to donate (β = 0.505, SD = 0.026, t = 19.440, p < 0.001), supporting H1 that higher levels of EI are associated with higher intentions to donate.

H2. To assess H2, the effect of EI and donor status on donation intentions was examined. The results were significant (β = −0.272, SD = 0.071, t = −3.836, p < 0.001). A simple slope analysis revealed that donor status significantly moderates the relationship between EI and donation intentions. This was further probed by undertaking a moderated regression analysis using PROCESS Macro extension of SPSS, specifically the Model 1 template with 5,000 bootstraps. It is important to note that PROCESS provides unstandardised betas (b), which should be considered when interpreting the results presented next.

The interaction effect tested in PROCESS was also significant (b = −0.531, SE = 0.12, t = −5.111, p < 0.001) and showed additional insights to the PLS-SEM analysis that the conditional effect of EI was significant and stronger for donors (b = 1.081, SE = 0.065, t = 16.455, p < 0.001) and dropped in strength whilst still being significant for non-donors (b = 0.550, SE = 0.104, t = 5.276, p < 0.001). These results demonstrate that higher EI is associated with higher intentions to donate, but this is likely to be significantly higher in donors than non-donors (see Figure 2).

Specifically, donors with high EI exhibited greater intentions to donate. However, at lower levels of EI, both donors and non-donors had comparable donation intentions. This highlights the importance of high EI in driving donation intentions while indicating that, when EI is low, donors and non-donors demonstrate similar donation behaviours.

H3. In H3, we examined the indirect effect of EI on intentions to donate when mediated by emotional exhaustion and compassion fatigue and moderated by donor status. The indirect effect was non-significant (β = 0.002, SD = 0.003, t = 0.727, p = 0.467) when not considering donor status as a moderator. To probe why, we examined the direct effect between EI, which was non-significant (β = −0.028, SE = 0.036, t = 0.765, p = 0.445).

In H3, we further considered that the mediation of emotional exhaustion and compassion fatigue for EI on intentions to donate may be conditional, specifically, being stronger for non-donors as opposed to donors. Prior to examining this effect, we looked to examine the moderated relationship of donor status and EI on emotional exhaustion, given that the direct effect as reported in H3 was non-significant for EI on emotional exhaustion. The interaction effect was significant (β = −0.154, SE = 0.074, t = 2.069, p = 0.039), which suggests that the effect of EI on emotional exhaustion is conditional on donor status. A simple slope analysis shows that donor’s emotional exhaustion does not vary based upon EI. Whereas, when observing non-donors, the simple slope analysis shows that emotional exhaustion is higher for those with lower levels of EI, and that this reduces as EI increases to the point of being lower than donors with high EI. Thus, when considering prior results, how EI is used in relation to emotional exhaustion is conditional, specifically, only significant for non-donors, not donors (see Figure 3).

To then test and probe the entirety of H3 we then examined the indirect effect of the interaction between EI and donor status in the PROCESS Macro extension of SPSS using the Model 83 template with 10,000 bootstraps. The index of moderated mediation was significant (index = 0.0253, SE = 0.013, LCI = 0.003, UCI = 0.0550). Specifically, the indirect effect for non-donors was significant (b = 0.024, SE = 0.001, LCI = 0.004, UCI = 0.053), suggesting that for higher levels of EI in this group, intentions to donate can be explained via the mediators of emotional exhaustion and compassion fatigue (these results are visualised in Figure 4). Whereas, for donors, the indirect effect was non-significant (b = −0.001, SE = 0.005, LCI = −0.011, UCI = 0.011). This provides evidence to support H3. A summary of the hypotheses' results is presented in Table 5 and Figure 4.

When considered together, these findings indicate that EI does not protect donors from emotional exhaustion or its subsequent impact on compassion fatigue and future donation intentions. In contrast, for non-donors, EI serves as a buffer against emotional exhaustion and its downstream effects. Consequently, non-donors with low EI who experience emotional exhaustion are unlikely to consider donating, whereas those with high EI can navigate emotional exhaustion and compassion fatigue, potentially maintaining their intention to donate.

In relation to the controls used, income (β = 0.249, SD = 0.032, t = 7.688, p < 0.001) and gender (1 = female, 2 = male; β = −0.118, SD = 0.032, t = −3.641, p < 0.001) had a significant effect on EI. Education had a significant influence on intentions to donate (β = 0.082, SD = 0.027, t = 3.024, p < 0.001). Income had a significant influence on compassion fatigue (β = −0.110, SD = 0.033, t = −3.00, p = 0.001). All other effects of the control variables were non-significant ≥ p.146.

To test the robustness of our model, we ran a rival specification reversing the mediation sequence, with compassion fatigue preceding emotional exhaustion. Donor status remained as the moderator, EI as the independent variable, and intentions to donate as the dependent variable. Results showed non-significant effects for both the moderation of donor status (β = −0.037, SD = 0.077, t = 0.479, p = 0.632) and the link between emotional exhaustion and intentions to donate (β = −0.058, SD = 0.039, t = 1.507, p = 0.132). These findings mitigate concerns about a competing reverse causality by showing that an alternative moderated mediation chain is not supported.

The current research has three aims that it seeks to empirically explore. Regarding the first aim, the findings indicate that higher EI is associated with stronger intentions to donate. However, considering the potential differences between donors and non-donors, this association is most pronounced among donors compared to non-donors. In relation to the second and third aims, the research provides the following evidence. Emotional exhaustion and compassion fatigue mediate the relationship between EI and donation intentions, but only for non-donors, not donors. Specifically, EI plays a key role in explaining variations in emotional exhaustion among non-donors, whereas no notable differences in emotional exhaustion were observed among donors based on their level of EI. The implications of these insights for the non-profit literature are discussed next.

This research contributes to the literature by applying COR theory (Hobfoll, 1989), a framework rarely used in non-profit studies, to explain variations in charitable giving. Unlike theories such as the Theory of Planned Behaviour (White et al., 2023) or the Warm Glow Model (Ranga et al., 2025), which emphasise attitudes, cognitions, or emotional arousal, this study considers conditions where emotions may be “switched off”. It examines how the presence or depletion of emotional resources shapes donation decisions, highlighting factors that hinder giving (emotional exhaustion, compassion fatigue) and those that encourage it (EI). Empirically, it demonstrates how emotional resource loss can trigger “loss spirals” and self-protective behaviours (Ford et al., 2023), reducing the likelihood of financial giving, particularly among non-donors. The key theoretical contribution is advocating for a broader application of COR theory, especially through an emotional lens, to capture overlooked dynamics of donation behaviour. By extending beyond arousal-based explanations to include conditions where emotions are muted, this study provides a more comprehensive understanding of how both the presence and absence of emotional engagement shape charitable giving.

This research also provides new insights into how non-profits can use and manage emotions to elicit donations. It identifies the mediating roles of emotional exhaustion and compassion fatigue, two concepts underexplored in non-profit research. Unlike prior studies that focus on emotional appeals (Sandoval and García-Madariaga, 2024) or fostering empathy (Chan and Septianto, 2022), this study investigates emotional exhaustion and compassion fatigue, states characterised by low emotional arousal and a diminished capacity to understand others’ emotions. While these concepts have been explored in fields like nursing (Ding and Wu, 2023), they are under-researched in the context of donation behaviours and marketing. This study evidenced that these aspects are central and explain emotional disengagement in non-donors as well as importantly, a boundary condition of when it doesn’t occur, which is in the case of donors. This study empirically demonstrates how these overlooked concepts explain emotional disengagement or dissociation due to resource depletion, ultimately affecting donation behaviour. Unlike lower emotional arousal or reduced empathy, emotional exhaustion and compassion fatigue indicate a more severe state where individuals “switch off” or actively avoid using emotions. In this state, emotional resources are depleted, prompting self-protective behaviours (Hobfoll, 1989; Ford et al., 2023) to shield against further distress in non-donors. Interestingly, our moderation findings based on donor status do not support a mediation pathway in which emotional resource depletion, via emotional exhaustion and compassion fatigue, explains how donors rely on EI to sustain giving. This suggests that, for donors, an alternative mediation mechanism may be at play. One possibility is that the protection of financial resources plays a more central role in shaping their giving behaviour, although this requires further theoretical development and empirical testing.

A third contribution of this research is in how it examines and explains emotional resource use and depletion, and its effect on donation intentions, extending previous studies that often-explored moderating factors like gender and age (Wymer et al., 2021; Lee and Park, 2025). By considering how emotional resource dynamics influence donation intentions across individuals with different donation histories, this study moves beyond moderation within a single group. The research reveals the heterogeneity of factors driving first-time versus repeat donations, highlighting how emotions and their management shape donation behaviour. Examining both donors and non-donors together provides a deeper understanding of the interaction between EI, emotional exhaustion and donation intentions. Recognising these differences offers valuable insights into how the principles of COR, particularly the emphasis on losses and gains and the utilisation of emotional resources, manifest differently based on donor status.

A key practical implication of this research is that non-profits should carefully consider the use of emotional appeals, as they may exacerbate emotional exhaustion. The findings suggest that non-profits, particularly when targeting new donors (non-donors), should be mindful of negative emotional appeals or other strategies that might inadvertently heighten emotional exhaustion and negatively impact donations. For non-donors, it is important to avoid negative emotional appeals, such as guilt and sadness linked to depictions of suffering. For example, campaigns that show distressing images of malnourished children with messages like “Every minute a child dies from hunger unless you help” may inadvertently deter non-donors, as the evidence in this research suggests such appeals could be at risk of triggering emotional exhaustion and compassion fatigue. By contrast, positive emotional appeals that emphasise hope and impact, such as “Your donation helps a child thrive in school with a full stomach and brighter future”, could motivate support without overwhelming potential donors.

Another important consideration for non-profits is the need to remain attuned to events that evoke widespread stress, such as natural disasters and economic downturns. These crises can increase emotional exhaustion and compassion fatigue, potentially affecting donor engagement and public support. To navigate these challenges, non-profits should proactively consider how to acknowledge and address these emotional burdens in their outreach and fundraising. This involves implementing systematic approaches to assess fluctuations in emotional exhaustion and compassion fatigue among their target audiences. Non-profits can use surveys, media monitoring, and social media analytics to gauge emotional exhaustion levels over time. By leveraging these insights, they can adjust their messaging, avoid overwhelming donors and non-donors, and adopt a context-sensitive approach to eliciting donations.

This research, like all studies, has limitations that present opportunities for future exploration. One key limitation is its focus on donors and non-donors within a single geographic area, specifically, in relation to an Australian-based charity that partnered in this study. Research suggests that emotional arousal and expressions can vary across cultures and impact cause-related marketing efforts differently, which should be considered when interpreting these findings in different cultural contexts, particularly those that differ significantly from Australia. Future research could also investigate whether the observed relationships vary across countries, particularly by comparing individualistic and collectivist cultures to assess whether the strengths of the pathways in the current model differ. Such cross-country research would be especially valuable for large international charities that operate and market across diverse cultural contexts, providing evidence on whether cultural sensitivities related to factors such as emotional exhaustion and compassion fatigue should be considered in their strategies. In doing so, this line of inquiry could not only enhance theoretical understanding of cultural influences on donor behaviour but also offer practical guidance for tailoring communication and engagement approaches to different cultural settings.

A key strength of this research is its demonstration of the link between EI, emotional exhaustion and compassion fatigue using a large donor and non-donor sample. However, causal inferences are limited, highlighting opportunities to explore additional factors shaping these psychological responses and their impact on donation. An interesting extension would be to examine how high-arousal negative emotional appeals, such as guilt, disgust or shame, contribute to emotional exhaustion or compassion fatigue. Future studies could use experimental designs with charity marketing stimuli to measure these effects and assess their role in donation behaviour. Integrating emotional exhaustion and compassion fatigue into research on emotional appeals could offer valuable insights into whether non-profits should exercise caution in certain strategies.

Portions of this manuscript were edited for language, clarity and readability using ChatGPT (OpenAI). The tool was used solely to enhance communication and presentation and did not contribute to the study design, data analysis, interpretation of results or generation of ideas. The authors take full responsibility for the content of the manuscript.

The authors wish to acknowledge the funding and assistance in data collection provided for this project by Wishlist, a hospital charity, supporting Nambour, Caloundra, Maleny, Gympie and Sunshine Coast University hospitals in Australia.