Most Efficient Message Framing to Promote Charitable Giving [Literature Review Report]

By Zoé Roy-Stang @ 2023-05-14T14:26 (+14)

The purpose of this literature review report is to provide insights on how to increase charitable giving in the general population. The central question that guided this research process was “What is the most efficient messaging for the promotion of charitable giving?” The evidence points to different ways of presenting and framing messages based on evidence of outcomes such as subjective intentions to donate, decision to donate, and donation amount. The last section provides 14 recommendations based on the reviewed research.

Methods

I conducted this literature search using various search engines, databases, and tools. The first tool I used was Elicit, an AI assistant that answers research questions based on an overview of key articles. I pasted the central research question from the previous paragraph into Elicit, and its output included a one-paragraph answer as well as summaries of findings from a few articles. I reviewed the seven cited articles it provided and included the relevant ones in this report. Then, I referred to the Carleton University Cognitive Science Databases and selected three databases to search through: PsycINFO, Google Scholar, and SpringerLink. Within the PsycINFO search, keywords included “framing effect on charitable giving”. There were 10 results, including one meta-analysis. I reviewed them all and rejected two that were unrelated to the research question and one that was a book (thus not within the scope of allocated resources for this project). In Google Scholar, the main search was “meta-analysis effect of framing on charitable giving”, with additional keywords including “meta-reviews”, “messaging”, “effect of messaging”, and “donations”. A few relevant meta-analyses and experimental studies were selected from the results. On SpringerLink, keywords included “message framing and charitable donations”, and the search yielded a few papers from various fields. I tried many keywords and combinations in a government database—Canada's open government portal—but could not find relevant sources of information. In total, I was able to find 20 relevant sources including one meta-metaanalysis, four meta-analyses, and many experimental studies. These articles come from a variety of domains including cognitive science, psychology, experimental economics, business, communication, and marketing science.

Review of Findings

Results from each of the meta-analyses provided different insights, but their findings were consistent with each other as they covered different aspects of the research question. The most important source was a meta-metaanalysis of 1339 studies by Saeri et al. (2022). They found four factors that increase charitable donations. The first was framing the donation as benefitting an individual; the second was to increase exposure to benevolent behaviours (including giving); the third was clarifying the donation’s impact; and the fourth was to make donations tax-deductible or inform people about tax-deductible donations. 

The next two meta-analyses covered specific factors related to charitable giving. The first, a meta-analysis of six studies (each containing multiple experiments), found that the way a donation request was framed influenced charitable giving amount based on whether the message was combined with positive or negative images (Genevsky & Knutson, 2018). Message framing of positive or negative valence had the most impact on charitable giving when combined with an image of matching valence. These six studies included hypothetical and real incentives as well as online and in-lab settings.  The second—a meta-analysis of 41 studies on the compassion fade effect—found that being exposed to larger groups of victims in need reduced helping behaviours and willingness to help. The degree of certainty and the threat level associated with adverse events modulated the effect of victim group size on charitable giving behaviours and intentions.

Finally, the remaining two meta-analyses pointed towards null findings for the factors they examined. The first was a meta-analysis of 27 studies (2020), and its results suggested that there was no significant difference in charitable giving when advertised messages were framed in terms of gains. The second was an experimental economics meta-study by Fromell et al. (2020). They examined the pattern of results across 17 experimental studies that investigated whether altering task demands to manipulate cognitive resources could modulate the switch between the slower deliberative system and the faster intuitive system (based on the Dual Systems theory of altruistic behaviour). The results were null, suggesting that changing task demands to increase activation of the intuitive system does not affect charitable giving; the authors discuss the implication that perhaps there is little interference between the intuitive and deliberative systems if they both lead to the same decision in most monetary giving trade-off scenarios.

The remaining studies can be grouped into two categories based on the outcomes they measure. The first set includes three experimental studies investigating the intention to donate whereas the second set includes 11 studies investigating donation decisions and amounts. Factors associated with each target outcome are identified and discussed. 

A few factors have been linked to donation intentions. For instance, Chang & Chen (2019) found that participants’ donation intentions increased when the charitable giving involved donating as a gift compared to a charity sale where the donation is also a purchase. A second study by Cao (2016) investigated the effect of framing charitable donation requests in terms of losses or gains associated with charitable giving or inaction. The results showed that when participants believed they were more likely to be affected as a result of their inaction, the loss-framed messages led to higher charitable donation intentions. Finally, a third study by Das & al. (2008) found that when the donation message was set in the context of attaining a charity goal, donation intentions were higher. They also found that the most effective charitable appeal strategies involved combinations of a positive message frame with anecdotal evidence and a negative frame with statistical evidence, replicating the findings of previous studies.

Eleven additional experimental studies measured giving behaviours—the most relevant outcome to the central question of this report. In a field experiment by Adena & Huck (2022), framing charitable giving as a donation rather than a reward-linked contribution was associated with more giving. A follow-up survey experiment suggested that gifts were perceived as more voluntary and the idea of giving had a positive correlation with emotions. A different field experiment by Sudhir et al. (2016) found that in an Indian population, charitable giving requests were most successful when they included appeals to sympathy biases. Examples of appeals containing sympathy biases included portraying an identifiable victim or depicting members of an in-group.  Please note that the identifiable victim effect is contested due to failure to replicate (Maier et al., 2023) [2024 edit]. Garinther & Arrow (2022) found inconclusive evidence that framing a donation appeal as benefitting an individual rather than a group was associated with higher donation amounts. Glazer & Konrad (1996) highlighted that people were still willing to donate even when the donation had no prosocial impact. This finding supported the signaling theory, which postulated that people may donate to signal wealth. Kulow (2016) found that people who reported having strong karmic beliefs donated more to charity when the messaging framed the donation in terms of gains for others as opposed to self-gain (compared to participants who reported weak karmic beliefs). Jeong et al. (2011) found that gain-framed messages worked best on people with a disposition to approach and loss-framed messages worked best to persuade people with a disposition for avoidance. This effect was observed both when the gains or losses were relevant to the self or others. Moche et al. (2020) found that people were less willing to donate when they were reminded of the opportunity cost (alternative uses of money). They found no effect of framing the opportunity cost as prosocial or proself. Smith & Berger (1996) found that anchors (i.e., pre-set suggestions of monetary donation size) influenced the amount people gave but did not influence their likelihood of donating. Sussman et al. (2015) found that when messages framed a donation as an exceptional (uncommon, infrequent) expense, participants were more willing to donate than when the messages were framed as ordinary (common, frequent). Participants did not consider the effect of the donation on their budget as much when it was framed as exceptional. White & Peloza (2009) found that donation requests framed in terms of benefits to others were associated with more funds raised when donations were public, whereas framing donations is terms of self-benefits was more efficient when donations were private. Diederich & al. (2022) found null results when comparing donation matching subsidies, rebate subsidies, and unit donation (donating to buy a good at a given price per item). However, they observed that rebates were associated with a higher proportion of donors whereas matching and discounts raised more funds per donor. 

Recommendations

Based on the current research reviewed in this report, the following 14 recommendations were made about the best ways to construct charitable giving messages intended for the general population: 

(1) Increase exposure to benevolence (e.g., by showcasing people’s giving behaviours). 

(2) Highlight an individual recipient of aid. 

(3) Clearly state the impact of donations. For example, “$50 can provide sight-restoring cataract surgery for an adult” (Seva Canada, 2023) follows the clear format ‘X amount is sufficient to purchase Y, which will have Z impact’. 

(4) Inform the public of the tax deductibility of donations. Providing clear and simple guidelines about the tax deduction process may be helpful. 

(5) Frame donations as gifts rather than purchases or reward-linked contributions. 

(6) Include images matching the valence of corresponding messages (i.e., use an image and message that are both associated with either negative or positive emotions). 

(7) Use statistics with negatively framed messages and anecdotes with positively framed messages. 

(8) When mentioning negative outcomes or losses associated with the choice not to donate, focus on the negative outcomes that can affect the donor themselves. 

(9) Showcase recipients of aid from the same group as a donor or highlight features from the recipients of aid that donors can identify with. 

(10) Use anchors (suggested donation amounts) to influence people to donate more. 

(11) Frame donations as exceptional expenses (e.g., ‘Your one-time donation will make a difference!’). 

(12) Highlight benefits to others when donations are public and benefits to the donor when donations are private. 

(13) Target sub-groups with dispositions to approach or avoidance; highlight gains (positive impact of donation) for one and losses (negative impact of not donating) for the other. 

(14) Avoid mentions of the opportunity cost of donations.

References

Adena, M., & Huck, S. (2022). Voluntary ‘donations’ versus reward-oriented ‘contributions’: two experiments on framing in funding mechanisms. Experimental Economics, 25(5), 1399-1417. https://doi.org/10.1007/s10683-022-09759-6 

Butts, M. M., Lunt, D. C., Freling, T. L., & Gabriel, A. S. (2019). Helping one or helping many? A theoretical integration and meta-analytic review of the compassion fade literature. Organizational Behavior and Human Decision Processes, 151, 16-33. https://doi.org/10.1016/j.obhdp.2018.12.006 

Cao, X. (2016). Framing charitable appeals: The effect of message framing and perceived susceptibility to the negative consequences of inaction on donation intention. International Journal of Nonprofit and Voluntary Sector Marketing, 21(1), 3-12. https://doi.org/10.1002/nvsm.1536

Chang, C. C., & Chen, P. Y. (2019). Which maximizes donations: Charitable giving as an incentive or incentives for charitable giving?. Journal of Business Research, 97, 65-75. https://doi.org/10.1016/J.JBUSRES.2018.12.046

Das, E., Kerkhof, P., & Kuiper, J. (2008). Improving the effectiveness of fundraising messages: The impact of charity goal attainment, message framing, and evidence on persuasion. Journal of Applied Communication Research, 36(2), 161-175. https://doi.org/10.1080/00909880801922854

Diederich, J., Eckel, C. C., Epperson, R., Goeschl, T., & Grossman, P. J. (2022). Subsidizing unit donations: Matches, rebates, and discounts compared. Experimental Economics, 25(2), 734-758. https://doi.org/10.1007/s10683-021-09732-9 

Fromell, H., Nosenzo, D., & Owens, T. (2020). Altruism, fast and slow? Evidence from a meta-analysis and a new experiment. Experimental Economics, 23(4), 979-1001. https:// doi.org/10.1007/s10683-020-09645-z

Garinther, A., & Arrow, H. (2022). Linguistic framing effects in business and refugee aid contexts: A replication and extension of Cooley et al. (2017). Journal of Experimental Psychology: General, 151(5), e1-e18. https://doi.org/10.1037/xge0000627 

Genevsky, A., Knutson, B., & Yoon, C. (2018). Request framing moderates the influence of affective images on charitable giving.

Glazer, A., & Konrad, K. A. (1996). A signaling explanation for charity. The American Economic Review, 86(4), 1019-1028. https://www.jstor.org/stable/2118317

Jeong, E. S., Shi, Y., Baazova, A., Chiu, C., Nahai, A., Moons, W. G., & Taylor, S. E. (2011). The relation of approach/avoidance motivation and message framing to the effectiveness of charitable appeals. Social Influence, 6(1), 15-21. https://doi.org/10.1080/15298868.2010.524369

Kulow, K.(2015). The moderating role of karma in the relationship between other-gain vs. self-gain appeal frames and Charitable Giving. (Doctoral dissertation). Retrieved from https://scholarcommons.sc.edu/etd/3163 

Maier, M., Wong, Y. C., & Feldman, G. (2023). Revisiting and rethinking the identifiable victim effect: Replication and extension of small, Loewenstein, and Slovic (2007). Collabra: Psychology, 9(1).

Moche, H., Erlandsson, A., Andersson, D., & Västfjäll, D. (2020). Opportunity cost in monetary donation decisions to non-identified and identified victims. Frontiers in Psychology, 10, 11. https://doi.org /10.3389/fpsyg.2019.03035

Ought; Elicit: The AI Research Assistant; https://elicit.org; accessed March 15, 2023

Restore sight and prevent blindness in low- and middle-income countries. Seva Canada. (n.d.) Retrieved March 30, 2023, from https://donate.seva.ca/ 

Saeri, A. K., Slattery, P., Lee, J., Houlden, T., Farr, N., Gelber, R. L., … & Zorker, M. (2022). What works to increase charitable donations? A meta-review with meta-meta-analysis. VOLUNTAS: International Journal of Voluntary and Nonprofit Organizations, 1-17. https://doi.org/10.1007/s11266-022-00499-y 

Smith, G. E., & Berger, P. D. (1996). The impact of direct marketing appeals on charitable marketing effectiveness. Journal of the Academy of Marketing Science, 24(3), 219-231.

Sudhir, K., Roy, S., & Cherian, M. (2016). Do sympathy biases induce charitable giving? The effects of advertising content. Marketing Science, 35(6), 849-869. https://doi.org/10.1287/mksc.2016.0989 

Sussman, A. B., Sharma, E., & Alter, A. L. (2015). Framing charitable donations as exceptional expenses increases giving. Journal of Experimental Psychology: Applied, 21(2), 130-139. https://doi.org/10.1037/xap0000047

White, K., & Peloza, J. (2009). Self-benefit versus other-benefit marketing appeals: Their effectiveness in generating charitable support. Journal of Marketing, 73(4), 109-124. https://doi.org/10.1509/jmkg.73.4.109

Xu, J., & Huang, G. (2020). The relative effectiveness of gain‐framed and loss‐framed messages in charity advertising: Meta‐analytic evidence and implications. International Journal of Nonprofit and Voluntary Sector Marketing, 25(4), e1675. https://doi.org/10.1002/nvsm.1675