Celebrity Endorsements and Donations: Empirical Investigation of Impact on Philanthropic Giving

Abstract

This study investigates the effects of celebrity endorsers on donations and views of nonprofits. Celebrity endorsers and their relationships with nonprofits evoke a series of emotions that greatly affect how people perceive and donate to nonprofit organizations. Past research demonstrates the tremendous growth in individual philanthropy and the importance of celebrities as influencers on social media. Data was obtained from a random sample of 277 individuals and subjected to statistical analyses. By studying two different celebrities that come from a different race, gender, and background, we found that our findings and significance values changed depending on which celebrity was tested. Our findings indicated that ads that evoke positive emotions are more likely to convert to donations. Demographic characteristics, such as age, gender, and income also affect donations. In some cases, the nonprofit’s cause and advertising message were powerful enough to override the impact of the celebrity. Our findings can help marketers style their ads to evoke certain emotions, by using celebrity endorsers that will help them increase donations. Furthermore, understanding what drives donors will help nonprofits tailor marketing strategies. It is expected that by targeting their donor base effectively, non-profits will then witness an increase in donations.

Keywords

Nonprofits philanthropy marketing strategy celebrity endorsements celebrity scandals

Introduction

There is something remarkable to be said about the overall impact celebrities have on the increase of donations for nonprofits. In fact, there has been significant growth in philanthropic giving in recent years. Indiana University’s The Philanthropy Outlook 2017 & 2018 reports a steady rise in philanthropic giving by individuals that is anticipated to follow an upward trajectory in foreseeable future. Empirically, this growth is reported as an estimated 18.7 percent growth between 2011 and 2015. Some researchers claim the reason for this is because of a younger, more socially aware audience (Bivin, 2017). Taking a deeper look into the root cause, research was carried out by studying the advertising message that is put out by marketers. When studying nonprofit advertisements, it is easy to note the face of the known celebrity smiling, or looking serious front and center on the page. Paired with the celebrity is the person, place, or thing he/she is supporting. A notable feature of nonprofit marketing is celebrity endorsement of their appeals for donations and celebrities acting as a “brand ambassador,” for the charitable cause. Organizations that have successfully used celebrity endorsers include well-recognized charities, such as Make-A-Wish and Silver Lining Foundation in service of children that need medical care. For the purpose of this research, we used two nonprofits that had two different and distinct agendas. First was St. Jude’s Children’s Hospital to measure the impact cancer research had on donors. By using St. Jude’s as one of the measurements, we hoped to evaluate emotions related to sadness, guilt, and empathy. The other nonprofit measured was the Tiger Woods Foundation, whose mission is to give students the tools they need to thrive in their education and beyond. Testing our hypotheses by using the Tiger Woods Foundation allowed us to measure the impact of celebrity endorsements of nonprofits on a less emotionally heavy topic.

Past research has established that celebrity endorses can influence how potential donors view and make contributions to a nonprofit organization (e.g. Harris, 2014). This research focuses on nonprofits that utilize celebrity endorsers in their marketing messages. We investigate celebrity endorsements effects on the views of nonprofits and likelihood of donating. Additionally, we examine how celebrity endorsers’ past scandals, the donor demographics, and emotions that were evoked by advertisements that included celebrity endorsers affected donations. We tested our hypotheses using random sample of 277 respondents.

In the following section, we discuss past research in the area, use these to develop our hypotheses. Subsequent sections deal with our findings and a discussion of our findings. We conclude by reiterating some of the managerial implications of our findings.

Literature Review

Charitable Giving

Over the past 40 years, charitable giving has substantially increased in the USA (The National Philanthropic Trust, 2017). The way in which people give is important for nonprofits to understand. During this time of philanthropic growth, it is crucial that nonprofits communicate effectively to their finite pool of donors. Between 2011 and 2015, there was a growth rate of 18.7 percent. American donors alone contributed $373.3 billion to 1.2 million US charities in 2015. Total giving in 2017 and 2018 is expected to increase by 3.6 percent and 3.8 percent, respectively (Bivin, Osili, McKitrick, Bergdoll & Claire, 2017).

Philanthropy

The profile of a nonprofit’s donor base can define who gives and how much they decide to give. It can also establish the difference between corporate donors and individuals, as well as high-income and low-income donors. The National Philanthropic Trust reports between 2014 and 2015 and philanthropic giving from the individual went up by 4.1 percent. It was also reported that 98.4 percent of high net worth households were given to charity (The National Philanthropic Trust, 2016). According to the National Philanthropic Trust, 63 percent of high net worth donors cite “giving back to the community as a chief motivation for giving” (The National Philanthropic Trust, 2016). For millennials, donations are viewed as an investment. They look for returns on their donations, and therefore report featuring on what is done with their money is a key for return donations (Bivin et al., 2017).

Celebrity Endorsers and Millennials

Studies show that traditionally celebrities bring attention to a product or service, especially on social media. Celebrities are often used as a way to break through the clutter, even though celebrity endorsements do not convert to purchasing. However, they can be strong assets when introducing a new or an unfamiliar product to the market (Knoll & Matthes, 2017). In this new online age, celebrities can lend familiarity and credibility to new products and services on the Internet (McCormick, 2016). The millennial demographic not only has a strong online presence but also often use social media platform to express their views and beliefs. Millennials are often affected by the way celebrities portray their beliefs and views, thus they are more likely to follow what their favorite celebrities support (McCormick, 2016).

Nonprofit Advertising

When creating a marketing strategy that focuses largely on persuading through effect (emotions), it is important for firms to know the foundational premise and framework behind such a message appeal. Emotions in promotional messages play an even more important role when trying to relate a not-for-profit to the audiences. Nonprofit advertising is further complicated in its unique feature, marketers are trying to attract potential donors to donate money to causes that may not serve the donors themselves. While the majority of consumers can be swayed based on the good of others, some might need more coaxing.

Theoretical Framework

Based on past research, key factors that formed the theoretical foundations of our hypotheses included emotional appeal used in an ad and their relationship with demographic characteristics of the audience. The actual persuasive power of the appeal to generate donations is important, whether the emotions in the message focus only on feelings of sadness and guilt or emotions of joy and happiness. Last, we looked at the specific celebrity used in a nonprofit advertisement. We wanted to see whether their background and any scandals in their lives could cause donations to increase or decrease. Perceptions of the celebrity endorser and consequently the perceived impact of scandals attached to such an endorser are important as well. We examined this through a hypothesis that looked at the scandals of Tiger Woods and Jennifer Aniston, both powerful figures in modern society. We tested negative emotions, positive emotions, and finally, all emotions evoked by a message. It was believed that audiences would not donate as much money when an ad used positive emotions to sway people.

Hypothesis 1: Ads that evoke positive emotions, such as hope, thankfulness, persuasion, and reflectiveness, will be less effective than negative emotions of guilt and shame in generating donations.

Since nonprofits depend on donations from everyday consumers, it is natural to see if the level of income affects how much they donate. We predicted that more income is directly related to how much they donate.

Hypothesis 2: The higher the level of income, the more likely a person is to donate.

When a marketer chooses a celebrity to represent the face of a campaign, they must take into account the celebrities past scandals and reputation. The way people react to the celebrity will decide if they donate or not. This is why we believe that the older the respondent the more awareness of celebrity and what they have done in their lives.

Hypothesis 3: The higher the age of a person the more likely they are to be aware of a celebrity’s past scandals, which results in the ad appeal to not drive the donor likelihood to donate.

Figure 1 provides a conceptual map of our framework and summarizes various hypotheses that we tested and hypotheses related to other demographic variables.

Figure 1.

Source: Created by authors.

Methodology

Methodology Selection

Through the use of a descriptive research design, a quantitative study was created and used to collect primary data. A descriptive research design was used to collect primary data for a quantitative study on the effects that celebrities have on nonprofit marketing. This design was selected because the questions that were being asked required respondents to answer using a seven-point Likert-type scale. The online survey method was selected as the vehicle for collecting the data because of the ability of it to be sent to diverse respondents and ease of responding that it offers them. It also allowed for easy and efficient data collection and analyses with the least amount of human errors.

We used past studies from the Handbook of Marketing Scales: Multi-Item Measures for Marketing and Consumer Research to develop our construct measures. An online questionnaire with six clear categories, demographics, psychographics, emotions associated with the advertisement, likelihood to donate, respondent perception of celebrity, and perception of the nonprofit, were used to collect the data. These categories assisted us in evaluating two different advertisements with each having a different celebrity and nonprofit. We measured the dependent variables that affect donations to nonprofits with celebrity endorsers that included whether or not an ad with a celebrity endorser evoked positive emotions (hope, thankfulness, persuasion, and reflectiveness), the level of income of potential donors, and the ages of potential donors.

The online survey method was used to implement the study. This method gave us several advantages. First, it assured that respondents’ answers would be recorded anonymously, this was crucial when asking respondents about sensitive topics such as donation habits, income, and race. The second advantage was that respondents were required to answer every question to complete the survey. The original distribution was done through random sampling, but then was continued through snowball sampling, the survey was sent to family and friends and passed along after that.

Consequently, 277 fully completed surveys were used for further analyses. We used structured, matrix type questions, some included images of ads that required responses. The scale measurement used was an interval seven-point Likert-type scale in order to measure precise differences between scale points. The two extreme anchors ranged from 1 = “Strongly Disagree” to 7 = “Strongly Agree”. Every question in our survey was a structured question, the respondent had to choose from the given predetermined responses. The survey was kept under 6 minutes to avoid response fatigue and to ensure higher response rate, since we provided no incentives to the respondent.

Online Survey Method

The online survey method allowed for responses to be anonymous and required participants to answer every question. This method also allowed us to evaluate how psychographics, demographics, and emotional reactions affect other variables such as how people donate and their perception of what they are donating to. The online survey method allowed for the ads that were being evaluated to be presented with each scale with relative ease. There was no scrolling or constant flipping done by the respondent. This allowed respondents to easily evaluate the ads with how having to put in a lot of effort to respond the questions.

Scale

The scale that was used is seven-point Likert-type scale. The scale was developed with seven points to allow the respondents to select varying degrees of the two extremes given “Strongly Disagree” and “Strongly Agree”. There was also a neutral choice allowing us to see what sorts of variables did not affect our respondents. If this scale had an even amount of points, then the respondents would be unable to select neutral. The Likert-type style of the scale allowed for respondents to answer questions based on agreeing or disagreeing. This style aligned with the way we wanted our data presented.

Our Sample consisted of 277 complete and usable surveys. The majority of our respondents were female (68.2%), while 28.83 percent were male, the remaining 3.33 percent indicated transgender and other categories. For meaningful data analysis, we measured the differences between the top two gender groups since the remainder categories had very small cell sizes. Above half (56.76%) of respondents were between 18 and 24 years old, placing them in the millennial generation. As anticipated due to student researchers being involved, the majority of respondents were students or students who were employed part-time totaling at 49.55 percent combined. An overwhelming number of respondents identified their race as white (81.08%). The largest income level of our respondents was greater than $100,000 (29.68%).

Table 1 shows the original scales and the corresponding Cronbach’s alpha used from previous research.

Table 1.

Cronbach’s Alpha

Original Scale	Original Scale Cronbach’s Alpha	Our Adaptation to Scale Cronbach’s Alpha	Sample Items
Positive and negative effect scales (PANAS)	0.72	0.874	x
Feeling towards ads	0.93	0.803	x
Expertise, trustworthiness, and attractiveness of celebrity endorsers	0.80	0.795	x
List of values: LOV	0.83	0.894	How important is sense of belonging to you?
Attitudes influencing monetary donations to charitable organizations	0.93	0.659	My image of charitable organizations is positive.

Source: Watson & Tellegan (2011), Edell & Burke (2011), Ohaniam (2011), Kahle (2011) & Webb & Brashear (2011).

The scales consist of Positive and Negative Effect Scales (PANAS) (Watson, Clark & Tellegen, 1988), Feelings Towards Ads (Edell & Burke, 1987), Expertise, Trustworthiness, and Attractiveness of Celebrity Endorsers (Ohanian, 1990), List of Values: LOV (Kahle, 1983), and Attitudes Influencing Monetary Donations to Charitable Organizations Webb, Green, Brashear (2000). We modified aspects of each scale in order to more clearly reflect our research questions. All of our adaptations to the scales, except for attitudes influencing monetary donations to charitable organizations, had alpha values of 0.70 or above indicating consistency.

Findings and Discussion

Table 2 summarizes our findings from ANOVA. All results were significant at the 0.10 level.

When evaluating our first hypothesis, we were testing to see whether positive emotions were a main driver for people donating. We ran multiple regressions to better understand the relationships between our variables. We found with each regression that the p-value was significant at the 0.05 level. Meaning, regardless of the nature of emotion evoked (positive, negative, or mixed), the respondent was willing to donate, as long as the money went to a good cause. The reaction to emotional appeal was so strong that the prior actions of the celebrity would not impact likelihood to donate. This finding was a fact that shocked us, we expected to see some sort of impact of donation when a scandalous celebrity was involved in the campaign. Taking it a step further, we compared the results of the emotions expressed toward a specific nonprofit, and how it affects donations.

When studying the demographic findings, the p-value for gender and education was significant when we tested it with Jennifer Aniston. However, the p-value for gender and education was not significant with Tiger Woods. It would be interesting to conduct further research and see why it was significant for a female celebrity compared to a male celebrity.

When we measured income against the two celebrities and the two non-profits, we found different results. The results had no significance when we tested the regression for Tiger Woods, but there was significance with Jennifer Aniston. The results were surprising since we thought level of someone’s income would drive how much they donated. Looking further into the workings of both non-profits, it is easy to see the possibility that someone would be more likely to donate a bigger sum of money to a children’s hospital, than to a community-based organization. This would explain why the mean for Jennifer Aniston was at 20.755 while Tiger Woods’s mean remained low at 9.529. Nonetheless, each organization was not affected by the amount of income a donator had, they were still likely to receive a donation.

Since using two celebrities with different demographics, it was important to see whether the gender of the donor would have any effect as to whether or not they viewed the celebrity as being a good role model. Once running the ANOVA, we concurred that the results of evaluating both celebrities were different. Jennifer Aniston’s mean was higher when compared to Tiger Woods’s mean. Meaning, more respondents believed Tiger Woods was a better role model when compared to Jennifer Aniston. There was expectation that the use of Tiger Woods and the controversy surrounding him in the late 2000s would cause donor push back. What would be even more enthralling to see is a retest of this research after Tiger Wood’s DUI arrest in late 2017. However, both celebrities are classified as being role models for different reasons. Therefore, they are not in direct competition with each other because they each target different markets, while it is still important to note that both celebrities have faced scandals and controversies in their past. We chose to test them because of their well-established careers in their fields.

Last, we looked to see if the higher the age of a person had any effect on whether or not they were aware of a celebrity’s past scandal(s). Once we ran the ANOVA, it became significant to both Jennifer Aniston and Tiger Woods. Notably, Tiger Woods was a little more significant most likely due to his more recent scandals in the past couple of years. When we made our hypothesis, we suspected that Tiger Woods might be a little more significant. This is because of a timeline of events. When putting them on a timeline, Jennifer Aniston was extremely popular in the early 2000s, while Tiger Woods’s rocket to stardom did not come until the later 2000s. One thing that would be interested to test now would be to see whether the results for Tiger Woods increased in significance due to his most recent arrest for a DUI (table 3).

Table 2.

Summary Findings (ANOVA)

Independent Variables	Dependent Variables	Mean	F-value	P-value
Positive Emotions	Likelihood to Donate
Jennifer Aniston		40.377	46.857	0.001
Tiger Woods		289.808	111.594	0.001
Higher Level of Income	Likelihood to Donate
Jennifer Aniston		20.755	6.12	0.014
Tiger Woods		9.549	2.506	0.115
Gender of Donor	Affects if Celebrity is Seen as a Good Role Model
Jennifer Aniston		10.120	2.902	0.036
Tiger Woods		3.379	0.302	0.824
Higher the Age of a Person	Awareness of Past Scandals
Jennifer Aniston	20.024	4.635	0.032
Tiger Woods	102.886	8.762	0.003
Being Mindful of Others	Believed if Celebrity Cared
Jennifer Aniston	13.201	3.046	0.082
Tiger Woods	37.75	3.289	0.071

Source: Created by authors.

Notes: (i) Agreement Scale: a seven-point scale ranging from 7 (strongly agree) to 1 (strongly disagree) and

(ii) All mean differences results are significant at the 0.10 level of significance.

Table 3.

Regression Analysis Results

Confirmed Hypotheses	Dependent Variable	R2	P-value
Ads that evoke positive emotions like hope, thankfulness, persuasion, and reflectiveness be a driver for people to donate.	Likelihood to donate
Jennifer Aniston		0.165	0.001
Tiger Woods		0.336	0.001
The higher the level of income, the more likely a person is to donate.	Likelihood to donate
Jennifer Aniston		0.027	0.014
Tiger Woods		Results not significant
The emotions expressed towards a specific nonprofit affects donations.	Affects nonprofit donations
Jennifer Aniston-St. Jude’s		0.119	0.001
Tiger Woods-Tiger Woods Foundation		0.292	0.001
The higher the age of a person the more likely they are to be aware of a celebrity’s past scandals.	Awareness of Past Scandals
Jennifer Aniston		0.021	0.032
Tiger Woods		0.038	0.003
A person who is mindful of others feelings is more likely to donate regardless of the nonprofit or the celebrity does not drive the likelihood to donate.	Believed if celebrity cared
Jennifer Aniston		0.013	0.082
Tiger Woods		0.015	0.071

Source: Created by authors.

Concluding Remarks

By conducting this survey on nonprofits, celebrities, and their donations patterns, we were able to obtain invaluable information about whether or not celebrity endorsers drive donations. Results revealed which type of celebrity endorsers do better with which type of donors. After further research, we found that the particular celebrity used in a nonprofit campaign had little effect on the donations received. While it is important to note that the popularity of the celebrity at the time of the campaign can affect how much money is donated. The overall consensus after completing this survey is that people are always willing to donate as long as it is to a good cause. However, this is not to say that using a particular celebrity can help increase how much each individual donates. There is always something to be said about the celebrity used in a nonprofit advertisement but this can be an uncontrollable factor. When testing how much is donated based on the particular celebrity it could be problematic. This is because there will always be changing preferences or appeal that a celebrity has on each respondent. The overall emotions portrayed in an advertisement have the power to move almost anyone to donate, regardless of the celebrity used. This is because nonprofits, such as St. Jude’s, are much greater in an importance than any grudge or dislike a donor has on a celebrity.

There are limitations to this research and results should be viewed with caution. We did not have sufficient demographic variety (race, for instance). As past research suggests, one should not expect participants to be honest with us when we asked questions about how they feel about nonprofits and donations. This dishonesty can come from participants being embarrassed about whether or not they donate. Age affects donations, if a potential donor is aware of a celebrity’s past scandal, it can affect the perception of a nonprofit organization. Advertisements with celebrities’ and nonprofits that evoke positive emotions do not affect how much money is donated.

A celebrity’s future in their respective field may go through ups and downs. Scandals (real, alleged, or perceived) can occur at any time and without warning. Trying to predict this is challenging and something nonprofits need to consider when working with celebrities.

In the future, more research should be carried out into celebrity endorsements cycles and scandal effects on donation patterns. There is lots of potential in this area of research. Donations are incredibly important to a nonprofit’s success. Most nonprofits depend heavily on them, and by understanding how people donate will help nonprofits serve their recipients more effectively.

Through our findings, we did begin to observe celebrities’ effects on whether or not a respondent would donate. Celebrities can bring awareness to a nonprofit, just like a product in the for-profit world, thus effecting donations (Knoll & Matthes, 2017). It is helpful for businesses and marketers to use this information when creating a marketing and advertising plan for nonprofits because of the celebrities they hire for the advertisement. When choosing a face to represent the nonprofit a marketer or manager needs to be aware of the past scandals they have had. A celebrity’s past and their unpredictable future have the chance to hurt their reputation as well as the organizations they work with. Trying to predict the future and choosing the appropriate celebrity that can be less of a risk can be further understood through future research. Our team believes the research conducted in this study is the first step in understanding more clearly what types of marketing tools can affect donations received. Therefore, the more study that is carried out on celebrities’ effects on donations patterns will help nonprofits leverage this information to increase their development. Further research is done for focuses on how and when people donate as well as how a company should choose the perfect celebrity.

Footnotes

Appendix A: Advertisements Used

References

Bivin

Osili

McKitrick

Bergdoll

Claire

(2017 January). The Philanthropy Outlook .

Devasagayam

(2014). Research Methods . USA: McGraw-Hill Education.

Edell & Burke. (2011). Reactions to marketing stimuli. In Handbook of marketing scales: Multi-item measures for marketing and consumer behavior research (3rd ed., pp. 317–320). Los Angeles: SAGE Publications.

Giving USA. (2016). The annual report on philanthropy for the Year 2015 . Chicago: Giving USA Foundation.

Harris

E. E.

Ruth

J. A.

(2014). Analysis of the value of celebrity affiliation to nonprofit contributions. Nonprofit and Voluntary Sector Quarterly , 44(5), 945–967.

Kahle. (2011). List of values: Love. In Handbook of marketing scales: Multi-item measures for marketing and consumer behavior research (3rd ed., p. 154). Los Angeles: SAGE Publications.

Knoll

Matthes

(2017). The effectiveness of celebrity endorsements: A meta-analysis. Journal of the Academy of Marketing Science , 45(1), 55–75.

McCormick

(2016). Celebrity endorsements: Influence of a product-endorser match on Millennials attitudes and purchase intentions. Journal of Retailing & Consumer Services , 32, 39–45.

The Non Profit Times. (2014, October 28). Nonprofits Worth $887.3 Billion to U.S. economy. Retrieved July 23, 2017, from www.thenonprofittimes.com/news-articles/nonprofits-worth-887-3-billion-u-s-economy/

10.

Ohanian. (2011). Reactions to marketing stimuli. In Handbook of marketing scales: Multi-item measures for marketing and consumer behavior research (3rd ed., p. 330). Los Angeles: SAGE Publications.

11.

The National Philanthropic Trust. (2016). 2016 Annual Report. National Philanthropic Trust . Retrieved April 23, 2017, from https://www.nptrust.org/daf-report/pdfs/donor-advised-fund-report-2016.pdf

12.

Watson

Tellegen . (2011). Positive negative affect scales (PANAS). In Handbook of marketing scales: Multi-item measures for marketing and consumer behavior research (3rd ed., p. 316). Los Angeles: SAGE Publications.

13.

Webb

Brashear . (2011). Attitudes influencing monetary donations to charitable organizations. In Handbook of marketing scales: Multi-item measures for marketing and consumer behavior research (3rd ed., p.165). Los Angeles: SAGE Publications.