Framing & Defaults on Teen’s Donation


The Influence of Defaults and Framing in Teenagers’ Donation Decision Making

Ruoqing Wang
Moonshot Academy 


In recent years, more teenagers in China are participating in and even leading charity donations. Research has shown that individuals’ donation decisions are affected by heuristic influence such as default settings and framing. This paper examines the impact of default donation options and the framing of charitable solicitations on Chinese teenagers’ online donation decisions. The author designed a 2 x 2 experiment to examine separate and interactive effects of message framing and default donation options for teenagers’ online donation decisions. Overall, 165 teenage participants returned valid questionnaires. Participants reported a higher rating on donation rate when the default options are lower than the default options are higher, but the higher default options enhanced teenagers’ average contribution.  Additionally, the framing of solicitation did not affect teenagers’ donation decisions since the donation rate and the amount of donation are indifferent. Finally, there was no interaction effect between the defaults and framing. The overall results indicate that, to enhance teenagers’ charitable contribution, we can increase the numerical size of default options to enhance teenagers’ contribution, but higher donation default options tend to reduce the donation rate. Moreover, teenagers are not sensitive to single goal framing. The results have significant implications for how charities such as NGOs, student clubs, public service/non-profit projects can motivate donations by applying default effect. 

Keywords: default effect, framing effect, online donation, teenagers

Related Work and Hypothesis Development

Default effect, namely that making an option as a default increases its possibility to be selected, has been shown to have a significant impact in many domains, including organ donation (Johnson & Goldstein, 2003), preference for green electricity (Pichert & Katsikopoulos, 2008), retirement planning (Madrian & Shea, 2000), auto insurance (Johnson et al, 2002) and corporate law (Listokin, 2009). Research suggests that default effect may be explained by people’s physical laziness (Johnson & Bellman, 2002). Other studies show that default effect may be affected by anchors (Tversky & Kahneman, 1974) which are the initially presented values that disproportionately influence decision makers’ judgements. Individuals may make decisions based on anchors they have previously been exposed to (Tversky & Kahneman, 1974). 

Similarly, default options, that is, a specified amount to be donated unless a donor actively changes to a different contribution level, are common features on charity websites and online fundraising platforms. One rationale for the study of defaults is that binding default donation level may serve as a reference point, or anchor (Tversky & Kahneman, 1974), which individuals may insufficiently adjust from. Therefore, a charity can take advantage of default contribution levels to enhance donations. Prior studies show that defaults have a strong impact on individuals’ donation behavior. However, according to three recent convincing papers by Goswami and Urminsky (2016), Fiala and Noussair (2017) and Altmann et al. (2018), the causal effect of default options on donation behavior remains ambiguous. Goswami and Urminsky (2016) demonstrated the effect of “choice-option” on donation rate, average donation, and the resulting revenue through eight experimental and natural field experiments. After investigating 1508 participants who made 2423 donation decisions, Goswami and Urminsky (2016) concluded three effects that default options would cause in people’s donation decisions: “low bar” effect (low defaults increase donation rate), “scale-back” effect (low defaults reduce average donation amounts), and a “default-distraction” effect (defaults reduces the impact of other cues).  However, Fiala and Noussair (2017) discovered no significant differences in overall contribution levels under various default options in their study with 270 bachelor’s and master’s students majoring in economics, business, and law. In addition, Altmann (2018) argue that defaults have no obvious influences on people’s average donation amount, contradicting Goswami and Urminsky’s “scale-back” effect. They implemented a field experiment on a large online platform for charitable donations and varied default options in two distinct choice dimensions: The main donation decision and an add-on choice. By collecting and examining data on approximately 680,000 donation-page visits and nearly 23,000 donations, these researchers found out that defaults served as an important role of attraction for donation in both decision dimensions. They claimed that default options had no significant impact on individuals’ donation amounts (Altmann, 2018). 

There is reason to conduct further studies for a more refined understanding of the default effect. First, a default option acts as a suggestion, which means it is considered as a candidate response that subjects entertain, at least as a temporary belief, and therefore influences people’s decision making (Jacowitz & Kahneman, 1995). Plus, even when people decide not to choose the defaults, the options might act as anchors to affect their decisions by nudging them to select an option similar to the default (Dhingra et al, 2012). Therefore, the author proposes the hypotheses:

H1: Teenagers’ donation rate will be higher when the default option is lower. 

H2: A higher default option will be more effective than a low default option when promoting teenagers’ average amount of donation.

Framing indicates the different presentation of a logical equivalent outcome to decision-makers. One way to construct framing is by presenting information positively or negatively. Positive framing appeals to the positive influences people’s donation will have. Negative framing emphasizes the consequences of people refusing to donate (Kerhof et al, 2003). Plus, framing can also be based on risk, attribute, or goals. Irwin Levin claims that there are three types of positive and negative framing: risky choice framing (set of choices with various risk levels), attribute framing (objects/events attribute for characteristics), and goal framing (consequences or implied goals of behaviors) (Levin et al, 1998). Goal framing has been applied and researched widely in the medical context, but it seems that goal framing “could be the most applicable to a fundraising context because it is geared towards persuasion” (Smyth & Macquillin, 2018). Previous studies have indicated that both positive and negative framing can elicit more donations than information present naturally. Chun-Tuan Chang and Yu-Kang Lee (2010) considered the effect of positive (‘with your donation, their life could become hopeful’) and negative (“without your donation, their life would be hopeless”) framing in the context of child welfare charity. The result supporting for negative framing serves as a better donation enhancer, due to loss aversion and negativity bias. Similarly, Xiaoxia Cao (2016) indicates that “a negative-gain (loss) frame enhances donation motivation among people who are more subjective to the negative consequences of not donating”. 

Additionally, it was also discovered that people from different age groups have a different perception of framing effect, but there are limited studies that have examined the framing effect on teenagers. Kim, Goldstein, Hasher, and Zacks (2005) found that older adults (aged 58 to 78 years) are more vulnerable to framing effect than younger adults (aged 17 to 28 years). They adopted two problems, the “fatal disease problem” from Wang, Simons, and Bredart (2001), and “cancer treatment” problem, from McNell and colleagues (1982) with a positive and negative frame. The result indicates that “older adults showed a significant framing effect for both fatal disease problems, and the cancer treatment problem” (Kim, 2005). In contrast, younger adults did not show a framing effect for either problem. The question that this paper tries to explore is which type of goal framing (positive or negative) is a more powerful enhancer for teenagers in promoting donation rate and contribution level. In light of this, the author postulates:

H3: Teenagers’ donation rate will be higher when the solicitation framing is negative.

H4: A negatively framed message will be more effective than a comparable positively framed message when promoting teenagers’ contribution levels.

The interaction effect of defaults and framing on individuals’ decision making has also been tested in various studies. For instance, Bahirat, (2018) empirically explored the roles of defaults and framing in an online privacy context by conducting a survey study with 1133 U.S.-based adult participants. In the survey, they presented 13 information-sharing scenarios and asked each participant a decision question. Each decision question consists of a positive or negative framing (“Would you enable this feature?” or “Would you disable this feature”) and a default option (Yes or No). The result suggests that, apart from separate influences, defaults and framing show additive effects in influencing the construction of preferences (Bahirat, 2018). Likewise, Johnson, Bellman, and Lohse conducted another research in the context of smartphone users in privacy-setting interfaces but found the interaction effects of defaults and framing not significant (Johnson & Bellman, 2002). The researchers investigated over 25,000 Web users of the Virtual Test Market (VTM) with four various questions (1. “Notify me about more health surveys” 2. “Do Not notify me about more health surveys” 3. “Notify me about more health surveys” with default 4. “Do Not notify me about more health surveys” with default). They found that default and framing appeared to be additive and did not interact (Johnson & Bellman, 2002). These two papers led the author to consider the possible interaction effect of defaults and framing on teenagers specifically in the charitable donation setting. Therefore, the author postulates:

H5: There would be no interaction effect of defaults and framing on teenagers’ donation decisions.


An experiment of adopting a 2 (goal message framing: positive vs negative) x 2 (default option: high vs low) was developed to examine which of the separate and comprehension effects of message framing and default options when promoting teenagers’ online donation behaviors. Care was taken to ensure that the positively and negatively framed versions of charitable solicitations provided the same quality and amount of information, except for the distinct high and low donation options and default option value. Four different questionnaires were developed. Doctor Panda Non-profit Project (aims to contribute to malaria prevention mainly through the approaches of donating long-lasting anti-malaria nets) was used as the setting of the donation activity (Online fundraising for long-lasting anti-malaria nets that would be donated to African children who are affected by malaria). Apart from information about the donation activity, all the questionnaires displayed the same general information about how the donation can make a difference to some African children’s lives. Then, the author used an f-test to assess the separate effect and interaction effect of defaults and framing.


The author combines a framing setting manipulation (positive versus negative) with a default manipulation (high defaults versus low defaults).

1.Default options

Moreover, the default effect was manipulated by the value of the default options.

  • High-default option: 90RMB
  • Low-default option: 18RMB

In this paper, the author adopted goal framing and manipulated the framing via a charitable solicitations wording.

·       Positive framing: “Donating long-lasting anti-malaria nets can effectively prevent malaria. According to the Against Malaria Foundation, with every 500 additional nets put into use, one more child’s life can be saved from malaria.”

·   Negative framing: “Donating long-lasting anti-malaria nets can effectively prevent malaria. According to the Against Malaria Foundation, when the number of mosquito nets donated and put into use drops by 500, one more child will lose his life due to malaria.”

 Framing Default 
Questionnaire 1Positive Low 
Questionnaire 2 Negative Low
Questionnaire 3Questionnaire 4Positive Negative HighHigh 

Table 1. The Default and Framing manipulations 
3.Final questionnaires

Each of the four comparable questionnairecontains two sections: (1) a brief introduction of the nonprofit project Doctor Panda and its donation activity, (2) the benefits of donating (negative consequences of not making a donation), and (3) a multiple-choice question of the contribution level (Apart from choosing default options or other options, participants can also fill in the amount of money they want to donate or choose not to donate). (4) participants’ personal information, including gender, age, experiences of online donations, and monthly allowance.


A total of 195 individuals participated in the main experiment. Because of incomplete responses and age below or exceed the prescribed age (teenagers aged between 12 to 19 years old), the number of usable questionnaires was reduced to 165, ages ranged from 12 to 19 years. There were 85 males and 82 females; 64.07% of participants have had previous online donating experiences.


An F-test was performed to assess whether the choice proportions were significantly different across the two frames and the two default options. Participants reported a higher rating on donation rate when the default options were lower (89.02%) than the default options higher (83.53%), indicating that the lower default value could facilitate participants’ response rate when making donation decisions. Thus, H1 was confirmed.

 Donation RateN
Low 0.890282
High 0.835385
Average 0.8628167

Table 2. The effect of defaults on donation rate 

Additionally, participants reported a higher average amount of donation when the default options are higher (M = 58.35, F (1, 166) = 88.947, p < 0.05) than the default options are lower (M = 16.524, (1, 166) = 88.947, p < 0.001). Consistent with our predictions, the effect of defaults on donors’ contribution level was significant. Hence, H2 was supported. 

 Mean Std. Deviation N
Low 16.52412.159782
High 58.35338.050685
Total 39.74735.2802167

Table 3. The effect of defaults on average donation amount 

Moreover, participants who read the positively framed solicitation showed a higher donation rate (91.95%) than those who received a questionnaire with negatively framed solicitation (82.5%). However, the difference was not significant. Therefore, H3 was not confirmed, since the result supports the opposite of the hypothesis. In addition, under a positiveframing message, the participants’ average donation amount was higher (M = 39.747), and the average donation amount under a negative framing message was slightly lower (M = 35.713). Again, this difference was not significant. Thus, H4 was not supported. 

 Donation RateN
Total 0.87225167

Table 4. The effect of framing on donation rate. 
 Mean Std. Deviation N
Positive 39.74735.998187
Negative 35.71334.585580
Total 37.81435.2802167

Table 5. The effect of framing on average donation amount.

Additionally, the results indicated no interaction effect between the defaults and framing ((1, 166) = 0.002, p = 0.966). Finally, a significant effect of previous donation experience on donation level was found, with participants who had donation experience selecting higher donation level (M = 41.85) than those who had no donation experience (M = 30.62, (1,166) =4.047, p < 0.05). Plus, there was no significant effect of gender, age or monthly allowance on teenager’s contribution levels. 

 Mean Std. Deviation N
Have 41.8537.211107
Do not have 30.6230.52960
Total 37.8135.280167

Table 6. The effect of donation experiences on average donation amount. 


In this study, the author found that default donation options can affect teenagers’ donation decisions. The participants showed a higher donation rate when the defaults were relatively low and showed a higher average donation amount when the defaults were high. Additionally, the result shows that the framing of solicitations did not affect teenagers’ donation decisions because the donation rate and the amount of donation did not differ. Finally, there was no interaction effect between the defaults and framing, which means that the existence of one effect did not affect the other. 

The author extended the finding of the research paper on the effect of defaults and framing in privacy policy (Johnson et al, 2002) to the realm of donation. The results are consistent with earlier work on the default effect by Goswami and Urminsky (2016). However, the results seem to contradict previous findings regarding framing effects: Chang and Lee (2010) found that negative frames were “maybe both more salient and therefore effective” in eliciting donations. One of the reasons for this contradiction might be that prior research has mainly recruited college students or older adults as participants, and the performance of teenagers has rarely been the central focus of the research. In accordance with Kim et al, “older adults (aged 58 to 78) are more likely to show framing effects than younger adults (aged 17 to 28) because they tended to rely on heuristic information processing more than younger adults. The author speculates that, because age positively affects teenagers’ sensitivity to frames, teenagers (aged 12-19) may be less likely to be influenced by framings. Additionally, the results demonstrated no interaction effect of defaults and framings because there was no significant framing effect in the experiment. 

The results add to a growing body of evidence suggesting that default options for charitable donation may elicit teenagers’ donation intention in surveys. Apart from its theoretical significance, the results show deep implications for how charities such as NGOs, student clubs, public service/non-profit projects can motivate donations through the application of default effects.

Limitations and direction for future research

The limitations inherent in this study present opportunities for future research. First, the findings are based on a single hypothetical research exposure to one of four online donation questionnaires, which do not have real-world consequences for respondents. Participants may feel less concerned about their decisions and reported higher willingness or amount for donation since they would not really donate the money. Thus, the research is open to debate in terms of whether the results will generalize to decisions with real financial consequences for donors. We encourage more future studies with actual donation decisions. It is possible that different results might be obtained if participants are facing real online donation scenarios, which require them to make real donation decisions. 

Secondly, broader and more diverse samples of teenagers should be examined in additional research. The sample size of this research includes 165 participants, which is relatively limited. The convenience sample that is collected through the author’s social networking tends to produce biases. Participants might have displayed a higher willingness to donate and donate a higher amount of money owing to their acquaintance with the author. Also, all the participants are Chinese teenagers who live in developed urban areas. Thus, the research is open to doubt whether the results will generalize to the donation intention of Chinese teenagers. Therefore, one proposition for additional study is to collect larger and more diverse samples and avoid convenience samples.

Finally, a worthwhile issue for additional research is to explore the effect of defaults and framing with vividness presentations, since in charitable contexts, such as crowd-funding platforms, vivid presentations are usually examined through case stories and images of recipients. Concrete personal stories could affect the donation decision of individuals. Moreover, Burt and Strongman (2005) found that, compared with photographs that expressed less emotion (both positive or negative), those generating fewer emotions were more effective at increasing donation. Burt and Strongman’s (2005) and Small and Verrochi’s (2009) findings both support that images expressing negative emotions more effectively promote donations. For instance, future researchers may choose to replicate this study by examining whether adding vivid presentations in experiments could further promote teenagers’ donation decisions. 


Altmann, S., Falk, A., Heidhues, P. Jayaraman, R., Teirlinck, M. (2018). Defaults and Donations: Evidence from a Field Experiment. DICE Discussion Paper, No. 294, ISBN 978-3-86304-293-6, Düsseldorf Institute for Competition Economics (DICE), Düsseldorf.

Bahirat, P., He, Y., Knijnenbur, B. P. (2018). Exploring Defaults and Framing effects on Privacy Decision Making in Smarthomes. USENIX Symposium on Usable Privacy and Security (SOUPS).

Burt, C. D., and Strongman, K. (2005). Use of images in charity advertising: Improving and compliance rates.International Journal of Organizational Behavior, 8(8), 571-580.

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.

Chang, C. T., and Lee, Y.K. (2010). Effects of message framing, vividness congruency and statistical framing on responses to charity advertising. International Journal of Advertising, 29(2), 195-220.

Das, E,. Kerhof, P., and 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

Dhingra, N., Gorn, Z., Kener, A., & Dana, J. (2012). The default pull: An experimental demonstration of subtle default effects on preferences. Judgment and Decision Making, 7(1), 69-76.

E. J. Johnson and D. Goldstein. Do Defaults Save Lives? Science, 302(5649):1338–1339, Nov. 2003.

Fiala, Lenka and Charles Noussair. 2017. “Charitable Giving, Emotions, and the Default Effect.” Economic Inquiry 55 (4):1792–1812.

Goswami, Indranil and Oleg Urminsky. 2016. “When Should the Ask Be a Nudge? The Effect of Default Amounts on Charitable Donations.” Journal of Marketing Research 53 (5):829–846. Haggag, Kareem and Giovanni Paci. 2014. “Default Tips.” American Economic Journal: Applied Economics 6 (3):1–19.

Jacowitz, K. R., & Kahneman, D. (1995). Measures of anchoring in estimation tasks. Personality and Social Psychology Bulletin, 21, 1161–1166.

Johnson, E. J., Bellman, S. & G. L. Lohse. “Defaults, framing and privacy: Why opting in-opting out.” Marketing Letters 13.1 (2002): 5-15.

Johnson, E. J., Hershey, J., Meszaros, J., & Kunreuther, H. (1993). Framing, probability distortions, and insurance decisions. Journal of Risk and Uncertainty7(1), 35–51.

Levin, D. K., Is Behavioral Economics Doomed? The Ordinary versus the Extraordinary, EUI MWP LS. European University Institute Research Repository.

Levin, I. P., Schneider, S. L., & Gaeth, G. J. (1998). All frames are not created equal: A typology and critical analysis of framing effects. Organizational Behavior and Human Decision Processes, 76, 149-188.

Listokin, Y. (2009). What do corporate default rules and menus do? An empirical examination. Journal of Empirical Legal Studies, 6(2), 279-308.

Madrain, B. C., & Shea, D.F. (2000). The power of suggestion: Inertia in 401 (k) participation and savings behavior.National Bureau of Economic Research.

Small, D. A., and Verrochi, N. M. (2009). The face of need: Facial emotion expression on charity advertisements. Journal of Marketing Research, 46(6), 777-787.

Smyth, R. and MacQuillin, I. (2018). You’ve Been Reframed – Discussion Paper 2: Positive & Negative Feedback v1.1. Plymouth: Rogare/Hartsook Centre for Sustainable Philanthropy 

Steffen; Falk, Armin; Heidhues, Paul; Jayaraman, Rajshri; Teirlinck, Marrit (2018) : Defaults and donations: Evidence from a field experiment, DICE Discussion Paper, No. 294, ISBN 978-3-86304-293-6, Düsseldorf Institute for Competition Economics (DICE), Düsseldorf.

Tversky, A., & Kahneman, D. (1974). Judgment under Uncertainty: Heuristics and Biases. Science, 185(4157), 1124-1131. doi:10.1126/science.185.4157.1124

Wang, X. T., Simons, F., & Bredart, S. (2001). Social cues and verbal framing in risky choice. Journal of Behavioral Decision Making, 14, 1–15.

Leave a Reply

Your email address will not be published. Required fields are marked *