How priming affected donation amounts by 96%

Intervention · Donaciones benéficas


A disaster affecting millions of people is distressing, but it rarely moves us as much as when we learn of a tragedy that befalls one person. Researchers recognized the difference in how we react to victims and give to charity, noting that donations would be more effective if they were better distributed, rather than disproportionately steered towards identifiable victims.

Teaching, framing, and priming Interventions did moderate generosity towards identifiable victims, but the adjustment did not significantly increase donations to statistical victims. The biggest impacts were seen with priming feeling over analytical modes of thought processing. Donors were 96% more generous towards identifiable victims when given an affect prime. Conversely, analytical priming only boosted donations to statistical victims by 5%.

Rating = 3/5 (Supports the hypotheses but does not increase welfare equity)

The impact of teaching, framing, priming interventions on donations to statistical and identifiable victims
Donations following identifiability lesson

7% increase to statistical victims
108% decrease to identifiable victims
Donations comparing alternate framing of identifiability lesson 
(e.g. ‘people give more to identifiable victims’ vs. people give less to statistical victims’) 
$1.12 - ‘more’
$1.43 - ‘less’
$1.00 - ‘more’
$1.37 - ‘less’
Donations to separate and joint presentation of victim types

$1.14 given to statistical victims
$2.38 given to identifiable victims
$1.43 given to identifiable victims with statistics
Donations comparing analytical and feeling processing primes

5% increase to statistical with an analytical prime
96% increase to identifiable victims with a feeling prime

Key Concepts 

Identifiable victim effect - The tendency to feel more sympathy and a greater urge to help victims of a tragedy when one or a few people are at the center. 

Affect heuristic - Emotion-driven decision-making, rather than relying on concrete information. This is a quicker, automatic thought system than the more analytical alternative. 

Priming - Exposure to a stimulus, like a word, that influences a person’s response or decision making, usually without them being aware of it. 

Framing - The way information is presented, or framed, that influences how we perceive or process it. 

The Problem

Charities struggle to raise enough money for worthy causes that affect thousands of victims, while by comparison individual victims elicit much greater generosity. This is known as the ‘identifiable victim effect’; they stimulate a more powerful emotional response than the vague concept of a large group of people. A famous example is “Baby Jessica” who fell into a well. People sent over $700,000 for her rescue. Comparatively, statistical victims - though equally deserving and more numerous - are difficult to imagine and therefore hard to feel sympathy for.  

Heart beats brain

Identifiability evokes the affect heuristic, part of a dual cognitive system which is driven by emotional, and more automatic than its counterpart, analytical mode. 

Related to the issue of identifiability related to charity is the “proportion of reference group effect”, the tendency to judge tragic events on a scale of proportion rather than absolutes. For example, an event that results in 10 deaths in a community of 200 is a disaster, while generally the same number of deaths in a community of a million is not. 


Researchers, Small, Loewenstein, and Slovic hypothesized that thinking analytically about the value of lives would reduce donations to an identifiable victim, but have no impact on giving to statistical victims. Across four studies, participants completed a short survey on their technology habits in exchange for $5.00. After completing the survey, the participant was invited to make a donation to the Save the Children charity. 


Their interventions were informed by the MINDSPACE Framework, a mnemonic that captures nine behavior drivers, also including Messenger, Incentives, Norms, Defaults, Salience, Commitments, and Ego. Priming manipulated an analytical or emotional thought processing mode, while affect tapped into subjects’ natural tendency to rely on feelings to make quick decisions. 

Teaching about identifiability

Each participant received a description of either a little girl, or a broad description of the problem of starvation in Africa. The intervention taught half of the participants about the identifiability effect, and how it impacts donations.

Reframing the intervention

To rule out the potential for the framing of the identifiability effect impacting the results of study 1 too heavily, for the second study researchers framed the problem in the opposite way. For example, the lesson about identifiability was rephrased from ‘people give more to identifiable victims’ to ‘people give less to statistical victims’. 

Subtly seeding identifiability

Where studies 1 and 2 explicitly drew attention to identifiability, study 3 was more subtle. Study three gave participants three donation options: donation to an identifiable victim; to statistical victims; or an identifiable victim with statistical information. 

Analytical vs. emotional

The final study did not educate participants about identifiability, instead priming calculation-based, or feeling-based thought. After completing the survey, half of the subjects received a series of math problems, priming analytical thought. The other half of the subjects were primed for emotion-driven reactions with an ‘impression questionnaire’. 

Results and Application

Learning about identifiability reduced donations to identifiable victims, but didn’t increase generosity towards statistic victims. Reframing the intervention for the second study didn’t have a significant effect on donations. The implicit lesson still resulted in the identifiable victim receiving more than double the donation value as statistical victims, although the identifiable victim presented with additional statistical data didn’t fare as well. As predicted, this additional data dampened the affective system, making subjects less emotionally reactive. Priming analytical thought processes was successful in decreasing generosity towards identifiable victims - by almost 100% less compared to their emotion-primed counterparts. However, neither prime had any significant effect on donations to statistical victims.

Education Teachers may be able to educate more effectively by tying concepts and events to an identifiable central character, creating an emotional connection to the subject.
Marketing / Business People respond more to storytelling than dry data. In a world that increasingly champions data, marketers can create greater appeal by turning to a people-driven narrative. 
Financial Services Banks and financial institutions may encourage customers to save or invest by experimenting with framing and priming of their messaging. 


  • While this intervention is concerned with welfare, its hypotheses didn’t support increasing equity.
  • It’s unclear whether the people included as examples of identifiable victims consented to being in the study. 
  • Culture or background of the participant is not explored as a factor in attitudes towards statistical victims compared to identifiable victims.

Does the intervention demonstrably improve the lives of those affected by it?
Room for Improvement
The intervention adjusted generosity towards identifiable victims, but did not increase donations to statistical victims.
Does the intervention respect the privacy (including the privacy of identity) of those it affects?
Room for Improvement
Participants in the intervention were chosen at random and not identified. 
It’s unclear how the examples of identifiable victims were chosen and whether they gave consent.
Does the intervention have a plan to monitor the safety, effectiveness, and validity of the intervention?
Not Applicable
There is no ongoing element to the intervention.

Does the intervention abide by a reasonable degree of consent?
Room for Improvement
Participants consented to completing a survey on their technology habits, but it’s unclear whether the identifiable victims consented to being included in the study. 
Does the intervention respect the ability of those it affects to make their own decisions?
Participants were given the option to donate some or all of their $5.00 reward. 
Does the intervention increase the number of choices available to those it affects?
Not Applicable
The number of available choices was not impacted. 

Does the intervention acknowledge the perspectives, interests, and preferences of everyone it affects, including traditionally marginalized groups?
Room for Improvement
Acknowledges the challenges of subconscious attitudes towards statistical victims, but does not note how attitudes may vary across different cultures, such as those from more collectivist societies. 
Are the participants diverse?
Insufficient Information
Diversity data was not included. 
Does the intervention help ensure a just, equitable distribution of welfare?
Room for Improvement
The intervention was concerned with welfare equity but doesn’t lead to increases where needed. 

Related TDL Content

Does the Quantified-Self Lead to Behavior Change? - Data is often used to convince, but often gets in the way of influencing decision-making. Learn more about the need for data refinement to motivate and maintain behavior-change.


Small, D. A., Loewenstein, G., & Slovic, P. (2007). Sympathy and callousness: The impact of deliberative thought on donations to identifiable and statistical victims. Organizational Behavior and Human Decision Processes, 102(2), 143-153.

Notes illustration

Eager to learn about how behavioral science can help your organization?