What motivates people to help others? Are they utilitarian like Mill and Bentham, or do they care about equity and desert, like Aristotle and Locke? We use a Dutch TV show to answer this question.

There are different normative views on giving. According to utilitarianism, proposed by Jeremy Bentham and John Stuart Mill, individuals should give to maximize societal well-being. This perspective forms the foundation for welfare economics and the effective altruism movement.

There are different normative views on giving. According to utilitarianism, proposed by Jeremy Bentham and John Stuart Mill, individuals should give to maximize societal well-being. This perspective forms the foundation for welfare economics and the effective altruism movement.

How do people actually give? Evidence on the role of utilitarian (or “impact”) considerations in giving is mixed, while there is strong support for that of equity and desert (“responsibility”). However, empirical studies have not directly compared the importance of the two views.

In our paper, we use data from the Dutch TV show Geld Maakt Gelukkig (“money buys happiness”) to compare how perceptions of impact and responsibility influence donation decisions in a quasi-controlled field setting with stakes that are consequential for the recipients.

In each episode, three candidates appeal to a studio audience of one hundred people for financial help to address a difficulty they face.

The cases are very diverse. For example, the candidates depicted below asked for money for breast augmentation following gender reassignment surgery, legal support to resolve issues caused by a legal guardian’s mistakes in managing debts, and a final family vacation due to a terminal illness.

The three candidates are introduced one by one, along with their financial difficulties and the amount of money they are requesting. This introduction is delivered through the combination of a short video documentary and a panel interview.

Each audience member is endowed with €100 and must donate this amount to one of the three candidates at the end of the episode.

In total, 10.000 euros is thus divided in each episode. Below is the outcome for the three candidates just introduced.

The show offers a unique opportunity to study giving: audience members donate under controlled conditions, choosing between three cases without personal connection. This setup resembles the impartial spectator design used in lab studies on moral preferences.

We have data on giving in 55 episodes. But how do we measure the perceived impact of giving, or the perceived responsibility of the candidate? We conduct laboratory sessions at Maastricht University where we ask Dutch participants to rate candidates on these dimensions!

We study how these ratings relate to giving behavior. This ternary plot shows the shares received by candidates who ranked highest, middle, or lowest on impact in an episode. Impact matters: on average, top-ranked candidates received 44%, middle 34%, and lowest 22% of the total.

Responsibility perceptions matter too. This figure shows the shares given when candidates are ranked by their control over their situation. More money went to those with less control: the most control received 24%, intermediate 36%, and least control 40% on average.

These were simple univariate analyses. However, impact and responsibility judgments are strongly negatively correlated. To separate their effects on donation decisions, we use Dirichlet-multinomial regression

Donations increase strongly with impact and decrease moderately with responsibility. A one-standard-deviation increase in impact leads to a 6.2 percentage point rise in the share received, while a similar increase in responsibility lowers it by 2.7 percentage points.

Impact thus matters more than responsibility. This is backed up by a dominance analysis that decomposes the overall fit into the additively separate contributions of each variable. Impact is approximately 2.4 times as important as responsibility for understanding giving behavior.

Interestingly, both the regression and the dominance analysis show no indication that people discriminate on age, gender, or attractiveness in their giving decisions.

These conclusions are highly robust across a wide range of different specifications.

To sum up, we examined giving behavior using a Dutch TV show and find that while both utilitarian and equity & desert considerations are important drivers of donation decisions, utilitarian considerations are considerably more important.

This paper is joint work with Martijn van den Assem, Paul Smeets, and Inka Everhardt Hiabu. The paper can be downloaded here .