1. Pope Francis on the Disabled
The attitude of humankind towards the disabled takes a central place in the thinking of Pope Francis. In his encyclical Fratelli Tutti[1] two commands coexist: inclusion and solidarity.
The Pope calls for “open societies that integrate everyone” (97). He deplores “hidden exiles” […]. Many persons with disabilities feel that they exist without belonging and without participating. Our concern should be not only to care for them but to ensure their active participation in the civil and ecclesial community, […] acknowledging each individual as a unique and unrepeatable person” (98). “Every human being has the right to live with dignity and to develop integrally […] even if they […] were born with or developed limitations” (107). This is a strong call for recognition.
Yet the Pope also notes: “Some people are born into economically stable families, receive a fine education, grow up well nourished, or naturally possess great talent. They will certainly not need a proactive state; they need only claim their freedom. Yet the same rule clearly does not apply to a disabled person” (109). “Solidarity finds concrete expression in service [… a]nd service in great part means caring […] for the vulnerable members of our families, our society, our people” (115). This is an equally strong call for assistance.
Actually, these two views should not be seen in opposition with each other. In conclusion, the Pope draws the picture of a world governed by “charity, [… which] is the best means of discovering effective paths of development for everyone” (183). “Charity is always a preferential love shown to those in greatest need; it undergirds everything we do on their behalf. […] Only a gaze transformed by charity can enable the dignity of others to be recognized and, as a consequence, the poor to be acknowledged and valued in their dignity” (187).
Or in short: recognition engenders assistance, and assistance enables the disabled to be recognized.
In this contribution, I draw on experimental evidence to cast light on the willingness of others to assist those in need. I have no direct evidence on recognition. It would be a worthwhile exercise to develop experimental designs that identify recognition. It would be even more important to test in which ways, and to which degree, assistance follows from recognition, and recognition follows from assistance.[2]
Very often, the disabled are needy. They need help from the abled, or from the healthy members of society. Some disabilities are so severe that the disabled person needs a special form of help: someone else must decide on her behalf. I report experimental evidence on both: charitable action, and decision making on behalf of another person. I am using the experimental evidence to gain a sense of the likelihood that persons with a disability receive the help they need, to lead a meaningful life.
2. Why experiments?
A person suffers from a severe physical or mental handicap: does this person receive the help she needs to lead a tolerable life? A student comes to the lab. He wants to earn a bit of pocket money. The experimenter gives him some extra money, and observes how he spends it: does he share some of the money with another, undisclosed participant in the same lab who has not received such an extra endowment? There is an obvious and wide gap between the aspect of the world this Plenary wants to understand (the situation of the disabled) and the phenomenon I have studied (allocation choices of experimental participants). Is there anything one can learn from the experiment for the real-world situation this Plenary wants to understand?
In the literature, this gap is discussed under the heading of external validity (Bracht and Glass 1968, Calder, Phillips et al. 1982, Lucas 2003, Mitchell 2012). Essentially, no experiment is fully externally valid. Learning from experimental evidence always requires a leap of faith. The phenomenon that one studies is only analogous to the phenomenon one wants to understand. Why would any researcher do that? There are alternative empirical methods. The most straightforward does always have appeal. One directly observes what one wants to understand. Yet, normally a researcher is not content with making a statement about one individual handicapped person and one other person on whom she depends. The researcher turns to empirical work because she wants to generalize: under which conditions are handicapped persons more likely to receive the help they deserve? Observing a single interaction does obviously not suffice to answer this question. Observing many instances of interaction is better. But one can never rule out that help for the needy results from a very positive selection from the population of potential help givers, or from the fact that the handicapped person is particularly nice, to mention only two of a myriad of possibilities (more from Angrist and Pischke 2008). Generalisations require causal claims, and isolating causality with observational data is a very challenging enterprise.
These concerns matter for the topic of the Plenary. There are plenty of worthy topics for empirical investigation: the impact of disability legislation; the rate of donations for disability; career choices in families with a disabled member. But all of them are fraught with confounds. Disability legislation is not a one-to-one mapping of the attitudes in the population towards the handicapped. Of course, in a democracy one hopes to find a relation between the desires of the people and the laws that are passed on their behalf. But there are many filters between the former and the latter. Not everything people say translates into what people do. In order for legislation to be adopted in Parliament, the ruling party (or the coalition) must find the issue sufficiently important to prioritize it. It must keep in mind what are likely reactions on the next election day. The donation rate likely differs by income and wealth. People do not only give for the disabled. If they have already donated for other worthy causes, the disabled are likely to get less money (for background see Ploner and Regner 2013). How donations are treated by the tax authorities is also likely to affect donation choices. Career decisions depend on many more things than having a disabled child or sibling. The rest of the family may just need the money. The potential caregiver may have a very attractive career opportunity that she does not want to let go. A tolerable professional care institution may be available, so that the potential caregiver from the family experiences less pressure.
This is where the experiment shines. The very fact that raises the external validity concern actually is its definitional strength. Precisely because the experiment is radically artificial, one may be sure that observed effects have a causal interpretation. If the experiment is properly designed, baseline and treatment differ in a single respect. A sufficiently large number of participants is randomly selected from the population of interest. Half of them are randomly assigned to the baseline, and half of them to the treatment. If the treatment group behaves in ways that systematically differ from the behaviour of participants in the baseline, one may confidently infer that the manipulation has caused the difference in behaviour.
Of course, the external validity gap persists. It depends on the research question whether it is acceptable. In this paper, I argue that the experimental findings are sufficiently related to the topic of the Plenary to make them useful. It certainly will not be possible to conclude from the experimental findings the exact degree of help that the disabled may hope to get. But with the help of the experiments, one learns about elements of human nature that a disabled person may hope to trigger.
3. Charitable Action
Economics is an individualistic discipline. It is interested in the choices of isolated individuals, in the interaction of such individuals, and in societal effects resulting from the decisions of individuals. In the economic textbook, it is assumed that an individual maximizes utility, assuming that everyone with whom the individual interacts does the same, and expecting the first individual to do so as well.[3] Outside economics, these assumptions are often misunderstood. They are conceptual building blocks, not normative commands. And utility need not be confined to income. This last qualification has spawned an entire subdiscipline. Behavioural economics theorizes, tests and catalogues behavioural regularities that induce decision makers to deviate from profit maximization (for a summary account see Dhami 2016). Many of these regularities are cognitive: individuals interpret the available information in a distorted way. In the present context, motivational effects are more important. Under which circumstances do some, many, or nearly all individuals refrain from being selfish?
As I have explained, the very reason for going to the lab is identification. One pays the price of artificiality since one wants to isolate cause and effect. This explains why (economic) experiments are usually radically simple. One such simple design is called the dictator game (the game has been invented by Kahneman, Knetsch et al. 1986, S290 f.). Two participants are randomly matched. One of them receives an endowment. The literature calls this person the dictator. The dictator is free to keep the endowment for himself, or to share any fraction with the other participant with whom she has been matched, and whom she knows has not received an endowment. This experiment has been repeated hundreds of times. There is always a fraction of participants who simply keep the money for themselves. But typically, the majority shares some of the money with the passive counterpart.
I have done a meta study of the dictator games published until 2009.[4] As the left panel of Figure 1 shows, which is taken from this meta study, if the recipient is just another participant of the same experiment, the most frequent individual decision is to keep the complete endowment. The second most frequent decision is the equal split: the dictator keeps 50% of the endowment, and gives the other half to the passive counterpart. It is very rare that a participant gives away more than half of the endowment. Given the design of the experiment, this distribution of choices makes intuitive sense. Yet as the right-hand side of the panel shows, the distribution of choices looks considerably different if the recipient is deserving. Now it not only becomes more frequent that the dictator gives away more than half of the endowment. Actually, giving everything even becomes the most frequent decision. This decision is even more prevalent than keeping the complete endowment. Also the average amount that dictators give away is considerably (and significantly) higher (Engel 2011, 594, Umer, Kurosaki et al. 2022).[5]
“”
Figure 1
Donation Rates in Dictator Games
Actually, the available information is even more fine-grained. Some experiments have manipulated the upfront endowment for the recipient. While in the standard design it is zero, in these experiments recipients have received a positive amount. This design feature can be interpreted as a manipulation of the deservingness. As Figure 2 shows, this has a clear effect. The higher the upfront endowment of the recipient, the less she receives. Actually the relationship is almost perfectly linear (Engel 2011, 595).[6]
Figure 2
Donations Conditional on the Deservigness of the Recipient
We had the opportunity to test prison inmates on the dictator game (Chmura, Engel et al. 2017). As Figure 3 shows, they were fairly generous when being in the role of dictator.[7] In the experiment, we have had each prisoner decide twice: once if the recipient is another inmate of the same prison, and once if the recipient is the well-known charity Brot für die Welt. Bubbles above the 45° line stand for prisoners who give more to charity than they give to another inmate. Most of them do.[8] Hence even those whom one might have thought to be least likely to do so are sensitive to the needs of those in dire circumstances.
Figure 3
Prisoners Giving to Each Other and to Charity
How much another person is in need of help is often not clearly defined. If a person has the ability to help, she quite often faces uncertainty. One might have thought that potential helpers use the uncertainty as an excuse. In other contexts, it has been shown that many individuals exploit the opportunity to hide behind uncertainty. They use “moral wiggle room” to get them both: they are selfish and avoid having bad conscience (Dana, Weber et al. 2007). Yet in a dictator game in which we manipulate the information about the recipient’s endowment, we find the opposite (Figure 4). With an increasing degree of uncertainty, dictators give more, not less. They give most if any information about the recipient’s endowment is ostensibly withheld (Engel and Goerg 2018) .
Figure 4
Dictator Giving if the Recipient’s Endowment is Uncertain
Much of charitable giving is indirect. The donor does not directly interact with the intended recipients. Rather she donates some money, or physical goods for that matter, to an organization that promises to use the donation to the benefit of the intended recipients. Such indirect generosity is inherently risky. The donor must trust the direct recipient that she will use the gift in the intended way. The risks are manifold: the direct recipient may simply embezzle money; she may be ineffective, so that money gets unnecessarily lost; she may not be careful enough in selecting the final recipients. In another experiment, we have addressed this concern, and have given donors[9] the possibility to invest an extra amount of money to insure their donation against such risks (Buijze, Engel et al. 2017). As we wanted to be able to isolate the motive, we have had three treatments. In the first treatment, there was a risk that the donation would not be effective. In the second treatment, the risk was exclusively affecting the donor herself: she could not be sure that she would receive a 50% refund from the tax authority. In the third treatment, we have combined both risks. As Figure 5 shows, participants are willing to insure themselves against either risk. We do not find any significant difference between treatments in this respect.
Figure 5
Willingness to Pay for Insuring a Donation
Yet, as Figure 6 shows, we do find clear treatment differences of the availability of insurance on the donation rate.[10] In the baseline, neither risk was present. In the treatments, there either was the risk of the donation not reaching its intended recipients, the tax refund not being granted, or either risk with 50% probability. The dark grey bars demonstrate that, in the absence of the insurance option, the risk that the donation is subverted has a dramatic effect on the willingness to donate, while the risk that, after the fact, the donation turns out more expensive only has a small effect. Yet once the insurance option is available, the donation rate jumps back to its original degree.[11] This is a remarkable finding. It shows that donors are not deterred by either risk as long as they can use extra money to make sure the risk is not materializing.
Figure 6
Effect of the Availability of Insurance on the Donation Rate
4. Acting on Behalf of Others
The disabled rightly care about their autonomy. They want to be treated with respect. They want to decide themselves about the kind of life they are living. But unfortunately, some forms of disability make it difficult, if not impossible, for the disabled person to choose herself. Someone else must step in and decide on her behalf. There is growing interest in behavioural economics in the determinants of decision-making on behalf of third parties (Frey, Herzog et al. 2018, Montinari and Rancan 2018, Eriksen, Kvaløy et al. 2020), including decision-making on behalf of incapacitated patients (Elliott, Gessert et al. 2007, Pope 2011). In the typical experimental implementation, the third party is another anonymous participant (see e.g. Cerrone and Engel 2019). This design is appropriate if one wants to understand the general effect of taking responsibility for other people’s lives. Yet for the present context, it is more important how natural representatives decide on behalf of a handicapped person. The most natural representative is a family member.
This is what Alexandra Fedorets, Olga Gorelkina and myself have studied in an experiment (Engel, Fedorets et al. 2023). We had the good fortune to run our experiment on the complete intervention sample of the German Socio-Economic Panel SOEP (Goebel, Grabka et al. 2019). In 494 households, interviewers randomly selected 2 household members. Theoretically this could have been one of the parents and one of the children. But effectively we have only very few (48) parent-child pairs, but 177 parent-parent pairs, and 269 child-child pairs. Due to concerns about prior consent, we only tested children who are 18 years old or older (but still live in the household of their parents).
Participants had to choose one of 11 lotteries. Each lottery has a high and a low outcome. Both outcomes are equally probable. The first lottery is actually a safe choice, as both outcomes are the same. Participants gain 10€ with certainty. All other lotteries expose participants to true risk. But there is a catch. Risk-taking is profitable. The good outcome increases twice as fast as the bad outcome deteriorates. Hence in the second lottery, the good outcome is 12, while the bad outcome is 9. The gap between the good and bad outcome increases in steps of 2 for the good and in steps of 1 for the bad outcome. In the final lottery, the good outcome is 30, while the bad outcome is 0. Hence the more a participant is averse to risk the smaller the spread between the good and the bad outcome.
In the present context, the condition is most important in which one family member is randomly given the active role, while the other is passive. The participant in the active role chooses a lottery on behalf of the passive participant. This decision has no material consequences for the decision maker. All effects are on the passive family member. Figure 7 summarizes the main result. It compares the decision made by the active family member with the decision that the passive family member had made in an earlier part of the experiment, when deciding on behalf of herself. The graph peaks at zero: by far the most frequent choice of the active family member is the choice that the passive family member had made herself before. There are deviations, but they are not symmetric. If the active member deviates from what she knows the passive member to wish, she is much more likely to deviate to a more cautious choice. This suggests that the active family member did not simply want to impose her own preferences or convictions; otherwise, deviations would have to be symmetric. Rather, if she deviates from the discernible will of the passive family member, the active member imposes a more cautious approach. This pattern is consistent with a responsibility effect: most people react much more negatively if they incur a loss, compared with foregoing a gain (cf. Branas-Garza, Durán et al. 2009, Pahlke, Strasser et al. 2015). It is a different matter whether a person freely decides to incur a risk, or whether the same risk is imposed on her by someone else, how benevolent this other person might be.
Figure 7
Making a Risky Investment on Behalf of Another Family Member
5. Tentative Implications for Living with the Disabled
At the beginning of this essay, I have stressed the wide gap between the experimental findings I report on and the way society, and its individual members, treat persons with disabilities. Any conclusions must be very tentative. But arguably. the disabled receive respect and help. Being respectful and helping is a human trait that transcends the attitude towards disability. Arguably, if we treat the disabled in the way they deserve, society must capitalize on behavioural traits of considerably broader scope. It is in this spirit that I have reported evidence from dictator games (about the willingness to help those in need) and about decision-making on behalf of third parties (about a respectful way of deciding about other people’s lives if they cannot do so themselves).
None of the reported effects is universal. But happily for the present policy problem, this is not a major concern. It certainly would be desirable that everybody comes to the rescue of a person who cannot help herself since she has been hit by a disability. But if the fraction of individuals who are willing to do so is sufficiently large, this already goes a long way towards alleviating the burden on the shoulders of a person with a disability. In that spirit I conclude on an optimistic tone: Misericordia is real.
References
Angrist, Joshua D and Jörn-Steffen Pischke (2008). Mostly Harmless Econometrics. An Empiricist’s Companion. Princeton, Princeton University Press.
Aumann, Robert J (1995). “Backward Induction and Common Knowledge of Rationality”. Games and Economic Behavior 8(1): 6-19.
Bracht, Glenn H and Gene V Glass (1968). “The External Validity of Experiments”. American Educational Research Journal 5(4): 437-474.
Branas-Garza, Pablo, Miguel A. Durán and María Paz Espinosa (2009). “The Role of Personal Involvement and Responsibility in Unfair Outcomes”. Rationality and Society 21: 225-248.
Buijze, Renate, Christoph Engel and Sigrid Hemels (2017). “Insuring Your Donation. An Experiment”. Journal of Empirical Legal Studies 14: 858-885.
Calder, Bobby J, Lynn W Phillips and Alice M Tybout (1982). “The Concept of External Validity”. Journal of Consumer Research 9(3): 240-244.
Cerrone, Claudia and Christoph Engel (2019). “Deciding on Behalf of Others Does Not Mitigate Selfishness: An Experiment”. Economics Letters 183: 108616.
Chmura, Thorsten, Christoph Engel and Markus Englerth (2017). “At the Mercy of a Prisoner. Three Dictator Experiments”. Applied Economics Letters 24: 774-778.
Cochard, Francois, Julie Le Gallo, Nikolaos Georgantzis and Jean-Christian Tisserand (2021). “Social Preferences across Different Populations. Meta-Analyses on the Ultimatum Game and Dictator Game”. Journal of Behavioral and Experimental Economics 90: 101613.
Dana, Jason, Roberto A Weber and Jason Xi Kuang (2007). “Exploiting Moral Wiggle Room. Experiments Demonstrating an Illusory Preference for Fairness”. Economic Theory 33(1): 67-80.
Dhami, Sanjit (2016). The Foundations of Behavioral Economic Analysis, Oxford University Press.
Doñate-Buendía, Anabel, Aurora García-Gallego and Marko Petrović (2022). “Gender and Other Moderators of Giving in the Dictator Game: A Meta-Analysis”. Journal of Economic Behavior & Organization 198: 280-301.
Elliott, Barbara A, Charles E Gessert and Cynthia Peden-McAlpine (2007). “Decision Making on Behalf of Elders with Advanced Cognitive Impairment. Family Transitions”. Alzheimer Disease & Associated Disorders 21(1): 49-54.
Engel, Christoph (2011). “Dictator Games. A Meta-Study”. Experimental Economics 14: 583-610.
Engel, Christoph, Alexandra Fedorets and Olga Gorelkina (2023). Risk Taking in the Household. Strategic Behavior, Social Preferences, or Interdependent Preferences?
Engel, Christoph and Sebastian J Goerg (2018). “If the Worst Comes to the Worst. Dictator Giving When Recipient’s Endowments Are Risky”. European Economic Review 105: 51-70.
Eriksen, Kristoffer W, Ola Kvaløy and Miguel Luzuriaga (2020). “Risk-Taking on Behalf of Others”. Journal of Behavioral and Experimental Finance 26: 100283.
Frey, Renato, Stefan M Herzog and Ralph Hertwig (2018). “Deciding on Behalf of Others: A Population Survey on Procedural Preferences for Surrogate Decision-Making”. BMJ open 8(7): e022289.
Goebel, Jan, Markus M Grabka, Stefan Liebig, Martin Kroh, David Richter, Carsten Schröder and Jürgen Schupp (2019). “The German Socio-Economic Panel (Soep)”. Jahrbücher für Nationalökonomie und Statistik 239(2): 345-360.
Kahneman, Daniel, Jack L. Knetsch and Richard Thaler (1986). “Fairness and the Assumptions of Economics”. Journal of Business 59: S285-S300.
Larney, Andrea, Amanda Rotella and Pat Barclay (2019). “Stake Size Effects in Ultimatum Game and Dictator Game Offers: A Meta-Analysis”. Organizational Behavior and Human Decision Processes 151: 61-72.
Lucas, Jeffrey W (2003). “Theory‐Testing, Generalization, and the Problem of External Validity”. Sociological Theory 21(3): 236-253.
Montinari, Natalia and Michela Rancan (2018). “Risk Taking on Behalf of Others. The Role of Social Distance”. Journal of Risk and Uncertainty 57(1): 81-109.
Pahlke, Julius, Sebastian Strasser and Ferdinand M Vieider (2015). “Responsibility Effects in Decision Making under Risk”. Journal of Risk and Uncertainty 51(2): 125-146.
Ploner, Matteo and Tobias Regner (2013). “Self-Image and Moral Balancing. An Experimental Analysis”. Journal of Economic Behavior & Organization 93: 374-383.
Pope, Thaddeus Mason (2011). “The Best Interest Standard. Both Guide and Limit to Medical Decision Making on Behalf of Incapacitated Patients”. Journal of Clinical Ethics 22(2): 134-138.
Umer, Hamza, Takashi Kurosaki and Ichiro Iwasaki (2022). “Unearned Endowment and Charity Recipient Lead to Higher Donations. A Meta-Analysis of the Dictator Game Lab Experiments”. Journal of Behavioral and Experimental Economics 97: 101827.
[1] https://www.vatican.va/content/francesco/en/encyclicals/documents/papa-francesco_20201003_enciclica-fratelli-tutti.html
[2] As a very weak substitute, I show that decision-making on behalf of others is influenced by knowledge about their autonomous wishes.
[3] In the jargon of the discipline, one assumes “common knowledge of rationality”; a classic is Aumann, Robert J (1995). “Backward Induction and Common Knowledge of Rationality”. Games and Economic Behavior 8(1): 6-19..
[4] In the meantime, a number of meta analyses of dictator games have been published: Larney, Andrea, Amanda Rotella and Pat Barclay (2019). “Stake Size Effects in Ultimatum Game and Dictator Game Offers: A Meta-Analysis” Organizational Behavior and Human Decision Processes 151: 61-72, Cochard, Francois, Julie Le Gallo, Nikolaos Georgantzis and Jean-Christian Tisserand (2021). “Social Preferences across Different Populations. Meta-Analyses on the Ultimatum Game and Dictator Game”. Journal of Behavioral and Experimental Economics 90: 101613, Doñate-Buendía, Anabel, Aurora García-Gallego and Marko Petrović (2022). “Gender and Other Moderators of Giving in the Dictator Game: A Meta-Analysis”. Journal of Economic Behavior & Organization 198: 280-301, Umer, Hamza, Takashi Kurosaki and Ichiro Iwasaki (2022). “Unearned Endowment and Charity Recipient Lead to Higher Donations. A Meta-Analysis of the Dictator Game Lab Experiments”. Journal of Behavioral and Experimental Economics 97: 101827.. But they are all confined to single design features, and do not give an overview over the entire literature.
[5] Meta-regression, dependent variable: fraction of endowment given, cons .261***, deserving recipient .115***, N = 445, adj.R2 .075.
[6] Meta-regression, dependent variable: fraction of endowment given, cons .291***, recipient endowment (as a fraction of the dictator’s endowment) −.213***, N = 445, adj.R2 .052.
[7] Actually, the amount prisoner dictators share is not significantly different from the amount that students give.
[8] Linear mixed effects regression, dependent variable: fraction of endowment given, cons .265***, recipient charity .110*, N = 124, p model .0308.
[9] To the charity “Unite for Basic Needs”, which serves orphans.
[10] Participants had the choice between keeping the complete endowment and donating 50% to Unite for Basic Needs.
[11] In a linear probability model, we explain the decision to donate 50% of the endowment (cons .607***) with the risk that the donation is subverted (-.223+), the availability of insurance (.034) and the interaction of both explanatory variables (.300*), N = 229. Hence the net effect of the risk that the donation is subverted plus the interaction effect is even positive, and not significantly different from zero. Hence the availability of insurance neutralizes the detrimental effect of the risk.