Solidarity can be defined, in behavioural terms, as the willingness to help people in need who belong to the same group and who are victims of negative events such as unforeseen illness, natural catastrophes, unemployment, workplace accidents, and so on. It involves a constitutive element of altruism—a willingness to help people in need—and an element that restricts this type of behaviour to a certain community of people, namely those who belong to the same group, as well as another element that restricts solidarity to certain situations, namely those in which the victim suffers from a negative event. While there are certainly other definitions of solidarity, for instance as a legal principle or a philosophical concept, a behavioural definition permits us to investigate the existence, extension, location, and contours of solidarity, which is necessary for any empirical investigation into the relationship between solidary behaviour and the law.
The first element of solidarity is altruism. There is considerable evidence that many individuals care for the wellbeing of others and are often willing to spend money, time, and effort to help others. While we have all certainly witnessed people donating money to charities, helping people in need, or devoting time and energy to the wellbeing of the community, these acts might have been caused not by a desire to help others, but by other external factors. These might include the desire to bolster one’s reputation, to induce reciprocity from others (ensuring their help in return), or even to profit monetarily (perhaps through tax benefits or by gaining new customers or business partners). How can we be sure that people act altruistically when these factors are absent, and that the desire to help others is the only possible motive guiding such types of behaviour?
Laboratory experiments that implement the famous ‘dictator game’ achieve this goal. In these experiments, subjects—most often students, although not necessarily—interact with each other anonymously through computer terminals. They are randomly matched in pairs, and while one person receives an endowment of EUR 10 (the ‘dictator’), the other receives nothing (the ‘recipient’). The game simply involves the dictator deciding how much of the EUR 10 to give to the recipient. Subjects then wait in their cabins until they are called to receive an amount in cash that corresponds to how much they kept for themselves (the dictators) or to how much was transferred to them (the recipients). This one-off anonymous interaction excludes by design reputational concerns, either with respect to one’s image before others outside the lab or before those with whom they interacted in the lab. Their transfer is not made public, and the recipient, although informed of how much she or he received, cannot know who made that transfer; and the subject making the transfer cannot bolster her or his reputation with that act. Secondly, since the interaction occurs just once, and the recipient cannot react in any way to the subject who just transferred (or did not transfer) some money to her or him, there cannot be any reciprocity. Thirdly, there is no tax benefit in sight, nor any way through which the subject making the transfer can profit monetarily from that act.
Participants transfer, on average, between two and three euros. This does not mean that everybody behaves partially altruistically. Roughly half of the people behave purely selfishly by not transferring anything at all. The other half transfers between one and four euros, displaying a degree of altruism. A small minority chooses to divide the endowment equally with the recipient, and virtually no one transfers more than half of the endowment. From this, we can conclude that even among strangers there is altruistic behaviour that is widespread and relevant, but not universal nor always present.
There have been hundreds, if not thousands, of dictator games that have taken place in very different countries in the past four decades. These results are robust, observed across different cultures and among students and non-students alike. While there might be some influence of culture on how far people behave altruistically in such a game, it is not very large, and a certain proportion of individuals across extremely diverse communities consistently display such behaviour.
The second element of solidary behaviour is group membership. People can act altruistically towards anyone, but they tend to be solidary particularly with those who belong to a group they identify with. Solidarity is expected from those connected with each other by something they share, such as a common history, geography, class, interest, problem, or feeling, and this particularistic dimension is inherent to solidarity. It is present, for instance, in the solidarity shared by workers of the labour movement, by women in the #MeToo movement, among many others. In fact, experimental evidence reveals that transfers in the dictator game are higher when the social distance between the dictator and the recipient is smaller. In a network of college students, people give more to a friend than to a stranger, and this does not depend on the fact that the person knows the identity of the recipient—people also give more to an unknown friend than to a stranger. People further give more to a friend than to a friend of a friend, and more to a friend of a friend than to a friend of a friend of a friend, an effect called the ‘inverse distance law of giving’.
There is hence a higher degree of altruism with members of the same group, or with people that are closer to oneself.
The last element of solidary behaviour is risk. People can be altruistic towards those in need, independent of the cause of the need. People can be altruistic towards those who have less than they have themselves. However, people can also be solidary with those who are unfortunate enough to suffer a negative event (sudden illness, unemployment, accident, etc).
Experimental evidence for this type of behaviour is observed in the solidarity game, a game that is very similar to the dictator game, but that involves risk. In this game, three participants can earn each a certain amount, for instance EUR 10 with a high probability, or they can earn nothing with a low probability. Before the risk materializes, they are asked how much they are willing to give, if by chance they are selected to earn the EUR 10, to those who by chance are selected to earn nothing. Results reveal that, remarkably, almost 80% of the participants choose to gift positive amounts. On average, people transferred around EUR 2.5 to a single loser, and EUR 1.5 to each of the two other participants if they both earned nothing. If, in a group of three people, only one is the victim of a negative event, then this person can expect to earn EUR 5 in total, with the other two members transferring EUR 2.5 each to this person while keeping EUR 7.5 for themselves in a very equitable final distribution (5, 7.5, and 7.5).
These results show that solidarity is perhaps more widespread than mere altruism, as 80% of the people made positive transfers in the solidarity game while around 50% of the people made positive transfers in the dictator game that involved no risk of income loss. The amount earned by the victim of the negative event, that is, by the one who by chance is selected to receive nothing, is also higher in the solidarity game.
There is therefore substantial evidence that many people are solidary even in the absence of a legal obligation to be so. In the dictator or solidarity game, there is no legal obligation to transfer any amount to the other person, nor any legal benefit from doing so. The law can, however, certainly interact with the tendency toward solidarity through different channels and in different forms. While there is no definitive answer to how this happens, some hypotheses can be gleaned.
First, the law can crowd out human solidarity. The social security system, with its institutionalized form of insurance, performs some of the functions performed by interpersonal human solidarity. It provides benefits to the members of the group—in this case, the insured individuals, normally workers in the same country—when they suffer from negative risks. Unemployment insurance provides them with a monthly payment once they lose their jobs. Workers’ compensation provides them with benefits if they suffer an accident in the workplace (or, depending on the country, also on the way to or back from the workplace). Health insurance provides them with medical treatment if they fall sick. Once the victim is to an extent taken care of by the social security system, there is less need for fellow citizens to take care of this person, as she or he is not as much in need as he or she would have been in the absence of the social security system.
In fact, when introducing formal insurance in a community, risk coverage in the community might even decrease. Insurance payouts often take the place of informal transfers, and reduce people’s willingness to provide informal support. While victims benefit from the insurance payment, they receive less informal support from other members of their community, and the result can be lower payments received by victims—as already observed in reality, albeit in rare cases where formal insurance is not well developed and functional. Can we then conclude that social insurance has, or can have, a negative and undesirable effect on interpersonal solidary behaviour? This is probably not the right question to ask since social insurance often makes solidary behaviour unnecessary, as many risks are covered, especially when the social insurance system is generous and widespread.
But does social insurance crowd out solidary behaviour in cases where social insurance cannot fully cover the risk, and where there is still need for solidary behaviour?
Consider the case of a person who, unable to work, receives a pension that, to some extent, maintains her income. If she has health problems, then health insurance will cover this other risk. If she or he suffers from, say, depression, she or he might well have access to adequate health care, but this can never perfectly restore her well-being unless other individuals are solidary and relate to this person, incurring costs in, for example, visiting the person, talking to her or him, going for walks together, etc. This residual risk requires solidary behaviour, as the law and the social insurance system can never fully cover all risks, nor dispense with the need for interpersonal solidary behaviour for the maximization of overall social welfare.
Second, the law can crowd in human solidarity. It can create incentives for people to exhibit solidarity more frequently, and it can express how solidarity is a value protected by the law, fostering solidary behaviour indirectly. It is probably easier for the law to directly incentivize solidary behaviour by providing tax incentives, direct monetary payments, or some other benefit to those helping others in need. But the law does not need to offer direct monetary payments; instead, it might simply make solidary acts easier, be it by reducing the costs associated with those acts or by increasing their reach and effectiveness. Whether the resulting behaviour can still be defined as solidarity is open to question, but it certainly benefits the victim. In the case of the person discussed above, people willing to help her or him could simply be paid an amount of money that, even if not sufficient to elicit helping behaviour by itself, would help to achieve this goal by ‘going the second half’. Further incentives could include free public transportation to and from the place where the person is, access to instruments or means helpful for the solidary act, or dissemination of information on how to help others suffering from certain specific events. Alternatively, the law can relieve those who devote time and energy to help others from military duties, electoral duties (such as serving as an official during elections), and so on.
It is a fact that the law, and especially the social security system, interacts with solidary behaviour that would have existed to some extent anyway in the absence of the law. This type of behaviour does not vanish once legal forms of solidarity are introduced, for the law can never fully perform this function. However, the law can crowd out natural solidary behaviour, potentially leaving the victim of a negative event in a worse position. An important avenue for research would be to investigate further the different cases in which the law takes on some of the functions of human solidarity: when this happens, to what extent, and with which consequences.