Blood donor motivation: what is ethical? What works?
Alois Stutzer, Department of Business and Economics, University of Basel, Peter Merian-Weg 6, CH-4002 Basel, Switzerland
The retention of previous donors and the recruitment of new donors is a serious challenge for many blood donation services in their effort to prevent blood shortages. More and more services make use of some sort of donation incentives. However, the use of (material) incentives to motivate blood donors is fiercely controversial, and there is a longstanding (ethical) debate about whether it should be allowed that donors receive material rewards. Interestingly, this debate is dealt with in almost complete absence of systematic empirical evidence on the effectiveness of material incentives in encouraging people to donate. In this article, we argue that the discussion on what is ethical in motivating blood donors should be enriched with empirical evidence based on field experiments. We confront the Titmuss controversy with recent results from an experiment administering lottery tickets as a motivation device. Moreover, we take up a neglected phenomenon in the study of blood donors: many nondonors are not principally against donating blood; they have just never made up their mind about becoming active blood donors. We propose active decisions as a mechanism to transform latent prosocial preferences into actual prosocial behaviour.
Many blood donation services rely on voluntary, nonremunerated donations and thus on the prosocial motivation of their donors.1 Despite the inherent free-rider problem of this policy, the arrangement so far seems to have worked satisfactorily most of the time. However, three problems have emerged recently that increase the risk of blood shortages in the future. First, innovations in surgery and oncological therapies have led to more aggressive medical treatments in cases that were previously thought inoperable or incurable, increasing the number of transfusions and requiring larger amounts of blood . Second, there is a general tightening of donation eligibility criteria, such as stepped-up travel restrictions or restrictions on past blood recipients because of Creutzfeldt–Jakob disease. Third, there are widespread seasonal shortages because of reduced supply, particularly in summer and early winter .
The question arises how additional people can be recruited as blood donors and how previous donors can be motivated to donate blood on a regular and frequent basis [5,6]. A substantial body of research has approached this question by studying the motives of blood donors [7–9] often with the idea to focus donor appeals on these motives. While according campaigns rarely raise ethical concerns, their impact is usually not assessed and less and less considered sufficient to guarantee an adequate supply of blood. Instead, more and more donation services make use of some sort of donation incentives. In contrast to appeals, the use of (material) incentives to motivate blood donors is fiercely controversial and there is a longstanding (ethical) debate about whether it should be allowed that donors receive material rewards. Interestingly, this debate is dealt with in almost complete absence of systematic empirical evidence on the effectiveness of material incentives in encouraging people to donate.
In this article, we argue that the discussion on what is ethical in motivating blood donors should be enriched with empirical evidence based on field experiments. In section Titmuss in the field, we confront the Titmuss controversy with recent results from an experiment administering lottery tickets as a motivation device. Section Active decisions and the never-thought-about-nondonors takes up a neglected phenomenon in the study of blood donors: many nondonors are not principally against donating blood; they have just never made up their mind about becoming active blood donors. We propose active decisions as a mechanism to transform latent prosocial preferences into actual prosocial behaviour.
Titmuss in the field: experimental evidence on the effects of material incentives
The Titmuss controversy
Common experience suggests that people are more likely to pursue an activity if it is rewarded with a selective incentive. Accordingly, not only many economists but also practitioners in the field predict that selective incentives increase blood donations. Yet there is a deep-rooted scepticism about using incentives in blood donations even on a temporary basis. That scepticism is based on the conjecture that using incentives may undermine the motivation to donate blood. It is associated with Richard Titmuss , who famously argued, ‘[f]rom our study of the private market in blood in the United States, we have concluded that the commercialization of blood donor relationships represses the expression of altruism, erodes the sense of community, […]’ (p. 245). In fact, his book ‘The Gift Relationship: From Human Blood to Social Policy’ still today is the starting point of almost any debate on the use of interventions to motivate potential blood donors. It serves as the moral ground of skeptics of the use of incentives for encouraging donations. When Titmuss wrote the book, he wanted to express his concerns about the commercialization of society . The comparison of the British and the American blood procurement system mainly served as a showcase to make his arguments. His book spurred a fruitful discussion on the ethics of markets vs. voluntarism from a communitarian perspective [12,13].
In sum, there are two countervailing hypotheses about the consequences of selective incentives. Rewards render blood donations more attractive and people donate more. In contrast, the Titmuss hypothesis predicts that pricing undermines people’s sense of community and cohesion because people are deprived of the opportunity to express altruism and no longer face the moral conflict and challenge to answer the question about their obligations to strangers. This latter argument is related to recent research in psychology and economics that proposes that incentives can lead to less prosocial behaviour. If rewards are perceived as controlling, incentives may undermine prosocial motivation. This is often referred to as the motivation crowding effect [14–16]. Accordingly, people donate less if rewards are offered.
There are two basic methods of studying the consequences of selective incentives on blood donations. The prominent method usually applied directly asks people about their attitudes towards specific motivation devices or whether they would change their intentions to donate blood if they would be offered some specific incentive. These questionnaire studies are relatively easy and cheap to administer and provide comparative results about attitudes towards different motivation devices. The alternative and rarely used method is to study the effectiveness of incentives in controlled natural field experiments.
A prominent study of the former kind is the Retrovirus Epidemiology Donor Study (REDS) . Conducted in 1998, it involved eight blood donation centres in the United States. Survey responses of more than 45 000 people were analysed. Previous donors were asked whether they would be encouraged, discouraged or remain neutral if they were offered some incentives. Offering a lottery ticket was found to lead to a net encouragement of between 1% (for donors older than 56) and 32% (for young donors aged 25 or less). The highest rate of net encouragement was 61% for blood credits and free cholesterol testing.
However, there are some fundamental difficulties in predicting the effectiveness of donation incentives based on these results. Responding to survey questions is ‘cheap talk’. Answering nonbinding questions anonymously involves no consequences. Responses might then be driven by social desirability and/or the expression of some general attitudes . Independent of strategic or conscious misreporting, it is often very difficult to imagine how one would behave in a situation when actually exposed to a specific incentive.
A case in point is free cholesterol testing as a motivation device. A substantial discrepancy in its effectiveness has been documented between the findings in the literature based on surveys (mentioned earlier) and the results of two large-scale field experiments . In both experiments, the first directed towards recruiting new donors and the second set-up for recalling previous donors, offering free cholesterol testing did not significantly increase blood donations overall.
We conclude that the impact of incentives on behaviour should be examined directly based on randomized trials.
Some evidence from field experiments
Studies that measure individual behaviour and apply a randomized design are rare, unfortunately. Moreover, they often involve nonrepresentative samples and do not analyse long-term consequences. From the studies we are aware of, no consensus finding emerges. One finds a doubling of the attendance rate in a campus blood drive if solicitors offered coupons for merchandise and participation in a raffle (number of subjects = 80) . In contrast, another study finds a negative effect on participation in a preparatory health examination when a cash payment was offered (number of subjects = 262) . The negative effect is statistically significant for female students.
There are only two studies based on large representative samples that address the causal effect of specific material incentives on donor retention so far. The first study applies an individually randomized trial design for a sample of 6919 first-time donors . The treatment group got a T-shirt offered while the control group solely got the standard invitation to donate a second time (recruited either by phone or by email). No economic difference in the return rates between the two groups is found. With incentives, 20·5% returned at least once during the test period, while without incentives the return rate was 20·6%.
The second study examines the effectiveness of lottery tickets as selective incentives . It is based on a pool of over 10 000 blood donors. Thereby, the short-term consequences when the incentives are in place are analysed as well as the long-term consequences after they are removed again. This is important because the motivation crowding effect might remain hidden as long as an incentive is offered (owing to an overcompensating effect). Any negative motivation crowding effect is uncovered once the incentives are removed.
The outcome variable captures whether people showed up at the blood donation centre after having received an invitation. In the experiment, one randomly chosen group of donors is offered a lottery ticket. One group received an appeal with the identical message as in the incentive treatment, but no reward is offered. Another group is not offered any incentives. It serves as a control. Importantly, the subjects were not aware that a study was being conducted. The invitations differentiated by treatment were mailed privately to the donors.
The overall picture that emerges from the experiment is that lottery tickets as material incentives have no general negative effects, neither short term nor long term. To the contrary, when looking at the overall experimental outcomes, offering a lottery ticket increases usable donations by 5% points over a baseline donation rate of 42%. Offering the lottery ticket is significantly more effective than appealing to donors without giving incentives. Moreover, the treatment effects vary between subsamples in interesting ways. It is mainly subjects who donated only infrequently in the past who respond to the lottery treatment and increase donations. There is essentially no incentive effect on donors who have shown a strong preference for donating blood in the past. Finally, no evidence is found that exposing donors to incentives would have reduced their prosocial motivation. After the experiment, donations generated from people who were incentivized during the experiment are no lower than those generated from people who received the rewards unexpectedly.
Active decisions and the never-thought-about-nondonors
Many potential donors have never contemplated becoming active donors. In a questionnaire study with young nondonors in the Netherlands , 61·7% report that they had never seriously considered becoming donors. In another survey of a regional blood donation centre in Switzerland , 42% of the nondonors had never thought about donating blood or did not feel addressed by the issue. This is surprising as in both countries there is no dearth of general appeals to donate blood. It suggests that some nondonors are not reached by the usual campaigns, not even in so far as they decide against becoming donors.
The idea of an active decision
How can specific prosocial behaviour be promoted if some people are not even against it but have just not considered it? In a recent study , we propose a decision framework where people are individually asked to either actively consent to or dissent from some prosocial behaviour. This is in stark contrast to some noncommittal appeals to behave prosocially that address everybody in the same way. Consider the issue of blood donation. With an active decision, people are confronted with a request to donate blood, to which they are expected to respond with either a ‘yes’ or a ‘no’. The active decision is supposed to induce people to deal with a specific prosocial behaviour and to make them aware of the social value of that particular behaviour. By doing so, active decisions are expected to contribute to the formation of issue-specific altruistic preferences, while simultaneously involving a commitment.
The hypothesis rests on insights in economics and psychology that preferences are partly formed in the process of decision making in unfamiliar choice situations. A pertinent example is the creation of nonuse values in contingent valuation studies . In active decisions, people are made aware of some particular issue and are induced to engage in cognitive evaluation and reasoning . In blood and post-mortem organ-donation decisions, the deliberation involves dealing with one’s own health and people are motivated to get over the denial and repression of their own mortality.
The behavioural consequences of confronting people with the decision whether to act prosocially may not be uniform but may depend on the degree of stability of the subjects’ altruistic preferences with regard to a specific issue. The effect of active decisions on behaviour, by its very nature, relies on the endogenous formation of preferences inherent in the process. We therefore expect that active decisions will be more effective in influencing preference formation when people are unaware of the importance of a specific prosocial engagement and when their latent motivation to donate or contribute would otherwise remain dormant. In contrast, if people are well aware of a public good and have already made up their minds about their contributions to it, preferences can be expected to be rather stable and little affected by active decisions.
Some first evidence
In a field experiment, we studied the potential of active decisions to motivate voluntary blood donations . Our study was incorporated in a Red Cross blood drive at the University of Zurich. In total, more than 1800 students participated. They were not aware that an experiment was taking place when they were approached in a lecture. A representative of the donation service made a brief announcement regarding the blood drive, and a survey was distributed. We implemented three experimental conditions: In the ‘strong active decision’ treatment, the survey contained a page at the end asking the subjects whether they were willing to donate blood at one of the time slots mentioned on the information sheet. They had two possible choices: yes or no. If they answered yes, they had to say when they would show up for the blood donation. In the ‘weak active decision’ treatment, the last page of the survey was nearly identical; we merely added a third possible option, allowing them to state that they did not want to make a decision. In a control condition, there was no such page at the end of the survey. However, along with the survey, all students also received an information sheet, listing dates and times in the week when they could come to donate blood. To obtain a measure of the latency of prosocial preferences, we asked the students in the survey whether they felt they were sufficiently informed about the importance of donating blood.
Our results lend support to the hypothesis that active decisions can uncover latent prosocial preferences. We find that among students who indicated that they were not sufficiently aware about the importance of donating blood, the strong active decision treatment increased blood donations in the following week significantly relative to the control condition.
We suggest that the normative discussion about the right measures to motivate blood donors should be relaunched. This holds in particular in relation to the use of material incentives to encourage potential blood donors. The core open questions should be precisely specified. We are convinced that some key issues in the ethical debate about interventions to motivate blood donors can be addressed with systematic empirical research.
In this article, we emphasize the concerns with regard to the Titmuss controversy, i.e. material incentives might not be perceived as valuable rewards but rather undermine donors’ prosocial motivation. Modern social science research in psychology and economics provides many insights that can be linked to the Titmuss controversy. Motivation and self-signalling theories provide a (psychological) underpinning of the possible interaction between material incentives and prosocial motivation. Moreover, there is a body of research documenting a motivation crowding-out effect of extrinsic incentives [28,29]. This indicates that prosocial motivation is a delicate little plant. However, so far the conditions are not well understood under which extrinsic incentives crowd out intrinsic motivation and those under which the two sources of motivation can co-exist and jointly encourage prosocial behaviour.
Despite the endurance of the Titmuss controversy, there is only limited evidence on the interaction of material incentives and prosocial motivation in the realm of blood donation. We are convinced that based on randomized trials most relevant insights are possible. The tools are available, and donation services seem ready and willing to run field experiments. While surveys on donors and nondonors attitudes towards different kind of interventions are a valuable source of inspiration, the relative effectiveness of different motivation devices should be tested in the field looking at people’s behaviour.
In our own study, referred to in this article, we find that lottery tickets can serve as a motivation device to overcome the risk of a blood shortage during summer. In particular, infrequent donors respond positively to this material reward. Moreover, also in the long run, no negative effects on blood donations are observed. This indicates that some material incentives can well co-exist with prosocial motivation and encourage donations.
There are for sure limits to motivating previous donors to donate more often. Moreover, it seems also more difficult to recruit new donors . Whether it is because of a sensory overload, campaigns seem to reach people less today. Interestingly, many nondonors have never thought about becoming active donors. One future challenge is to attract this group of people.
We propose active decisions as an intervention to make people think about a specific prosocial activity. First evidence from a field experiment indicates that people who report that they are unaware of the importance of donating blood are more likely to donate if approached by an active decision intervention.
There are, of course, many more hypotheses about what works in motivating blood donors. Some of them are based on recent insights in cross-disciplinary research in psychology and economics; others are based on gut feeling of experienced practitioners in the field. We propose randomized trials to scrutinize them. New insights about motivating blood donors are necessary and possible to secure an adequate blood supply in the future.
While whole-blood donations are generally unpaid in developed countries , there are some countries where primarily plasma donations but also some whole-blood donations (e.g. in the USA and Germany) are paid .
No conflicts to declare.