Autonomy and payment
One of the core values behind the imperative of non-exploitation in research is the value of human autonomy. The principle of respect for autonomy is a cornerstone of modern research ethics (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research
1979; Faden and Beauchamp
1986; Emanuel et al.
2000; Beauchamp and Childress
2001). It is also a bedrock of a fundamental ethical requirement for research, namely informed consent that serves to “ensure not only that individuals control whether or not they enroll in clinical studies”, but also that “they participate only when doing so is consistent with their values and interests” (Emanuel et al.
2000, p. 2706).
The principle of respect for autonomy has a high relevance for the practice of payment for research subjects as one of the most commonly expressed concerns is that payment can be coercive or constitute an undue inducement (undue influence), thus compromising the validity of informed consent (e.g., Macklin
1981; Faden and Beauchamp
1986; McNeill
1997; Dickert and Grady
1999,
2008; Largent et al.
2013; Gelinas et al.
2018,
2020). It is claimed that very high or (for other reasons) overly attractive payment may undermine the capacity of individuals to make autonomous decisions regarding study participation, by compromising their voluntariness, ability to adequate understand and assess research risks and benefits, or by “forcing” them to make choices against their “better judgment or deeply held beliefs” (Council for International Organizations of Medical Sciences
2016, p. 54).
Although there is still a substantial disagreement in the literature about whether an offer of payment may be perceived as coercion, what exactly constitutes undue inducement, and whether money can distort or compromise autonomous decision-making (Macklin
1981,
1989; Wilkinson and Moore
1997; Grady 1999,
2001,
2005; Grant and Sugarman
2004; Emanuel
2004,
2005; Wertheimer and Miller
2008; Klitzman
2013; Largent et al.
2013; Resnik
2015,
2019; Belfrage
2016; Largent and Lynch
2017a,
2017b; Millum and Garnett
2019), existing guidelines and regulatory documents disallow payment, which may unduly seduce people to consent for participation (Largent and Lynch
2017a). Thus, even though overly attractive payment offers could enhance timely recruitment by “alluring” a sufficient number of adequate individuals to join and stay in a study even against their better judgment, there is a regulatory consensus that such payment should not be offered.
The ethical unacceptability of such an overly attractive payment stems from the respect for autonomy. However, it also finds support in the principle of social beneficence. Common sense and empirical evidence, although still limited, suggest that attractive payments may have a negative impact on scientific value of research, as it may encourage potential and/or actual participants to withhold or misrepresent information, which are critical for their recruitment eligibility or continuation in research (Bentley and Thacker
2004; Dresser
2013; Dickert
2013; Devine et al.
2013,
2015; Largent and Lynch
2017a; McManus and Fisher
2018; Lynch et al.
2019)
. Such a concealment, fabrication or falsification by participants create risks for participants, but also for research resources and the integrity of research data as it can bias the results and undermine the validity of a study (Lee et al.
2018). And—if it occurs frequently enough—it may jeopardize the whole research enterprise. Therefore, assuming—what stills needs to be explored empirically—a positive correlation between the prevalence of deceptive behaviors and the attractiveness of payment (which depends not only on the amount, but also on payment method and timing), the principle of social beneficence provides ethical reasons for employing payment strategies which do not involve overly attractive payment schemes.
The principle of respect for autonomy has fueled much of scholarly discussion on unethical nature of overly attractive payment. However, little attention has been paid to that principle in the context of no-payment and underpayment. Undoubtedly, the lack of resources to cover direct or indirect costs associated with the participation may constitute a barrier for individuals, who otherwise would be willing to take part in research (American Cancer Society Cancer Action Network
2018; Nipp et al.
2019a; Largent and Lynch
2018; Bierer et al.
2021). The removal of these barriers by covering the relevant expenses is important for promoting potential subjects’ autonomy, because it enables individuals to exercise their free will to contribute to the development of science by serving as research subjects. Thus, the principle of respect for autonomy supports paying subjects recompenses, which make the participation in research a cost-free and revenue-neutral activity.
Finally, what is rarely observed, the principle of respect for autonomy provides a general support for public policies which allow remuneration of research subjects for their contributions—both in the form of reward and price—as it calls for respecting people’s right to decide freely in what practices and activates they what to engage in for the sake of earning their living. As Wilkinson and Moore note, “denying people the option of taking inducements reduces their freedom, since it removes an option that they prefer to the alternatives” (Wilkinson and Moore
1997, p. 377). Admittedly, biomedical research enterprise does not aim at broadening the scope of subjects’ freedom or autonomy. Nevertheless, it should not restrict people’s choices without a good reason. Participation in socially valuable biomedical research, whether or not one is paid for it, is neither morally wrong nor bad for the person concerned. On the contrary, it is considered at least as good as engaging in any other socially valuable and risky service or work. Therefore, there are no ethical grounds (either paternalistic or non-paternalistic) for depriving competent individuals an opportunity to serve as research subjects in exchange for money, provided that their decision to participate in a given study is autonomous (i.e., based on comprehensive and adequately understood information, and free from unduly controlling influences).
Moreover, some studies suggest that remuneration can in fact play a positive role in prospective subjects’ consent process by reducing the therapeutic misconception and highlighting research risks. Money might send a message to participants that they take risks and burdens for the sake of the benefit of science and society and “should be compensated for it, which would not occur if they were … expected to benefit from it.” (Glannon
2006, p. 252; also Dickert and Grady
1999; Grady
2001,
2005; Menikoff
2001; Largent and Lynch
2017a). Additionally, some studies suggest that payment can enhance autonomous decision-making by drawing prospective subjects’ attention to research risks and inconveniences (Cryder et al.
2010; Largent and Lynch
2017a; Fisher et al.
2019).
Justice, fairness and payment
Another ethical principle that has a direct relevance for the practice of paying research subjects is the principle of justice, both in its distributive and commutative dimensions. Many scholars worry that payment may be more attractive to individuals of lower SES, and thus offering payment for participation may result in unfair distribution of research benefits and burdens across the general population (e.g., Maclin
1981,
1989; Faden and Beauchamp
1986; Ackerman
1989; McNeill
1997; Grady
2005; Dickert and Grady
2008; Gelinas et al.
2018). The worse-off will shoulder a disproportionate share of burdens of research, while the benefits will accrue primarily to the better-off. A different, yet linked, concern relates to the risk of exploitation of research participants, especially those of low SES (Lemmens and Elliott
1999; Shamoo and Resnik
2006; Elliott
2008; Elliott and Abadie
2008; Abadie
2010,
2015; Stones and McMillan
2010; Phillips
2011a; Resnik
2015,
2019; Largent and Lynch
2017b; Gelinas et al.
2020; Bierer et al.
2021; MacKay and Walker
2021). Exploitation in research is about the unfair distribution of goods that arise from interaction between researchers and participants (Wertheimer
2008). “One party gets too little, while the other gets too much. Often, but not always, the unfair distribution arises because one party to the interaction is in a weak position, due to poverty, ignorance, or extreme urgency, which the other party can take advantage of, offering few benefits” (Emanuel
2004, p. 101). Economically disadvantaged individuals are in need of money and they tend to value a specific amount of payment more. For many of them accepting an offer of unfairly low but still
en bloc beneficial payment for participation might be a reasonable choice. After all, even small payment is better than no payment at all, although it constitutes a (mutually beneficial and consensual) exploitation.
These concerns are highly important, but—as it is shown below—they do not exhaust the role of justice/fairness in shaping an ethically sound payment practice. Moreover, in order to fully understand their consequences for ethics of paying research subjects, it is essential to note two things.
Firstly, biomedical research is a social practice embedded in specific social reality shaped by its historical and cultural roots, reigning power- and economic relations, and normative fabrics. This social reality is marked by profound social and economic inequities (within concrete societies as well as at the level of international community), which should be mitigated, and ideally eliminated by broad social reforms. Research enterprise should not entrench or exacerbate these background social injustices. However, it is neither designed for nor capable of actively fighting them, especially against participants’ poverty, unemployment or lack of access to high-quality healthcare (Fisher
2019). Justice in research requires treating all research participants fairly and equitably, also when it comes to payment for their participation. But it does not require paying participants of low SES in order to alleviate or remove hardships of their position. Such a positive impact of research payment on subjects’ social or economic condition is laudable, and even desirable, but it is not a demand of justice.
Moreover, contrary to views of some commentators, the commutative justice per se does not require offering payments to research subjects. Largent, Emanuel and Lynch claim that “when goods and services are not indented as gifts, failure to pay for them is a problem: we call it theft” (
2019, p. 1), thus suggesting that this is exactly what happens when participants are not fairly paid for their contribution to the common good. Despite its rhetorical attractiveness, this claim cannot be accepted as it rests on two mistakes. Firstly, it ignores the power of consent. Valid consent of a prospective subject for using her body for research purposes transforms theft into gift, lease, rent, work or other consensual relation with a researcher. Second, it forgets that it takes two willing parties to change provision of goods or services into transaction, i.e., exchange of goods and services in return for money. Thus, when a researcher is unable (e.g., due to budgetary constraints) or uninterested in offering payment for prospective participants (e.g., due to availability of sufficient number of unpaid subjects), or when a prospective participant is genuinely willing to contribute to the development of science without any remuneration and reimbursement, there is no ground for transaction. And there is nothing essentially unfair in allowing researchers and altruistically motivated participants to engage in scientifically and socially valuable biomedical research.
Payment is a recruitment incentive justified and prima facie required by social beneficence, not by justice or fairness. However, when it is to be offered, the offer should satisfy the requirement of distributive and commutative justice.
It is commonly accepted that the principle of justice requires distributing burdens and benefits of study participation in such a way that no segment of the population is unduly burdened by research or denied its potential or actual benefits. Recruitment criteria should reflect the scientific purpose of the study, not target populations which are considered “easy to recruit” “simply because of their easy availability, their compromised position, or their manipulability” (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research.
1979). Thus, researchers must neither exploit the vulnerable, in particular economically disadvantaged, nor exclude without good reason those who stand to benefit from study participation. They should strive to recruit an adequate cross-section of the population in order to spread research burdens and benefits fairly across the population. In other words, an adequate diversity of gender, race, ethnicity, age and SES in research should be sought (Geller et al.
2011; Kwiatkowski et al.
2013; Heller et al.
2014; Winter et al.
2018; cf. Dickert
2009; Bierer et al.
2021). It is worth noting that the last postulate finds an additional support in the principle of social beneficence as it enables the generalization of knowledge to be gained in research, thereby enhancing research scientific and social value (Wilkinson and Moore
1997; Grady
2005; Resnik
2015; Largent and Lynch
2017b). Thus, justice encourages payment schemes which have a potential of making distribution of research burdens fairer. And justice is against schemes that can deepen social inequalities, lead to unfair social distribution of research burdens and benefits or exploitation.
What should research subjects be paid for and in which amount to make payment consistent with demands of justice? Firstly, justice-related considerations provide an ethical justification for paying research subjects a recompense for direct costs related to research participation. By making participation in research a cost-free activity, recompense removes—at least some—economic barriers for participation, thus equalizing opportunities for all willing individuals to contribute to the development of science, no matter their SES. Moreover, as studies already referred to suggest (Unger et al.
2013,
2016,
2019; American Cancer Society Cancer Action Network
2018; Nipp et al.
2019b; Chino and Zafar
2019), reimbursement of direct expenses may improve access to potentially beneficial studies for patients, who otherwise would not be able to participate due to their low SES or disease-caused greater sensitivity to economic burdens associated with research. Thus, the reimbursement of direct costs has a potential to enhance justice in research not only by reducing inequities in access to research for those who otherwise could not afford it, but also by contributing to the fair distribution of clinical benefits associated with participation.
Secondly, the principle of justice provides a strong ethical reason for disallowing recompenses for the loss of reasonably expected profits, especially lost wages. Such recompenses lead to differential payments between participants—individuals who receive higher wages get higher recompense; those who are lower paid by employees—lower recompense, and those who are unemployed—no recompense at all. This is consistent with a norm of equity, as it makes participation a revenue-neutral activity for all, no matter how much they earn. Nevertheless, it is very likely to reinforce unfair distribution of research risks and benefits between different social strata by prompting researchers to make savings by drawing research participants from “cheaper” populations, especially from the unemployed and the low-paid (Dickert and Grady
1999; Resnik
2015).
Thirdly, justice supports remuneration for participants’ contributions to the development of science and society, provided it is fair, i.e., equitable, adequate, and non-exploitive. Remuneration is equitable when it does not violate the norm “equal pay for equal work” (Dickert and Grady
1999; Resnik
2015; Gelinas et al.
2018,
2020; Persad et al.
2019). Equal time, efforts, burdens and risks associated with participation deserve equal remuneration (measured in market value, rather than in nominal value in case of multi-site studies conducted in different settings). However, when different groups of participants in a study are expected to make different contributions, equitable remuneration should reflect differences in their input (Persad et al.
2019). This is consistent with equal respect and concern for each and every individual and it prevents discrimination.
Remuneration is adequate when it is proportionate in value to the value of participants’ contribution to the study. The latter necessarily depends on various factors which determine how time- and effort-consuming, burdensome and risky give research project is. For example, when a study involves 2 visits, each lasting 60-min and requires filling a questionnaire and giving a blood sample for further analysis, and another project involves 4 visits of the same length and level of associated risks and burdens, participants of the latter study—ceteris paribus—should be offered remuneration of double the remuneration offered to the participant of the first project.
All guidelines and regulatory documents referred to above follow this normative logic by recommending or permitting paying research participants an “appropriate” or “proportionate” or “just and fair” remuneration for the time spent and other inconveniences resulting from the study participation (Council of Europe
2005b, sec. 64; Health Research Authority
2014, par. 3.4; Council for International Organizations of Medical Sciences
2016, p. 53; Food and Drug Administration
2018; National Health and Medical Research Council
2019, par. 1.1). Only a few, however, mention the level of research risks among factors which should be taken into consideration when determining an adequate amount of payment. Moreover, those that do so provide conflicting instructions on this matter. For example, the CIOMS Guidelines
expressis verbis state that “the level of compensation should not be related to the level of risk that participants agree to undertake” (
2016, Commentary on Guideline 13). In contract, the Australian National Health and Medical Research Council’s guideline reads: “In cases where risk may be considered as a factor in determining payment, payment of participants based on the degree of risk associated with the research is not prohibited, so long as there is evidence that a participant’s ability to provide valid consent is not likely to be compromised” (
2019, par. 1.3.). Also the UK Health Research Authority “sympathises with the view that not to allow payments on the basis of risk would be unduly paternalistic in the absence of evidence that the participants’ ability to provide valid consent would be compromised” (
2014, par. 3.1.).
These regulatory variations mirror the lack of consensus regarding risk-based payments among research ethicists (Grady
2001; Menikoff
2001; Jones and Liddell
2009; Saunders
2009; Grimwade et al.
2020; Lynch and Largent
2020; Jamrozik and Selgelid
2020). Although a detailed analysis of this issue goes beyond the scope of this analysis, three following arguments provide strong reasons for arguing that adequate remuneration for research subjects should be proportionate also to the level of risk involved in participation: (i) an argument from consistency, which notices that risk-based remuneration is accepted in many non-research contexts, e.g., in high-risk professions (Menikoff
2001; Jones and Liddell
2009); (ii) an argument from the nature of human “guinea pigging” which claims that the assumption of risk is often an essential contribution of research participants (Menikoff
2001; Różyńska
2018; Malmqvist
2019); (iii) an argument for public trust, which argues that paying proportionally to the incurred risk meets expectations regarding the fair treatment of a significant portion of prospective subjects as well as researchers, REC/IRB members, and other members of the research community (Czarny et al.
2010; Ripley et al.
2010; Largent et al.
2012; Grimwade et al.
2020), thereby enhances public trust in research at large.
Finally, remuneration is non-exploitive when it is not lower than a socially accepted payment which is (or would be) offered for a similarly time- and effort-consuming, burdensome and risky activity, outside research context in the same setting (Gelinas et al.
2018; Largent and Lynch
2017b). Although space does not permit discussing in details how a proper amount of such defined payment should be calculated, a proposal widely advocated in the literature should be mentioned here, namely a minimum hourly wage benchmark. Numerous authors and guidelines suggest that the amount should be based on the minimum hourly wage in the region or country as a point of reference (Council for International Organizations of Medical Sciences
2016, Commentary to Guideline 15, 53; National Health and Medical Research Council
2019, Appendix 1, 7) with augmentations for particularly burdensome procedures (Ackerman
1989; Dickert and Grady
1999; Grady
2005; Gelinas et al.
2020), risks involved (Menikoff
2001), and even other additional benefits (cf. Anderson and Weijer
2002). This proposal has three advantages. It is relatively easy to implement as it provides a clear method of setting the baseline amount of reimbursement for research participants. It allows for keeping payments sensitive to specific features of the project and subjects’ contributions by accepting relevant payment augmentations. And—since it sets the reference value of an non-exploitive remuneration relatively low—“more researchers could afford to pay a fair wage and fewer would be inconvenienced by a prohibition on unfairly low wages” (Phillips
2011a, p. 219). However, it also faces challenges. Offers of payment calculated in such a way would be attractive for prospective subjects with low SES, who have no reasonable alternatives to get engaged in better paid activities. But these offers would likely have no impact on recruitment of individuals who are better-off, i.e., who have better paid jobs or capacity to make more money outside research context, thus potentially biasing the subjects’ recruitment. In contrast, the average-wage benchmark for non-exploitive remuneration, mentioned by some authors (Phillips
2011a) would seem fairer as it would make payment offers reasonably attractive to both individuals of SES and to those better-off, thus promoting a fair distribution of research risks and benefits between different social strata.
Fourthly, the principle of justice does not prohibit remunerations going beyond what constitutes equitable, adequate, and minimally non-exploitive payment—i.e., remunerations driven by a market-driven forces. Although such payments are conceptualized as “price” for desired services, in order to avoid exploitation, they should not be lower than remuneration viewed as “reward”. They may, however, be higher than “rewards” and they may vary within the same study on the basis of salient characteristics of particular groups of participants (i.e., their age, sex, race, ethnicity, rare clinical status etc.). Offering disproportionally high and/or differential remuneration does not violate requirements of justice insofar as the payment is designed to help the study to meet its social and scientific goals by enhancing recruitment and retention of the necessary category of subjects, and it does not reproduce or reinforce wider social inequities and injustices, e.g., racial biases or class differences (Persad et al.
2019).