Reciprocity and the duty to stay

ABSTRACT Some restrictionist arguments justifying the duty to stay as a means of addressing medical brain drain have relied on reciprocity as the moral basis for their policy proposals. In this essay, I argue that such reciprocity-based justifications for the duty to stay ignore crucial conditions of fittingness as relates to the funding of medical training.


Introduction 1
In recent debates on the medical brain drain, philosophers have focused on the moral underpinnings of various policy proposals intended to address the disparate health outcomes in wealthy developed nations and poor developing nations (Brock 2013;Brock andBlake 2017, 2014;Oberman 2013;Wellman and Cole 2011). These proposals suggest that developed nations ought to restrict immigration to prevent the harmful effects of mass departures of medical doctors from developing nations. Additionally, philosophers defend these proposals in various arguments for medical doctors' duty to stay in their countries of training when their services are desperately needed. These restrictionist arguments for the duty to stay variably rely on some widely accepted account of reciprocity. As it pertains to political communities, the more interesting cases of reciprocity are indirect or dispersed obligations owed to all of society.
In this paper, I reject Ferracioli and De Lora (2015) restrictionist arguments regarding the migration of medical doctors. Ferracioli and DeLora adopt Becker's (2005) account of reciprocity in defence of proposals for contracted medical training in developing countries. They conclude that medical students have a reciprocity-based duty to stay (temporarily) and serve their compatriots with their skills after graduation. I here endorse Becker's conception of reciprocity, but I reject the restrictionists' application of it. Thus, my paper proceeds as follows: § 1 outlines Ferracioli and DeLora's two reciprocity-based arguments for the duty to stay and for contractually reinforcing that duty. § 2 outlines Becker's account of reciprocity, four oversimplifications he identifies, and standards of fittingness and proportionality integral to understanding direct and indirect reciprocity. § 3 then explains how the restrictionists appeal to reciprocity in arguing for the duty to stay and the proposed contracts. Next, § 4 considers the moral significance of the errors restrictionists allude to, before § 5-9 turn to my main rejection of the restrictionists' application of Becker's account. I focus on the case of publicly-funded and privately-funded medical students in those sections. I argue that the restrictionists' application of Becker's reciprocity fails to meet a standard of fittingness. 2 First, some terminological housekeeping. I use 'source countries' to refer to the countries where these medical doctors complete their medical training and 'source states' to refer to the governing bodies of these countries. Thus, a typical example of a source country could be Ghana (Dare, Onajin-Obembe, and Makasa 2018). While I refer generally to theorists who defend the duty to stay as 'restrictionists' and any argument for that view as a 'restrictionist' argument, in this paper I specifically mean Ferracioli and De Lora. Furthermore, I use 'doctors' to refer to 'medical students' and 'medical doctors' interchangeably since the duty to stay maps onto both groups. I will only distinguish where a nuance in the restrictionist argument requires it.

Two reciprocity-based arguments
Restrictionists provide what I consider two distinct reciprocity-based arguments for the duty to stay. The error argument states that medical students and doctors, in their efforts to provide care, routinely inflict pain on their patients. Restrictionists allege that medical students are more prone to expose their patients to risks and highlight surgical procedures that inflict significant levels of pain and risk on patients. On this basis, restrictionists conclude that doctors, upon completion of their training, bear obligations of reciprocity. Since restrictionists acknowledge the permissibility of these exposures to risks, pains, and harms as inevitable features of training (Ferracioli and De Lora 2015, 606), I designate them as 'errors,' in contrast to malpractice or wrongings, and thus dub this the error argument. 3 This obligation of reciprocity is owed to their compatriots who might need medical care in the future, after their training (Ferracioli and De Lora 2015, pp. 605-606). From the claims of reciprocity, the error argument holds that source states must ensure that these doctors discharge their obligations of reciprocity, as by conditioning medical training on a contract to stay in the source country. The duration of required stay, restrictionists argue, should be no more than four years (Ferracioli and De Lora 2015, 610). They also state that although source states should institute the contracts, recipient states ought to be the enforcers of the contracts by denying entry to doctors who have not met the terms of their contracts.
In the second argument, restrictionists stipulate that a country institutes medical training to protect the basic rights to health care of citizens. They further note that when doctors emigrate, their talents and skills become unavailable to their compatriots left behind. Based on this empirical claim, restrictionists argue that doctors who migrate after their medical training harm their compatriots by leaving them vulnerable to health challenges. Therefore, doctors have obligations of reciprocity to their compatriots to use their talents and skills in securing their basic rights to health care. Henceforth, I refer to this argument as the 'departure' argument. As with the error argument, restrictionists argue in the departure argument that source states must institute contracts ensuring that doctors discharge their obligations of reciprocity. Providing medical training only with contracts to stay would enable source states to ensure that these doctors fulfill their obligations of reciprocity.
It is worth noting that the background conditions for restrictionists' claims in both arguments is the dearth of doctors due to migration and the negative impacts on their compatriots left behind. 4 Their arguments may not support policies to restrict emigration from countries not facing similar medical shortages. With that said, I summarize the core claims of the two arguments below and then § 2 explains the account of reciprocity grounding these restrictionists arguments.
The error argument has the following claims: (1) During training which helps them refine their skills, doctors make errors.
(2) Since doctors benefit from medical practice which involves these errors, they obtain obligations of reciprocity to their compatriots. (3) Source states ought to ensure that doctors fulfill their obligations of reciprocity.
(4) The best or only way for source states to ensure that doctors fulfill their obligations of reciprocity is for doctors to temporarily practice medicine before emigrating. (5) Therefore, source states must condition medical training on contractual agreement to stay temporarily after training.
The departure argument has the following claims: (1) The purpose of training doctors in source countries is to protect citizens' basic rights to health care. (2) Thus, trained doctors have obligations of reciprocity for the training received to protect their compatriots' basic rights to health care. (3) Source states ought to ensure that doctors fulfill their obligations of reciprocity. (4) The best or only way for source states to ensure that doctors fulfill their obligations of reciprocity is for them to temporarily practice medicine before emigrating. (5) Therefore, source states must make medical training conditional on a contract to stay temporarily after training.

Becker's account of reciprocity
Becker aims to ensure that social interactions in political communities result in human flourishing. To this end, he presents basic features for an expansive conception of reciprocity. He emphasizes that commonly accepted direct, one-to-one, reciprocal exchanges are only a subset of an appropriate conception of reciprocity, so some conceptions of reciprocity account only for this small subset of cases while making four oversimplifications precluding recognition of the more expansive forms of reciprocity. Regarding these oversimplifications, Becker states that people commonly assume that (1) reciprocity involves direct, one-to-one exchanges, (2) reciprocity must be in-kind return, (3) the scope of reciprocity is limited to voluntary relationships, and (4) reciprocal exchanges must have goods of equal or comparable absolute value. Becker claims that a correct concept of reciprocity must encompass direct reciprocity, as when 'you scratch my back and I will scratch yours,' and indirect reciprocity, as when you scratch my back and I scratch that of a third person. Specifically, Becker argues that a concept of reciprocity, in order to overcome oppositions to indirect (dispersed) reciprocity must meet standards of proportionality and fittingness. The underlying thought behind the fittingness standard is with regards to the return of good for good received or the giving of good to correct for bad received. Becker believes an appropriate conception of reciprocity, by meeting this fittingness standard, successfully applies to cases of direct and indirect reciprocity. Ultimately, the goal of his proposed conception of reciprocity is to satisfy the aims of mutually advantageous social arrangements which promote human flourishing. According to Becker (2005, 23-24), his account of reciprocity helps 'avoid defeating our pursuit of that aim . . . .'

Restrictionist application of Becker's account
Ferracioli and De Lora use the oversimplifications Becker highlights to respond to three opposing claims to their arguments for the duty to stay and their proposed contract. They consider these responses as a demonstration that their arguments for the duty to stay avoid oversimplifications about reciprocity and meet the fittingness standard. First, they believe it is an oversimplification to claim that reciprocity is only justified when there is a perfect match between the patients on the receiving end of the errors made by medical students and the patients who are treated by the students after the completion of their training. To this claim, they respond that reciprocity may justifiably be conceived as an indirect one-to-many exchange.
Second, restrictionists note the oversimplification stipulating that medical doctors staying in the source country is not equal compensation for the errors committed over the course of their medical training. Essentially, the claim is that this form of compensation is incommensurate in value with the errors made. To this, restrictionists respond that reciprocity neither requires an exchange on equal terms nor an exchange of goods of equal value. Sometimes, it is acceptable to reciprocate with something less valuable.
The last oversimplification restrictionists address regards the involuntary nature of the obligations of reciprocity in this context. They respond to the objection that compelling medical doctors to reciprocate conflicts with actual reciprocity. They respond that we endorse many reciprocal obligations that are not fully voluntary such as the obligations of reciprocity in filial relationships (Becker 2005, 20-22;Ferracioli and De Lora 2015, 607).

The moral significance of errors and harms
In support of the error argument, restrictionists claim that doctors, in order to heal, 'inflict pain on a daily basis' and 'impose high health risks and harms' (Ferracioli and De Lora 2015, 605-606). While this seems true at face level, it does not provide compelling reason to accept the error argument. First, when medical students and doctors treat patients and make permissible errors, they do so only after informing patients about potential risks and receiving the patients' consent. As a rule, law and morality demand that medical practitioners inform patients about the nature of their procedures and receive their consent for undertaking these procedures. At the very least, in hospitals, patients must consent before procedures are undertaken. The errors associated with medical practice are justified unless these doctors engage in procedures to which their patients have not consented. This seems to be the case in malpractice, which restrictionists note as part of their evidence in support of the error argument.
Since these errors are justified and result from consented to risks, any obligations of reciprocity that stem from these errors are misplaced. When doctors offer a service and make clear the risks involved, they owe patients the agreed upon. They are at the same time not liable, in exercise of their due diligence, for any realization of the risks patients were informed of. Importantly, they do not, based on errors made, owe any further obligations of reciprocity to all of society. Indirect duties of reciprocity are not generated because doctors do not act morally wrongly when patients provide informed consent to treatment with knowledge of the risks involved.
To be fair, restrictionists state that medical students likely have no malicious intent to make errors. They concede that medical practice and its associated errors aim to meet patients' health needs. However, they narrow their argumentative focus on the errors made during training and argue that such errors in training call out for special compensation. Accordingly, they present an analogy for the conclusion that doctors have obligations to society because of their training. In comparison to an analogy pointing out that the patients are not being simply used as means for the doctors ends, Ferracioli and De Lora (2015, 608) argue that the medical training case is analogous to a case in which 'someone who is taking driving lessons but gives you a lift to your medical appointment in order to learn how to deal with stressful driving situations. In that case, the student does owe you compensation because a better alternative was available, namely, the driving instructor.' I believe this illustration does not improve the error argument. The complicated and false picture here is that of doctors receiving training because of their patients giving them their bodies or of their patients' bodies being used wrongly by the medical institutions for the refinement of their medical trainees' skills. But if this is the case, medical students who harm patients should rightly be suspended or charged with malpractice. Medical institutions should equally be liable for these exposure to unnecessary and unjustified risks and harms. A more accurate picture is that patients receive care that involves some risks. Therefore, the patients, at least ex ante, benefit from access to medical care even though there are risks involved.
As I have noted, these errors are justified. Indeed, restrictionists acknowledge that these errors are not negligent, intentional wrongs. 5 To insist on anything morally stronger would mean that the reciprocity owed is in fact based on something morally troubling as in a case of medical malpractice. This would amount to reciprocity owed because of harm. But careful attention to a standard philosophical account of harm is in order here. On Joel Feinberg's account of harm, harms are defined as wrongful setbacks of interests. He (Feinberg 1987, 36) writes, 'only setbacks of interests that are wrongs, and wrongs that are setbacks to interest, are to count as harms in the appropriate sense. ' We can understand interests in this context to capture welfare, well-being, or flourishing in relation to an individual's health. If we consider Ferracioli and De Lora's phrasing of 'exposure to risks, actually harming, and rendering vulnerable' to fall under this broad notion of harm, a wrongful setback of interests would mean that there is no appropriate justification for the harm caused and that the medical care does not benefit the patients.
If restrictionists conclude that these errors are actual harms, then the error argument amounts to stating that harmful actions validate further reason to harm. If indeed this form of harm is the norm, the required solution is not further reciprocal obligations. To suggest that the fitting response to this is to assign duties of reciprocity to society makes a mockery of moral significance of medical harms. If medical doctors harm their patients, they deserve legal punishment and not re-assigned obligations to society.
I suspect that restrictionists would prefer to disassociate from this conclusion. As I will show in later sections, such a view does little service to restrictionists' moral defence of the duty to stay. On the other hand, to accept the moral insignificance of these errors means they do not provide morally interesting grounds for the restrictionists appeal to indirect reciprocity. Therefore, I believe arguments based on obligations of reciprocity for doctors can only be fitting if there is an additional factor beyond patients' consented to risks. I believe this factor is the source of funding for their medical education.

Public versus private funding
I reject the error and departure arguments because they do not meet the fittingness standard for reciprocity. I will demonstrate this with reference to restrictionists' undifferentiated application of obligations of reciprocity to all medical students irrespective of the source of their funding for their education. The following sections contain my rejection of the reciprocity-based arguments based on what I consider to be a lack of fittingness.
Philosophers invoke the general idea of reciprocity in political communities as large as countries based on the view that individuals participate in a mutually advantageous cooperative scheme. Beyond their direct obligations of reciprocity, individuals owe indirect obligations of reciprocity to all members of the country. These obligations arise even when one does not directly benefit from most of one's compatriots. This kind of indirect reciprocity appears to be one of the features of national tax regimes. Taxes are collected to ensure that there is funding available for the benefit of all citizens. This includes those who cannot afford some goods or services and therefore need governmental support through public funds and those who might never need the public funds generated.
In large contemporary political communities, citizens are required to discharge their obligations of reciprocity to maintain the government's ability to secure basic human needs and rights. Therefore, beneficiaries and non-beneficiaries are expected to contribute their quota in some form. 6 That people bear such obligations of indirect, one-tomany reciprocity seems uncontroversial. Nevertheless, restrictionists misapply this conception of reciprocity in their undifferentiated argument for all medical students. They fail to discriminate between obligations of reciprocity for medical students who privately fund their education and medical students who depend on public funding. The closest they come to acknowledging the normative import of this distinction is when they write, 'Moreover, because our account does not appeal to the financial costs of medical education, governments are justified in requiring private medical schools to offer similar contracts to their own applicants.' (Ferracioli and De Lora 2015, 610) Thus, restrictionists' disinterest in considering this important distinction rests on their belief that the appeal to reciprocity in their arguments are independent of education funding considerations. However, we must immediately note that the above-quoted statement is not directed to the issue of fitting obligations of reciprocity but rather to the role of private schools in joining governments to implement their reciprocity-based medical education policy. Moreover, we must not lose sight of the fact that considerations about private schools and about privately-funded education are related but logically distinct. One may attend a private school or public school on either public or private funds. Thus, the relevant normative distinction here concerns whether one's education is privately or publicly funded.
Privately-funded medical students in source states are no armchair speculation. Empirical studies find an increasing turn from purely public-funded medical education to dual track systems which allow for public and private funding of higher education. We must assess the claims of reciprocity for publicly-funded and privately-funded medical education in the light of these findings. If we attend to dual track systems in higher education in African countries, for instance, we must be more circumspect about the justifications for obligating medical students. Dual track systems of higher education allow public funding and private funding for all educational costs. Studies attribute this primarily to countries with taxation difficulties (which are common in source countries) and increasingly long lists of public needs requiring funding. Extensive research by Teferra and Altbachl (2004, 29) shows that 'in virtually all cases, researchers observe the constant decline of direct and indirect resources allocated for higher education by governments.' This is consistent with a recent case study of Kogi State in Nigeria where ninety-six percent of respondents report not receiving governmental 6 As Ferracioli and De Lora (2015, 608) note: 'the fact that some of us might not actually need health care in the future does not cancel the moral obligation to support the educational system that makes the acquisition of medical skills possible.' support for their higher education (Durowaiye and Khan 2017). A recent UNESCOfunded study by Varghese (2004, 63-66) reveals that private sources dominate the funds for higher education in Ghana. 7

Violating fittingness: undifferentiated obligations
I believe the obligations of reciprocity for doctors who do not rely on public funding for their education are different from the obligations of their colleagues who benefit from public funding, yet restrictionists make no such distinction. According to restrictionists, both groups of doctors have reciprocity-based obligations because of the errors involved in their training and the harm from their departure. This conclusion however demands an evaluation of whether conditioning medical training on contract, despite these differences, meets the standard of fittingness of an appropriate conception of reciprocity. Contrary to restrictionists, I believe the source of funding for their training makes a difference to the obligations of reciprocity doctors owe. I assess this with regards to the error argument. For doctors trained with public funding, we should identify and track the good received straightforwardly, the refinement of skills and public funding for the acquisition of these skills. The relation between the two benefits received from the public contributions mean that these doctors' compatriots can reasonably demand medical service in return on the grounds of indirect reciprocity. Governments fund the medical training of the doctors with public contributions; therefore, doctors' obligations of reciprocity are indirect. Additionally, the willing receipt of public funding signals their approval or endorsement of this form of cooperative scheme and its demands. To focus specifically on the error argument, funding for the education and the permissible errors during training can together be deemed goods received for which temporary stay is a fitting return. Staying to serve one's compatriots who paid for one's education is a plausibly fair return under the circumstances. Thus, fitting return for the good received justifies indirect reciprocity. 8 The considerations significantly differ for privately funded training in which the received benefit boils down to the permissible errors during training. Without the benefit of public funding, medical education seems to generate only obligations of direct reciprocity related to the circumstances of the errors. Further obligations of indirect reciprocity are unjustified and unfair to these doctors. Moreover, such obligations of indirect reciprocity do not meet the fittingness standard. 9 I will summarize my rejection in this section as follows. First, as a baseline, both publicly-funded and privately-funded doctors may have obligations of direct reciprocity 7 Overall, there appears to be reduced public financing of higher education and increasing private funding in source countries. 8 Here, I agree with Dwyer (2007) that, 'People are free not to study; they are free not to exercise their skills; they are even free to leave a society. But at least in many circumstances, when people choose to acquire professional skills and rely on public resources and institutions to achieve that goal, they also acquire some social responsibilities.' 9 A relevant qualification may for instance appeal to public goods enjoyed by these doctors. The often-used example is that of a secure environment in which they have received their training. But even here, I object that we would have to make the strange and absurd claim that public goods are provided with the sole aim of motivating obligations of reciprocity. I do not know of any serious understanding of public goods that support such a view. Ultimately, the appeal to public goods is quite a demanding justification to make for the sort of obligations recommended by the reciprocity-based arguments in preceding sections.
stemming from the medical education they receive. If, as restrictionists claim, doctors benefit from the errors they make on their way to become professionals, doctors must make sure their consenting patients benefit too by doing their due diligence in spite of the risks of those errors. This qualifies as a form of direct reciprocity that applies to all doctors. As a result, we need an explanation for moving to indirect reciprocity. I posit that the relevant explanation must appeal to the source of funding for the medical training since such funding makes the training possible. That is, the benefit of refining skills and gaining qualification is received in the first-place rests upon the source of funding for medical training. The difference in sources of funding generates differences in forms of reciprocity. Note that this is not an oversimplification, for I am not claiming one-to-one exchanges are the sole form of reciprocity. Instead, my point is that expanding obligations to one-to-many requires additional explanations. We cannot conclude that merely because one-to-one exchanges do not exhaust reciprocity, one-to-many exchanges apply in all circumstances.
Consider an illustration. Betty belongs to a coding club. One of the rules of the club is that each member must contribute to organize celebrations during special days. All members endorse this norm and its associated demands. On Betty's birthday, members of the club contribute to organize a birthday party for her. Clearly, Betty owes it to the rest of the members to contribute when next there is a celebration of a special occasion or contribute in some other comparable form. She has a one-to-many mapping of obligations of reciprocity to her club members. In contrast, consider the scenario where Bob, a member of the club, organizes Betty's birthday with his own resources. In such a case, Betty owes reciprocity to him alone. She has a one-to-one mapping of obligation of reciprocity to Bob. Distinguishing these cases of direct and indirect reciprocity is not oversimplification. I reiterate that if a medical doctor benefits from public funds for her education, we have a straightforward case of her having to pay back to the society that funded her education, under the right limits and specifications. However, a student who privately funds her medical education does not gain the same indirectly reciprocal obligations as her publicly-funded colleague. 10

Violating fittingness: involuntary contracts
Though Becker rightly identifies that some obligations of reciprocity stem from nonvoluntary relations, restrictionists' arguments for the duty to stay and their proposed contracts violate Becker's fittingness standard. I agree that some involuntary relations of reciprocity, such as those in vital filial relations, are justified. The obvious analogy is that children have reciprocal obligations even though they do not have a choice regarding which parents they are born to. But this alone does little to justify further involuntary reciprocal obligations. I find the analogy unpersuasive for the duty to stay. The state of health care in a country doctors did not voluntarily choose membership in should not dictate artificially restricting their access to medical training. Neither should it dictate how they can reciprocate for the training they receive. Even if we accept medical training and the permission to make errors as well as a population on which to refine skills as benefits, we must reject the pre-emptive contracts of reciprocity.
We should never endorse such involuntary obligations while neglecting the values at stake. Bestowing benefits under such contractual conditions does not appropriately justify obligations of reciprocity. To suggest otherwise violates the fittingness standard. It is analogous to the cases of parents who care for their child only because they want to be cared for later in life. We should not endorse filial relations where care for children is provided by parents with the sole aim of being able to cash in later. The same applies to source countries where the duty to stay and medical education on contract is justified by such a conception of reciprocity.
Moreover, we are not typically troubled by reciprocity in nonvoluntary filial relationships because we recognize significant values in such cases, particularly in that we expect parental love and care for children. We value parent-child relations because we believe and expect that they promote children's welfare and development. So, even though children do not choose their parents, we consider reciprocal obligations that meet these expectations fitting. In contrast, we tend to deny obligations of reciprocity in filial relationships characterized by parental abuse. Under such circumstances, we rightly doubt the prospects of children's welfare and development. In such cases, the fact that children do not choose their parents instead means that we do not have to conclude that they owe them obligations of reciprocity.
The same reasoning applies to the restrictionists' arguments. The individual wishing to become a medical doctor was involuntarily born in a society with deficient health care provision. There is an uncontroversial arbitrariness to one's country of birth. This not only impacts one's opportunities but also under what conditions one can take advantage of those opportunities. Citizens who desire a career in medicine do not freely choose to be born in source countries with unfortunate conditions of health care. These citizens cannot choose their place of birth and original citizenship. The contracts suggest that these governments may conditionally bestow these benefits and can alter the lives of doctors by taking advantage of an involuntary relation. 11 This undermines the value of reciprocity in such relations and runs afoul of the standard of fittingness. 12

Violating fittingness: asymmetrical freedoms
This section continues my rejection of the restrictionists' account of fitting obligations of reciprocity for medical doctors in source countries. In this section, I consider a caveat in the restrictionists' justification for the use of state power to generate doctors' duty to stay. In spite of the conclusions of their moral argument, restrictionists place 11 One might object that finding oneself in an involuntary relation does not invalidate the duties stemming from those relations. Fair enough. My point is that involuntary relations involving bestowal of benefits with morally wrong motivations reduce the force of duties of reciprocity for the recipients of those benefits. This is again why I believe we should not endorse reciprocity claims in some filial relations. More to the point, the proposed contracts undergirding government support for medical education should not be motivated by a future claim or credit on reciprocity. Thanks to an anonymous reviewer at Ethics and Global Politics for encouraging me to respond to this objection. 12 In their response to an oversimplification, restrictionists appear to have themselves fallen prey to an oversimplification! They assume that we accept involuntary relations without a deeper justification. This is to oversimplify why we accept these obligations.
a limit on justified state coercion when they consider the moral force of medical students' freedom of occupation. According to restrictionists, freedom of occupation requires that states permit medical students to switch careers upon graduation, if they desire (Ferracioli and De Lora 2015, 615). This means that we have one condition under which state power cannot be employed to achieve the ends of restrictionists' reciprocity arguments. I believe this conditional limit is inconsistent with the error and departure arguments and I will show why by analysing their concession to the moral force of freedom of occupation while failing to appreciate the moral force of freedom of movement.
Generally, individual freedoms are foundational and instrumental for individual interests. Freedom of occupation, for instance, allows for an individual to secure and realize her interests as it pertains to careers or jobs. Similarly, freedom of movement permits an individual to satisfy her interests in various geographical locations of her choice. These two specific freedoms share the core importance of freedoms for individual interests. Therefore, if restrictionists are willing to concede the force of the freedom of occupation against their argument from reciprocity, then it stands to reason that they should similarly concede the force of freedom of movement. However, the restrictionist account treats these freedoms asymmetrically. According to them, the force of freedom of occupation means that obligations of reciprocity stop short of forcing graduates into a medical profession. They claim that one may still be trained as a doctor and choose to leave medicine for another career. This claim is an exemption from the obligations of doctors in source countries despite the restrictionists' reciprocity arguments.
This concession to freedom of occupation and without similar for freedom of movement betrays an inconsistency in the restrictionists account. To begin with, when restrictionists permit freedom of occupation they further weaken the premises of their error argument for the duty to stay. This is because, to claim that doctors owe obligations of reciprocity based on the errors they make during training is to imply an obligation deficit when a trained doctor switches careers after her training. To argue that this reciprocity deficit does not arise for a career-switching doctor is to dismiss the rationale behind the error argument. In doing so, restrictionists devalue the errors to which they appeal on their account of reciprocity. They do not further present an argument to suggest that once the benefits of medical training are received, there is some other factor to distinguish the obligations of doctors who wish to exercise their freedom of occupation from that of doctors who wish to exercise their freedom of movement. These considerable inconsistencies in the restrictionists further prove the implausibility of the moral argument for the proposed restrictions.
Restrictionists might respond that doctors who take up another occupation in the same country (unlike their compatriots who leave after their medical training) are still able to make some restitution for the investment the state has made in them. For instance, they might pay taxes or work a job that has comparable social benefits locally. Unfortunately, this response sufficiently weakens the force of the error argument to make it irrelevant to the restrictionists' claims of reciprocity. The point stands that restrictionists invoke claims of reciprocity because of the errors. Thus, permitting a newly minted doctor to leave for another occupation (irrespective of location) means she does not provide the medical care restrictionists deem fitting restitution for the errors she has made over the course of her training. For the restrictionists' account, the fitting act of reciprocity is to provide medical care locally. Where this is substituted for any other act, a reciprocity deficit remains. 13 Restrictionists might further contend that freedom of occupation operates on the same reasoning as freedom of exit from one's country (A subcategory of freedom of movement). In both instances, a major reason against state enforcement of the obligations of reciprocity is the risk of state abuse. Freedom of occupation, on this reasoning, prevents such unjustified limits on the life plans of doctors. After all, restrictionists ultimately claim that it is recipient states who ought to refuse to include doctors from source countries when they have not fulfilled their obligations of reciprocity. Moreover, restrictionists may claim that these recipient states merely exercise exclusion against freedom to enter which is asymmetric to freedom to exit because recipient states have a right to exercise discretion over their immigration arrangements. But this is only a partial picture of the predominantly formal nature of high-skilled migration to be expected of doctors. Through embassies and consulates in source states, a recipient states' acts of exclusion effectively block doctors' freedom of exit. They can be prevented from leaving through visa denials. And in practice being denied a visa in one's country is dissimilar to being turned away at the border of a destination country. More importantly, the error and departure arguments will have no force if we prescribe this passive approach of source states. 14 I now return to restrictionists' attempt to make the duty to stay less burdensome for doctors who intend to depart after training. Ferracioli and De Lora (2015, 610) aptly distinguish between 'what persons are owed from the point of view of morality and what morality can reasonably demand from individual moral agents.' This is one reason they resort to temporary stay in their proposed contracts. To this end, they state (Ferracioli and De Lora 2015, 610), 'But although it is true that the duty to stay cannot fully mitigate the harms associated with the brain drain in some parts of the world, it is important to recognize that even a few years can significantly increase the provision of health care services where the ratio of doctors per population is considerably low.' Thus, the success conditions for their reciprocity-based arguments do not demand total protection of the health needs of vulnerable populations.
Nevertheless, restrictionists fail to acknowledge how freedom of occupation and freedom of movement operate in the same moral terrain. Both freedoms aim to secure valuable interests and often work in connection to secure individual interests. The life plans may include a change in career immediately after their training and may similarly include departure to another country to practice medicine after medical training. Moreover, it is plausible that a doctor may choose to switch careers after emigrating. In such a case, restricting freedom of movement is tantamount to restricting freedom of occupation. A doctor should be free to leave her country of training if her departure is 13 In any case, nothing in the restrictionists' response above precludes doctors who leave after their training from paying some restitution through financial remittances. It might be true that securing restitution is easier locally than internationally, but this does not suggest that it is impossible for doctors who leave to make restitution. Therefore, the point stands that there is an inconsistency in how restrictionists apply freedom of occupation and freedom of movement in their account. Thanks to an anonymous reviewer at Ethics and Global Politics for encouraging me to respond to this counterargument. 14 These considerations border on issues of state legitimacy, territorial right, justified coercion, and various aspects of freedom of movement. Those issues are beyond the scope of this paper.
to exercise her freedom of occupation in another country. So, we have further reason to apply similar considerations in upholding both freedoms. Both freedoms should similarly matter for their effects on the life plans of doctors. It is true that sometimes, we restrict some freedoms so other freedoms can be secured or while others remain unrestricted. But restrictionists cannot rely on their account of reciprocity for that argumentative move. Because they generate reciprocity from errors and potential harms, their proposal must restrict freedom of both movement and occupation or leave both unrestricted.

Violating fittingness: recycling reciprocity
I must concede that there is at least one ground to commend attempting to justify limiting the duration of the duty to stay. In so circumscribing the duty to stay, restrictionists avoid the reasonable objection to permanently stay on grounds of individuals right to leave their country of birth. 15 But notice that at this point in their account, the stipulation of temporary stay does not fundamentally rest on the error argument but on the departure argument instead. While it is true that in both cases, the recommended duration of stay is temporary, it is the departure argument that truly points to the harms of departure and makes room for this important limit on the demandingness of the duty to stay. This is not the case for the error argument. If the limit were from the error argument, medical doctors will have even less burdensome duties to stay since then it is not their life plans we ought to value but the errors they have benefited from. Restrictionists prescribe protecting the health care rights of citizens in source countries. This, as we have seen in the error argument, draws on some features of medical practice. These same features persist after medical students graduate from medical school and so the restrictionists' error argument opens the prospects of indirect obligations of reciprocity even during the years of temporary stay after graduation. Unsurprisingly, restrictionists present no defined threshold for how much error is permissible. I believe this opens further considerations about errors committed after medical training. Since they provide no basis for making a definitive statement about what sorts of errors generate reciprocity, restrictionists seem committed to concluding that errors committed after medical training generate further obligations of reciprocity. Following the restrictionists' consequentialist reasoning, a major condition for reciprocity requires that no residual obligations stemming from benefits received is left undischarged. We can infer this from their description of the errors and of the patients on which these errors are made as benefits to these medical doctors and their future patients. Restrictionists thus may invoke the duty to stay on a rolling basis. No matter the stage of career for a medical doctor, they are bound to, as restrictionists acknowledge, inflict pain on a daily basis as part of routine medical practice. 16 What restrictionists do not 15 See Whelan (1981). Although there is no necessary connection between the country in which medical doctors receive their training and where they hold citizenship, I am relying on the prevalence of this connection. 16 Ferracioli and De Lora (2015, 605) write: 'So, in order to heal, trained health care professionals inflict pain on a daily basis (think about surgical procedures but also in vaccination, biopsies, colonoscopies, the eliciting of reflexes, and so on). In a previous stage, when medical students are in the process of becoming licensed practitioners, we contend that they inflict even more harm and subject the patients to even greater risks than their more experienced colleagues.' further acknowledge is the extent to which the errors of physicians, unlike their counterparts in training, go unreported. This seems to be a factor behind claims that medical students or physicians in training are more prone to errors. However, evidence shows that experienced physicians are less inclined to share their errors. Medical students, being under the supervision of physicians, are not similarly able to avoid scrutiny. 17 A series of studies highlight internal and external barriers against more experienced physicians disclosing their errors. Patients' desire for disclosure of errors runs against physicians' reluctance because disclosures pose risks to their careers and professional status (Gallagher and Levinson 2005). Fears of lawsuits and associated shame generate internal barriers to disclosure. On the other hand, physicians display uncertainty and cautiousness regarding what to say when communicating errors to patients. Researchers attribute this to doctors' lack of training on disclosure methods (Gallagher et al. 2006(Gallagher et al. , 1585. Moreover, doctors avoid the word 'error' and often it is health care workers who point out errors the patient would not be aware of (Gallagher 2003).
The phenomenon of physician burnout further evidences the significance of errors made by physicians. Nationwide studies in the USA, for example, point out the independent contributions of burnout to physicians' medical errors (Tawfik et al. 2018). In source countries, evidence indicates even greater physician burnout. This is somewhat unsurprising given severe workload pressures and suboptimal working conditions which in turn contribute to frequent errors. 18 As noted above, restrictionists cannot evaluate these errors differently without diminishing the moral significance of the errors. Just as when they were medical students, medical doctors make errors. Moreover, we ought to weigh the errors more significantly at the latter stage of their medical careers. When they become professionals, we are justified in holding them to higher expectations due to their qualifications. Furthermore, it seems more likely that where there are both medical students and medical doctors, the medical doctors will oversee procedures that involve the most risk of harm. Medical students are less likely to be placed in charge of highly risky procedures unless there is a lack of more experienced medical doctors. As a result, the restrictionist appeal to error would generate duties on a rolling basis. Again, it is a repudiation of the error argument for them to state that these doctors have paid their dues. Without defining and defending what errors can be excused, the logical implications in their arguments for the duty to stay applies.

Conclusion
I will recap my critical view presented in this paper against restrictionists' arguments for the duty to stay. I have argued that in the case of privately funded medical students, the appeal to indirect obligations of reciprocity like that borne by their public-funded colleagues violates the fittingness condition for obligations of reciprocity. It therefore does not justify any legal demands for privately-funded medical students to fulfill obligations of reciprocity by staying temporarily in source countries. Consequently, 17 See, for instance, Kaldjian et al. (2008, 721). 18 Ayisi-Boateng et al. (2020)) provide an illuminating discussion of burnout among Ghanaian physicians. I object to the recommendations for contracts to restrict the migration of all medical students without regard for whether they privately funded their training.
While there are negative effects of mass migrations of medical doctors from sending countries, Ferracioli and DeLora do not present convincing arguments for their restrictionist proposals. In their attempts to present a novel normative defence of the duty to stay, they miss important reasons to limit such duty's scope and demandingness. We cannot demand undifferentiated obligations of reciprocity for all medical students (no matter the means of funding their education) and at the same time satisfy the fittingness condition for an account of reciprocity.

Disclosure statement
No potential conflict of interest was reported by the author(s).