Discerning the nature of freedom is indispensable to the philosopher, or anyone else who seeks to understand the relationship between individuals and society. Intuitively, most people hold that freedom is a good thing and that restricting the freedom of an individual is usually immoral. At the same time, most people will hold that some restrictions placed upon individuals are justified owing to their safeguarding the freedom of other individuals or enabling other virtues, such as justice or safety. Somewhere in between these two intuitions, and their subsequent theoretical elaborations, we find the origins of most discussions of freedom, as well as political ideologies. But while people hold different views on what exactly is meant by freedom, it can be agreed on that unjustified restrictions of the freedom of individuals are wrong. Thus, coercion has become a theme closely related to freedom, often, though not always, described as the antithesis of freedom.

One of the most challenging debates regarding coercion has been on the role of threats and offers. In general, it is accepted that threats are coercive as they force the receiver of the threat to act against her will. Some, however, argue that offers too can be coercive and ultimately restrict the freedom of the person receiving the offer. In order to claim that offers can restrict freedom, two things must be demonstrated.

First, it must be established that an act performed due to coercion can be said to be unfree. Second, it must be demonstrated that coercive offers exist. Examining the validity of the latter is the main task of this essay.

I argue that despite numerous attempts, scholars have failed to demonstrate the possibility of coercive offers, either by being unable to successfully replace competing views of coercion with more useful frameworks or as a result of focusing on the perlocutionary effect of a proposal on the receiver of the proposal rather than the illocutionary force of the offer by itself. Thus, while threats and other forms of coercion certainly restrict freedom, offers cannot.

Freedom and Coercion

The difficulties of establishing which kinds of coercion restrict freedom begin already as we try to explain the relationship between freedom and coercion. Coercion can be said to take place when a person performs an action not in her own but someone else’s interest. As Hayek ([1960] 2007: 254-55) points out, coercion still implies that the person is acting, but that she bases her reasons off of the manipulative options that someone else has created for her. Thus, we are not talking about coercion if someone physically manipulates her hand to trace her signature, which is such an obvious restriction of freedom that it is uninteresting to conceptually examine, at least for the purposes of this essay. Nor are we referring to all instances where one person’s actions or behaviour alters that of another to bring about certain acts, such as when someone has taken my seat in the library and I am forced to sit somewhere else. Rather, coercion refers to instances where a person expresses that she intends to inflict harm on another person if that person does not act according to her will (Hayek 2007). Bay (1970: 16-17), Lucas (1966: 57) and MacCormick (1982: 235) argue for similar views, that is, that there is no real choice to be made, as one offer is by a significant margin more desirable than the one inflicting the most harm. But some scholars have attempted to argue that such views are wrong, or at least inaccurate in capturing how the individual’s freedom is being restricted.

Murray and Dudrick (1995) argue that in any situation where a proposal is presented, the individual has a choice, regardless of how disproportionate the options may look to an observer. Our preferences arise from freely cultivated dispositions and coerced acts should be regarded as free because they are similar to other free acts, as we are always constrained in our choosing, be that by threateners or natural circumstances. They emphasise that an act that would by other views of coercion be morally excused may also be morally excused if it were the “morally appropriate or at least morally defensible act” (p. 121), yet this should not be mistaken for unfreedom.

There are several problems with this view, but I will only touch upon a few here. The first objection is intuitive and exposes a logical error in their argument: if my preferences arise from freely cultivated dispositions and become, as they argue, part of my habits and consolidated world view, does it logically follow that I am free when I am faced with a coercive situation? By their logic I am being held responsible for my choice in one given situation based upon former choices (as well as circumstances over which I had no influence at all). I may be held morally responsible for my previous choices, but that does not make my situation any less coercive, for it is my will that is being coerced above all. Secondly, their distinction between coercive proposals and morally excused choices seems to neglect one of the key reasons for discussing the relationship between coercion and freedom to begin with, namely that coercion often implies exclusion of moral obligations (Frankfurt 1988). Finally, they seem to ignore the effects of the intentional coercion inferred by another person, which, as will be demonstrated below, is key to the discussion. As Hayek points out, in what seems like both a utilitarian and Kantian point at once, “[c]oercion (…) is bad because it prevents a person from using his mental powers to the full and consequently from making the greatest contribution that he is capable of to the community” (Hayek 2007:255). The arguments for coercion as restricting freedom are convincing, then, but the challenge of conceptualising coercion still requires explaining.

Robert Nozick ([1969] 2007:261-62) offered a framework for how to think about coercion that most writers seem either to have agreed with or responded to since its first presentation. According to his account, person P coerces person Q into not doing act A if and only if:

• P threatens to do something if Q does A (and P knows he’s making this threat)
• This threat renders Q’s doing A substantially less eligible as a course of conduct than not doing A.
• P makes this threat in order to get Q not to do A, intending that Q realise he’s been threatened by P.
• Q does not do A.
• P’s words or deeds are part of Q’s reason for not doing A.
• Q knows that P has threatened to do the something mentioned in 1, if he, Q, does A.
• Q believes that, and P believes that Q believes that, P’s threatened consequence would leave Q worse off, having done A, than if Q didn’t do A and P didn’t bring about the consequence. (Nozick 2007: 261-62).

From this account, coercion implies that there is always a proposal involved in coercion where the coercee is forced to choose between two or more options (to do or not to do A in the main account presented above). The account also conceptualises coercion as only occurring if the coercee in fact chooses to (4) … not do A. And finally, his account holds that Q (the coercee) must be in a disposition that makes P’s (the coercer) actions coercing. The last two implications of his account have received much attention and criticism, and help to capture some of the main differences in views on coercion, and bring us closer to how and why we ask whether or not offers can restrict freedom.

Distinguishing Threats from Offers: Predicted and Moral Baselines

Nozick’s criteria for coercion place much emphasis on the notion of threatening without qualifying what makes for a threat. To illustrate the issue at hand, Nozick gives the following example: if P offers Q higher remuneration to work for him than Q’s current employer, and Q accepts the offer because he is incentivised by higher remuneration for his work, has P coerced Q into accepting his offer? Some might argue that P has in fact coerced Q because the offer might as well be stated by saying that if Q does not come work for P, Q will not receive the remuneration, but as he points out, that would make all employers coercers by default, and ultimately reduce all offers of improved conditions to threats (Nozick 2007). To avoid this pitfall, Nozick suggests that whether the proposal made to Q is a threat depends on the extent to which the consequence implied by accepting or refusing the proposal changes the course of Q’ actions from what they would have been in a normal or expected case. If the consequences implied by the proposal make Q’s situation better, Q has received an offer; if they are made worse, a threat. Thus it makes sense to talk about a baseline, relative to Q’s preferences, from which Q would want to move from a pre-offer situation to an offer situation, but not a pre-threat situation to a threat situation. The baseline is determined by a predicted and moral account, where the predicted course of events is what would normally happen if there were no intervention in terms of a proposal and the moral account refers to what is morally expected. In some cases, however, the predicted and moral expectations may differ. Nozick considers the following example: a slave master who beats his slave every day offers not to beat the slave in exchange for some favour. Given the normal course of events, this is an offer because the slave most likely wishes to move from the pre-offer situation to the offer situation voluntarily. Yet, it seems to be a coercive offer because it deviates from the morally expected situation of no beating and slavery at all. In situations like these, Nozick argues, we should consult the preferences of the person receiving the proposal.

In a similar manner, Hillel Steiner ([1994] 2007: 282-91) states that to distinguish between threats and offers we need to acknowledge that the accession-consequences of offers and the non-accession-consequences of threats ensure absolute deviations from the baseline, not just relative increments and decrements of wellbeing. Thus we can confirm that we can in fact identify what constitutes a threat. From these criteria, we can determine that only threats can be coercive, because person Q would not voluntarily move from a pre-threat situation to a threat situation, whereas Q would want to move from a pre-offer situation to an offer situation. Distinguishing threats from offers by referring to baselines focusing on what the normal course of events is, however, susceptible to at least two problems: one is to reduce coercion to a moral concept that is dependent on a reference to prior rights and wrongs, and the other is that the baseline approach may in fact be open to an offer being interpreted as coercive after all. The first problem is addressed in the next section and the second problem in the section following that.

Zimmerman’s Non-Moral Framework and Coercive Offers

David Zimmerman (1981) argues that Nozick, by considering questions of rightness or wrongness, makes coercion into a moral concept. The drawback of such an approach is that if coercion were a moral concept all proposals would need to be assessed in light of the rightness or wrongness of the conditions under which they are proposed. For example, assessing whether a wage proposal – such as the one in the example of P offering Q a new job in the previous section – would inevitably refer to the question of whether or not capitalism as an economic system is morally acceptable. If Q’s preferences render him to prefer the morally expected course of events as his baseline, which in his opinion a capitalist system deviates from, what Nozick sees as an offer becomes a threat. Another issue with moralising coercion is that it deviates from seeing coercion as the restriction of freedom and starts referring to concepts such as justice and rights. As Zimmerman attempts to convince his reader by developing a non-moral framework, not only does a non-moral account of coercion serve better to distinguish threats from offers, but it also opens up for the possibility of so-called coercive offers.

Zimmerman suggests that the baseline for a pre-proposal situation should be the expected course of events in all cases, but that the framework should be expanded in order to assess whether an offer is coercive or not. In the case of the slave who is beaten by his owner, the slave owner is in fact making an offer and not a threat, as Nozick would claim (the threat being “don’t take my offer and I’ll beat you; and even if you accept the offer, you are still my slave”). As demonstrated above, based on the predicted course of events, the slave owner is making an offer. Yet, the offer is coercive. To demonstrate this, Zimmerman adds to the framework an alternative pre-proposal situation that the offeree would prefer to the actual one. His hypothesis is thus that “an offer is coercive only if Q would prefer to move from the normally expected pre-proposal situation to the proposal situation, but he would strongly prefer even more to move from the actual pre-proposal situation to some alternative pre-proposal situation” (Zimmerman 1981: 132, italics in original). To avoid all possible alternative pre-proposal situations, but still replace a moral baseline, Zimmerman adds that there must be a requirement of historical, technological and economic possibility for the alternative situation. While the historical and technological requirements are highly intuitive, he says that the economic requirement raises an enquiry about balance between the preferred alternative pre-proposal situation and the cost this would inflict on the person giving the offer. Sticking with the example of P offering Q a new job, the question concerns when P’s cost makes Q’s preference unfeasible. The reason this is an important enquiry is because P’s offer becomes coercive not by virtue of the offer being unattractive yet nonetheless the more viable option for Q, but rather that P is actively preventing Q from being in the pre-proposal situation that Q wishes to be in. Such proposals, then, are not threats, but offers that nevertheless are coercive. In one example, Zimmerman (1981: 134-35) asks the reader to imagine an island with only two factories, owned by A and B, respectively, and no other possibilities for employment. Q is kidnapped by A from the mainland, where Q has ample job opportunities she would prefer, and taken to the island. Both A and B offer jobs to Q and Q needs to choose between starving and taking a job. By Zimmerman’s framework, we can determine that A has proposed a coercive offer, whereas B has not. The fact that B is taking advantage of Q’s situation does not restrict Q’s freedom.

Zimmerman manages to successfully expose the issue of moralising coercion by showing how that would make coercion ‘parasitic’ on prior rightness or wrongness, but his own account has several drawbacks. According to Alexander (1983), the separation of preventing Q from doing something and not producing the conditions for Q to do something, can only be made in three ways, neither of which are able to justify Zimmerman’s framework. The first would be to distinguish between acts and omissions, although this would mean that Zimmerman would have to disregard that B acts to prevent Q from acquiring the alternative pre-proposal situation she wishes to by, for example, not allowing Q access to the factory’s money and transport. This somewhat exaggerated example is likely meant to illustrate the complexities of distinguishing between acts and omissions. The second possibility of distinguishing might be by arguing that B has a moral right not to share property with Q, thereby rendering Q’s preferred alternative pre-proposal situation economically impossible. That would, however, turn Zimmerman’s non-moral account into a question of prior moral rights. The third and last way of distinguishing between prevention and not producing is by claiming that B’s offer does not prevent Q from attaining whichever resources B has, because B’s resources cannot exist without B. If that is not the case, then surely B, by holding on to those resources, is acting to prevent Q from obtaining her preferred alternative pre-proposal situation. Zimmerman answers to these critiques in a later article by suggesting that one way of distinguishing between preventing Q from obtaining something and not producing the conditions for Q to attain that thing is by placing constraints on how P came to be in her own position before time T, so that P only prevents Q from obtaining something only if P acted before T to obtain those resources. In contrast, if P naturally finds herself with those resources, she is simply not producing the conditions Q seeks, thus certainly exploiting Q but not coercing her. It is surprising that Zimmerman does not recognise how he himself ends up relying on prior moral rightness and wrongness by providing such an explanation, as it is assumed that the acquisition of resources before time T somehow makes the act coercive, which implies that it carries certain moral associations we must consult in order to make our assessment. As he fails to do that, he fails to justify his own reason for abandoning a moral baseline for assessing threats and offers and ends up with a similar but less fruitful framework than the one he sought to replace.

Threats and Coercive Offers as Moral Equivalents
Some scholars have kept with Nozick’s framework, albeit with modifications, for predicted and moral baselines, but nevertheless concluded that offers can restrict freedom. It is key to Nozick’s argument that a threat involves worsening the situation of the person receiving the proposal and that the opposite is true for an offer, and that this is viewed in light of what is expected. In his examples this is often illustrated by what the parties to the proposal used to do. Harry Frankfurt (1988) criticises this view, claiming that we should rather consider how the course of events would unfold now were it not for the intervening proposal. And that cannot be done by referring to old terms based on how things used to be done, but rather by assessing whether or not the person receiving the proposal is now better or worse off than she would be without it. By adopting a position like that, we assume a more nuanced view where an offer does not become a threat only because it deviates from an expected course of events, but rather an unattractive offer that is coercive in nature, albeit still an offer. When an offer assumes this function, however, it restricts the freedom of an individual by virtue of the same mechanisms as a threat. An example of such an offer is a proposal of withholding a threatened penalty. In Frankfurt’s words: “an offer is coercive (…) when the person who receives it is moved into compliance by a desire which is not only irresistible but which he would overcome if he could” (1988: 42). Viewed like that, an offer can violate an individual’s autonomy in a manner similar to threats.

Stevens (1988) goes further and argues that offers can be coercive for the same reasons as threats, and that, in fact, threats are coercive if and only if offers are. His argument is that if threats and offers were not both coercive, there would be a relevant moral difference between two proposals that are similar in all respects, except that one is a threat and one is an offer. Because there is not, threats are coercive if and only if offers are. The brevity of this essay does not allow for a full recount of Stevens’ structured and well-defined argument, but in short it demonstrates that there is no moral difference between threats and offers because which is which depends on the pre-proposal situation of the person receiving the proposal, which in many cases will render the expected outcome of the proposal the same. Consider the example of highwayman P pointing a gun to Q’s head and saying: “give me your money or I’ll kill you!” and as opposed to a case where P points a gun to Q’s head and says: “I’ll kill you”, before waiting some time and saying: “give me your money and I won’t kill you!” In the latter case, an offer is proposed, because it offers to improve Q’s situation significantly; yet, there is no moral difference between the two cases. As the examination of threats and offers stands now, then, it appears that we are forced to choose between Nozick’s initial account and rely on prior notions of morality, thus confusing the discussion of coercion as a restriction on freedom, or we must choose an account where a situation that intuitively strikes us as threatening becomes a (coercive) offer.

Refocusing the Lens: Threats and Offers as Illocutionary Acts
One feature of all the accounts examined in this essay is that they all attempt to distinguish threats from offers by focusing on how they affect the person receiving the proposal. Christine Swanton (1989) describes this as treating threats and offers as perlocutionary acts. Establishing whether a proposal is a threat or an offer cannot depend on the preferences of the receiver, because preferences only determine the perlocutionary effect of a proposal, but not the characteristics of the proposal in and of itself. Imagine the example of the highwayman above, but this time Q is suicidal and wishes to die. The highwayman is thus giving Q an offer in all cases if we are to consult Q’s preferences. But the reason the highwayman is proposing a threat is by virtue of the proposal’s illocutionary feature. This view ties well in with the original account of coercion provided above, where coercion is an act by another person intended to cause harm and manipulate the will of the subject, and avoids removing the relationship between coercion and freedom. Additionally, this view does not force us to abandon our intuitive understanding of threats as restricting our freedom and offers as expanding it. One issue with this view is that it may make it harder to establish moral responsibility, which is, as I have argued above, one of the key reasons of distinguishing threats from offers, but that problem cannot serve as a reason to keep frameworks that are shown to be conceptually flawed.

Conclusion
In this essay I have examined the relationship between coercion and freedom, and demonstrated that coercive acts certainly restrict freedom, both from a utilitarian and Kantian perspective. I have placed the debate on how to distinguish between threats and offers in perspective without exhausting the literature, but rather by focusing on some key writers to highlight the main issues in the debate. From my discussion, it becomes clear that attempts at demonstrating the possibility of coercive offers have suffered several conceptual shortcomings that leave these efforts unsuccessful. The issue of abandoning a moral baseline proved unsuccessful, perhaps because any assessment of threats and offers inevitably comes back to what we deem to be moral. And the error of only focusing on the perlocutionary effects of proposals has led scholars to mistakenly see offers as sometimes coercive. While beyond the scope of this essay, I suspect that a successful account of how exactly a proposal can restrict freedom must cautiously involve both a perlocutionary and illocutionary assessment, without deviating from seeing coercion as the wrongful restriction of freedom.

Bibliography
Alexander, L. A. (1983) “Zimmerman on Coercive Wage Offers” Philosophy & Public Affairs, 12(2), pp. 160-164.

Bay, C. (1970) The Structure of Freedom (Stanford: Stanford University Press), pp. 16-17.

Frankfurt, H. G. (1988) The Importance of What We Care About, (Cambridge: Cambridge University Press), pp. 26-46.

Hayek, F. A. (2007) from The Constitution of Liberty in Ian Carter, Matthew H. Kramer, and Hillel Steiner (eds) Freedom: A Philosophical Anthology (Oxford: Blackwell), pp. 254-260.

Lucas, J. R. (1966) The Principles of Politics (Oxford: The Clarendon Press), p. 57.

MacCormick, N. (1982) Legal Right and Social Democracy (Oxford: The Clarendon Press), p. 235.

Murray, M. J. and Dudrick, D. F. (1995) “Are Coerced Acts Free?” American Philosophical Quarterly, 32(2), pp. 109-123.

Nozick, R. (2007) from “Coercion” in Ian Carter, Matthew H. Kramer, and Hillel Steiner (eds) Freedom: A Philosophical Anthology (Oxford: Blackwell), pp. 261-277.

Steiner, H. (2007) from in Ian Carter, Matthew H. Kramer, and Hillel Steiner (eds) Freedom: A Philosophical Anthology (Oxford: Blackwell), pp. 283-292.

Stevens, R. (1988) “Coercive offers” Australasian Journal of Philosophy, 66(1), pp. 83-95.

Swanton, C. (1989) “Robert Stevens on offers” Australasian Journal of Philosophy, 67(4), pp. 472-475.

Zimmerman, D. (1981) “Coercive Wage Offers” Philosophy & Public Affairs, 10(2), pp. 121-145.

Zimmerman, D. (1983) “More on Coercive Wage Offers: A Reply to Alexander” Philosophy & Public Affairs, 12(2), pp. 165-171.

Written by: Pouya H. Jafari
Written at: The University of Edinburgh
Written for: Dr Lynn Dobson
Date written: December 2015

Further Reading on E-International Relations

• Is “One Man’s Terrorist Another Man’s Freedom Fighter”?
• Can the Use of Torture in Intelligence Gathering Be Justified?  
• Terrorists or Freedom Fighters: A Case Study of ETA
• Can Soldiers Refuse to Fight? The Limitations of Just War Theory
• Was There a Soviet-led Menace to Global Stability and Freedom in the Late 1940s?
• Can International Organisations Become “Autonomous Sites of Authority”?