Christopher Heath Wellman’s article in the journal Ethics, Immigration and Freedom of Association, (pdf) is generally well-regarded, even among advocates of open borders. Jason Brennan and Bas van der Vassen have both expressed plans to pen their own responses to it. In the essay, Wellman attempts to demonstrate a presumptive right of legitimate states to limit immigration on the basis of freedom of association, and then argues that neither egalitarian nor libertarian arguments can overcome this presumptive right to close borders.* I argue here that Wellman fails to establish this presumptive right, and that freedom of association suggests instead a presumptive right of individuals to migrate across national borders.

Wellman correctly points out that the freedom of association includes the freedom to not associate with certain others–the right to exclude–and this freedom exists for both individuals and groups. His workhorse analogies are marriage and private clubs. A person may unobjectionably reject suitors for marriage and a married couple is not required to open their marriage to outsiders. A private association is permitted to restrict its membership. The simple extension of this idea is that, likewise, a nation may restrict its membership by prohibiting immigration. Wellman illustrates the strength of the presumption that a group can exclude members by pointing to the examples of the Boy Scouts, who have been under fire for excluding gays and atheists from leadership roles, and the Augusta National Golf Club, which has similarly suffered scrutiny  for excluding women. Wellman correctly points out that even in these controversial cases, the burden of argument lies with those who seek to abridge the group’s freedom to determine its membership.

Marriage doesn’t work as a load-bearing analogy for other forms of association. A marriage requires and sustains unanimity both for its formation and for any subsequent membership changes. Unanimous decisions to associate are not distinguishable from individual decisions to associate in any interesting way. And neither private groups nor nations typically enjoy unanimity, so I won’t discuss the marriage analogy further.

The first objection Wellman anticipates is that states are different in morally important ways from other groups in that, for instance, “political states do not owe their membership to the autonomous choices of their constituents.” Instead of addressing this point on theoretical grounds, he deploys an interesting reductio ad absurdum.

[Here] I would like merely to highlight some of the unpalatable implications that follow from denying a country’s right to freedom of association. In particular, consider the moral dynamics of regional associations like the North American Free Trade Agreement (NAFTA) or the European Union (EU). If legitimate states did not enjoy a right to freedom of association—a right which entitles them to decline invitations to associate with others—then they would not be in a position to either accept or reject the terms of these regional associations. Think of Canada’s choice to join NAFTA, or Slovenia’s decision to enter the EU, for instance. No one believes that it would be permissible to force Canada into NAFTA or to coerce Slovenia to join the EU. (Of course, nor may Canada or Slovenia unilaterally insert themselves into these associations!) And the reason it is wrong to forcibly include these countries is because Canada’s and Slovenia’s rights to self-determination entitle them to associate (or not) with other countries as they see fit. Put plainly, if one denies that legitimate states like Canada and Slovenia have a right to freedom of association, one could not explain why they would be righteously aggrieved at being forced into these mergers.

It’s true that states are generally given ample room to negotiate these sorts of interstate associations, but consider what these arrangements really mean for the people involved. An association agreement is just a package of policies, usually including reductions of trade barriers, regulatory harmonization, easier immigration, and sometimes the establishment of limited, shared governance bodies. These are the policies that impact the lives of citizens. The ‘association’ is arguably just a symbolic bow to tie it all together, no more meaningful than when two municipalities half  a world away adopt one another as ‘sister cities’. Whether these policies are justified or not is a series of distinct questions, and they should be treated as such.

From the impacted citizen’s perspective, it makes little sense to speak of “forcing” a state to enter such a regional association. That would just mean some foreign government implementing new policies–which still need justification individually–instead of her own government. This may or may not be acceptable, but the question is more capably handled by democratic theory rather than a ponderous interpretation of the freedom of association. Incidentally, because immigration laws impact potential migrants so forcefully, there is a good case to be made that they should be included in that democratic decision.

Note also that regional association agreements between nations in most cases expand freedom of association at the individual level, so there’s no conflict caused by the state exercising its freedom of association in this way.  This would be like the Augusta National Golf Club deciding to merge with a women’s golf club, thus admitting women into the new club and eliminating the source of conflict. Wellman acknowledges the distinction between expanding and limiting association, but maintains his argument by providing the example of an essentially uncontested secession of one nation from another, namely Norway’s 1905 secession from Sweden.

In this case, more than 99 percent of the Norwegians voted in favor of political divorce and Sweden as a country did not resist the separation. Whatever one thinks about the justifiability of statebreaking, this seems like a paradigmatic case of permissible secession. If each individual’s right to freedom of association trumps the state’s right to self-determination in those cases in which the group as a whole seeks to disassociate from others, however, then Norway’s secession was unjustified; it was impermissible because every last Norwegian (if not also each Swede) had the right unilaterally to veto the political divorce and the plebiscite in favor of separation did not garner unanimous consent. Again, I presume without argument that this position is implausible. And if an individual’s claim to freedom of association does not trump her state’s right in the case of secession, there seems good reason to believe that an individual’s right would be equally impotent in the realm of immigration.

There are a couple of problems with this argument. First, Wellman tilts at a straw man when he speaks of one individual’s right to freedom of association trumping the state’s right to self-determination. The question isn’t “Does each individual’s right to association have equal weight to a state’s handling of international affairs?” The closure of a nation’s borders violates every individual’s freedom of association en masse, impacting every individual on both sides of the border. A more germane question is “How does the general freedom of association for individuals compare in terms of moral weight to a state’s freedom of association in international affairs?”

The second problem is that the example of secession still does not grapple with the distinct issues involved in a political secession. My quick reading of the 1905 split between Norway and Sweden suggests Norway was already effectively self-governed, complete with its own political and legal institutions, when it broke from Sweden. I am not aware of how migration between the two nations evolved at the time, but this information is critical for understanding the full moral dynamics of the divorce.

Consider the hypothetical case of Scotland seceding from the United Kingdom. Most likely free movement across the border and individual freedom of association more generally would remain unimpeded and the political division would be benign. But suppose instead that one of the countries closed the border, severing presently existing and potential relationships between British and Scottish citizens. Regardless of how any plebiscite decided, I submit that the two scenarios would clearly not be equally just or permissible. The moral issues involved are not adequately characterized by the Scottish (or British) state’s freedom of association. It appears in this case that the severing of associations caused by closing the border would surpass any value gained from the state’s exercise of its freedom of association (or divorce). This is what closed borders around the world do now as a matter of course. This Scottish example shows that there is a morally significant difference between cases where state freedom of association and individual freedom of association converge and where they diverge. In the case of divergence state freedom of association does not obviously carry more moral weight than individual freedom of association, suggesting the former should not be treated as presumptive.

Wellman raises the stakes of the regional association example by suggesting that the forcible annexation of one nation by another cannot sensibly be condemned without appealing to a state’s freedom of association.

Imagine, for instance, that a series of plebiscites revealed both that an overwhelming majority of Americans wanted to merge with Canada and that an equally high proportion of Canadians preferred to maintain their independence. Would it be permissible for the United States to forcibly annex Canada? I assume without argument that, even if the United States could execute this unilateral merger without disrupting the peace or violating the individual rights of any Canadians, this hostile takeover would be impermissible.

Again, an annexation of a population into another state, complete with its already established institutions, is best understood not in terms of freedom of association, but in alternative frameworks, like democratic theory, which could more meaningfully grapple with the loss of democratic representation accompanying sudden, momentous policy shifts. Moreover, the assumption that such an annexation could be accomplished without the threat of violence is simply implausible. In the case of violence or the threat of violence, it becomes clear that the unilateral merger is impermissible for many reasons, including basic humanitarian concern.

It’s worth stepping back now to consider some basic questions. In the examples above I have appealed to a conflict between the state’s and the individual’s right to association, implying the moral concerns are different between states and other groups, which I have conceded do have a presumptive right to exclude. But there is conflict when a non-state group excludes individuals. When an individual’s right to association is violated, she may be forced to interact with someone against her will. When that right is compromised for a group, an individual within the group will often still be able to avoid direct interaction with ‘interlopers’, and may only suffer indirect interactions with them (seeing them at meetings or hearing about their activities within the group). Simultaneously, some individuals within the group–not to mention erstwhile outsiders–enjoy an expanded freedom of association. It must be remembered that dissenting gay- and atheist-friendly Boy Scouts, and non-sexist Augusta National golfers, would prefer their respective groups did not exercise the right to exclude.

Still, Wellman is correct to point out that “if no one doubts that golf clubs have a presumptive right to exclude others, then there seems no reason to suspect that a group of citizens cannot also have the right to freedom of association …” Why then do I balk at allowing the same group freedom to prevail at the national level? Wellman’s argument pivots on the analogy of states to garden variety private clubs. This analogy fails because of the scope and nonvoluntary nature of states.

The scope of an individual’s belonging to a nation is vast. All of an individuals actions occur within the context of her nation.  Every exercise of liberty and every association an individual has occurs within that nation and under the guidelines set out by the state. And an individual is born into a nation, having no choice among alternatives. Any option to join an alternative nation is severely limited (indeed the justice of this fact is the subject of dispute). In contrast, an individual is typically free to choose what private organizations to join, and those organizations each serve distinct and limited purposes.

Suppose Billy the Boy Scout would enjoy scouting with Ahmed the gay atheist, and Ahmed likewise would like to join the Boy Scouts. The exclusive policy of the Boy Scouts clearly exacts a toll on these individuals. But at least Billy and Ahmed are free to interact with each other outside the Boy Scouts. They may be colleagues, or one may employ the other, or they may live together, join the same hobbyist clubs, or they may simply be friends who enjoy one another’s company. The dynamics of national membership are radically different. If Billy’s nation excludes Ahmed then our starstruck protagonists are precluded from interacting in a wide range of capacities, including those above. This is a severe curtailment of the individual freedom of association in favor of group freedom of association. This severe loss of freedom  does not obtain in the case of private groups because of the many alternative possibilities remaining for individual association. This drastic difference in outcomes between private group exclusion and state exclusion calls into question any presumptions based on the analogy between the two.

There are other miscellaneous problems with the analogy of nations to private groups. Many private groups, for instance, have rules providing for the expulsion of members, an authority not readily granted to liberal states vis-à-vis citizens. Meanwhile, membership of most private associations is not hereditary or granted at birth, as is citizenship. On that point, the state’s right to association seems equally applicable to restricting births among the native population as it is to restricting immigration, another unsavory implication. These issues may have satisfactory answers. Perhaps Wellman would argue that reproductive freedom is so important that it simply defeats the presumptive right of states to exclude that he has outlined. But these considerations suggest the analogy of nations to private groups is not straightforward.

Wellman acknowledges that a state’s right to exclude immigrants directly limits its subjects’ freedom to invite and interact with foreigners. He addresses this while discussing whether immigration restrictions violate the property rights of citizens, but I think the pertinent problems arise even if we just stick to freedom of association and forego a discussion of property rights. Wellman appeals to the argument from political externalities that people have good reason to care about–and control–their nation’s immigration policy.

And if there is nothing mysterious about people caring about who are (or could become) members of their golf clubs, there is certainly nothing irrational about people being heavily invested in their country’s immigration policy. Again, to note the lack of intimacy among compatriots is to miss an important part of the story. It is no good to tell citizens that they need not personally (let alone intimately) associate with any fellow citizens they happen to dislike because fellow citizens nonetheless remain political associates; the country’s course will be charted by the members of this civic association. The point is that people rightly care very deeply about their countries, and, as a consequence, they rightly care about those policies which will effect how these political communities evolve. And since a country’s immigration policy affects who will share in controlling the country’s future, it is a matter of considerable importance.

I won’t get bogged down in the nuances of political externalities here (start here if you’re curious), since my goal is limited to disabling Wellman’s establishment of a presumptive right of states to exclude immigrants. Above I hinted at a distinction between direct and indirect associations among individuals. Direct associations are those such as friendship, work or employment relationships, religious or community fellowship, teacher/pupil relationships, etc. Roughly, one might think of these sorts of associations as those in which two individuals know each other’s names. Indirect associations include the other folks in line at the checkout line, passers-by on the street, fellow commuters, and indeed, one’s civic associates. The existence and actions of these others do impact an individual, and vice versa, as is readily felt during rush hour traffic, say, or when one’s political values are affirmed or rejected on election day. But on an individual basis, these indirect associations have far less impact than direct associations. Perhaps they have important aggregate effects, but such effects would need to be very significant and very frightening to justify interfering with intuitively valuable direct relationships. Just as importantly, any such effects must be demonstrated. They are not the stuff of presumption.

To summarize, Wellman seeks to establish a presumptive right of states to exclude immigrants by making the analogy of states to private associations, which do enjoy a presumptive right to exclude. But exclusion by states severely limits individual freedom in a way that private exclusion does not; options still remain for individual association when private groups exclude. Wellman’s provocative examples of states exercising the freedom of association seem harmless only because they also expand individual freedom of association. I have attempted to show that there is an unavoidable conflict between a state’s right to exclude and a general right of association among individuals, and that this individual right has greater moral significance than the alleged right of states to exclude. If this is true, then, contra Wellman, individual freedom of association actually suggests a presumptive right to migrate across national boundaries. Even if I have failed to make this stronger claim, I believe these considerations still show that Wellman’s arguments fail to establish a presumptive right of states to exclude. In this case, the state’s and the individual’s freedom of association simply annihilate one another, and the ethics of migration must be settled by other arguments.

*Wellman stipulates “legitimate” states throughout his essay, but never defines the term (other than providing Canada and Slovenia as examples); I’ll leave off this modifier. I am uncomfortable with the language of states having rights. Throughout this piece, when I say “state’s right” I mean the “collective right of the citizens of a nation, as expressed or enforced by their democratically elected governments,” the legitimacy of which I’m also assuming here.