High stakes instrumentalism and billionaire philanthropy
Many have argued that billionaire philanthropy is objectionably undemocratic. For example, Anand Giridharadas writes:
“When a society helps people through its shared democratic institutions [as opposed to private charitable foundations], it does so on behalf of all, and in a context of equality. Those institutions, representing free and equal citizens, are making a collective choice of whom to help and how. Those who receive help are not only objects of the transaction, but also subjects of it—citizens with agency. When help is moved into the private sphere, no matter how efficient we are told it is, the context of the helping is a relationship of inequality: the giver and the taker, the helper and the helped, the donor and the recipient.”
There have been other criticisms of these arguments by Scott Alexander and from Cullen O’Keefe. I think one additional point is worth making. The point of view expressed by Giridharadas, Rob Reich and others has almost no support from within mainstream political philosophy, at least as a criticism of effective altruist billionaire philanthropy. The reason for this is that almost all mainstream political philosophers endorse a view I call High Stakes Instrumentalism, which permits the use of undemocratic procedures, such as billionaire philanthropy, in order to avoid high stakes errors.
[Cross-post from my blog]
High Stakes Instrumentalism
When studying for my doctoral thesis, I set out to argue for an instrumentalist defence of political procedures: I argued that we should use democratic procedures if and only if doing so produced the best results. Upon approaching the topic for the first time, I expected to find the field to be split broadly into two camps: pure instrumentalists and pure intrinsic proceduralists who argued that democracy is intrinsically valuable or intrinsically just and so should be used even if it does not produce the best results.
However, what I actually found was that political philosophers were broadly divided into pure instrumentalists and proponents of hybrids of instrumentalism and intrinsic proceduralism. Even proponents of the latter type of view endorse a theory I call High Stakes Instrumentalism:
High Stakes Instrumentalism = For all cases in which we can feasibly use either a political procedure in the set of procedures S1, or a procedure in the set of procedures S2, and all procedures in the set S1 would produce high stakes errors if used, but those in S2 would not, we ought, for instrumentalist reasons, to use a procedure in S2 rather than S1.
More informally, High Stakes Instrumentalism is the idea that if we can use undemocratic procedures to prevent high stakes political errors, then we ought to do so, even though those procedures are undemocratic.
The boundary of high stakes error is usually drawn somewhat fuzzily at policies that violate rights to subsistence or an economic minimum, or basic liberal rights. For example, Joshua Cohen argues,
“Decisions should also be substantively just, according to some reasonable conception of justice, and effective at advancing the general welfare. But a principle of political equality states norms that will normally override other considerations, apart from the most fundamental requirements of justice.”[^1]
And Tom Christiano writes:
“[Democratic] institutions are partly evaluated by whether they manage to protect democracy, liberal rights, and the economic minimum. But beyond these there is no agreement on justice in law and policy in terms of which we can evaluate democracy from the egalitarian standpoint. Therefore, with the exception of these, democracy will be entirely intrinsically justified from the egalitarian standpoint.”[^2]
I discussed the popularity of High Stakes Instrumentalism in a paper in Ethical Theory and Moral Practice. The only prominent political philosopher to deny High Stakes Instrumentalism was arguably the Jeremy Waldron of Law and Disagreement, but even he explicitly came to accept High Stakes Instrumentalism a few years later.[^3] In general, it seems as though High Stakes Instrumentalism is a principle which any prima facie plausible democratic theory ought to accept. It appears to be very difficult to defend the view that majorities have the right to violate fundamental rights or to be tyrannous to minorities. Thus, while denying High Stakes Instrumentalism is an option, it is not a palatable one. Indeed, the fact that almost all democratic theorists accept High Stakes Instrumentalism provides some indication of its intuitive strength.
Does political theory condemn billionaire philanthropy?
Billionaire philanthropy is undoubtedly politically inegalitarian and undemocratic. When Bill Gates decides to spend $1m on vaccinations in poor countries, he has unequal influence over that decision. If he were to make this decision democratically, he would put it in a fund and have the US electorate (or maybe all global citizens) vote on what to do with the money.
However, if High Stakes Instrumentalism is true, Gates’ influence over how this money is spent is not objectionable, as long as his control over it prevents high stakes errors. Gates spends substantial portions of his philanthropic money on direct global health aid and on global health research, such as research into vaccines. Open Philanthropy spends its money on saving and improving the lives of people in extremely poor countries; reducing the risk of pandemics; campaigning against horrific abuse of animals in factory farms, and so on. If this money and that of other impact-focused philanthropists were instead under the control of the American democratic system, it would not be spent on these priorities. Some would be spent on farm subsidies, some on wars in the Middle East, some on income support for people in high-income countries, and so on. There would, in short, be far more high stakes errors if this money were under democratic control.
Thus, while billionaire philanthropy may well be undemocratic, it would be incorrect to conclude that a substantial fraction of political philosophers believe it is therefore necessarily illegitimate. In fact, almost all democratic theorists accept that billionaire philanthropy is morally required, provided the money is spent wisely.
The same argument cannot be made for ineffective or harmful billionaire philanthropy. Many billionaire philanthropists donate money to projects with negligible social benefit, such as concert halls at their old university. Others, such as the Koch brothers attempt to cast doubt on the science on climate change. But this should not indict billionaire philanthropists who spend their money effectively on pressing global problems, such as Gates, Open Philanthropy, Hewlett, Children’s Investment Fund, and Bloomberg.
As Rob Reich has argued, billionaire philanthropy does deserve scrutiny in a democratic society. But this scrutiny does not mean that billionaire philanthropy should be placed under democratic control. Rather it should be focused on convincing philanthropists that the world is not, as most of them seem to think, a canvas on which to paint their personality, but is something full of huge problems that they ought to help solve by spending their resources in a careful and rational way.
[^1]: Joshua Cohen, Philosophy, Politics, Democracy : Selected Essays (Cambridge, Mass; London: Harvard University Press, 2009), 271–72.
[^2]: [Thomas Christiano, The Constitution of Equality: Democratic Authority and Its Limits (Oxford: Oxford University Press, 2008), 73.
[^3]: Jeremy Waldron, “Disagreement and Response,” Israel Law Review 39 (2006): 64–65.
Thanks for writing this. I like articles that showcase examples were academic research we might be unaware of is important to an issue we care about.
I agree with this comment—thanks! A follow up: can you say why political theorists accept high stakes instrumentalism (as opposed to stating that they do)? It sounds like this is effectively a re-run of familiar debates between consequentialists and non-consequentialists (e.g. “can you kill one to save five? what about killing one to save a million?”), just wrapped in different language, so I’m wondering if something else is going on. I suppose I’m a bit surprised the view has no detractors—I imagine there are some (Kant?) who would hold the seemingly equivalent view you can never kill one to save any number of others.
Hi Michael, I was also surprised! I think the basic intuition is that people want to avoid sanctioning the tyranny of the majority. Philosophers think it would be unjust for the majority to violate the basic rights of minorities. As I argue in the paper, democracies do this all the time. Lots of philosophers who are at pains to emphasise their democrat credentials are also in favour of the US Supreme Court for this reason—it doesn’t get much more undemocratic than that.
The two main arguments for intrinsic proceduralism are from autonomy and from equal respect/basic equality. It is usually held that people’s rights to act autonomously stop at the rights of others. No-one holds that people have a right grounded in autonomy to pursue the project of killing minorities. On equal respect/basic equality, I have a hard time understanding how it is an argument rather than merely a restatement of the claim that intrinsic proceduralism is true: in spite of basic equality, there are lots of rights and powers we distribute unequally, such as the right to practice medicine or drive.
I really appreciate this post too!
That said, I’m confused on why you believe this post is an example of “academic research we might be unaware of is important to an issue we care about.” If it were the case that the majority of political philosophers oppose High Stakes Instrumentalism, I think I would waiver little in my belief that EA billionaire philanthropy is justified and even laudable. For example, I would not suddenly believe that Gateses are wrong to prevent low-income children from dying instead of gifting it to the US federal government for reallocation, nor that Tuna and Moskovitz are wrong to want to reduce the risk of future pandemics and AGI catastrophes.
(This, to me, is akin to why I do not update strongly on there being slightly more deontologists in expert surveys of philosophers than consequentialists)
Given this, it directly follows that expert consensus should not be a strong update in favor of my preferred hypothesis, if my preferred hypothesis was not strongly motivated by expert deference to begin with.
It’s possible that I just have a lot less epistemic humility/expert deference in this domain than you do? But from reading your public writings and our occasional conversations, I do not believe that this is true in other domains, so I’d be interested in knowing what is different here.
Good post! Thanks for linking to my work. I also agree with Larks that it’s nice to have academic political theory brought in here.
High-stakes instrumentalism could still be consistent with objecting to billionaire philanthropy if policies are available that (i) also prevent high-stakes errors and (ii) are ‘more democratic’ than billionaire philanthropy. These would be ‘intermediate policies’ in between the extremes you’ve considered: billionaire philanthropy on one hand, and the other hand funds under full democratic control.
There are plausible examples: e.g. it could be stipulated by law that a fund must be used for a certain cause area such as global health, with democratic control only applying within those bounds.
This seems broadly similar to arguments for representative rather than direct democracy.
Thanks, very interesting.
I’m curious whether, however, some political philosophers effectively limit the scope of relevant high-stakes errors to people within the same polity. If, as you suggest in a comment, their main intuition is to avoid a tyranny of the majority, it’s not so clear if they’re moved to consider errors affecting the citizens of other nations or non-human animals.
If so, this could mean that some of these philosophers would disagree that even EA-ish billionaire philanthropy does in fact prevent high-stakes errors, and thus they might object to it after all.
Indeed, if your characterization of their views is correct, then it seems that these philosophers would have to endorse a utilitarian dictator who implements policies of massive expropriation and redistribution of resources. Surely they would think this proves too much? (But maybe you think they are just making an empirical mistake, or that this is fact what they’d have to accept.)
More broadly, my loose impression was that non-consequentialist philosophers often think there are important distinctions between members of one’s own polity and others. E.g. my impression was that it’s quite controversial how to apply Rawls’s theory of justice to between-nation issues, and certainly many people deny that his leximin principle applies globally. (But I’ve read barely any political philosophy, so correct me if I’m wrong.)
Hi, this is a good question. On within-nation rights only, firstly, I think almost all of the democratic theorists think there are at least limited cosmopolitan duties of justice to benefit the very badly off in poor countries at least up to some arbitrary sufficientarian standard. Hardcore nationalism also isn’t that popular in political philosophy, and the moderate intrinsic proceduralists out there don’t believe in it. Secondly, for theorists who endorse an act/omission distinction with respect to national responsibility, there is still the Thomas Pogge point that the global system of institutions in part causes poverty. in that context, aid is a corrective to an injustice.
Thirdly, this raises the question of how we should define the demos. One common view is that decisions that affect all should be decided by all. But then decisions in the US about budgetary spending also affect people in poor countries, animals and future generations, so they should on this standard have a say in the decision or a proxy should have a say for them. This speaks for global enfranchisement rather than US national democracy. Another view is that decisions that coerce others should be decided democratically. But then immigration restrictions coerce prospective immigrants, so again this speaks in favour of global enfranchisement.
On animals, it is very uncommon to hold that animals have no moral status at all, and that abuse of animals could never be classed as a high stakes error. Most philosophers are committed to the moral status of severely disabled humans, dogs and cats, but this clearly implies that chickens and pigs also have moral status.
On Rawls, I’m a bit allergic to Rawls exegesis after years of doing political philosophy, so I’ll leave that one.
Do you have a sense of where numerically the line for what counts as “High Stakes Instrumentalism” should be among political theorists who endorse hybrid theories? Ie, what line is reasonable for where consensus would be:
Presumably, most philosophers and political theorists (are political theorists philosophers?) would agree that averting genuine risks to human extinction and large losses of human life are sufficient reason to avoid proceduralist positions. I expect most philosophers (though maybe not most laymen) would also consider factory farming to potentially be in that category.
But where is the line (typically) drawn? Presumably hybrid theorists will not be satisfied with a position like:
I read your post and skimmed the accompanying paper, and I see positive examples but not negative examples of what’s sufficiently large to count as High Stakes Instrumentalism.
Now, I think it’s possible/likely that there’s no clear dividing line, such that we can say with certainty that the consensus position is to take the democratic view if it’s A% inefficient but not A+1% inefficient. However, right now my uncertainty for where the line might naturally be ranges over several orders of magnitude!
I apologize if this comment comes across as aggressive or silly. This is a field I’m genuinely very ignorant of, so it’s quite likely that the question is simple or perhaps conceptually confused.
Defining the line of what counts as severe injustice, a high stakes error or a violation of a basic right, is not done precisely in the literature and is in my view impossible to do in a theoretically satisfying way. I think this is true for all nonconsequentialist thresholds. The point of nonconsequentialism is to avoid having to say how good something is, which makes it difficult/impossible to know how to trade-off different nonconsequentialist elements against each other. What do I do if I have to choose between the right to free speech and the economic minimum? If I don’t know how good these things are, I don’t see how I can compare and weigh them. Equally, what do I do if I have a 10% chance of violating someone’s right to free association and a 20% chance of violating someone’s right to an economic minimum? If you don’t know how good these outcomes are, probability weighing them isn’t much use when you’re deciding how to act.
Ultimately, the boundary of what counts as a high stakes error is defined fuzzily and arbitrarily.