Jane English on Rawls and duties to future generations
Jane English
I recently learned[1] about Jane English, a brilliant and regrettably little-known philosopher from the middle of the 20th century. This post is mainly about a contribution of hers to a longtermism-relevant debate, but first I’ll say a little about her life, which was pretty remarkable.
English was born in Cleveland in 1947 and did her PhD in philosophy at Harvard, where she studied with Quine, Putnam and Rawls. She graduated in 1973, was promptly hired by UNC Chapel Hill, and earned tenure in 1977. She died the next year while climbing the Matterhorn, at age 31.
During her five years of work as a professional philosopher, English published important papers on theoretical concepts, underdetermination in science, the morality of abortion, gender equality in sports, and duties to future generations in a contractarian framework. Pretty amazing! All the signs suggest she was destined for philosophical greatness. She was also a serious runner, swimmer and tennis player. (You can read a bit more in this memorial note written by her colleagues at UNC.)
The English paper I want to talk about here is “Justice Between Generations” (1977, Philosophical Studies 31, 91-104). The paper deals with Rawls’ views on intergenerational savings principles. English presents several powerful objections to Rawls, offers a sensible alternative proposal, and says interesting things about what kinds of intergenerational duties we can hope to derive in a Rawlsian contractarian framework. These ideas should be relevant to those longtermists who want to constructively engage with non-consequentialist traditions.
Rawls and the veil
First, some Rawls.
Lots of people consider Rawls’ A Theory of Justice (1971) to be the 20th century’s most important work of political philosophy. In the famous contractarian thought experiment at its heart, Rawls imagines the members of society choosing principles of justice from behind a “veil of ignorance”, where nobody knows their own demographic categories, personal aptitudes or conception of the good.
The self-interested agents in this “original position”, Rawls thinks, would choose principles which grant everyone the greatest possible liberties (compatible with everyone else having those same liberties), and which maximize the well-being of the worst-off person in society (since, for all you know, that person might be you).
A natural way of understanding the original position is to assume that: (1) Everyone in the OP is motivated by individual self-interest. (2) Everyone in the OP belongs to the same generation. As Rawls notices, though, this pair of assumptions causes problems. Under these conditions, the participants in the OP would assign all available resources to themselves to use for their own purposes, leaving nothing for future generations—not quite the perfectly just society we’d hoped to derive.
Rawls’ fix for this is to modify condition (1). Rather than purely self-interested individuals, the participants in the OP should be heads of households, who want what’s best not just for themselves but for their whole families. Under this assumption, we get a nonzero rate of savings, since everyone wants good lives for their children.
English on Rawls
English’s criticisms
Enter Jane English. As English points out, Rawls’ solution to the savings problem is unsatisfactory for several reasons:
First: Throughout AToJ, Rawls emphasizes his intention to assume as little as possible about the contingent motivations and feelings of the participants in the OP. Turning the participants into heads of households full of concern about their children’s futures is a pretty big deviation from that ideal. (And it also seems to spoil the theory’s universalist ambitions. What if the society in question doesn’t have anything like households in this sense? Or did Rawls think these were somehow natural and inevitable?)
Second: Rawls’ solution doesn’t tell us anything about how much heads of households would or should value the next generation’s interests relative to their own. “Unless we know how strongly they care, we do not know which principles will be chosen. Are heads willing to starve to death to send their children to college, or are they only willing to drive a Datsun instead of a Buick to achieve that end?” (94)
You might try to solve this problem in a principled way by taking heads of households to be overall-family-welfare-maximizers. But this could lead the OP participants to accept unfair family arrangement principles. For instance, it might maximize total family welfare to have a strict sexist division of labor, or to determine inheritance by arbitrary principles of primogeniture. Since these are exactly the kinds of utilitarian bargains that Rawls rejects on the social level, it’s hard to see why they’d be OK within families.
Another possible solution is to imagine a “veil-within-the-veil”, where family members choose savings principles in ignorance of who belongs to which generation. But this won’t work either. Any positive savings rate requires the older generation to sacrifice for the benefit of the younger generation. A priori, the younger generation isn’t worse off than the older. So this violates the “difference principle” which parties in the OP are supposed to accept, which says that inequalities are only permissible when they benefit the worst-off.
Third: Not everyone has children to save for, and not every child has household members to save for them. So Rawls’ approach can’t explain why current people in general should save on behalf of future people in general.
English concludes, I think correctly, that the head-of-household strategy is a bad way to motivate care for future generations in a Rawlsian contractualist framework.
English’s proposal
Her own solution? Get rid of assumption (2), and assume instead that the original position contains participants from different generations. Since some later people will be badly off if earlier people don’t save, and since anyone in the OP might turn out to be a later person themselves, this version of the OP will yield an intergenerational savings principle.
Why is this better than Rawls’ approach?
Most obviously, it avoids all the ad hoc stuff about households and families. This immediately solves the first and third problems above.
It also seems very much in line with the spirit of the OP. The whole point of the thought experiment is to consider which principles of justice we’d choose if we knew as little as possible about our contingent situations and interests. Surely the generation to which we belong is one of these contingent factors. It deserves to be hidden behind the veil.
But you might worry that English’s proposal runs into the same difficulty as the “veil-within-a-veil” idea mentioned above. A priori, we don’t know that later generations will be worse off than earlier ones—on the contrary, they might be much better off. So a savings principle might have the effect of forcing poor early people to sacrifice on behalf of rich later people. This violates the difference principle, so it should be unacceptable.
As English points out, we can avoid this problem if we only require the best-off members of earlier generations to save for the worst-off members of future generations (after they’ve already done as much as they can to help the currently worst-off people). The only way this strategy would run afoul of the difference principle is if every member of some early generation is worse off than every member of all future generations. This might be possible in theory, but it seems exceedingly unlikely in practice.
As we saw, English criticizes Rawls’s head-of-household approach for not giving any guidance about the magnitude of a just savings rate. What kind of rate do we get on her view?
The rate of saving called for… would be a relatively slow one, of course. Under some adverse empirical circumstances, namely when the worst-off in the first generation absorb so many goods that the worst-off in the second are no better off than their predecessors, no progress at all would result. Typically, however, progress is generated from several sources. First, innovation, invention and the accumulation of knowledge alone contribute to making later generations better off. Second, the co-existence of several generations produces short-term saving, as I have argued above. Third, the amount of sacrifice required to preserve and pass on knowledge or machines is small compared to the expense required to produce them anew. This means that relatively small sacrifices on the part of the better off in the first generation (such as oiling the machines and recording the knowledge) will tend to improve the lot of their successors significantly. (101)
Another virtue of English’s approach is that “it also calls for temporally reversed ‘saving’, e.g., through deficit spending, from those who are better off in later generations to aid those worse off in earlier ones” (103).
Happy ending: Rawls was persuaded by these criticisms and abandoned the head-of-household approach in his later Justice as Fairness.
English, contractarianism and longtermism
Longtermists often motivate their views by appealing to the rights and interests of future generations. And these views often involve “grand futures”, in which our descendants are long-enduring, numerous, wealthy and happy. Can English’s version of contractarianism explain why current generations should sacrifice to make such futures possible? Can any broadly Rawlsian theory do so?
English herself seems skeptical of some of these obligations, at least from the viewpoint of intergenerational justice:
It might be the case, on my account, that a society starting out at a subsistence level, under adverse empirical circumstances, would never be required by justice to begin the accumulation necessary to lead to centuries of progress and culture. But this seems to be correct. Of course, progress and culture are also goods, but generating them does not seem to be entirely a question of justice. We might want this progress for perfectionistic or other reasons, especially if the sacrifice on the part of the worst-off was small. Then this would be a saving chosen for reasons that override considerations of justice between generations. (103-4)
Besides bequeathing “progress and culture” to our descendants, though, many longtermists are concerned about the existence and size of future generations. That is, longtermists often favor:
Existential risk reduction, i.e., earlier generations sacrificing so that future generations have a better chance to exist.
Plenitudinous futures containing vast numbers of our descendants.
The best contractarian hope for justifying saving toward these desiderata, it seems to me, is to include possible as well as actual future people as bargainers in the original position. We’d also have to assume that possible future people prefer (sufficiently high-quality) existence over nonexistence. Given this setup, it’s not hard to see how the parties to the OP would agree to try to maximize the number of future people (to the degree allowed by the difference principle).
Is this a reasonable amendment to the OP? I can imagine arguments for both sides, but the idea doesn’t seem completely indefensible. The principles of justice governing social and political organization should plausibly cover some questions of population ethics: for instance, it seems unjust for one generation to reproduce in such large numbers that the next generation has no hope of sustaining itself. It’s natural to explain this in terms of the principles one would’ve accepted if they were, for all they knew, one of those possible future people.
If contractarianism has space for this kind of reasoning, it should also be able to accommodate longtermist population imperatives. I’d like to think Jane English would agree.[2]
I think Rawls’ difference principle would endorse extinction, unless a) we can guarantee no future lives worse than nonexistence some other way, b) the process of extinction brings about worse lives than all future counterfactual lives if we don’t go extinct, or c) our continued existence prevents worse lives in other populations (nonhuman animals if they don’t go extinct with us, reemergent life, or aliens).
That’s interesting! The path from the difference principle to extinction (or any sort of existential conclusion) isn’t obvious to me, so maybe you can say more? Is the thought that creating more people won’t generally benefit the worst off, because some of the new people will have very bad lives?
If so, I still wonder whether the difference principle applies. On its face, the DP forbids socio-economic inequalities that don’t benefit the worst-off. But there a few different ways of taking this.
On a very broad reading, the DP forbids any action which partly causes a state of impermissible inequality, regardless of the nature of that action. (So, having kids who turn out to be poorly off would count.)
On a narrower reading—which I guess is the way I’d read it—the DP is only forbidding something like socio-political systems which predictably lead to impermissible inequalities. It’s not a moral principle meant to apply to all actions whatsoever, but a principle of justice meant to apply to laws and institutions.
The narrower reading seems better because, among other reasons, the broad reading threatens to make almost every action impermissible. (A Rawlsian Paralysis Argument!) But maybe none of this is what you had in mind?
Mostly it’s because some future people will probably have bad lives, so it’s better that they don’t come to exist. If we go extinct, they won’t exist. You’ll have fewer people near the bottom levels of welfare.