Cohen on the Desirability and Feasibility of Socialism

In “Why Not Socialism?”1 G.A. Cohen explains what is objectionable about capitalism, what is desirable about socialism, and why it’s an open question whether socialism is feasible. Part of what’s desirable about socialism is that it realizes a radical kind of equal opportunity. What interests me in this post is the other reason socialism is desirable, which is that it realizes community, and in particular a non-instrumental, non-market form of reciprocity.2 I accept Cohen’s account of the tension between capitalism and reciprocity, but I think it raises problems for his analysis of the feasibility of socialism.

Cohen identifies two obstacles to feasibility: motivation and information (“social technology”). He rejects the motivation problem; people are not too selfish by nature for socialism, though capitalism does train us to operate on the basis of greed and fear. The information problem is more challenging, he thinks. We need to make decisions about resource allocation based on information that doesn’t exist in any one place, but is dispersed throughout the population (Hayek). Each of us has special knowledge of our own preferences, capabilities, and local environment. In a system of prices based on supply and demand, this information gets aggregated in an economical form. I don’t need to know everything everyone else knows, I just need to recognize that the end result of their choices is that the price of some product has risen. That’s enough to tell me that I should turn to some close substitute – or that I should consider producing the good that is in short supply. Letting prices fluctuate with supply and demand allows each individual to choose what to consume and produce, while ensuring that we have enough miners and farmers, not all musicians and hockey players. Unfortunately markets tend to generate unjust inequalities, and encourage us to instrumentalize each other. Cohen admits that we don’t at present know how to realize personal choice and prosperity on a large scale in a manner that is consistent with equality and community. Yet he also insists that we don’t know that we will never know how to do this. So he leaves the feasibility of socialism as an open question. I think he’s missed a motivational problem, however, one that arises from his own account of the value of community.

To explain what’s bad about capitalism, Cohen imagines a camping trip organized on a capitalist basis.

“You could imagine a camping trip where everyone asserts her rights over the pieces of equipment and the talents that she brings, and where bargaining proceeds with respect to who is going to pay what to whom to be allowed, for example, to use a knife to peel the potatoes, and how much he is then going to charge others for those now peeled potatoes which he bought in an unpeeled condition from another camper, and so on. You could base a camping trip on the principles of market exchange and strictly private ownership of the required facilities. Most people would hate that.”

“Harry loves fishing, and Harry is very good at fishing. Consequently, he brings back more fish than others do. Harry says: ‘It’s unfair, how we’re running things. I should have better fish when we dine. I should have only perch, not the mix of perch and catfish that we’ve all been having.’ But his fellow campers say: ‘Oh, for heaven’s sake, Harry, don’t be such a shmuck. You sweat and strain no more than the rest of us do. So, you’re very good at fishing. We don’t begrudge you that special endowment, which is, quite properly, a source of satisfaction to you, but why should we reward that pre-eminence?” (cf Mill).

These passages are meant to show that when on a camping trip, we accept the principle that everyone should share roughly equally in both the benefits and the burdens of the trip, weighted by need on the benefits side, and by ability on the burdens side (Harry shouldn’t expect to work less than others just because he can catch fish more quickly). 

Cohen also thinks that the camping trip involves an attractive form of non-market reciprocity, as an element of community. By ‘community’, Cohen means the shared valuing of mutual concern. In other words, I care about you and you care about me, and we each care that the other cares, etc. (the analogue of common knowledge, but for values rather than beliefs). The normal way shared valuing manifests itself is as non-instrumental reciprocity. 

“I serve you not because of what I can get in return but because you need my service, and you, for the same reason, serve me.”

If the sentence stopped after “need by service” it would express unilateral altruism. Cohen’s idea is that people want to serve others who serve them in turn, because they value such relationships for their own sake, on top of whatever further benefits they may bring. They want to serve not just anyone, but those others who want to serve others. 

“I do not want to be a sucker who serves you regardless of whether you are going to serve me (unless you are unable to), but I nevertheless find value in each part of the conjunction – I serve you and you serve me – and in that conjunction itself I do not regard the first part – I serve you – as simply a means to my real end, which is that you serve me” (67).

This non-instrumental form of reciprocity contrasts with what Cohen calls “market reciprocity,” in which I serve others only to induce them to serve me. Marketplace competition pressures us into treating each other merely as means to our own ends, he claims.

“The immediate motive to productive activity in a market society is typically some mixture of greed and fear, in proportions that vary with the details of a person’s market position and personal character. In greed, other people are seen as possible sources of enrichment, and in fear they are seen as threats. These are horrible ways of seeing other people, however much we have become habituated and inured to them, as a result of centuries of capitalist civilization” (66; also 71).

I agree that reciprocity is in tension with the organization of society on a competitive basis (though Jason Kuznicki expresses some reasonable doubts, based on cases of apparently legitimate market behaviour within friendships). Reciprocity is a promising basis for socialism, as it is a widely shared value. Yet reciprocity also poses an important obstacle to socialism’s feasibility.

Cohen discusses Joseph Carens’ proposal for a socialist society that consciously uses the market for the sake of allocative efficiency, while using the tax and transfer system to cancel out all of the inequalities the market generates. People act so as to maximize their pre-tax income (subject to some prerogative for personal time and projects), while fully understanding that post-tax they will not end up richer than anyone else who works equally hard (incomes will legitimately differ based on preferences for leisure vs the products of labour). From the point of view of reciprocity, this system faces a major hurdle. What assurance does anyone have that their fellow citizens are fulfilling their productive duties? For example, it might be my duty to take the most lucrative employment I can find (pre-tax), even if I like it less than some lower paying job, knowing that post-tax my income will be the same. Are others willing to do the same for me? In any pluralistic society, many people will not share the socialist ethos. And even if the ethos is widely shared, uncertainty can drive noncompliance. Even if I know that we both want to comply, I may be uncertain about whether you know that I know this, and so wonder if you might not comply due to this doubt. In short, if community is based on reciprocity, and if reciprocity has a conditional structure – if reciprocity is distinct from unilateral altruism, as Cohen seems to accept – then there are motivational challenges to socialism’s feasibility, for large-scale societies, not just informational challenges (though information is involved too).

I’ll conclude by mentioning two further issues. First, it’s not clear that this problem is simply one of feasibility, as opposed to desirability. For reciprocity is an element of community, which is a positive value.

Second, Cohen thinks that community goes beyond justice, in the sense that it imposes requirements on us that are in addition to the requirements of justice. Yet reciprocity is closely related to justice. I think we could be committed to reciprocity between citizens without being committed to a strong form of political community, i.e. without always putting first the shared valuing of mutual concern at the level of the polity as a whole. We can be committed to community at different levels, after all.

  1. This essay is available as a short book, but I am referring to the version in the edited collection Democratic Equality: What Went Wrong? For a summary of the book, see Herbert Gintis’s Amazon review. What I say here builds on his points about disagreement and free-riding in large groups, and his other work on reciprocity. ↩︎
  2. In “Publicity, Reciprocity, and Incentives” I used reciprocity to defend the basic structure objection to the socialist critique of productive incentives, but I didn’t discuss “Why Not Socialism?” ↩︎

Review of Moller’s Governing Least

I’ve reviewed Dan Moller’s Governing Least: A New England Libertarianism at Notre Dame Philosophical Reviews. For other responses, see :

Bryan Caplan

That’s the only response I can find, at the moment, but here’s some related material from Moller:

podcast on

Bleeding Heart Libertarian post

plus a one-sentence-per-chapter summary of the book that I wish I’d read first.

Reciprocity and Populism

Reciprocity and Populism

Samuel Scheffler asks whether Rawls’s theory of justice has anything to say about the rise of Trump and populism. Rawls’s theory is primarily about what the right principles of justice are, not what explains people’s behaviour. But Rawls did discuss the stability of conceptions of justice, and in doing so he relied on the empirical hypothesis of motivation by reciprocity, which may help explain recent events.

Stability: Imagine that a given a set of principles were successfully implemented in a society’s public institutions. Would that society have a tendency to correct any subsequent deviations from justice that might arise? The answer depends in part on whether people growing up under those institutions would develop an effective and appropriate sense of justice. Rawls thought that his own conception of justice was more stable than utilitarianism because it fit better with reciprocity, the tendency to respond in kind. He viewed motivation by reciprocity as a deep trait of human nature, constructed by evolution rather than culture and social institutions. Utilitarianism permits inequalities if they maximize the aggregate, which could require that some accept less than others so that those others can have even more than they would otherwise have, e.g. [2,4] < [1,7], because 6 < 8. In contrast, Rawls’s “difference principle” requires that inequalities benefit all, which in a simple two-position scenario means that inequalities must increase the size of the smaller share (hence [2,4] > [1,7] because 2 > 1).

Scheffler draws on this connection between reciprocity and the difference principle to suggest that what accounts for the rise of Trump and populism is in part the United States’ “failure to achieve – or even to strive seriously to achieve – an ideal of reciprocity.” During the last 40 years, the US has experienced “skyrocketing economic inequality.” It is almost as if the US has been intent upon maximizing the size of the biggest share rather than the smallest. No surprise, then, that we are witnessing a backlash, the consequence of “resentments inspired by sustained inequality.”

In a recently published response to Scheffler, Paul Weithman argues that resentment at rising inequality is a misdiagnosis of the causes of populism. Weithman draws on Arlie Hocschild’s study of Tea Party supporters in Louisiana. In a widely quoted passage, Hochschild uses the metaphor of people waiting in line to realize the American Dream, while others, aided by government, cut in line. Hochschild argues that support for Trump and populism is motivated by a sense that the government is taking from ordinary hard-working folks to give a leg up to people who may be in need but who aren’t deserving, the line-cutters, people who feel no shame in taking government handouts when they could work, handouts paid for by the taxes of others who are willing to work. “The government has become an instrument for redistributing your money to the undeserving” (Hochschild, quoted by Weithman). Weithman points out that the resentment Hochschild describes is not directed at the great gains made by the better off. Instead it is directed downward.1

“Hochschild’s interviewees clearly think their government violates a demand of reciprocity. But they do not think it has done so by allowing objectionable gains to the best-off. They think it has done so by allowing objectionable gains to the worstoff.”

Weithman has two official explanations of the problem. The first has to do with the fact that we can measure mutual benefit with respect to different points of comparison or “benchmarks.” Hochschild’s respondents look at what they had in the past, and what they would have had had they not been taxed so heavily to pay for benefits for those below them. In contrast, in Rawls’s theory equality is the benchmark, so these comparisons to historical baselines are not relevant. The second explanation has to do with the fact that mutual benefit from inequality gets complicated when there are multiple social positions, instead of just two. In a multi-position scenario, the situation of the least well-off can be improved at the expense of those who are just above them, the second-worst off, who are still far from the best off.2 Weithman suggests that we might reduce downward-oriented resentment by adopting universal programs that would reestablish “chain connection,” which was Rawls’s terms for the situation in which gains to the worst off also benefit the next worst off, and so on up the line.

There is more obvious failure of reciprocity, however, in the perceptions of Hochschild’s respondents. It’s right there in Hochschild’s quote; desert. Weithman does mention “merit and hard work.” And he notes that Hochschild’s respondents wouldn’t accept Rawls’s baseline of equality because they don’t think that the ‘takers’ are making the same contributive effort as are the ‘makers’. Weithman doesn’t say anything further about this perceived violation of norms of contribution and personal responsibility, however.

What are people resentful about? Is it that the worst off have gained while they, the next worst off, or those in the middle, have lost? Not exactly. If we accept Hochschild’s results, what people resent is at that the worst off have gained unfairly at their expense, by cutting in line – this is the perception she reports – while they (the next worst off) have played by the rules. The rules in question are rules of personal responsibility and social contribution. The problem with the government programs her respondents object to (and the taxes that support them) is not that only some people get them; as Martin Gilens argues, that isn’t perceived as a problem if the people are blamelessly in need, as in the case of medical care in old age. No one can avoid getting old. The perceived problem is that people can modify their behaviour to make themselves recipients by avoiding work (i.e. moral hazard).

For me, the question Hochschild’s study raises is whether Rawls hasn’t missed the boat on desert. The idea of individual deservingness plays an ambiguous role in his theory of justice. On the one hand, inherited inequalities of income and wealth count as morally arbitrary because they are undeserved; no one chooses their parents. These inequalities therefore have to be justifiable to everyone, in order to be consistent with mutual respect in a society of equals. On the other hand, Rawls went to argue that ideas about desert were out of place when invoked as standards of assessment for basic social institutions. (By acting within the rules people would gain entitlements of various kinds, but the system of entitlements was not designed as a means to rewarding the deserving, nor of punishing the undeserving).

Many people have questioned Rawls’s theory of justice on the grounds that it assigns no fundamental role to desert (e.g. David Miller and Michael Sandel). Popular thinking does seem to assign an important place to desert. So it is striking to read Rawlsians debate the bases of populist resentment without explicitly engaging the issue. The debate is about whether resentment results from the outsized gains of the better off, which don’t seem to be benefitting everyone else, or from the small gains of the worst off, which are perceived to come at the expense of the next worst off or of those in the middle. But if the real issue is about perceived violations of reciprocity at an individual level, these issues about changes in relative standing of different parts of the income distribution are not fundamental. One might even argue that the Rawls’s theory is not helpful for thinking about our current situation, precisely because it rejects desert as a principle of justice.

However, I think that would be a mistake. For reciprocity can itself be a ground of desert. We say that one good turn deserves another, after all. The thought is not that benefiting others may be a good way of eliciting benefits from them, though this is often true. It’s that in many contexts it is fitting to respond to receipt of benefits by benefitting others in turn, even apart from any prospect of future gains. From this perspective, government programs that provide assistance unconditionally may be perceived to undermine reciprocity, since they make it possible for some who are able to benefit others not to contribute but simply to receive. (Reciprocity can also be understood as a limit on general duties, rather than a duty to return benefits).

There has been a lot of discussion of this objection to unconditionality, the objection from reciprocity or exploitation. The issue is complicated for a number of reasons, in particular:

  1. the reciprocity objection might be valid other things equal but not decisive all things considered, when balanced against other considerations, e.g. expanding people’s effective range of choice.
  2. reciprocity might speak against unconditionality in some respects but *for* it in others, e.g. if it would help mitigate a lack of reciprocity between men and women with respect to household labour (see pp.98-, the section on “free-riding and household labour”).
  3. the demand for reciprocity might apply only if surrounding institutions are not too unjust, making conditionality itself conditional.

I’d like to make a more general point about markets and reciprocity. In thinking about reciprocity and unconditionality of government assistance, we need to think about the alternative, which is market determination of income. And we should recognize that a competitive economic system encourages non-reciprocal behaviour. It is true that if I want something you own, I have to give you something I own in return. That is a kind of reciprocity, if one compares it to the alternative of my taking what you have by force (paying the iron price, as they say in the Game of Thrones’ Iron Islands, rather than the gold price). However, a competitive system puts pressure on people not to reciprocate good behaviour in the past, but to maximize future returns. When in a downturn or due to technological change firms let go workers whose skills are superfluous, they will often have to ignore the fact that these individuals worked conscientiously at their allotted tasks. There is a social benefit to ignoring reciprocity, in such cases, which is that these choices send signals about scarcity of resources relative to preferences, but this is an argument based on efficiency, not fairness. (For an illustration of this point, see G.A. Cohen’s discussion of the way his father’s career in a Montreal clothing factory ended, at pp.180-1 of If you’re an egalitarian, how come you are so rich?)

  1. In Scheffler’s defence, it might be said that even if resentment is not consciously directed at large gains of the better off, it may still be caused by it. Stagnation of median wages along with large gains at the top end can make it feel that the line is getting longer, getting stretched out. And then when some are perceived to be cutting in line, the reaction is stronger than it might otherwise be. 
  2. For example, with four positions, consider A:[1,4,7,10] vs B:[2,3,8,15]. B > A by maximin because 2 > 1, but there is arguably more inequality in B, and it’s not benefitting everyone, since 3 < 4.