Wednesday, June 27, 2012

Does Strong Reciprocity Support a Darwinian Left?

In 1998, the political left was dispirited.  The collapse of the Marxist regimes in the Soviet Union and Eastern Europe, the turn in Communist China from Maoism to markets, the unpopularity of welfare state programs, and the global expansion of capitalism spurred by international free trade--all these events seemed to show the triumph of capitalist libertarianism and the defeat of socialist egalitarianism.

In that year, two essays were published arguing that the revivification of the left would require a turn to Darwinian science.  "The left needs a new paradigm," Peter Singer declared in the British journal Prospect, and that new paradigm should be based on a Darwinian understanding of human nature, which would replace the traditional utopian vision of the left as based on the assumption of human beings as a blank slate that was malleable and perfectible.  Singer's essay was expanded into a short book--A Darwinian Left: Politics, Evolution, and Cooperation.

The other essay was written by Samuel Bowles and Herbert Gintis and published in the Boston Review, along with commentaries by thirteen authors.  (Large parts of this essay were published elsewhere: Bowles and Gintis [1998b] and Fong, Bowles, and Gintis [2005].)  Bowles and Gintis argued that the left could promote a new egalitarian politics based on a Darwinian science of human cooperation that would support egalitarian policies.  They wrote: "It is not self-interest that opposes the welfare state, nor unconditional generosity that supports it.  We will show that there is a solid foundation for cooperation and sharing in two basic human motives--we call them strong reciprocity and basic needs generosity.  Moreover, we argue that hostility to contemporary forms of egalitarianism is evidence for, not against, that deep foundation, and that new egalitarian initiatives are fully compatible with it."

"Basic needs generosity" was identified as the propensity to share with others to assure them some minimal level of the means of subsistence.  "Strong reciprocity" was identified as "a propensity to cooperate and share with others similarly disposed, and a willingness to punish those who violate cooperative and other norms--even when such sharing and punishing is personally costly." 

Bowles and Gintis then sketched the reasoning for explaining these and other grounds of cooperation as innate propensities of the human species as shaped by both genetic and cultural evolution under the influence of group selection through war, so that more cooperatve groups tended to prevail over less cooperative groups.

The essay in 1998 was an ideological manifesto or intellectual prospectus for Bowles and Gintis that has guided their writing over the past twelve years.  Their latest book--A Cooperative Species: Human Reciprocity and Its Evolution--continues to elaborate the arguments sketched in their 1998 essay.  Reading that essay and comparing it with Singer's essay illuminates their fundamental project in promoting a Darwinian left.

The similarities in the arguments of these two essays is striking.  Like Singer, Bowles and Gintis cite the work of Petr Kropotkin in Mutual Aid on the Darwinian evolution of cooperation as supporting a natural human disposition to generosity and reciprocity, in contrast to the mistaken assumption that Darwinism promotes a view of life as selfishly competitive.  Kropotkin thought that a socialist anarchism could be rooted in a Darwinian understanding of human beings as evolved to be naturally social animals.

Like Singer, Bowles and Gintis cite the research in behavioral game theory that explains the evolution of cooperation in ways that challenge the rational egoism of Homo economicus.  This research indicates that some people in some situations really are purely selfish beings.  But they are generally in the minority.  Some other people--also a minority--are pure altruists who cooperate unconditionally.  But the majority of people are conditional altruists--people who cooperate as long as they think others are cooperating and who are willing to punish cheaters who don't cooperate.

Like Singer, Bowles and Gintis think that Darwinian science can explain much of human cooperation as driven by kinship (cooperating with relatives) and reciprocal altruism (cooperating with non-kin as long as there is some expectation of future reciprocation of cooperation to repay the altruist).   The Darwinian logic of kin-selection has been elaborated by William Hamilton.  The Darwinian logic of reciprocal altruism has been elaborated by Robert Trivers.  The further elaboration of reciprocal altruism has included the logic of indirect reciprocity, developed by Richard Alexander, Martin Nowak, and others, which explains how individuals might cooperate if their reputations for being cooperative benefit them in the future, and if those who fail to cooperate suffer punishment from their bad reputations.

What is distinctive about the argument of Bowles and Gintis, however, is their claim that another evolved motivation is required to fully explain human cooperation--strong reciprocity.  Unlike the "weak reciprocity" of Trivers, "strong reciprocity" is the disposition to be cooperative and to punish cheaters even when cooperation and punishment is costly in ways that can never be repaid to the cooperator or the punisher.  Actually, as I have indicated in a previous post, "strong reciprocity" is a bad choice of words because by definition there is no future reciprocation. 

That's why the Darwinian explanation for how such a disposition could evolve requires group selection, so that the benefits for the whole group compensate for the sacrifices suffered by individual members of the group.  This is most likely to be explained through an evolutionary history of warfare, as Bowles and Gintis argue, in which individuals might sacrifice their own lives for the sake of their group, and those groups without such self-sacrificing individuals would tend to be eliminated.  This would favor the evolution of what Bowles and Gintis call "parochial altruism"--being cooperative within one's own group in competition with those outside the group.

Although Singer does not develop this idea of strong reciprocity in the way Bowles and Gintis do, he does come close to it when he says that to fully explain altruism, we need to explain not just "kin altruism" and "reciprocal altruism," but "genuine altruism toward strangers."  His illustration of such genuine altruism is the study by Richard Titmuss (The Gift Relationship) of the blood banks in Great Britain, which relied on voluntary blood donation with no possibility for the recipients to repay the donors (A Darwinian Left, 54-59).   Those on the left need to understand such genuine altruism so that they can look for ways to promote it as the basis for truly altruistic cooperation.  Singer wonders whether the left might need to encourage such altruism through social policies that reward this behavior with enhanced popularity and social status.  Of course, Bowles and Gintis would have to reject this, because the genuine altruism of strong reciprocity does not require any payback, and the rewarding of reciprocity with a good reputation would turn this into the reciprocal altruism of Trivers rather than the strong reciprocity of Bowles and Gintis.

And yet, as already indicated, even the strong reciprocity of Bowles and Gintis is not an unconditional altruism, because even the strong reciprocator will stop cooperating with cheators, and he will want to punish them.  Moreover, even the strong reciprocator--as a parochial altruist--is naturally tribalistic in favoring those within his group over those outside his group.  Even strong reciprocity falls far short of an utterly disinterested humanitarianism.

I am not yet fully persuaded that Bowles and Gintis have made a convincing case for the evolution of strong reciprocity.  But even if I were persuaded to accept their account of strong reciprocity, I don't see how this--along with the rest of their account of the evolution of cooperation--supports a Darwinian left, because what they say would seem to provide more support for a Darwinian libertarianism (or classical liberalism).  This came up a couple of years ago, when I wrote an essay for "Cato Unbound" on "Darwinian Liberalism," and Herbert Gintis was one of the commentators.  Although he seemed to largely agree with most of my points, he had to reject my general conclusion that a Darwinian science of human cooperation supported classical liberalism.  But I couldn't exactly understand his disagreement.

In their Boston Review essay of 1998, Bowles and Gintis argue that the left must give up its utopian view of human beings as a blank slate open to perfectionist visions of social transformation.  They thus seem to agree with Singer that a Darwinian left would require "a sharply deflated vision of the left, its utopian ideas replaced by a coolly realistic view of what can be achieved" (62). 

In fact, much of this "deflated" leftism would be acceptable to libertarians, who have long assumed that conformity to human nature and a "coolly realistic view of what can be achieved" must guide social policy.  For example, Singer agrees with Adam Smith about the benefits of a market economy in channelling the selfish motivations of human nature in ways that serve the public good; and he also agrees with Smith about the danger in trusting those in government to exercise concentrated coercive power over the economy and society.  Again, Bowles and Gintis seem to agree with Singer about this.

And yet while Singer generally rejects the utopianism of the traditional left, he concludes his essay and his book by suggesting that "our capacity to reason can, in the long run, take us beyond the conventional Darwinian constraints on the degree of altruism that a society may be able to foster," which could lead us to "the idea of an impartial concern for all of our fellow humans, or, better still, for all sentient beings" (62-63).  Here Singer takes us back to the utopian vision of a pure, disinterested altruism--a universal love for all human beings and even for all sentient beings, which would support Singer's argument for "animal liberation."  As far as I can tell, Bowles and Gintis never fall back into such utopianism, because even their strong reciprocity is a conditional and parochial altruism that creates tragic conflicts of interest that can never be perfectly resolved.

Bowles and Gintis reject the utopian hope of the traditional left that central planning through governmental coercion can achieve perfect social and economic justice and equality.  But even as they reject the naive confidence of the left in governmental planning as a substitute for markets, they also reject what they take to be the naive confidence of libertarians in the spontaneous order of free markets.  Bowles and Gintis argue for recognizing the importance of "community governance" in solving problems of collective action through the enforcement of social norms where both government and markets would fail.

But then Bowles and Gintis fail to see that what they say about "communities" or "community governance" corresponds to what libertarians would say about "civil society."  In my Cato Unbound essay, I argued that "evolutionary psychology has confirmed and deepened the Darwinian understanding of moral order that arises in civil society through the spontaneous order of human action rather than the coercive order of governmental design."  Gintis responded by insisting that there was no evidence for this because "every known society has a collective mechanism that deals with the establishment of social values and that regulates the treatment of individuals who violate social norms," and he called this "collective governance."  But what he is calling "collective governance" or "community governance" is exactly what I am calling "civil society." 

The point is that human social order can arise not just from governmental planning or market exchange but also from the natural (familial) and voluntary associations of life--families, neighborhoods, churches, schools, clubs, and all kinds of commercial and charitable enterprises.  Darwinian science explains how this is possible by explaining the genetic and cultural evolution of the social propensities that underlie these social arrangements.  This supports a Darwinian libertarianism that recognizes the proper functioning of all three sources of social order--governments, markets, and communities.

Although Bowles and Gintis come close to accepting this position, they can't quite bring themselves to give up the traditional left's reliance on governmental coercion as superior both to markets and to communities.  One can see this in their exchange in the Boston Review with Kevin McCabe and Vernon Smith.

McCabe and Smith are famous as economic experimentalists employing behavioral game theory.  Smith won the Nobel Prize in 2002 for his work in experimental economics.  Their work is cited by Bowles and Gintis.

McCabe and Smith generally agree with Bowles and Gintis in interpreting their experimental work as showing the inadequacies of the Homo economicus model and the need to understand the other-regarding character of human cooperation.  But while Bowles and Gintis think this research shows the need for governmental welfare programs, McCabe and Smith think this shows the need for cultivating the natural tendencies to charitable reciprocity among individuals and social institutions to take the place of governmental welfare programs.

McCabe and Smith write:
We think the widespread experimental and other evidence for the innate practice of reciprocity by some, but not all agents (roughly half our subjects choose reciprocity, and half choose noncooperative play when given a choice) in decentralized processes of social exchange, suggest that the problem of distributional transfers might best be handled by individuals, acting alone or through social institutions.  That is, the alleviation of poverty should largely be considered none of the government's business, and in this sense belongs in the same category with drugs, abortion, health care, population size, prostitution, schooling, and religion. . . . Centralized power yields abuse as certain as death, taxes, and clumsy government intervention.  Basically, people cannot trust governments to do their bidding because each is bidding differently, and the aggregation of the bids faces implementational infeasibility.  Information on the fine structure of distributional preferences is dispersed, and no central authority can deploy resources to effectively satisfy those preferences.  These ancient truths have been well expressed by Bastiat, Hume, Locke, and others.
McCabe and Smith would prefer to end the governmental welfare state completely and allow private charity to take its place.  But if a more gradual reform is required, they recommend a system of tax credits to provide incentives for charitable reciprocity.  You could allow taxpayers to make direct contributions to poor individuals or to charitable organizations helping the poor, and these contributions would count as 100% deductions from their taxable income. Over time, these individual contributions would take the place of the governmental welfare budget.   In fact, most of the most effective programs even today for helping poor people are privately administered and privately funded.  But the rising budgets of governmental welfare programs tend to crowd out these private programs.  Eliminating the governmental welfare system would allow the social preferences for reciprocity to fully express themselves through private charitable activity to benefit poor people.

What McCabe and Smith recommend has been elaborated by libertarians like Michael Tanner (2003) who shows how the governmental welfare state could be replaced by private charity.  He concludes:
Civil society would rely on a reinvigorated network of private charity.  An enormous amount of evidence and experience shows tht private charities are far more effective than government welfare programs.  While welfare provides incentives for counterproductive behavior, private charities can use their aid to encourage self-sufficiency, self-improvement, and independence.  Private charity can individualize their approaches and target the specific problems that are holding people in poverty.
Most important, private charity is given out of a true sense of compassion, which forms a moral bond between giver and receiver.  Private charity enriches the lives of everyone involved and helps to nurture the true tendrils of community. (107) 
Oddly enough, in response to McCabe and Smith, Bowles and Gintis insist that "community governance" and the motivation to reciprocity cannot be trusted to do this, because only government can be trusted to alleviate poverty.  "The absence of poverty is a public good, like environmental quality or public safety, the provision of which through private markets is bound to be insufficient.  Voters have strong preferences to live in a society which protects all of its members against dire poverty, and these preferences simply cannot be implemented by private acts of charity."

But if Bowles and Gintis really believe that human evolution has endowed us--at least many of us--with propensities to "strong reciprocity" and "basic needs generosity," then why shouldn't they believe that helping the poor is best done through communal institutions that enforce social norms of fairness and benevolence? 

If they were to fully embrace the Darwinian insight into how social cooperation arises as an unintended emergent order of our evolved social instincts, they might then have to finally reject the position of the traditional left in favor of Darwinian libertarianism.

Some previous posts on related themes can be found here, here., here, and here.


Bowles, Samuel, and Herbert Gintis.  1998a.  "Is Equality Passe?"  Boston Review, November.

Bowles, Samuel, and Herbert Gintis.  1998b.  "Recasting Egalitarianism."  In Erik Olin Wright, ed., Recasting Egalitarianism: New Rules for Communities, States, and Markets, 361-97.  London: Verso.

Fong, Christina M., Samuel Bowles, and Herbert Gintis.  2005.  "Reciprocity and the Welfare State."  In Herbert Gintis, Samuel Bowles, Robert Boyd, and Ernst Fehr, eds., Moral Sentiments and Material Interests: The Foundations of Cooperation in Economic Life, 277-302.  Cambridge, MA: MIT Press

Tanner, Michael D.  2003.  The Poverty of Welfare: Helping Others in Civil Society.  Washington, DC: Cato Institute.


Empedocles said...

I wonder if you could say more about which areas of life are best left to governments, which to markets, and which to communities.

W. Bond said...

Gintis does not strike me asleftist, but more as a former leftist.

Most of his positions lead to a position in the modern world of classicliberalism. I suspect, however, he maypolitically favor, e.g. the so-called economic “happiness”literature as a way to intellectually justify a broader welfare state.

By analogy to Darwin’s grandfather’s quote about Unitarianismas being a “featherbed for falling Christians,” this rather suspect“happiness” literature strikes me as being a featherbed for fallingprogressives.

D. McCloskey has a nice recent piece in the NewRepublic titled: “Happyism.”

Larry Arnhart said...


Yes, I think you're right: Gintis is a "former leftist."

It would be helpful, however, if Gintis and Bowles could explain their change of mind, and how Darwinian science contributed to that change.

Actually, even Singer's DARWINIAN LEFT gives up so much of the traditional vision of the left that one might wonder--as many readers of his book have--what's left of the left in the "Darwinian left"?

To me this indicates how any commitment to the Darwinian science of human cooperation tends to push one towards classical liberalism or traditionalist conservatism.

Larry Arnhart said...

As I have suggested in my posts, the change that Bowles and Gintis have undergone over the past 14 years can be quickly seen by comparing what they say in the December 1998 issue of the BOSTON REVIEW with what they say in the May/June issue.

One should also note that in 1998--in their RECASTING EGALITARIANISM-- they were arguing for guild socialism (worker-owned firms)established through a coercive governmental redistribution of assets, while now they don't talk about that.

Anonymous said...

It seems to me that once you drop the notion of egalitarianism and the blank slate (which has always been the left's faith-based ideology), accept the fact that behavior is largely heritable, and that kin selection is (and must be to exist) the basis of altruism, it is difficult to understand how the left can recast itself.

Thus in order for the left to survive, it must ignore these Darwinian principles and instead adopt the notion of group selection, which is indeed a dangerous idea, especially for the left.

Kent Guida said...

Thank you for taking the time to examine Gintis' thought in such detail. I have been puzzled about that ever since his comment in Cato Unbound, which made no sense to me.

One more thing. Darwinian Conservatism, Darwinian Liberalism, Darwinian Libertarianism -- they all mean exactly the same thing, right? Isn't this a great source of confusion and real marketing problem? There are many people who would never, under any circumstances, accociate themselves with anything carrying the 'conservative' label. Ditto for the other two labels. Each is a red for someone. I think this has a large effect on how your thought is understood and received. I wish I had a solution to offer.

Larry Arnhart said...


Yes, you're right about my use of labels.

I still believe that contemporary American and British conservatism is rightly understood as a fusion of classical liberalism and traditionalist conservatism. But I know that the tension between these two strands of thought is confusing.

I tried to make my argument for Darwinian science as supporting this fusionist view of conservatism in my lecture to the Philadelphia Society some years ago.

My general impression is that my arguments find a somewhat warmer reception with the libertarians/classical liberals than with the conservative traditionalists.

Troy Camplin said...

Here's a thought: if humans exhibit "strong reciprocity," then there is absolutely no need to have governments force people to be generous.

Larry Arnhart said...


I agree.