Behavioural politics: social insurance and mutual benefit in Robert Sugden’s The Community of Advantage

Abstract

In The Community of Advantage, Robert Sugden advocates a system of universal social insurance to ensure that all citizens enjoy the benefits of market institutions. The Community of Advantage however does not shed light on the extent of redistribution that social insurance should provide. While the criterion of mutual advantage seems to licence only a minimal level of insurance, people’s expectations and collective bargaining power are likely to lead to a much more extensive redistribution. This is not necessarily a bad thing, since a well-functioning society is more than a well-functioning market economy. An extensive social insurance protects the rich, the poor, and, above all, the cooperative attitudes without which a democratic society cannot function well.

1 Introduction

The Community of Advantage (2018) (TCOA, from now on) is Robert Sugden’s first major book since The Economics of Rights, Cooperation and Welfare (1986). After thirty-two years, Sugden owed his readers a monograph integrating the research that he has done across economics and philosophy. TCOA lives up to the expectations: it includes the most advanced contractarian justification of the market system since Gauthier’s Morals by Agreement (1985); a devastating critique of Libertarian Paternalism; and a new interpretation of the findings of behavioural economics for the design of social policy.

There is enough material in TCOA for several commentaries, so I will have to be selective. I will try to keep my eyes on the forest rather than on a single tree, metaphorically speaking, focusing on the general implications of Sugden’s contractarianism. In the second section, I will highlight the Humean roots of his approach, in particular Sugden’s commitment to design institutions that will be acceptable to people as they are. Section 3 is devoted to a central problem of Sugden’s theory, namely the problem of defining the extent of redistribution or the amount of social insurance that is required to compensate people for their brute bad luck and ensure that market institutions work to everyone’s advantage. Sugden argues that mutual advantage must be measured relative to a baseline, and that the baseline is the status quo. In Sect. 4 I will argue that this idea may conflict with the Humean principle of taking people as they are: in the real world, the baseline will be determined in large part by people’s expectations, as well as their relative bargaining power. Section 5 tries to reconcile this fact—and the large amount of redistribution that it implies—with the ideal of democratic equality. In a well-functioning society, redistribution is needed not only to ensure that everyone benefits from market institutions, but also to protect the rich from the poor, the poor from the rich, and more generally to safeguard the cooperative attitudes that are necessary for the well-functioning of institutions.

2 Contractarianism and behavioural politics

There are many ways to be a contractarian. Robert Sugden is a contractarian in the tradition of David Hume. First and foremost, he is a methodological Humean: he believes that social philosophy must be based on the careful observation of men and women as they are, not as philosophers (or social scientists) would like them to be. There is a strong anti-utopian spirit in TCOA, best exemplified by Sugden’s critique of the “view from nowhere” that runs throughout the book.

There are also affinities with James Buchanan’s constitutionalism, which has always exerted an influence on Sugden’s work. Sugden’s contractarianism differs from Hume’s and Buchanan’s in one important respect, though. In the Essays, Hume famously mentions the “maxim” that “in contriving any system of government, and fixing the several checks and controls of the constitution, every man ought to be supposed a knave, and to have no other end, in all his actions, than private interest” (Hume 1777). The knavery principle has played a central role in the work of Buchanan and other contractarians.¹ But behavioural economists, including Sugden, have seriously discredited the idea that people are always—or for the most part—rational maximizers of their own material interests.

Hume himself recognized that “it appears somewhat strange, that a maxim should be true in politics which is false in fact”. Although Hume’s remark was based on casual observation, half a century of systematic research in psychology and economics have shown that people are not predominantly selfish, and even when they are, they are often unable to pursue their own self-interest in a coherent manner. These deviations from self-interest and rationality surely must be taken into account by legislators, economists, and political philosophers. A genuine Humean must subscribe to the “realism” rather than to the “knavery” principle.

Something like the realism principle imbues the work of another giant of contractarianism. In A Theory of Justice, John Rawls rejects any metaphysical justification of the constitutional principles that must regulate a “well-ordered” society. He claims that “conceptions of justice must be justified by the conditions of our life as we know it” (Rawls 1971, p. 454). This implies, according to Sugden, that “the governing principles of the society must be understood and accepted by its members. But if a conception of justice is to be public, the hypothesis that individuals accept its principles must be consistent with the facts of human psychology” (TCOA, p. 174).

Rawls was not a social scientist, so unsurprisingly the realism principle plays a marginal role in A Theory of Justice. But as a prominent behavioural economist, Sugden is in a much better position to explore its implications. It is a gigantic task: taking the realism principle seriously means nothing less than building a new scientific discipline. We may call it “behavioural politics”, in analogy with behavioural economics. It is both an exciting and a scary analogy: unlike behavioural economics, behavioural politics is in its infancy. There are many scattered pieces of research on political psychology, but no unifying framework. And its normative implications remain largely unexplored.²

TCOA may be a seen as a foundational contribution to this project. Unsurprisingly, Sugden raises as many problems as he tries to solve. There is work for a whole generation of researchers, and he must be praised for pointing the way forward. In the following sections, I will try to follow his footsteps, highlighting some unresolved issues and tensions of his contractarian proposal. In particular, I am unsure that Sugden’s proposal is really “consistent with the facts of human psychology”. Perhaps it does not have “psychological stability” in Rawls’ sense. And in general, I cannot find in TCOA a precise answer to the key question of the extent of redistribution that is necessary to make market institutions acceptable to all the members of society. As I will try to explain, these are important questions not just theoretically but also practically, in the strange climate of today’s politics.

3 Contractarianism and fairness

TCOA has the great merit of emphasizing how peculiar contractarianism is, in the context of modern political philosophy. A contractarian advises people, not philosopher-kings or dictators. She does not take an impartial perspective. She aims at identifying institutions that it is in people’s interest to choose, regardless of whether they are “right” or “good” from an external viewpoint.

A central thesis in TCOA is that the market is an institution that people have an interest to choose. The market clearly is not the only institution that calls for a justification, but TCOA is very market-centred. The justification of market-based allocations takes a large amount of space in the book, and other institutions are mainly presented as necessary complements to the market system. This market-centrism may be explained by Sugden’s desire to subvert a way of thinking about markets that has become engrained both in economics and in political philosophy. But it does lead to demote some important issues, as we will see shortly.

TCOA is not an apology of free market liberalism. Unlike Buchanan, Sugden is the sort of libertarian that is liked by people on the left. The requirement that the social contract must be for everyone’s advantage puts a considerable veto power in the hands of the least lucky citizens. The claim that a viable contract must include some form of social insurance against “brute luck”, in particular, is an important theme in chapter eight, which is one of my favourite chapters of TCOA.

Sugden’s concern about brute luck, however, is not egalitarian in spirit. It stems from a well-known tension: although markets are extremely good at generating wealth (or, in Sugden’s preferred currency, opportunities), they cannot be fair. As Friedrich Hayek and others explained a long time ago, uncertainty makes markets indispensable. But it also disrupts fairness, because implementing a fair distribution would destroy the system of incentives upon which the market system is based.

It is often contended that it is unjust to let the burden of such unforeseeable changes fall on people who could not foresee them, and that, if such risks are unavoidable, they ought to be pooled and the losses equally born by all. It can, however, hardly be known whether any particular change was unforeseeable for all. The whole system rests on providing inducements for all to use their skill to find out particular circumstances in order to anticipate impending changes as accurately as possible. This incentive would be removed if each decision did not carry the risk of loss, or if an authority had to decide whether a particular error in anticipation was excusable or not. (Hayek 1976: 125)

This is why, according to Sugden, market institutions cannot be evaluated with respect to a fair baseline. They must be evaluated using the principle of mutual advantage:

there is a fundamental tension between the idea of a fair baseline and the principles by which market economies work. Psychological stability has to rest on continuing expectations of mutual benefit, defined relative to a baseline that evolves over time and that cannot be justified in terms of abstract principles of fairness. (TCOA, p. 174)

Sugden recognizes that beliefs about fairness do, as a matter of fact, inform people’s expectations. But such beliefs are a variable that may affect people’s behaviour, rather than an external reference point for the evaluation of social policy. A Humean contractarian, as we shall see, cannot completely ignore what people think about fairness. But she is not bound to design her policies so as to approximate the ideal of social fairness, whatever it may be. Mutual advantage ultimately determines what is or is not acceptable to the members of a society.

It is useful to see how these abstract issues bear on the mundane but pressing political problems that we currently face. After the crisis of 2007–2010, many citizens of the world’s wealthiest, democratically most advanced countries have started to question market institutions. Sugden’s theory offers a simple and plausible explanation: these citizens cannot see what’s in it for them. They ask: if others (the super-rich, the bankers, the élite) are benefiting from the current institutional set-up, but I am not, then why should I accept the rules of this game? It is a legitimate question for a contractarian, and a very troubling one. What is at stake is the future of market-based capitalism, and perhaps of liberal democracy itself.

So the social contract must be evaluated for its capacity to generate mutual advantage. Mutual advantage is relative, in the sense that it must be measured with respect to a reference point or baseline. I will address the problem of identifying the baseline shortly. But first I would like to ask another question: even in the best circumstances (when the market does generate mutual advantage for everyone) what do people think about market capitalism? The evidence indicates that many people do not perceive it as a cooperative enterprise.³ So their view of the market is at odds with the one that Sugden aims at promoting. The market is perceived as a competitive arena in which ambitious, hard-working individuals use their skills and labour to acquire wealth, primarily for themselves and their families. Self-interest, competition, and meritocracy are seen as essential ingredients of the way the market works.

As every entrepreneur knows, cooperation is an important ingredient, especially in ongoing relationships with suppliers and customers. But so is competition: entrepreneurs use competition to their advantage, while constantly trying to protect their business from it—via innovation, specialization, acquisition, patenting, and so forth. And most people are not entrepreneurs anyway, so they see mostly the negative side of competition, the one that erodes salaries and destroys jobs.

Sugden acknowledges this. There is a brave—and rare—discussion of “pecuniary externalities” in TCOA (p.189). A pecuniary externality is the effect that a change in market prices may have on a third party who is not directly involved in the transaction. The effect may be positive, when for example a decrease in demand allows a formerly priced-out buyer to enter the market; but it may also be negative, when a number of suppliers enter the market and reduce the profits of an incumbent firm. Pecuniary externalities differ from standard externalities because their effects are conveyed through market pricing, rather than outside of the market (as is the case with, e.g. pollution). For this reason, economists typically do not consider pecuniary externalities as a problem that ought to be fixed.

And yet, pecuniary externalities are a major source of dissatisfaction, perhaps indeed the main reason why markets are perceived as unfair. Workers in Western countries feel the threat of competition from cheap labour in developing countries, and call for protection. They think that protection is due in part because the threat of competition is not uniformly distributed: it harms different groups to different degrees. While some people benefit from the low price of unskilled labour, others suffer from it. Who benefits and who suffers is, moreover, a matter of brute luck.

Blaming the market is a category mistake, according to Sugden. The market cannot be criticized for being unfair, since it is not supposed to be fair in the first place. Still, the realism principle comes into effect at this point: people’s feeling of being treated unfairly cannot be ignored. And the mutual benefit principle also calls for action, since a viable social contract ought to be advantageous for all. We need other institutions to fix the problems that a market system may create. Sugden is an advocate of universal social insurance, a system of compulsory taxation and redistribution that is able to compensate for brute bad luck.

4 The baseline of the social contract

Sugden’s view of social policy does not differ radically from that of a social democrat. Free markets and redistribution are its main pillars. The devil is in the detail—in the balance between market freedom and redistribution that is licenced by contractarianism. Unlike Rawls, Sugden does not take equality as the baseline for redistribution. He does not believe that all the resources generated by a market system are potentially up for grabs, because for him market inequalities are not a necessary evil—a means to incentivize the productivity of the most talented members of society—that redistribution must try to mitigate. The balance, in Sugden’s picture, is weighed in the opposite direction: the necessary evil is redistribution, a price that we have to pay to ensure that a market-based social system works to everyone’s advantage.

Yet, we are not told what the right balance is. There is some vagueness in Sugden’s social contract. Two questions require a precise answer: first, How much compensation is in order? And compensation with respect to what?

If redistribution is a form of compensation for bad luck, we need to specify a baseline or reference point. Since there is no room for a fair baseline in Sugden’ contractarianism, the only possible baseline is the status quo:

a contractarian recommendation in favour of a market economy needs to show each participant, looking ahead from where she is now, that she can expect the institutions of the market to work to her benefit. It needs to be able to do this, not just at some specially tailored starting line, but whenever “now” happens to be. And the recommendation must acknowledge the unpredictability of markets and the brute luck that is an unavoidable component of the rewards they give. (TCOA, p. 195)

The important qualification is “looking ahead from where she is now”. There are two key concepts: now (the status quo) and looking ahead (expectations). Depending on how we interpret these two concepts, we get different forms of contractarian social policy. According to one version—which I think is Sugden’s—the outlook is entirely objective. Each individual must take an unbiased look at the situation, and judge whether the social contract is advantageous for her. The past does not matter; in particular, one’s evaluation of the future should not be affected by one’s habits or the affluence she has gotten used to.

This interpretation fits Sugden’s stern philosophy of life. It is consistent with the “never complain and never explain” attitude, the idea that people must take responsibility for their choices, that informs TCOA’s defence of the opportunity criterion (see, e.g. pp. 105–6). I have nothing to say against this attitude, from a normative viewpoint. Living in a country of notorious moaners and complainers, I actually find Sugden’s stern philosophy deplorably uncommon. The problem is that a Humean should take people as they are, not as he would like them to be.

The hypothesis that people’s judgements and expectations are affected by past experience must be taken seriously (Hume, the philosopher of habit and custom, would have approved). Let us take a fictional character, Jack, who has been working in a car factory for many years, earning an annual salary of roughly twenty thousand euros. Jack discovers that cars will be manufactured in Romania from now on. It is a classic case of brute bad luck: Jack’s job and welfare would have been more secure had he chosen decades ago to work in, say, IT. But Jack cannot be faulted for not having predicted future political and technological developments forty years ago. So compensation is due.

It is a fact of human psychology that in asking for compensation Jack will not take his status of unemployed person as a reference point. His past experience—as Jack, the car factory worker with a twenty thousand-euro salary—will affect his expectations, i.e. the just level of compensation that he thinks the other members of society owe him. Of course we might tell Jack that his claim has no basis. No one has the right to enjoy a certain level of well-being in the future merely because he has enjoyed it in the past. But this normative reply does not fit Sugden’s contractarianism neatly: what rights people should or should not have is what the social contract is supposed to define. One cannot prejudge the issue by ruling out some concessions by fiat.

So how much compensation is Jack entitled to? For a Humean, this question cannot be answered independently from the question “How much compensation does he think that he is entitled to?”. Jack’s psychology—as well as everyone else’s, of course—is one of the variables that eventually will determine which social contract is accepted. We know very little about behavioural politics, so it is impossible to give a list of the social and psychological factors that are likely to play a role. But some well-documented mechanisms surely ought to be taken into account. Jack probably suffers from loss aversion for example. Realistically, people are not going to like contracts that will make them worse off with respect to what they have become used to. Although minimal unemployment benefits are objectively advantageous compared to no income, for Jack they may not be enough. And a Humean contractarian cannot ignore such feelings: the baseline is Jack as he really is; Jack “looking ahead from here”, Jack affected by loss aversion; Jack willing to reject the liberal social contract because it frustrates his (biased) expectations.

The situation sketched here is a sort of bargaining game between Jack and the rest of society. The empirical evidence indicates that many factors—commitments, emotions, moral beliefs, conventions, as well as sheer bargaining power—jointly determine the solution of these games.⁴ So how much is good enough for the “Jacks” of society is likely to matter. And this may have unpleasant consequences.

Suppose that Jack has already received significant subsidies, before he takes unemployment benefits. In most developed countries, about one-third of the population pays most of the income tax.⁵ In Italy taxpayers in the income bracket from zero to twenty thousand euro (24 million people) contribute only 13% of tax revenue, for example. If we add those who do not pay taxes at all (19 million citizens), we find that roughly two thirds of the population enjoy healthcare, education, protection, democracy, and the services of the public administration for free or at a highly subsidized rate. Jack would be one of them.

I am not suggesting that this situation is wrong. I am just assuming that most people (Jack included) take this baseline—a situation in which they enjoy considerable benefits—for granted. It is a psychological fact, and behavioural politics must accept the facts. Sugden in fact mentions “normal expectations” as a stabilizer of the social contract:

Principle of mutual benefit. When participating with others in a voluntary interaction, and for as long as others’ behaviour in that interaction is consistent with this very principle, behave in such a way that the other participants are able to satisfy normal expectations about the consequences of the interaction with them. (TCOA, p. 162)

This principle is introduced in the context of Sugden’s justification of market institutions. But surely it cannot be limited to that. It must play a role in the justification of every institution, including social insurance. But what if expectations vary? Should we take a difference in expectations at face value? Or should we dismiss some of them as unjustified?

Let us introduce a new character: Jack’s niece, Jill. Born in the 1980s, Jill has rarely had a permanent job, and has never earned twenty thousand euros per year. As a consequence, Jill’s expectations are more frugal than Jack’s. Should her social insurance be poorer, only because she has a lower baseline?⁶ I take that the intuitive answer is “no”. But what is the basis for such an intuition?

5 Social insurance and democratic equality

Social contracts crumble if too many people are dissatisfied. This may be the situation we are facing now. A powerful coalition of disappointed citizens is questioning institutions that they perceive as dysfunctional, or at least as not serving their own interests. To form a coalition of these citizens is not easy, but several politicians are working on it. Such a coalition, if and when it will take shape, may be able to negotiate a new social contract using the weight of sheer numbers. And perhaps behavioural politics boils down to this: social contracts (and the extent of redistribution) reflect bargaining power.

Sugden dismisses this possibility,⁷ perhaps too quickly. The use of bargaining power sits uneasily with Sugden’s idea of society as a cooperative enterprise, to be sure. But the concept of mutual advantage leaves too many questions unanswered as we have seen. In particular, it leaves the issue of the extent of redistribution indeterminate, so something is needed to fill that gap.

Having dismissed the protection money account, Sugden quotes Arrow approvingly: “Social insurance can be seen as an integral part of a system of economic organization based on mutual advantage. The underlying intuition is well expressed by Kenneth Arrow”. Here is Arrow’s idea, quoted by Sugden:

My own view is that in some deep sense there are increasing returns to scale. … There are significant gains to social interaction above and beyond what individuals and subgroups can achieve on their own. The owners of scarce personal assets do not have substantial private use of these assets; it is only their value in a large system which makes these assets valuable. Hence, there is a surplus created by the existence of society as such which is available for redistribution. (Arrow, 1984: 188) (quoted in TCOA, p. 281)

It is a surprising quotation: Arrow, like Rawls and other egalitarians, seems to say that the surplus generated through social cooperation does not rightfully belong to particular individuals, but to society as a whole. If so, then all the surplus is potentially available for redistribution. Sugden’s choice of quoting this paragraph is surprising because the egalitarians rely on a normative premise—about the just ownership of surplus—that a contractarian like him should reject: no prior notion of justice should be taken for granted when we draft the contract.

Arrow’s principle moreover would justify redistributions that are much more radical than anything that Sugden seems willing to accept in the rest of the book. Suppose that the “existence of society” accounts for 99% of Lionel Messi’s income (because, without the fans, his football skills would be worth almost nothing). Then, according to Arrow, 99% of his income, minus what is strictly necessary for incentive purposes, belongs in the common pool that is up for redistribution. But this reasoning is diametrically opposite to Sugden’s justification of social insurance. Social insurance is not justified because the wealth has been produced “by society as a whole”. It is rather a residual portion of the surplus, that we pay to make the social contract acceptable to everyone, especially the unlucky ones.

It seems strange that Sugden should follow Arrow down the egalitarian road. But then, he is left with the protection money account. What is possibly wrong with seeing social insurance as the price that people pay for living in a free, rich society? The price may vary depending on the balance of power, but it would guarantee that everyone benefits from the social contract, at least to some extent. Moreover, bargaining is a fact of life that a coherent practitioner of behavioural politics cannot ignore.

The expression “protection money” sounds distasteful if it is interpreted narrowly, as a form of extortion that the majority of citizens impose on a productive or just lucky minority. But if it is interpreted broadly, it may become more palatable. First of all, social insurance should not just protect the rich, but also the poor. And second, more generally, it should protect the social mechanisms that make cooperation possible.

I suspect that a broader conception of protection money would lead to a justification of redistribution similar to the one offered by “democratic equality” theorists (e.g. Miller 1997, 1999; Anderson 1999, 2008). This may seem odd, given Sugden’s starting point. His principal concern is the justification of markets, not of democracy or equality. But Sugden and democratic egalitarians share a deep dissatisfaction with classic arguments for redistribution, in particular the idea that the effects of brute bad luck must be erased or at least mitigated for the sake of fairness. Elizabeth Anderson has claimed that in tackling these issues.

recent egalitarian writing has lost sight of the distinctively political aims of egalitarianism. The proper negative aim of egalitarian justice is not to eliminate the impact of brute luck from human affairs, but to end oppression, which by definition is socially imposed. Its proper positive aim is not to ensure that everyone gets what they morally deserve, but to create a community in which people stand in relations of equality to others. (Anderson 1999, pp. 288–9)

The obvious question is why—what is so valuable or important about standing in relations of equality to others?

If … we invoked social equality, we would argue that income differences on the current scale unavoidably translated themselves into social divisions. We would argue, for instance, that those whose incomes fell below one-half of the mean income (say) were very likely to feel alienated and excluded from social life, would experience an immense social gulf between themselves and those living a comfortable middle-class lifestyle and so on. But even though the burdens of inequality may be seen as weighing most heavily on those who have least, the argument in this … case is not centrally about unfairness in distribution; the claim is that our society should not be like that. (Miller 1997, pp. 224–5)

David Miller points out that modern democracies are built on a strong anti-hierarchical, anti-aristocratic ethos. Aristocratic statuses create pervasive, lasting, immutable asymmetries of power, which tend to spill from one domain to another. Hierarchies then may become a source of oppression, humiliation, and loss of self-esteem. And they will undermine solidarity, cooperation, trust—the virtues that support a well-functioning market society.

A democratic society, then, should provide three kinds of protection:

• 1. Protection of lucky citizens against the frustrated expectations of unlucky ones. This corresponds to the narrow form of protection money that Sugden dismisses in TOAC. The reality of bargaining power, however, should induce a Humean contractarian to take it seriously.

  It should also provide

• 2. Protection of unlucky citizens against the unconstrained ambitions of lucky ones. People are ambitious and productive to different degrees, and hence will accumulate different amounts of resources. But even if they were equally competitive and productive, brute luck would generate unequal allocations through the market process. Such inequalities would probably be cumulative over time, and unfortunately the unconstrained accumulation of wealth in the hands of a few individuals would create the possibility of using those resources to oppress the majority of citizens. Capping the resources of rich citizens is a way of preventing democracy from sliding towards oligarchy or even tyranny.⁸

• 3. Protection of the cooperative attitudes without which a market society cannot thrive. This is the most general and also the least measurable kind of protection that a well-functioning society should provide. Massive inequalities of wealth tend to promote social segregation. If lucky and unlucky, rich and poor citizens hardly ever meet, trust is undermined. Suspicion thrives in a segregated society, misunderstandings are rife, and cooperative attitudes tend to fade. Order can be kept in such a society, of course, but at a great financial and social cost. In general, a considerable amount of violence has to be exercised, making it difficult for both rich and poor citizens to enjoy the opportunities and wealth they have created.

6 Concluding remarks

Behavioural politics is almost non-existent today. If it will develop, it will have to answer the questions that Sugden has identified in TCOA. I have noticed that TCOA does not shed light on the key issue of the extent of redistribution that social insurance should provide. I have also pointed out that a Humean contractarian is bound to the realism principle, and that this principle may have problematic implications. In particular, people’s expectations and collective bargaining power are important factors that shape the social contract: in the end they may lead to much more extensive redistribution than Sugden seems willing to accept. But in the last section, I have also argued that a well-functioning society is much more than a well-functioning market. For this reason, we should reject the narrow view of protection money and endorse a democratic view of redistribution. Social insurance offers protections of different kinds: protection for the rich, for the poor, and, above all, protection of the cooperative attitudes without which the market—or any other institution, for that matter—cannot function well.