I. Context

R. H. Tawney’s Equality, published in 1931, passionately sets out the case that Britain of the time was afflicted by morally unacceptable inequalities. Like many writers today, Tawney was disturbed by the great disparities between those with the greatest wealth and income and those who were struggling even to maintain a minimally healthy diet. Like thinkers to follow, Tawney pointed out the steep inequalities in the mortality rate between affluent and impoverished districts even within the same city. Material inequality and its effects are, Tawney argued, of great concern. But, it seems, Tawney thought that there was something worse, or rather, that material inequalities were a symptom of something worse: something he called “the religion of inequality.” Above all, Tawney argued against rigid class distinctions, in which aristocratic privilege was protected and reproduced over the generations, while those at the bottom of the scale were deprived of education, opportunity, and political influence even though they were the great majority. People of different social classes barely met or mixed, even as children. For Tawney the evil that a society of equals must overcome is clear: it is class division and its reproduction.^{Footnote 1}

I share much of Tawney’s methodological approach. First, I reject the limited idea that equality should be regarded as an exclusively distributional concern. Rather, like Tawney and now many others, I want to ask what it would be to create a society in which each person regards and treats each other person as an equal. Second, like Tawney, and again like many others, I believe that the only sensible way to tackle this problem is first to understand the social inequalities that afflict us in contemporary society. Engaged political philosophy should attempt to arrive at solutions that emerge from reflection on problems. Rather than proceeding by formulating a theory of justice, we should identify injustices in the world and use whatever conceptual resources we have on hand to help us think our way through, and out of, such problems.^{Footnote 2}

However, there is one vital respect in which I need to depart from Tawney, and that difference provides the motivation for this essay. While Tawney was able to concentrate on social class as the dominant barrier to social equality, the world has moved on. We now need to consider barriers of gender, sexual orientation, race, disability, age, religion, region, and national identity, as well as class. In a full treatment, we would even look too at more voluntary or informal groups, such as gangs or people who adopt a certain style of dress and music. It may still be that social class is the most important element, and it could even be that many other categories reduce to economic class, but we cannot take that for granted in advance.

Although I have argued elsewhere that having a strong social network can be very advantageous for an individual,^{Footnote 3} I want here to look at the other side of group membership and how it can lead to forms of social exclusion, or, better, “differential social inclusion.”^{Footnote 4} Given the complexities of the modern world, the term “social exclusion” does not seem quite right, as it connotes a binary opposition of inclusion and exclusion. There are at least four reasons why a binary division is unhelpful. First, a given individual will generally fall into many groups and can be included in one respect (as a middle class professional) and excluded in another (being a member of low status religion or caste).^{Footnote 5} Second, exclusion implies the existence of a privileged “mainstream,” but there can be several potentially competing mainstreams, and a group or individual can be included in one but not another. Third, groups that are themselves excluded can in turn exclude others by, for example, monopolizing a particular type of employment, and indeed such exclusion can be mutual. Finally, exclusion is a matter of degree, from the subtle ways in which conversational groups form at social gatherings to sustained and organized violence. Hence we live in a world in which groups can be both included and excluded at the same time, and in different ways, and any individual can be a member of many such groups. While it is still necessary to talk about social exclusion in particular cases, the macro-picture is one of differential social inclusion, with a complex network of inclusions and exclusions. The basic idea of social exclusion is simply that members of some groups will be less able to “fit in” with a privileged mainstream of society than others.^{Footnote 6}

Brian Barry has argued that social exclusion typically creates two types of problems for egalitarianism: one related to the unfairly diminished opportunities experienced by people in marginalized subgroups, the other related to damaged social solidarity.^{Footnote 7} I don’t say that the tensions created by the relationship between sub-groups and the mainstream(s) are the only barriers to social equality, as forms of hierarchy and domination often exist even within groups, and independently of group membership, such as within a workplace. Nevertheless, social exclusion is a problem that deserves consideration in the context of philosophical study of social equality. We have to consider the possibility that the social forces that create and perpetuate exclusion are so strong that they set limits on the possibility of overcoming social inequality in the modern world.

II. The Problems

I want to explore what subgroups are, how and why they form, what (good and bad) effects they have and why they have them, and what intellectual and policy response is appropriate from the point of view of social equality. We can break this inquiry down into seven questions, as follows, which will be the subject of the following sections:

1. a) What is to be understood by a group?

2. b) Why do groups form?

3. c) How do groups bind themselves together?

4. d) What are the disadvantages of group membership?

5. e) What are the advantages of group membership?

6. f) Why is differential social inclusion a problem?

7. g) What policy responses are possible?

III. What Is To Be Understood by a Group?

First of all, what is a group? For many purposes it is possible to rest on an intuitive understanding, but for the sake of later analysis, it will be helpful to add some substance to the idea, if not a strict definition. I start from some insightful remarks of Erving Goffman, in relation to what he calls “stigma categories.” Goffman points out that people with a stigma in common:

This brief account needs a degree of supplementation. First, there is no reason to restrict the analysis to stigmatized groups: similar accounts can be given of members of Mensa, or elite athletes, or indeed academics. Second, there is no reason to insist that people need a conscious belief that they belong to a group. All they need is a vague understanding, not necessarily articulated, that “these are people (more) like me (than many other people).” And third, I would wish to generalize “treatment of each other” to behavior more generally.

Consider, for example, an exercise class. It is likely that if the class only contains women — whether designed this way or not, and whether this is in fact consciously observed by the participants — it is plausible that at least some of the women will behave differently than they would have done if men were present. Or consider the often-noticed phenomenon of a lively conversation falling into silence when someone not of the group (a boss, a woman, a policeman, a non-philosopher, a very famous philosopher) enters the room. Groups can be reinforced by patterns of language use, involving distinctive accents, idioms, and topics of conversation.^{Footnote 9} And outward facing behavior matters too. Richard Sennett defines “tribalism” in terms of showing solidarity to insiders but aggression to outsiders.^{Footnote 10} Furthermore group membership will affect individual behavior in the sense of what you are more likely to do in the company of people with whom you feel more comfortable. Indeed the behavior can be very substantial, including acquiring (or deliberately failing to acquire) traits, skills, or qualification.

As in many attempts to give an account of a group, there is something circular about this approach, as it appears to require people to recognize themselves as members of a group in order to engage in behaviors that mark them out as a member of the group. But what I want to take from this analysis is simply that someone may recognize particular others as being “more like me” than the general population, and in the presence of these people may be more likely to behave in ways that either come more naturally in this company, or that test the members’ loyalty to each other, thereby reinforcing both its internal identity and its difference from those who do not regard themselves in that way. Whether the group has a sense of identity, or a name for itself, or is named by others, seems less important, although these will all intensify the group’s existence.

Typically the notion of an excluded subgroup requires contrast with “mainstream” society. How do we understand that notion? It is not the same as “the majority,” for in societies such as apartheid South Africa the mainstream was numerically the minority.^{Footnote 11} However it is not the same as “the hyper-elite” for as Barry also argues there may be groups in society who split themselves off from the mainstream at the top end, living in gated communities with their own security and making little or any use of public education, health or transport. Although his focus is not on equality as such, Goffman suggests that in the United States of the 1950s there was a type of default nonstigmatised male identity, the

It would be wrong to restrict the mainstream to the group of people exactly like that. But many societies, historically, have had a focal or aspirational point not unlike this, and the mainstream in a narrow sense marks out this realm of privilege. It includes people who aspire to this model, their families, and now, but perhaps not in the 1950s, their female equivalents. How much it also includes people of different races, ages, religions, and physical characteristics will differ from society to society. They are, or are on their way to becoming, economically successful, and are also socially at ease having the feeling that they “belong.” Although they may have to work hard to earn the money they want, socially they “fit in” without any special effort.

Yet there are two complications with this vision. First, in addition to this elite notion of the mainstream, there is a broader idea of the mainstream too, which comes closer to the idea of being part of the more prestigious end of the formal economy. A cleaner or kitchen worker employed by a city law firm is part of mainstream society in the broad sense, but not in the narrow sense. In contrast, a wealthy person who has taken up the lifestyle of a bohemian artist is likely to be part of the mainstream in the narrow sense, but not in the broad sense. But a second complication is that it is, of course, too simple to suppose that there is only one mainstream in modern society. There may be competing mainstreams, for parallel to the commercial world there could be a mainstream political or intellectual or religious elite, or other elite groups formed around the world of entertainment or sport, all of which have their narrow and broad understandings. This, of course, adds to the complexity of the overall picture, and this complexity is one reason why “differential social inclusion” is a better term than “social exclusion.” Members of a group — the black political elite, for example — will be included in some “mainstreams” but perhaps relatively excluded from others.

IV. Why Do Groups Form?

The question of why groups form takes us deep into sociology, anthropology and psychology, and naturally there will be diverse and sometimes competing explanations. Here I want to proceed in a somewhat more superficial way, looking at examples from the literature rather than pushing the analysis to its most fundamental level. I also wish to restrict the scope of the discussion to socially excluded groups, which is to say groups that may face reduced opportunity or prospects for solidarity, rather than groups that suffer no disadvantages or even gain advantages. Broadly, we can distinguish “push” factors, which lead mainstream society to exclude a group, and “pull” factors that may lead a group to isolate itself. Yet we need to keep in mind that a “pull” factor may often be a response to, or an anticipation of, a “push” factor: it could be existing discrimination that makes a group forge a collective identity.^{Footnote 13} It is probably rather common that much social exclusion combines push and pull factors together.

The most obvious “pull factor” is the preservation, or less commonly the invention, of cultural identity, which leads to a rejection of some of values of mainstream society, or, more mildly, a celebration of some values that are not valued by the mainstream. For example, immigrant groups may have language, religion, food and other traditions that split them off from the mainstream. And this can be true for immigrant groups that have maintained an identity over many generations, especially those that differ in physical appearance from the mainstream. It is likely that such groups form and are reinforced in their identity both by the pull factor of wishing to maintain cultural identity, and push factors (to be described shortly) that keep the group out of the mainstream. Hence the existence and reproduction of such groups is, on the face of it, a threat, of varying seriousness, to social solidarity. Whether members of such groups also face reduced opportunity will vary. In some cases there may be no loss of economic opportunity. In other cases a person’s appearance, or manner of speech, or, on an application form, a name and an address, can be very revealing. Any of these factors could be enough to make it the case that a recruiter, consciously or unconsciously, does not shortlist a candidate for a job or for university admission.^{Footnote 14}

A second type of pull factor also, arguably, interacts with push factors. Paul Willis, in Learning to Labour focused his study on a group of male teenagers soon to leave school, who rejected academic values and were deliberately very disruptive to school discipline. The study took place in the British midlands in the early 1970s. These boys, known as “the lads” were of varying intelligence and it seems some of them would have had the capacity to do reasonably well in their schoolwork, but chose not to. They did not take exams, and left school at the earliest opportunity, to take whatever manual labor they could find. They then worked for a wage that will barely change in real terms over the course of their working lives. Other boys were more diligent in their schoolwork, and sought apprenticeships, which while low paid at first will, if all goes well, lead to high-skilled manual work and good pay and prospects. There is a type of reciprocal rejection between “the lads” and mainstream society. Of course the factory work for which they are destined is in a sense part of mainstream society, but not close to the “aspirational” part. Willis regards the behavior of the lads as a type of politics of resistance, or a protest against mainstream norms.^{Footnote 15}

In one way, of course, the behavior of the lads seems a perverse type of protest, as it seems to reinforce the low-income existence that they are protesting against. Yet Willis sees a subtlety, or at least a dilemma, in the boys’ outlook. Once the lads see themselves as a group, and as part of a wide working class protest movement (in an attitudinal, rather than political sense), individual advancement becomes highly problematic:

A related observation is made by Richard Hoggart: “when working-class people are asked to become foremen or N.C.O.’s [noncommissioned officers] they hesitate. Whatever their motives, they will be regarded now as on the side of ‘Them’”.^{Footnote 17} Even if groups are created through the push factors considered below, once they exist loyalty can hold the group together. Few want to be regarded as a class or race traitor.

So far, then, we have seen two mechanisms of group formation that have at least some element of pull to them: cultural identity, and the politics of resistance. Push factors, presumably, will largely be related to the idea that some groups are “different,” most likely inferior, and in the worst cases “polluting,” as in the excesses of the caste system. Ambedkar, for example, points out that in some places the lowest caste members in India were required to tie brooms to the their waists so that their footprints were swept away as they walked.^{Footnote 18} The term “untouchable” is, of course, intimately related to this idea of pollution. Is it a mystery how these ideas develop and are reinforced?

In some cases, where one group encounters another for the first time, it is not unreasonable to think that fear could enter the equation. Rousseau, in the discussion of the origin of language claims: “Upon encountering others, a savage man will at first be afraid. His fright will make him see those men as taller and stronger than himself. He will give them the name Giants.”^{Footnote 19} Rousseau also suggests, though, that this will be a temporary phenomenon, and after many encounters the savage will realize that the strangers are no stronger and taller, and therefore will invent a new name — “Man” — to express commonality rather than difference.

Rousseau may be over-optimistic. The initial sense of fear, strangeness or difference may or may not fade. Or rather fear can take a different form. Segregation into groups, with a clear relation of superiority and inferiority, can have great advantages for the advantaged groups. As Goffman reports:

Goffman’s words should strike a chord in relation to current political debates about immigration, for example. Opposition to immigration is almost always explicated not in racist or discriminatory terms, but in terms, in effect, of pressure on resources and opportunities, which in turn is an implicit plea for protection of self-interest in terms of preserving one’s own access to those resources. The debate concerns competition — for jobs, housing, school places, medical care and welfare benefits — and is conducted in terms of whether such worries are well-founded or ill-founded (for example whether the long-term benefits of immigration outweigh temporary adjustment problems). That stigmatization is a form of protection of relative privilege from competition has long been apparent to those on the receiving end.^{Footnote 21} In this connection it is worth commenting on an argument from Robert Putnam that genuine equality of opportunity would give a substantial boost to the U.S. economy.^{Footnote 22} Even if Putnam is correct, some, mostly lower-skilled workers, may well lose even if society as a whole gains, which may explain different patterns of attitudes to immigration at different social levels.

I have suggested that groups form through preservation of cultural identity, or from protest or resistance, but also through members of the mainstream regarding people with certain attributes as strange or threatening, especially in terms of competition. I do not want to argue that these are the only relevant mechanisms, but they are surely candidates for being among the most important.

V. How Do Groups Bind Themselves Together?

There are many mechanisms by means of which groups hold themselves together. I have made expected patterns of behavior a central feature of groups, in that people who are members of a group are likely to modify their behavior in front of other members of the group, relaxing, feeling more comfortable, or at least feeling expected to respond in particular ways to each other, and to participate in collective behavior toward outsiders. As noted above, use of language is a well-known marker of difference. Wider patterns of discourse — who can speak, who is taken seriously — are increasingly understood to mark boundaries of inclusion and exclusion.^{Footnote 23} Gossip has also been argued to be a powerful binding and differentiating force.^{Footnote 24} Dress, decoration, and other aspects of appearance also bind groups together, and mark them out from others. These attributes, as well as aesthetic judgments, can fall under the rather ubiquitous heading of “taste,” which is, of course, documented as a measure of “distinction.”^{Footnote 25}

The distinction between conscious and unconscious behavior is important, although it is too simplistic to regard this as a distinction rather than a continuum. At the most conscious extreme, there are forms of behavior deliberately undertaken in order to reinforce group activity in a way that is designed to mark the group off from the mainstream. Examples might include “the lads” in Willis’s study collectively playing truant from school even when doing so is more boring than being at school, or the anti-authoritarian attitudes and behavior of African Americans living in what Tommie Shelby calls “the Dark Ghetto.”^{Footnote 26} At near the other end of the scale are the patterns of play of poor and working class children studied by Annette Lareau. These children, she reports, play in informal mixed aged groups, including other family members (especially cousins), under their own initiative and not under the supervision of adults.^{Footnote 27} While this play is in its way highly appealing and valuable, these children do not thereby receive the training in discipline, time keeping, team work, and relations with adults experienced by the much more regimented play of middle class children. In this way, so it appears, they are disadvantaged when looking for and holding down jobs, college places and positions of responsibility. Although the play is, of course, conscious, its long-term effect of solidifying disadvantaged class position is not recognized, though highly fateful for future opportunities. In between these examples of actions deliberately undertaken to reinforce “counterculture” and those that are not even recognized as group-reinforcing behavior, there are examples such as Hoggart’s factory worker who refuses a position as foreman. This is taken out of group loyalty, but without necessarily being an act of protest (although it could be).

And here I should add a further dimension that I have discussed in more detail elsewhere.^{Footnote 28} One of the binding mechanism of a group, I noted above, is dress and decoration: wearing the right brands of clothes and shoes, and having the right hair and nail styles, for example. And it is also likely to mean enjoying social pleasures of mixing with other group members in places where money is spent, such as bars, clubs, gambling premises, hairdressers, and nail bars. It can be increasingly important for those who are excluded from mainstream society to make a special effort to fit in with local expectations, both in how one looks and how one spends one’s time (and money) with others. Doing so can have several functions, such as signaling that one is a member of the group, or that one does not bear the “shame” of poverty (even if in fact the costs of such social displays and interactions lead to considerable financial difficulty in other aspects of life).^{Footnote 29} This type of binding is a defense mechanism against exclusion from a group one cares most about, while at the same time emphasizes difference, and uses time and money in particular ways, all of which can intensify exclusion from other groups.

VI. What Are the Possible Disadvantages of Membership of a Subgroup?

Being a member of a subgroup can have a range of disadvantages, many of which are very well-known and have been touched on already. The most obvious is that one can be a victim of direct discrimination, in terms of employment, education, housing, and other resources. And in these cases Goffman’s speculation that stigmatization and hence discrimination reduces competition seems apt: it is a way of preserving the chances of a relatively privileged existence for those occupying mainstream positions. Direct discrimination is comparatively easy to confront, even if it can be very stubborn to remove. More difficult to detect and deal with is indirect discrimination, which comes in many well-known forms.^{Footnote 30} The general point is clear: that membership of a group can make some people much more able and other people much less able to fit into, or flourish within, mainstream institutions.

VII. What Are the Advantages of Group Membership?

If being a member of a group leads to a range of disadvantages, why do marginalized groups persist? To some degree we have seen an answer to this question: it can be in the interests of the mainstream to keep others out of competition and the creation and maintenance of subgroups serves this function. Once excluded a group may develop attitudes and behavior that reinforce exclusion.

It is not difficult, then, to understand why the mainstream “constructs” marginalized groups. And once groups have formed, being a member of a relatively privileged group can bring what Elias has called “gratifying euphoria.”^{Footnote 31} But very often members even of marginalized groups reinforce the group’s existence. Partly this will be through unconscious or semi-conscious patterns of behavior and response. Still, that is not enough to explain why individuals value group membership even when their membership reduces their life prospects.

The first and most obvious thing to say is that it appears that human beings simply have a basic need to “fit in.” They will, in general, be happier and less anxious if they feel part of a social group. This can also have beneficial effects. For example, it is well known that those who are active in religion have better health and live longer than those who are not. The reasons for this may be complex. Religious belief, or at least practice, will often encourage healthier life choices, or it could be that there is an indirect health effect of being an active member of a group. For example, those with a supportive social network will be more likely to be able to take complete rest when suffering from illness, rather than having to continue to work for income or to maintain a household, and thereby delaying, or never achieving, recovery. Religions are often very good at dealing with illness. Finally, there could simply be a health effect of group membership that arises from the reduction in chronic stress.^{Footnote 32} Some of these benefits would be available to members of more damaging groups, such as gangs, or a countercultural underclass, although these positive effects are likely to be swamped by the adverse health consequences of the associated lifestyle.

VIII. Why Is Differential Social Inclusion a Problem?

We saw that there are two central difficulties presented by social exclusion: reduced economic opportunity and reduced social solidarity. In fact, we can now see that these problems are very closely related. The fact that there is social solidarity within, but not across, groups reinforces a social pattern that creates reduced opportunity for marginalized groups. Charles Tilly has called this “opportunity hoarding” by the dominant group.^{Footnote 33} And it may be that the experience of reduced opportunity leads to a set of defensive actions that further reduce social solidarity: marginalization becomes self-fueling. In fact, as Tilly argues, opportunity hoarding can take place at all levels in society, when, for example, a manual trade is monopolized by members of one racial or national group to the exclusion of others. This, I noted in the introduction, is a further reason why “differential social inclusion” is a better descriptive term than “social exclusion.” We will explore what these patterns of exclusion mean for policy in the next section. What I want to do now is to look at the importance of opportunity and solidarity from the standpoint of social equality.

Opportunity is normally understood in terms of access to valued resources such as education and the workplace. In respect to the workplace, it is important, though, not to reduce opportunity to income. If kitchen porters and waiters received the same income, and, in fact, even if the conditions of work for the two jobs were comparable, there would still be a prima facie problem of equality of opportunity if all the waiters in a city were of one race and all the kitchen porters another. In this sense, then, notions of opportunity seem to straddle distributive and relational concerns. It is discriminatory or insulting to be refused one type of opportunity, even if other opportunities are equally well paid.

Nevertheless, it is worth keeping the ideas of economic opportunity (work and income) and status apart. In much of what I have outlined above I have talked about the diminished economic opportunity of people in marginalized groups. It is undeniable that this is a distributive matter. Yet it may fairly be asked whether a social or relational approach to equality should care so much about economic opportunity. Is that not to revert to a concern with distributional equality?

A first reply is to distinguish a concern for distribution with a concern for distributional equality. No one attracted to equality can ignore issues of distribution, such as poverty. Poverty is bad in itself and also, if avoidable, shows a contempt for the people who are allowed to live in poverty. To go a little further, anyone who is interested in equality of any sort can agree that a fairly high level of material sufficiency is a prerequisite for a society of equals. But it does not follow from this that material equality is essential for social equality. What is needed is enough to lead a rich and flourishing life.^{Footnote 34} And this point transfers to opportunity. From a purely economic point of view it is not necessary for opportunities to be strictly equal. It is enough if everyone has the opportunity to acquire the means to a rich, fulfilling, poverty-free life.

Saying this, though, does not in itself mean that inequality of opportunity above a threshold should always be tolerated, and this is where issues of status and solidarity come back into play. Patterning in employment is commonplace and should raise our suspicions. Suppose it turns out that there is a particular correlation between certain jobs and ethnic origins. For example, in the UK in 2011, 17.4 percent of all registered pharmacists were of Indian origin, 6.3 percent Pakistani, and 0.9 percent Bangladeshi.^{Footnote 35} This contrasts with the presence of these groups in the general population at the following levels: 2.5 percent Indian, 2 percent Pakistani and 0.75 percent Bangladeshi. Hence those of South Asian origin are over-represented, as are those of black and mixed-race ethnicity. In fact, the only group that is significantly under-represented among pharmacists are people of White British origin, who make up 80 percent of the population but only 51 percent of pharmacists.^{Footnote 36}

In this case it is not hard to offer an explanation. Being a pharmacist is relatively well paid (though not spectacularly so) and is a high status position, yet requires hard work and application over many years in order to qualify. Hence, it is an aspirational route for the socially mobile, attracting disproportionate numbers of talented individuals from previously disadvantaged groups. In the case of other prestigious positions where the training is substantially less arduous and the rewards are even higher, it may well be that the proportions switch in the other direction through opportunity hoarding of the elite.

Suppose, counterfactually, we had created a world in which people of all ethnic and other marginalized groups have sufficient economic opportunities to live rich and fulfilling lives, free from immediate need. Yet it still remains that the best housing, the best jobs, and the best leisure opportunities tend to go to members of the elite group. This would seem to indicate a failure of social solidarity between the elite and the rest. What does that mean? To make progress it would seem that we need an account of the requirements of social solidarity in order to understand the claim.

Failures of social solidarity come in various strengths. In severe cases, members of the marginalized group are treated as ripe for domination (for example they should have little say about their working arrangements) and exploitation (low wages, menial jobs). More moderate cases involve unarticulated assumptions about hierarchy: that people in other groups or from other backgrounds have to understand “their place.” This is likely to involve forms of snobbery and looking down, with the expectation of servile behavior. It will also include assumptions that members of certain groups have claims that are worth less, or are improper, or can be ignored. By contrast, one aspect of social solidarity, as I understand it, is the negation of these negations: not to look up or down, but to treat all “on the same level,” worthy of equal concern. Interestingly, from the point of view of social equality we seem to be pushed toward policies of equality of economic opportunity, not because economic equality is important in itself, but because policies that allow for unequal opportunities seem to offend against solidarity, treating some groups as worth less than others.

I should add, though, that there is another, stronger, idea of solidarity that is also important for the present analysis: regarding oneself as a potential collective actor with others, especially in the face of adversity.^{Footnote 37} The two notions are distinct: an army exemplifies solidarity in the “collective action” sense, but not the “all on a level” sense, as might a business fighting for survival in the face of severe economic competition in which the workers see their jobs under threat if profits are not restored. Equally, there could be solidarity in the “all on a level” sense without any occasion for collective action, as, for example, in a well-functioning neighborhood. This collective action sense of solidarity once more helps us understand that a binary notion such as social exclusion is too simple. The workers may act in solidarity with the bosses for some purposes, but in solidarity with each other against the bosses for other purposes. Furthermore, the two notions of solidarity reinforce the general point that different coalitions form for different purposes. Members of some ethnic groups will be invited to join the Chamber of Commerce, where all act together to campaign against increased business taxes, for example, but may be overlooked when the Country Club seeks new members. The collective action sense of solidarity may often be pragmatic, based on mutual self-interest, although it can also be based on mutual concern. However, it is the “all on a level” sense that seems to exemplify the idea of egalitarian solidarity.

IX. Policy Responses

Social exclusion is, in short, a matter of treating people as in some respect inferior because of their group membership. The theory of differential social inclusion accepts that a given individual can be a member of many different groups that form for different purposes, and in virtue of membership of different groups can sometimes be treated as included, and sometimes as excluded. In addition, there can also be different elite formations each competing to be “the mainstream.” Inclusion and exclusion can take place at all different levels in the social hierarchy, and, finally, come in different degrees. Where exclusion takes place it can be self-fueling in that it can lead to or encourage in-group behavior that reinforces expectations and patterns. Reduced social solidarity, which is problematic in itself, leads to reduced opportunity, which can reinforce poverty, but also solidifies ideas of difference. It goes without saying that a society that aspires to regard all of its citizens as equals needs to overcome differential inclusion understood in these terms, when it has the detrimental effects described above. But what routes are there to overcome social exclusion? In considering each possible strategy, it will be helpful to keep four questions in mind, although I will not pursue them mechanically in what follows. First, would the strategy help overcome problems of opportunity and solidarity? Second, if pursued with vigor and resources, how far could it go? Third, does it have other disadvantages from the point of social inequality? And fourth, critically, who can be the “agents of change,” bearing in mind that those with the greatest political power may well have a vested interest in preserving the status quo?

One possibility is that, over time, group difference will fade away simply through natural development. In the passage quoted from Rousseau, giants became men through familiarity. Goffman mentions the fact that categories can lose their stigmatized status. His examples are those of divorce and Irish ethnicity,^{Footnote 38} although he does not speculate on how this happens, or how permanent it will be. The casual anti-Semitism of pre-war writing has disappeared, but has anti-Semitism gone with it? Or consider the position for those of Japanese descent in the United States. By the standards of wealth and health, Japanese Americans are perhaps among the most privileged groups that have ever lived, but do they still suffer some forms of prejudice? Whatever we say about these examples, the real problem is that Goffman’s examples don’t seem to have generalized. Hence the proposal, in effect, to let things look after themselves, may be very limited and will offer no protection against adverse developments.

Goffman also presents another response to stigmatization, which is in effect to invert the relationship, through assertive self-affirmation:

Groups may contain a political or radical movement within them that attempts to pioneer this attitude. The Black Power movement, Gay Liberation, and Radical Feminism can be seen in this light. If successful, this response will improve members of the marginalized groups’ self-image, and reduce servility. Yet the degree to which this response overcomes other aspects of social exclusion will depend on the response of the mainstream, which can harden when challenged, especially in the short term, and so can intensify ridicule and hostile attitudes. In the longer term, assertive self-affirmation may be more be effective, albeit most likely as one tactic in a wider portfolio.

A third approach, integration, has been powerfully defended by Elizabeth Anderson. Anderson suggests that in the case of racial discrimination against African-Americans in the United States, there is an “imperative of integration,” writing “[I]f segregation is a fundamental cause of social inequality and undemocratic practices, then integration promotes greater equality and democracy. Hence, it is an imperative of justice.”^{Footnote 40} Later she says, “If racial segregation is the problem, it stands to reason that racial integration is the remedy.”^{Footnote 41}

Now, there is one understanding of these claims that makes them tautologically true. If racial integration is simply defined as the solution that social egalitarians seek, then the claim that it is the remedy cannot be contested. Yet in itself it yields no policy prescriptions. Alternatively Anderson’s claims, especially the first, can be understood causally, but then more will be needed to establish them. For, although the claims are tempting, there are two questions to ask. First, the degree to which geographic and social segregation is a cause, rather than a symptom, of social inequality and undemocratic practices is up for debate. But second, even if Anderson is correct that segregation is the root cause of other problems, it does not follow that we can revert to a previous situation by reversing the cause. Smoking causes lung cancer, but, sadly, giving up smoking does not cure lung cancer, for smoking has started a process that progresses independently of smoking. Similarly segregation may have put in place other mechanisms that cannot be undone by integration alone. This is an empirical question, not something that can be settled by what, apparently, stands to reason. Even if segregation causes inequality it does not follow that integration (either geographic or social) will end inequality. The damage may be too deep. And in fairness, Anderson is well aware of this, devoting considerable attention to the conditions under which integration does and does not lead to improved outcomes.^{Footnote 42}

But most importantly, even if Anderson is entirely right about how to overcome discrimination against Blacks in the United States, it is unclear how her proposals can generalize to differential social inclusion (I do not suggest that she claims that they will: she is looking at a very particular and important problem). In the incredibly diverse large cities of the world, with many groups held together internally, but divided from each other on racial, religious, ethnic, national, and even generational, bases, intersecting with issues of class, gender, disability, sexual orientation, and no doubt other factors, differentially included in multiple mainstreams, what does integration mean in practice? Integration yields clear policy proposals in societies divided on simple fault lines, such as Tawney’s class-divided Britain of the early-twentieth century, but in contemporary societies it is more limited in its promise. This is not to argue against integration, which is a needed and highly positive step to deal with some extreme situations, such as those discussed by Anderson, but it appears to have limited wider application. And it can also be argued that we would lose something very valuable in big cities if there were not districts associated with particular national or ethnic groups, with distinctive shops, restaurants, and leisure facilities, constituting an attractive place to settle for people with common culture or ancestry, provided, of course, that such differences were not associated with differential stigma and prestige.

A more modest response, which has been touched upon at various times above, is to try to develop stronger possibilities of social mobility. If individuals are unable to advance, economically, we need to understand why this is. In some cases the structure of an economy in itself creates few opportunities, and so the only socially mobile people will be those who are exceptionally able in some valued respect. In other cases, formally speaking, the opportunities are there, but the individuals concerned refuse to take them (as in Willis’s and Hoggart’s examples, discussed above). In further cases the individuals simply are not prepared in the right way to take advantage of the opportunities and need special training or advice. For example, one law teacher in Namibia insists that students in her (mostly black) class dress for business, in order to get them used to wearing the type of clothes they will need to wear for interviews and jobs, so they learn to “look the part,” when competing against white candidates who will be much more used to formal dress. A recent social campaign from a British bank includes a training program for disadvantaged teenagers in which they are taught to stand up straight, cultivate a firm handshake, and look their interviewer in the eye.

The idea of developing social mobility in this way really goes hand in hand with what we can call “cultivated assimilation.” Consider again Goffman’s examples of divorced people and those of Irish ethnicity. Why was membership in these groups ever a problem? It could have been pure prejudice. But it could also have been that these characteristics were associated with others that presented a problem. Perhaps for divorced men certain types of middle-class occupations were problematic, if, for example, it was common practice to have the boss around for dinner, or if other forms of social interactions related to the workplace were premised on the idea that all men would have a wife. For a divorced woman, in addition to reduced economic status, her social life would suffer too, given that connections via the husband would disappear and that on other occasions invitations might dry up out of fear that she might want to “catch” one of the married men who would be there. For those of Irish descent, it may be that early generations had not internalized the Protestant work ethic of U.S. society and were regarded as poor workers. Over time, attitudes, customs, and habits change, and the stigma disappears. One possibility is to try to accelerate these processes of assimilation via social policy. Through schools and education and other forms of training, individuals can come to develop the skills and attributes that they need in order to succeed. While people can have their own religions and cultural traditions, on this approach social policy needs to tame and assimilate them so that they are no bar to success in the workplace and in other competitive realms. Essentially this proposal accepts, even glories in, difference in the private sphere while trying to reduce it in the public sphere, so that all can compete on equal terms. It is possible that training individuals so that they are a better fit with the mainstream will not only improve their opportunities but be beneficial for their social groups by demonstrating that members of such groups are just as able as others. Mere social contact could also be beneficial from this perspective, although the point is controversial. This indeed is part of the idea behind integration.

However, there are several obvious difficulties, or at least limitations, to these policies. The extent to which the public/private distinction can be made to work will vary. If, for example, no accommodations are made for religious holidays then those who are orthodox in their faith will be no better off. But if accommodations are made, then the orthodox will, typically, be less desirable as employees because they will be able to work fewer days. (We may see the same with policies of generous maternity leave.)

Furthermore, those concerned with social equality may have mixed feelings about programs to improve social mobility. On the one hand, for the individuals involved, such programs could be highly beneficial. On the other hand, they take place within a structure in which there are fewer prized positions than people who seek them. Therefore, the proposal appears to be one of encouraging a fairer competition for unfair outcomes. As I noted above, Paul Willis claims that social mobility for the disadvantaged as a whole is possible only if the system is overcome. Similarly, Anderson discusses the successful Gautreaux program in Chicago, in which more than seven thousand poor black families were given the opportunity to move to integrated neighborhoods in Chicago. Although the program is beneficial on a fairly large scale, hundreds of thousands of poor black families in the region were nevertheless not given the same opportunity, and there are simply not enough houses in integrated neighborhoods to offer the program to everyone who could benefit. Again very large-scale social change is needed to create genuinely fair opportunities for all. Given, though, that this alternative is unlikely to be a realistic prospect, we simply have to do the best we can.

An alternative, which combines perhaps a more realistic and a more utopian element asks us to look again at marginalized subgroups, and to recognize their value. Rather than change the people to fit in with the mainstream, perhaps we should attempt to change the world so that the mainstream can be enlarged to incorporate cultural difference. Or perhaps a range of new “mainstreams” could be created. This is to use insights from the social model of disability to look at other forms of exclusion.^{Footnote 43} The first element of this approach is already happening all over the world: to bring the music, literature, dance, food, art, and other cultural traditions of marginalized groups to broader attention, so that others can understand their appeal and join in the celebration of a multicultural world. This may mean public subsidies, major cultural venues having festivals to celebrate diversity, and public broadcasting companies devoting time and resources to lead the appreciation of the unfamiliar. This is to tackle solidarity exclusion head on, by encouraging a broader set of values. Critically, if successful it will also devalue the importance of what was formerly the mainstream, reducing opportunity exclusion by removing a layer of prejudice. However, on its own it will not address the lack of fit between individuals and opportunities if those individuals have been brought up in ways that create a problem, such as the poor and working class in Lareau’s study who did not internalize the type of discipline required in the workplace. We have seen that one possible response to this is training programs to fill the gap. But a more solidaristic response is to change the workplace so it does not privilege one set of norms over another. This is what I referred to as the more utopian element in the package. Are we ready for workplaces where it does not matter if you cannot be relied on to show up on time, or meet a deadline? Perhaps new technologies that enable remote working do allow for more flexibility, but the norms of what will count as successful work practices may be very stubborn, and, many will argue, for good reason. For if changing workplace practices reduces productivity, the change will also reduce the tax base by which reform is financed.

Still, it may be that radical solutions are needed. Solidarity exclusion can best be directly countered by cultural policies that in effect refuse to accept the unique importance of the previous mainstream. The comparative model for dealing with opportunity exclusion is to reduce the dominance of the mainstream workplace and economy. One way of doing that, we saw, is to encourage alternative ways of making a living, as new technologies are already doing. Another is to ensure that individuals do not need to rely on the formal economy in order to acquire the level of resources they need for meeting the required level of sufficiency. In other words, if essential goods are free at the point of consumption, funded by taxation, then most people can have a decent life even if their incomes are relatively modest. This, of course, is the dream of the welfare state: free education, free health care, subsidized social housing, subsidized high quality public transport, and free or subsidized access to some cultural, sporting and leisure opportunities. Taxes would need to rise to pay for such things, which is politically highly problematic, but rises in taxes could have some benefits from the point of view of social equality, as well as economic equality, such as making it harder for the very rich to isolate themselves from society.

X. Conclusions

My goal in this essay has been to try to understand the nature of differential social inclusion, and to appreciate the difficulties it presents for social egalitarians. There are mechanisms that, if left to themselves, will exert a type of centripetal force, creating and reinforcing marginalization of subgroups, who will suffer reduced economic opportunity and diminished social solidarity.

I have reviewed some approaches to these problems: natural development; assertive self-affirmation; geographic and social integration; assimilation and social mobility; cultural validation; growth of the public sector; and changing workplace norms. Some of these approaches will overlap to some degree, but there are complex interrelations, and hard choices will need to be made. For example, social mobility may detract from the validation of minority cultures and also reinforce the exclusions of those left behind. Indeed those “moving up” could find themselves struggling to feel comfortable anywhere. The policies that seem most promising combine validation of the culture of marginalized groups with the development of new forms of work and the provision of a public sector that allows a comfortable life even for those who do not conform to mainstream norms. How effective will these policies be? And who would have the motivation and power to bring them into being, especially when those with the greatest influence may have much to gain from the status quo? Earlier I raised the question whether differential social inclusion is so stubborn that it makes the idea of social equality an impossible goal in the modern world. If so, then philosophical egalitarians may need to turn back to distributive equality in order to have a realistic aspiration. However, in my view it is too soon to draw this pessimistic conclusion.