1The aim of this article is to present a theoretical interpretation of important recent changes in the semantics of political communication about welfare issues. We believe that the essence of these changes is captured in the distinction between equal outcome and equal opportunity. In struggles relating to welfare issues, such as the ongoing battle for improved social integration of people with disabilities, arguments based on the idea of equal outcome appear to have lost a good deal of their force, whereas arguments based on equal opportunity have gained in strength. Our theoretical point of reference is the German sociologist’s Niklas Luhmann’s theory of social systems (Luhmann 1995). In this theory, communication is given a central role, as society is observed as a “system” composed of subsystems that ultimately consist of communicative “operations”. In the case of the political subsystem, such communications include (but are not limited to): parliamentary debates, passing of laws and regulations, publicly expressed claims to the state’s fiscal resources, calls for political action, and so on. Thus, when systems theorists speak of “communication”, they refer to an extremely broad phenomenon. They believe that this phenomenon is of key social importance insofar as communications are sequentially linked. For example, a contribution to a political debate influences and thereby constrains what can be said next. It is in the context of these theoretical assumptions that our article maintains that the mentioned changes in the semantics of political communication are of social (and sociological) importance.
2In political and medial communication about welfare issues, participants can hardly plead a stronger value than equality. The rhetorical force of equality crosses political boundaries, making it attractive not only to progressives, but even to conservatives – witness US President George W. Bush’s famous promise to “Leave no child behind”. Today, in virtually any public discourse that focuses on welfare politics, it seems difficult to find explicit rejections of equal treatment, equal access, equal rights or equal opportunity. Equality has become a value it is impossible to argue against – in a sense, a communicative offer one cannot refuse.
3Considered historically, the triumph of equality over values such as rank, origin, nobility or wealth is relatively recent. It was not before Rousseau (1755) that the idea that all men are born equal and that all inequalities are a product of society entered successfully into public debate. Instead of demanding equality of non-equals, Rousseau criticised the inequality of equals. He thus reversed one of the central starting points of debates on social justice, turning the presumption of men’s social inequality (which had long been justified in terms of equality before God) into a presumption of equality.
4More than 200 years after Rousseau, claims to equality have acquired near-universal applicability and appear regularly in communication about virtually any welfare-related issue. Nonetheless, over the last 20 to 30 years, an important change has taken place that affects how the concept of equality is used in such communication. Semantically, this change is reflected in the rise of the concept of equal opportunity and the parallel decline of demands for equal outcome. Since the 1980s, many European welfare states have increasingly replaced the goals of provision and compensation with the goals of deregulation and privatisation. Especially in the field of social policy, it can now be observed that the state increasingly abstains from costly interferences. Along the way, welfare beneficiaries have been promoted from passive clients to active interest-maximisers. Instead of merely receiving compensations, they are increasingly expected to exploit opportunities. In his most recent speeches, British Prime Minister Tony Blair has summarised these goals by painting a bright vision of the upcoming “opportunity society” (Blair 2004). In essence, its aim seems to be to create perfect equality through perfect competition, relying (somewhat inconsequently) on merit as a key discriminating principle.
5In this article, we attempt a theoretical and conceptual analysis of the prerequisites and functions of the contemporary rise of equal opportunity. Our focus is the communicative usage of this concept, specifically, its application in political communication about welfare issues. Thus, we are neither interested in equality as an allocative principle, nor in the question of whether (measurable) social inequality is on the rise, nor in the idea that equality is a social value and therefore “good” (compare Habermas 2001). Taking a systems-theoretical stance (Luhmann 1995), we observe equal opportunity as an integral component of the semantics of welfare debates. In the context of such debates, the concept is crucial because it structures participants” chances to express claims that are either accepted or rejected.
6As we will try to show, the communicative relevance of equal opportunity rests on two foundations. First, it rests on a semantic foundation. The rise of the concept of equal opportunity is accompanied by the fall of the idea of equal outcome, an idea that was strongly associated with the traditional welfare state that tried to create social equality through compensation. In contrast to equal outcome, equal opportunity presupposes that value is only produced to the extent that individuals actively pursue their opportunities. Second, it also rests on a social-structural foundation. At least since the 1970s, the political systems of Western welfare states had to find new solutions in the face of an ever-growing environmental pressure: a rising stream of claims to welfare resources from social groups that had succeeded in being defined as “underprivileged” or even “marginalised” (elderly, physically disabled or mentally ill individuals, immigrants, ethnic minorities, etc.). In all cases, such claims could be (and often were) justified by drawing on the value of social equality.
The central thesis of this article is based on these two aforementioned foundations. In our view, equal opportunity is a device for limiting and controlling the communicability of claims to social equality in welfare debates, especially claims to equal outcome through compensation. Equal opportunity implies a model of man as a rational being capable of pursuing opportunities in order to achieve socially sanctioned goals. Hence the concept presupposes that man can be held accountable for his success (or failure) in pursuing opportunities. In effect, this logic allows the political system  to manage the constantly increasing stream of claims for compensation more efficiently – and to do so without rejecting claims to equality in general, but only by limiting and controlling their legitimate scope.
In section 2, we will provide an analysis of the semantics of equality with respect to two of its instances: equal outcome and equal opportunity. Here, the analysis will be decontextualised, thereby ignoring specific applications of these two concepts in political communication on welfare issues. Section 3 focuses on such applications, arguing that claims to equality have a fundamental argumentative advantage over claims to inequality, an advantage that “coerces” the process of social differentiation to evolve means for delegitimising at least some claims to equality. Section 4 attempts to demonstrate that different stages in the process of social differentiation provided different means for this purpose. The current stage – functional differentiation – now appears to rely strongly on the idea of equal opportunity in order to limit and control claims to equality. In section 5, we will try to summarise some consequences of the broad contemporary success of the concept of equal opportunity.
Observation in terms of equal opportunity
7Equal opportunity represents a specific instantiation of the concept of social equality and may thus be contrasted to other instantiations, especially to equal outcome. As this section will try to show, both equal opportunity and equal outcome depend on the particular semantic properties of the concept of equality. Social equality/inequality are constructs that serve to observe social relations by way of comparison. In logical terms, equality can be distinguished from identity. Whereas identity/non-identity is an absolute relation, equality/inequality exists only as equality/inequality with respect to X (Sen 1992). Whoever applies the concept of (in)equality is thus forced to select the X with respect to which the comparison is being made. For this reason, any concrete application of the concept of (in)equality involves a moment of choice and is therefore contingent. In Luhmann’s (1995) terms, this amounts to saying that such an application constitutes an observation.
8We do not reject the idea that certain differences between individuals are so striking that an observer may usually feel forced to notice them (size, skin colour, physical/ intellectual abilities, property, etc.). However, we do assume that in the context of welfare debates, there are no distinctions between individuals which must be observed. Here, observation of equality/inequality is always a choice because there is always more than one applicable dimension of comparison. For instance, it is up to the observer to decide whether to see two individuals as, say, unequal in terms of skin colour or equal in terms of citizenship or unequal in terms of wealth, and so on. For this article, this means that we do not have any presumptions regarding the fundamental equality or inequality of social groups. We are not interested in observing (in)equalities of any social groups as such, but in observing how social groups are observed in welfare debates.
9The fact that any application of equality/inequality is a contingent observation has several interesting consequences. First, it means that any two nonidentical objects can be observed as either equal or unequal – the observer decides at will by way of choosing X. Second, it entails that any observation of inequality can be countered by an observation of equality (and vice versa). On a logical level, these two contrary observations are equivalent – which of them will gain social influence will thus depend on context factors. Third, it means that society cannot sanction all possible claims to equality. The legitimacy of such claims is a potentially contested, hence scarce, resource. As we shall try to show below, a key social device for limiting and channelling claims to equality is the generalised preference of one specific instantiation of social equality over another – in recent years, of equal opportunity over equal outcome.
10If used in critical diagnoses of social inequality (or, equivalently, in affirmative diagnoses of social equality), the concepts of equal outcome and equal opportunity commit the observer to specific and momentous assumptions about the role of individuals in the accumulation of value, respectively. A plain diagnosis of unequal outcome suggests that inequality is a state for which the disadvantaged observees cannot be held responsible. From their perspective, inequality is something they experience rather than cause (on the distinction, see Luhmann 1981) – a social calamity that should be repaired by social rather than individual efforts. For this reason, critical diagnoses of unequal outcome tend to lead to calls for compensation (see Swedish Government 1994/95: 164). In normal cases, such diagnoses are allowed to proceed “objectively”, that is, without asking the disadvantaged observees whether they actually consider themselves disadvantaged. The concept of equal outcome thus encourages the observer to act as custodian of the observee’s interests. In addition, the disadvantaged observees are observed as entitled to compensation independent of the question of whether/how they observe this measure. For example, within the logic of equal outcome, an unemployed individual is observed as entitled to unemployment benefits even if he chose his unemployment in order to receive such benefits. Obviously, this does not mean that the entity responsible for carrying out compensations – regarding welfare issues: the state – actually assumes that entitled individuals never pursue their outcome actively. It only means that this entity is expected to act as if individuals do not act in this way, that is, as if they were essentially passive welfare recipients.
A diagnosis of unequal opportunity, on the other hand, does imply that observees with lower levels of achieved value can sometimes be held (fully or partially) responsible for their situation. In some way or another, they may have contributed causally to its emergence. By implication, all observees are granted a more active role in the general accumulation of value than in the previous case. They are assumed to be not just passive recipients of benefits, but agents. They are observed as if they possessed the freedom not to achieve the highest value possible (this being the reason why the social value of equal opportunity is insolubly tied to the social value of individual freedom). In order to actually accumulate value, individuals need opportunities – for instance, basic rights, education, resources and access to infrastructure. If it can be established that observees with lower levels of achieved value lacked necessary opportunities, some third party – again, regarding welfare issues: the state – can be held accountable for this deficit.  Otherwise, the observees themselves can be held accountable. As in the case of equal outcome, it is obvious that observation in terms of equal opportunity is extremely schematic and based on an enormous reduction of complexity. Although observees may differ with respect to accumulated value, they are expected to be essentially similar in their capacity to freely pursue opportunities. By default, they are assumed to be equally capable and willing to maximise their interests. Being exempted from this assumption requires special reasons, for instance, official categorisation as sick, elderly or disabled (Stone 1985).
The problem: unbounded claims
11Political communication, including communication about welfare issues, is one of the primary application contexts of the concept of social equality. To a considerable extent, political communication is centred around claims, that is, expressed and subjectively legitimate desires for certain social goods. The fulfilment of a desire that is expressed through a claim depends to some extent on the addressee. Put abstractly, this means that the claim can be either accepted or rejected. However, whether any specific claim is going to be accepted or rejected by the addressee is often not an entirely contingent matter. For example, a claim that is plausibly framed in terms of a socially sanctioned value has a higher chance of meeting acceptance than a claim that is not.
12In this section, we argue that due to the logical properties of the concept of equality, distributive claims to equality have a fundamental argumentative advantage over claims to inequality. We shall conclude that this poses a problem that any non-primitive society needs to address. To begin, consider that any distributive claim that draws on the value of social equality involves a comparison of two (or more) parties. In essence, such a claim states
13[Premise 1] that party A and party B are unequal with respect to X (e.g. salary), but
14[Premise 2] that this is not legitimate because
15[Premise 3] A and B are equal with respect to Y (e.g. qualification),
16[Conclusion] therefore, that A’s and B’s share of (access to, ownership of, stake in, etc.) X should be balanced.
17If the proponent of the claim is cautious, it will be difficult for the addressee of the claim to reject premises 1 and 3. Thus, the rejectability of the conclusion depends essentially on the rejectability of premise 2. We argue that in absence of social-structural support, the addressee will find it difficult to reject premise 2 and find social acceptance for this rejection. The reason for this is chiefly logical. In any situation of conflict or argument, a rejection of an opponent’s argumentative position can be observed as an implicit claim that one’s own position is right (Reich 2003, chapter 4.3). Thus, by stating that premise 2 is wrong, the addressee of a claim to equality can be observed as making a claim to inequality. In order to back his claim, he will need to show that A’s and B’s inequality is legitimate. Warrants for this claim can have only one of two forms. Either they are tautological, stating that A and B are legitimately unequal in terms of X because they are unequal in terms of X, or they are not tautological, stating that A and B are legitimately unequal in terms of X because they are unequal in terms of Z. The first form is rejectable because it begs the question. The second form, on the other hand, can lead the proponent of the original claim to equality to ask for another warrant. Other things being equal, this will cause an infinite regress, and it can therefore not lead to uncontroversial justifications. We note that a claim to equality is not subject to a similar regress. Criticising inequality with respect to X by way of diagnosing equality with respect toY opens up no distinction between A and B that the opponent can use in order to request another warrant. In order to reject the claim, the opponent is forced to draw another distinction and state that A and B are unequal with respect to Z. This distinction, however, is his own – he, not the proponent of the original claim to equality, will be responsible for defending it.
18While claims to equality have an intrinsic argumentative advantage over claims to inequality, the process of social differentiation feeds communication about social equality with an ever-increasing range of opportunities for observing inequality. Whether social differentiation is understood according to Simmel (i.e. as a process of increased individualisation), Durkheim (i.e. as a process of increased division of labour) or according to Parsons/Luhmann (i.e. as evolution of functionally more and more autonomous subsystems, e.g. law, politics, economy, etc.), it involves necessarily some form of differentiation at the level of individuals. For example, the evolution of specialised roles (e.g. rulers, judges or healers) requires the co-evolution of what Parsons (1951) calls complementary roles (here: subjects, defendants or patients). For this reason, social differentiation increases individuals” chances to observe themselves and others as unequal with respect to some X. In fact, the more complex and differentiated a society, the more such chances exist (Luhmann 1997; a similar argument can perhaps be attributed to Bourdieu 1989).
These reflections allow us to diagnose a problem with which the process of social differentiation has been – and continues to be – confronted. To the extent that it creates durable distinctions between individuals, social differentiation needs to provide structural support for legitimising these distinctions against a potentially growing pressure of claims to equality. In retrospect, the process of differentiation must have “dealt” with the fact that claims to equality are, ceteris paribus, more defensible than claims to inequality.
The solution: legitimising inequality
19In this section, we will try to show that, historically, three different stages of social differentiation (segmented family clans, stratified classes, autonomous function systems, see Luhmann 1995) evolved different structural means for legitimising fundamental forms of inequality between individuals. To be sure, the most primitive form of social differentiation, segmentary differentiation (i.e. differentiation into families, tribes or clans), dœs not appear to require such structural means. Pure physical strength may have the last word as to who is to be leader, who is to be privileged, etc. Inequalities within and between the segments can thus remain purely factual in nature, receiving normative support only to the extent that they are factually upheld. In addition, experience, hence age, almost automatically creates observable individual characteristics that are advantageous for leadership positions, thereby at once instituting and legitimising gerontocracy (arguably the predominant type of rule in segmentarily differentiated societies).
20In Western Europe, the transition from segmentary to stratified differentiation (i.e. differentiation into estates or classes) is accompanied by the dissociation of leadership roles from personal qualities such as fighting skills and belligerence. A distinction evolves between being powerful and being a good fighter. For the new ruling class, this necessitates the ability to reject claims to equality with words rather than violence. Justification of fundamental and durable inequality is provided by religion: the hierarchical order between men mirrors the divine order, and God’s will cannot be questioned or even interpreted, only accepted. By implication, men cannot be held accountable for their social status, and they cannot be permitted to change it. Claims based on inter-strata comparisons are thus a priori illegitimate – there is simply no way to compare peasants with kings or knights with clergymen.
21Toward the beginning of the Renaissance, the growing operational autonomy of the economy vis-à-vis politics, law and religion initiated the emergence of a new form of societal differentiation, that is, differentiation into function systems. The unifying principle of these systems is orientation to a specific medium of communication: money in the economic system, power in the political system, and so on. Relative to this orientation, the reach of all function systems is, in principle, universal. All payments take place in the economic system, all trials in the legal system, etc. For this reason, all function systems are also all-inclusive, presupposing that any individual who can participate in social communication can also participate in function-system specific communication. However, inclusion into function systems remains a purely operative, fleeting phenomenon. For instance, by making a payment, one does not become a “member” of the economy. Not being “assigned” to specific social systems as they were assigned to social classes, individuals also lose their fixed place in the social hierarchy. In this sense, they are equals, and it seems hardly surprising that equality and freedom become two fundamental social values. The presumption of inequality that could be upheld in a stratified society turns into a presumption of equality under which function-system specific inequalities (e.g. property, rights, level of education) also require function-system specific legitimation (Luhmann and Schorr 2000, p. 255).
22In Sweden, it was not before the 1970s that the political system responded broadly to the ever-growing number of claims to welfare resources that were motivated by this structural change. An initial reaction was to declare a change in the goal of social welfare: from expansion to maintenance of the current level (Michailakis 1995, p. 192). However, rather than actually controlling claims, this change only communicated that resources were henceforth to be seen as scarce, thus accessible only by way of even fiercer competition. A more successful change was the introduction of the concept of “equal opportunity” into debates on welfare distribution. 
As mentioned above, the concept is intrinsically suited to limiting claims that are based on the premise that welfare is a form of value one receives rather than achieves. Achievement, on the other hand, presupposes an active and accountable contribution from the individual. Any observable inequality of outcomes is under the potential suspicion of having been caused by laziness or incompetence, that is, inadequate handling of one’s personal freedom. While retaining already widely acknowledged social objectives such as equal access and participation, the concept of equal opportunity thus shifts the burden of proving inequalities to be illegitimate from the welfare state to the authors of claims.
Consequences of equal opportunity
23The semantics of equal opportunity now permeates political and even legal communication on welfare issues in many Western countries. For instance, in Sweden, the concept plays a decisive role in policy regarding disabled persons, such as the Special Services Act of 1985 (SFS 1985: 568) and the Disability Act of 1993 (SFS 1993: 387). Put bluntly, Swedish social politics has turned disability from a category that justifies compensation into a category that delegitimises discrimination. In the United Kingdom, the concept lies at the core of Prime Minister Tony Blair’s recent vision of an “opportunity society” (Blair 2004). And the United Nations has adopted it in many policy recommendations that address social justice (see UN 1994). Of course, by endorsing equal opportunity, the political system renders itself unable to create social equality, as social equality is an abstraction that can be instantiated in a variety of incompatible ways, such as equal opportunity vs. equal outcome. And yet, for reasons that we have tried to detail in this article, the concept of equal opportunity helps the political system in solving a specific problem. Confronted with a growing number of claims to welfare resources, the political system applies the concept in order to reject (or simply be able to reject) certain claims as illegitimate. Although equality has become a key social value of contemporary society, its reformulation as equal opportunity dissociates it from the exclusive responsibility of the political system. As many authors have observed, this brings enormous economic relief as well as a diminishing need for bureaucratic administration (e.g. Roemer 1998). In order to be maximally acceptable, claims to welfare resources now need to be framed as claims to equal opportunity rather than equal outcome. This narrows the accessible range of potentially desirable welfare resources to encompass only thosewhich can be observed as means necessary for attaining more or less equal quality of life: basic human and citizen rights, access to public infrastructure, access to education, basic participation in political decisions, legal assistance, basic medical assistance, social insurances, among other things.
24Of course, the concept of equal opportunity is only one expression of the semantic changes that have actually taken place. Related concepts that have played important roles in communication on welfare issues include: empowerment, enablement, responsibility, opportunity, freedom of choice, self-determination, “focus on the individual”, civil rights, and many more. Popular explanations of these changes are based on ideas such as ideology, mentality or collective consciousness. The changes are seen as reflecting a move toward neoliberal thinking, with growing focus on the value of freedom and diminishing focus on the value of social equality. In our point of view, the problem of explanations in terms of ideology is not only that they arementalistic, hence unsociological, but also that they are empirically inadequate. If endorsement of concepts such as equal opportunity or empowerment were a matter of individual conviction, one should expect left-wing parties to reject them. However, this is not the case – in Sweden, the United States and many other countries, many left-wing parties now strongly endorse several of these notions. As mentioned, even equal opportunity is an instantiation of equality, but one that fits particularly well with the complementary value of freedom and its relevance to goals that modern leftwing parties can endorse, such as civil rights, empowerment or rejection of paternalistic government.
25Trapped amidst these semantic and structural changes, the individual finds that he is not only the object of new and enthusiastic political concerns for his social opportunities, but also accountable for making use of these. Disadvantaged life situations are under the universal suspicion of having been caused by mismanaged personal freedom. The burden of proof is on the individual. In order to complain about inequalities of outcome, he must prove that he never had sufficient opportunities in order to achieve equal outcome in the first place. He thus learns that the focus on equal opportunity re-introduces the old distinction between deserved/undeserved fates through the back door. In addition, he learns that factually unequal opportunity does not automatically warrant claims for compensation. As full equality in all respects is a practical impossibility, equality in all its instantiations (equal outcome, equal opportunity, etc.) is never more than a “reflection concept” (Luhmann and Schorr 2000, p. 256). This allows participants in debates on welfare issues to presume that social opportunities are normally more or less evenly distributed. Over the course of a lifetime, everyone has a “sufficient” amount of opportunities – unless he can prove otherwise (e.g. through claiming membership of a special social category, such as that of disabled individuals).
As a normative concept, the idea of equal opportunity may serve to make society more just while also “empowering” the individual – or it may not. Psychologically, the concept may inspire individuals with hope and motivation as well as politically and economically auspicious incentives – or it may not. These questions are not up to us to decide. In this article, we have tried to assume the role of a distanced observer of contemporary political communication on welfare issues. All communication leaves physical traces in the form of messages – in the case of political communication, they include transcripts or broadcasts of parliamentary debates, passed laws and regulations, newspaper articles, political interviews, press releases by lobby organisations, and so on. The ideas and theses we have put forward in this article are thus, in principle, open to empirical scrutiny.
Wendelin Reich: Assistant Professor, University of Gävle, Sweden. Dimitris Michailakis: Associate Professor, University of Gävle, Sweden.
In line with Luhmann (2000), from now on we use this word in the singular, assuming that there is only one (world) political system which is “segmented” into country-specific subsystems.
For an example attribution of accountability, see Lewin (1998), who deems the Swedish Disability Act of 1993 a fiasco and attributes its failure to municipal authorities.
For the case of disability, its introduction was endorsed in the publication of the United Nations’ “Standard Rules on the Equalization of Opportunities for Persons with Disabilities” (UN 1994), which were in part initiated by Sweden and Denmark.