Expressive Complexities of Constitutional Equality
Keynote Lecture for International Association of Constitutional Law Roundtable
Certosa di Pontignano, Italy, June 15, 2023
Introduction
My topic today is equality, which a famously complex and elusive idea. Equality is a bit like freedom, in that everyone wants it and agrees that it is a good thing, but no one can quite say what it looks like, and it is most notable by its absence. As an idea, equality is as powerful as it is vague. It motivates people to mobilize and engage in collective action for freedom. It also motivates backlash: many of our current populist and anti-immigrant movements are based on claims for equal status by those left behind by globalization. For them, status is like territory—a zero sum game in which group’s gain is another’s loss.
Equality is one of very few things that can be considered truly essential to written constitutions. Constitutions always say something about the amendment rule, and about choosing the head of state. Equality is now a part of this constitutional “core.” 99% of constitutions in force today protect it, with only Brunei and Israel lacking such a guarantee. Equality constitutes constitutions, just as constitutions constitute equality.
The U.S. Declaration of Independence begins with the famous phrase, “We hold these truths to be self-evident, that all men are created equal.” Yet the more self-evident proposition is that we are created unequal. Some of us are born with money while others are not; we genetic predispositions to be large or small, dark or light, tall or short, smart or dumb, beautiful and less beautiful, all of which matter for social outcomes. Which do we address with our limited constitutional tools? We cannot feasibly address every inequality, and so need criteria to figure out which dimensions to prioritize.
Material and Expressive Dimensions of Constitutions
I first want to introduce the idea that constitutional equality has both a material and expressive dimension. By material, I mean that constitutions are meant to change things, in the real world, and meant to make things more equal across socially significant dimensions of difference. By expressive, I mean that constitutions are meant to say things. They are designed to communicate status, to say something about the kind of society we want and to reflect demands of groups which have traditionally been excluded. Constitutional drafting processes of course involve expressions of ideas and dreams, as well as negotiations about rules of governance. The distinction between material and expressive dimensions roughly corresponds to what Alberto Simpser and I have characterized as the distinction between (i) constitutions as operating manuals, designed to provide a working guide to government, and (ii) constitutions as blueprints, meant to project a vision of a building not yet built.[1]
The distinction overlaps with trends in equality theory. Traditionally, discussions of equality start with Aristotle’s idea that “persons who are equal should have assigned to them equal things.” But this simple formulation raises questions that Aristotle himself recognized in his question, “equal and unequal in what?” Our concerns today are which asymmetries among us count? Are we interested in equality of outcome or opportunity? In what domains: education, health, wealth or others? Are we concerned with equality across groups or individuals? The questions are not exclusively concerned with material equality, but Aristotle’s “equal things” implies material distribution.
A second ancient source for thinking about equality is not justice, but the biblical idea of equal human dignity, that we are all created in the image of God. This is to jump up the ladder of abstraction, to the thing we all share, rather than the things that divide us. Seen from a certain level, we are all the same: we are not fish or plants, and we are also not God. From this source, we have duties to each other, captured first in the biblical laws of Noah, which were owed by and to every person. This notion of equality as dignity is expressive—it goes to our status.
In our contemporary debates in law, the trend is to move away from formal Aristotelian equality toward theories focused on status. As Ronald Dworkin put it, “the right to treatment as an equal more important than the right to equal treatment.” This is sometimes characterized as a shift away from formal equality to substantive equality, or in the US context a shift from anticlassification to antisubordination theories of equality.[2] Affirmative action and “temporary special measures”, for example, treat like people differently for the purpose of making up for past subordination. In a society stratified by race and gender, such an approach is needed to move toward equal status in face of entrenched social inequalities. This brings us back to the point made at the outset: equality is easiest to see in its absence, and its antonym is not so much inequality as it is hierarchy.[3]
In addressing the material dimensions of equality, the law is a clumsy instrument. The law works by using generalizations that inherently group unalike things, and sometimes treat unalike things alike. I am not sure what the speed limit is here in Italy but it seemed to be about 200 km per hour coming from the airport at midnight. A speed limit treats driving at 250 km or 201 km per hour as the same, but treats the difference between 199 km and 201 km per hour as significant. (Americans will recognize this example as a problem from the literature on rules and standards.)[4]
Equality law also reinforces and sometimes creates categories for protection; once formed, “categories may become the basis for evaluating group members in ways far beyond originally targeted. Reforms reproduce rather than surmount tensions between sameness and difference.”[5]
And then there is the sheer inefficacy of law. Black students in the United States are more likely to go to a segregated school today than at the time Brown v. Board of Education was decided.[6] At the same time, the poverty rate among African Americans has gone down and life expectancies are converging with whites, outcomes for which the law has at best an indirect effect. Inequality of status can sometimes be addressed through addressing economic inequality directly, but the reverse is not really true.
The Growing Articulation of Equality in Constitutions
One great trend in the drafting is a trend toward differentiation in the groups explicitly singled out for equality protection. As Figure 1 shows, the average constitution now names at least six categories for protection, roughly tripling since 1950.[7]
[Figure 1 here]
Perhaps an extreme case is that of Bolivia (2009), which announces that no less than 20 specific categories of persons are protected:
“The State prohibits and punishes all forms of discrimination based on sex, color, age, sexual orientation, gender identity, origin, culture, nationality, citizenship, language, religious belief, ideology, political affiliation or philosophy, civil status, economic or social condition, type of occupation, level of education, disability, pregnancy, and any other discrimination that attempts to or results in the annulment of or harm to the equal recognition, enjoyment or exercise of the rights of all people.”[8]
Nepal’s 2015 Constitution featured an equality clause that was hard to negotiate, and produced a complex scheme. The final version of Article 18(2) prohibits discrimination on the grounds of origin, religion, race, caste, tribe, sex, physical conditions, disability, health condition, matrimonial status, pregnancy, economic condition, language or geographical region, or ideology or any other such grounds. The next sub-article goes on to prohibit state discrimination on most, but not all, of these bases.[9] It then allows affirmative action for “women lagging behind socially and culturally, Dalits, Adibasi, Madhesi, Tharus, Muslims, oppressed class, backward communities, minorities, marginalized groups, peasants, laborers, youths, children, senior citizens, sexual minorities, persons with disability, pregnant, incapacitated and the helpless persons, and of the citizens who belong to backward regions and financially deprived citizens including the Khas Arya.”[10] Some but not all of these groups are further guaranteed to participate in state bodies on the basis of “proportionate inclusion.”[11] All this reflected both demands for status for traditionally backward and excluded groups, as well as some interest group politics, and reflects a trend to articulation of more and more categories, but is internally complex because of the different listings in different places.
The trend toward finer grained articulation seems likely to accelerate because of a combination of new awareness of injustice and interest group behavior. Looking around the world, we see that gender is the most universally prohibited basis of discrimination, and a recent addition is sexual orientation.
[Figure 2 here]
Some of the more specific categories reflect local conditions. For example, in Afghanistan, nomads are singled out for protection, and caste is mentioned in Nepal and India. The very expansive Chilean draft Constitution, rejected by voters in the Fall of 2021, distinguished gender from biological sex, and named the neuro-divergent as a protected group. Armenia’s constitution adds genetic features, while Cote D’Ivoire and Colombia mention philosophical opinions as a basis of protection.
Many constitutions leave the categories open ended, so that as new biases of discrimination are identified, they might be redressed. Some of these clauses are very broad indeed, such as that of Algeria which includes “all other conditions or personal or social circumstances.”[12] Such clauses do very well on the expressive dimension of making everyone feeling included. But material implementation becomes incoherent. The project brings to mind the Kurt Vonnegut novel Harrison Bergeron, in which a series of U.S. Constitutional amendments in the year 2050 declare all Americans equal, and prohibit anyone from being smarter, better looking or more physically able than others. What would it really mean for the state and society to correct every inequality?
Three critiques
From an expressive point of view, growing articulation would seem to be a good thing, given our increasing attention to entrenched hierarchies. But I want to raise three notes of caution. First, the more categories there are, the more internal conflicts between status groups, and the anti-hierarchy principle becomes harder to deploy. Second, the reification of categories by the state constructs identities, and reinforces boundaries that makes them difficult to overcome. Third, and relatedly, this creates an ever-expanding incentive to create new groups, without addressing the deepest sources of inequality, which in our era are driven by economic policy.
First, the more groups there are, there is more internal conflict among categories. The protection of religious diversity might interfere with rights of women; the right to non-discrimination on basis of property can hinder socio-economic redistribution; and the listing of groups creates a complex calculus of intersectionality which is ill-suited to contexts in which ordinal rankings are required. In university admissions in the US, is a disabled, gay white man more or less worthy than a Peruvian-Norwegian woman who is poor? Is a mid-caste woman in Nepal more or less worthy than a poor Madhesi man? These kinds of balancings are undertaken by administrators behind closed doors, largely on the basis of their own decisions about social justice. Not only are the various dimensions incommensurable in terms of hierarchy, but the policy invites everyone in society to search for elements of their complex multiple identities to appeal to the bureaucrat, regardless of whether they have experienced actual disadvantage. In the United States, to give one example, the “Latinx” category includes white people from Latin America, and people with origins in Spain, but not Italy.[13] The simple Black-White paradigm, which in the United States presents (to my mind) a clear justification for corrective justice to address the legacy of slavery, has given way to a distributive justice free-for-all with implicit values attached to different identities.
Second, no category ever disappears once it is named in the Constitution. Identities become reified and do not fade away. Caste in India is a good example, in which the categories of Scheduled Castes, Scheduled Tribes and Other Backward Castes have expanded to the point to where reservations threaten to constitute more than 50% of allocated seats and jobs. Malaysia’s “New Economic Policy”, initially adopted for 15 years in 1971, has become a permanent feature of that country’s life, ensuring that the numeric majority enjoys benefits at the expense of minorities. Would a more liberal regime encourage intermarriage among groups, and thus the fading of the cleavage? We will not find out. The law has the effect of freezing society, even as it intervenes within it to redistribute across lines.
Third, we face the challenge of populating the category of “other bases” for countries with open-ended equality clauses. We thus need a basis of deciding which groups to include. This goes back to the original question posed by Aristotle: equal and unequal in what?
In the Kenyan case of Eric Gitari v. NGO Coordination Board [2015), the High Court found that the 2010 Constitution’s list of the categories for protection was not exclusive, because of the use of word “including” before the articulation of the list. The case was brought by an LGBTI group whose petition to form an organization was rejected by the government on the grounds that homosexuality remained criminalized and “repugnant to the teachings, cultural values and morality of the Kenyan people.”[14] The Court held that the group had a right to association even if the underlying activity was illegal. The refusal to register the organization was discriminatory, and sexual orientation was a protected class for this purpose.[15]
The interpretive challenge for the court was a familiar one concerning constitutional silence. The Committee of Experts that had drafted the text did not include homosexuality as a basis for protection, possibly because of the deep unpopularity of LGBT rights in the country. But did the silence mean that the category could not be added later by courts? This evoked the 1998 case of Vriend v. Alberta in Canada, holding that Alberta’s rejection of sexual orientation under its Human Rights Act was unconstitutional under Charter of Rights and Freedoms. Although the national Charter does not mention sexual orientation, the Supreme Court of Canada held the list to be nonexclusive, and found that justice required reading the category into the set of protected classes.
In some cases, adding categories can give rise to new equality tradeoffs. In the 1999 case of Corbiere v. Canada, the Supreme Court added aboriginal non-residents to the protected class, giving them the right to vote in tribal elections held on reserves. This decision, sounding in liberal equality, undermined the demands of First Nations groups to determine their own internal rules of governance. Individual equality undermined the collective demand for sovereign equality. This illustrates the paradox that articulating groups introduces new asymmetries to overcome.
Adding Categories
What other categories might we add? In its 1989 decision of Law Society of BC v Andrews, the Canadian Supreme Court defined discrimination as “a distinction whether intentional or not but based on grounds relating to personal characteristics of the individual or group, which has the effect of imposing burdens, obligations or disadvantages, on such individual or group not imposed on others.” This is an extremely broad definition that challenges us to address every case of morally unjustified difference.
Consider several possibilities. I inhabit a body that is tall, male and Caucasian in appearance. Statistically this is a good thing in my country and here in Italy. Yet there are many contexts in which these things are a disadvantage. When I lived in Japan, I would wake up in our small apartment and hit my head, leading my children to rebuke me for cursing. I do not want to overstate the point, but being tall in Japan opens ones eyes slightly to the challenges posed by disability in a world made for able-bodied. The things the rest of us take for granted—the size of chairs, the amount of light produced by bulbs, the width of doors—can remind some of our bodies that they are unusual. This points to a fundamental truth about the notion of equality: it is relational and contextual.
Should we add categories like height? Consider that 30% of Fortune 500 CEOs are 6-foot-2 and taller, as compared with just 4% of all men in the United States. Should we compensate the short? Similarly, there is a well-established positive relationship between physical beauty and earnings.[16] Should we penalize the good-looking? There is also an “obesity penalty,” a negative correlation between weight and wages.[17] The obesity penalty affects women more than men, poor more than rich, old more than young, and white women more than Black.[18] Should we compensate the large and penalize the thin? And then within that, should we calibrate the compensation by membership in the other intersectional categories?
I’m pleased to speak here today as a member of the bald community, a diverse vibrant group that includes the very young, the very old, and people from every continent and creed. We're not a well-recognized minority, but we are here among you nevertheless. We are your friends and your neighbors and sometimes the rest of you don't appreciate the difficulties we have of living among you.
Bald men are a much bigger slice of the general population than the tall. The International Society of Hair Restoration Surgery estimates that 50% of Caucasian men older than 45 and 60% older than 60 have clinical balding. But only 20% of the Fortune 500 CEOs are bald. Stress can cause hair to fall out, so all things being equal, the percentage of bald leaders might be expected to be a little higher than average.
Before you go crying tears for the bald, consider that only 8% of the CEOs are women. 86% of them are Caucasian men. Thus, while bald white men are underrepresented vis-à-vis those with hair, they are not underrepresented relative to their percentage in society. This illustrates the general problem of choosing a reference group in making equality judgements.
Finally, consider the “Furries”, a group of people that dress in animal costumes and socialize with each other. Some 30% of Furries identify with the species of their chosen animal. Some even choose to make animal noises in response to queries in the classroom and wish to be addressed as a member of their species-identity. If an employer decides not to hire someone who insists on wearing an animal costume, is this discrimination? If this is their identity, who are we to challenge them?
One traditional way to think about which categories to protect is to ask whether the quality is immutable or can be changed. Race is the paradigm of immutability, though even here the old phenomenon of “passing” suggests that a degree of manipulation is possible.[19] Sex used to be seen as immutable, but no longer is. Baldness is mutable, either through wearing a hairpiece, or the modern technology of hair restoration. The immutability test seems to lose power in a world in which technology can intervene to change status. The Furries seem to be choosing to identify as rabbits, but they would assert that in fact it is their inner true self that is coming out.
Another criterion often put forward is a history of disadvantage. Here, the transgender have a claim that is different from the Furries, as the latter is a new group. The bald are out, but the short might be in. In a USA TODAY survey of a panel of CEOs and executives, 95% said, if given a choice, they would rather be bald than short. Clearly, to the disappointment of Larry David, baldness is not a protected category.
A third criterion is human dignity. If the social meaning of exclusion or disadvantage implies less capability or recognition, a group might be considered for benefit.[20] One source for this idea was the dissent by Justices O’Regan and Sachs in South Africa v. Jordan (2002) arguing that that the law couldn’t punish sex workers without punishing their clients. To do otherwise was to punish the weaker party in a transaction, giving life to stereotypes and antiwoman bias.
I like the human dignity criterion best and it accords with an expressive approach to the purpose of constitutional equality. There may be material differences in terms of earning power between tall and short, beautiful and less beautiful, bald and non-bald. But no one thinks that any disadvantage implies less capability or status as a general matter. Similarly, the Furries probably would not get protection. Most obviously, animals are not entitled to human dignity, but even if one overcomes that objection by emphasizing the human inside the rabbit-suit, dignity does not seem to be at issue. In contrast, the obese do suffer from a perception that their condition is due to a lack of discipline. The physical act of squeezing into a small chair while overweight is embarrassing and humiliating, in a way that was not true of me sitting uncomfortably in a chair in Japan.
Conclusion
Status-based theories of discrimination save us from the dystopia of Harrison Bergeron. They provide a basis both for figuring out which criteria we ought to consider in correcting past inequalities, and which groups deserve inclusion in an open-ended list of potential categories for protection. They tell us who to grant expressive benefit. But as yet, their record in motivating material change is fairly limited in my reading. I continue to believe that economic policy addressed at general inequality will in fact be the most effective way of addressing status or dignity based harms. But maybe that is just me.
[1] Tom Ginsburg and Alberto Simpser, Introduction, in Constitutions in Authoritarian Regimes (Ginsburg and Simpser, eds., 2014).
[2] Owen Fiss, Groups and the Equal Protection Clause, 5 Philosophy and Public Affairs 107 (1976); Jack Balkin and Reva Siegel, The American Civil Rights Tradition: Anticlassification or Antisubordination? U. Miami Law Review (2003-2004); Derrick Bell, And We Are Not Saved: The Elusive Quest For Racial Justice (1987).
[3] Catherine MacKinnon in Daedalus 2019.
[4] Fred Schauer, Profiles, Probabilities, and Stereotypes (2003).
[5] Jenson et al, Daedalus 2019.
[7] This is derived from the Comparative Constitutions Project survey that asks about 16 different categories. The list is under inclusive.
[8] Art. 14(II).
[9] Physical conditions, disability, health condition, matrimonial status, pregnancy are excluded.
[10] Art. 18(3).
[11] Art. 42(1).
[12] See also Morocco (“Whatever personal circumstances that may be.”)
[13] David Bernstein, Classified: The Untold Story of Racial Classification in America (2022).
[14] Victoria Miyandazi, Equality in Kenya’s 2010 Constitution 55 (2021).
[15] Miyandazi.
[16] John Karl Scholz and Kamil Sicinski, “Facial Attractiveness and Lifetime Earnings: Evidence from a Cohort Study,” The review of economics and statistics 97, no. 1 (2015): 15.
[17] John Cawley, "The Impact of Obesity on Wages," The Journal of Human Resources 39, no.2
(2004): 457; Michael T. Owyang and E. Katarina Vermann, “Worth Your Weight? Re-examining the Link between Obesity and Wages,” Regional Economist (2011).
[18] Ibid.
[19] Cheryl Harris, Whiteness as Property (1993).
[20] L’Hearueu-Dube in Egan, quoted Miyandazi at 71.