Democracy’s ruling hand

Constitutional quintessentials

In Constitutional Essentials, Frank Michelman puts the liberal political theory of John Rawls in dialogue with the traditional debates among constitutional lawyers (Michelman 2022, 1–2). Michelman’s goal is to illuminate a line of inquiry he has been pursuing explicitly for 35 years. ² When, in a legal order founded on the consent of the governed, is coercion of dissenters to be considered legitimate? Michelman does not promise to solve this problem – that is, to provide constitutional essentials that would fulfill Rawls’s liberal principle of legitimacy – nor to resolve the constitutional lawyers’ debates. He acknowledges that his discussion ‘will surface sundry doubts and concerns’ and warns that he ‘does not expect to set those doubts all to rest’. (Michelman 2022, 3–4) In his well-honed and familiar style, Michelman pushes earnestly and relentlessly on the difficult issues attendant on the project he calls justification-by-constitution. This may drive him to ever more qualified statements, but one cannot fault him for the resulting uncertainties. One can only admire his characteristic rigor and unfailing intellectual honesty.

There are some fundamental reasons the liberal principle of legitimacy cannot be cashed out. I am going to focus on the structural and conceptual (though we could also call them sociological or epistemological). ³ To put the point bluntly: A constitution resolves the problem of pluralism the same way a fig leaf averts the allure of sexuality. ⁴ The desideratum is a constitutional package to which members of a society, composed of free and equal citizens possessed with the capacity to reason, should agree provides a platform for stable, democratic life under conditions of reasonable pluralism (Michelman 2022, 7–8). ⁵ But this project rests on two related presuppositions that cannot be redeemed. First, it assumes a pluralism that, though intractable, nevertheless remains amenable to reason. ⁶ Second, it assumes that reason is a capacity independent of the profound differences of meaning, value, and forms of life that shape those intractable disagreements. Neither assumption is true. ⁷ ‘Reasonableness’ is a contingent phenomenon sensitive to the practices and conditions of a given society. ⁸ And it could not be otherwise: Differences in norms, values, and forms of life inevitably undergird and structure differences in meaning, perception, and interpretation. ⁹ Indeed, constitutionalism has always had a fatal circularity problem. ¹⁰ The most a constitutional package can supply – though it is no small thing – is a platform for disagreement.

I am going to sketch an alternative account of justification-by-constitution that will, at least initially, depart radically from Michelman’s Rawlsian account. This ‘constitution’ is not a ‘text-like object’, ¹¹ but the democratic nomos of situated social practices and attendant values. ¹² It is non-dualist, non-deflecting, and not scripted. On my view, the constitution (however conceived) belongs in usufruct to the living (Jefferson 1979, 382–88). However formally represented – that is, even when written and accompanied by judicial review or some institutionalized equivalent – it will necessarily unfold in an ongoing process of political cooperation and contestation (Winter 1990, 1494–1500). Even so, I argue that this alternative account is both the same as Michelman’s and the only one possible given the pragmatics of the human condition. The residual differences will emerge as we peel back the text-like object to reveal the sinews of democracy upon which legitimacy depends.

What the Laws say

I start with a line of argument that runs from Protagoras through Crito. In the former, Socrates claims that virtue and the excellence (aretē) required for governance cannot be taught. Protagoras’s reply – which goes largely unrebutted – is that the laws of the polis provide the civic education for citizenship: ‘Just as, when a child is still learning to write, the teacher draws lines on his book with a pencil and then makes him write the letters following the lines, so the city lays down the laws . . . for our guidance, and makes us rule and be ruled according to them’ (Plato 2009, 326d). To ‘rule and be ruled’ is the heart of the Athenian conception of democracy: the idea that the people both make and live under the rules that govern their lives together. Entailed is the requirement that every law apply equally to all citizens – a much more exacting version of the principle we now refer to as the ‘rule of law’ (Winter 2022b, 662–63).

In the Crito, Socrates radicalizes Protagoras’s argument. He recounts an imagined dialogue with the Laws (Plato 1956, 455–59). The Laws make the argument (later taken up by the American Legal Realists to whom Michelman refers in Chapter 14) ¹³ that the taken-for-granted preconditions of everyday life are the artifacts of constitutive legal rules: Without the marriage laws of Athens, Socrates’s own birth would not have happened; without the law’s mandate for nurturance and education, his own development would not have taken place (Plato 1956, 455). To be sure, Socrates explicitly grounds his duty to obey in the idea of contract (Plato 1956, 456–57). ¹⁴ But it is his respect for the law as a citizen that ultimately drives his choice. ¹⁵ The Laws say, ‘if you go to one of the cities nearest, . . . you will come as an enemy, Socrates, to their constitution, and whoever have care for their own cities will think you a destroyer of laws’. To escape his sentence, Socrates explains, would be to confirm the jury’s judgment. For, ‘whoever is a destroyer of laws would surely be thought to be a corrupter of young men’ (Plato 1956, 458).

If, as Protagoras says, the law is the great teacher that makes good citizens, then a destroyer of laws is a corruptor of youth. ¹⁶ But the social constructivist argument made by the Laws is not just a claim about proper character development. The claim, rather, is that the law constitutes the free and equal citizens whom we all – liberals and strong democrats – take as essential. As Hannah Arendt astutely observes, it is the ‘artificial institution’ of the polis that makes them equal. ¹⁷ The law, of course, protects the weak from the strong and the vulnerable from the predatory. More importantly, citizens are constituted as equals by the political rules that establish the equal participation of each citizen in the making of the law – expressed by the Greek terms isonomia, isēgoria, and isokratia (Ostwald 1969, 158–60; Vlastos 1995, 103–07). So, too, the rule-of-law principle establishes the equality of citizens before the law regardless of status. The Protagorean argument is not simply that the law teaches one how to behave. It is, rather, that the law teaches one how to be a free and equal citizen (i.e., to ‘rule and be ruled’). In modern terms, democracy and the rule of law are the social practices that found the equal dignity of all persons. They do so by prescribing as the constitutional essential the idea of self-government – that is, as Michelman elsewhere says, the idea of group of people ‘who take themselves to be committed to the project of governing themselves by a law that they themselves have uttered’ (Ferrara and Michelman 2021, 30). ¹⁸

‘Deep’ democracy

Democracy is the practice of collective self-governance defined by equal voice, equal power, and equal law. Majority rule is entailed because it is the only political practice that meets this standard (Barber 1996, 362; Winter 2022b, 658–59). But, in the absence of unanimity, democracy necessarily produces outcomes disagreeable to some. Its outputs are legitimate nonetheless because they issue from a process that gives what Michelman (2022, 5) calls ‘institutional form’ to the governance of social life under conditions that realistically accord equal recognition and respect to all participants. At the end of the day, this democratic practice is the only thing that can.

The ideal of collective decision-making under conditions of mutual recognition and respect is what I call (with all intended irony) ‘the minimalist constitution’. ¹⁹ It is fully procedural (what Michelman calls ‘processual’) and, as I shall explain in the final paragraphs, only nominally dualist. It is normatively ‘thick’ or (better) ‘deep’. It requires not the arms-length virtues of tolerance and reasonability, but the deeper, interpersonal skills of empathy, respect, negotiation, and compromise. If this seems too demanding, recall that Michelman’s liberalism ends up in the same place: ‘Call it sacrifice, gift, graciousness: this motivation to collaboration . . . must stand as a requisite final component in the Rawlsian realistic-utopian conception of political reasonability’ (Michelman 2022, 101). To be sustainable as a mode of collective self-governance, democracy cannot escape the need for recognitive relations.

This problem of ‘thin’ versus ‘deep’ goes to the heart of liberal constitutionalism. The central claim of liberal constitutionalism is that a text-like object or a ‘diplomatically abstract’ set of principles can work a deflection of disagreements within a pluralist polity (Michelman 2022, 43). This leads, in turn, to the Goldilocks problem: If the constitutional essentials are to work as a response to reasonable pluralism, ‘they must not be too thin; they have to say something beyond vaporous or trivial about matters people care deeply about. But neither may they be too thick. They must stop short of foreclosure of questions of fundamental import to some citizens, over which reasonable citizens divide’ (Michelman 2022, 42).

With many of the others in this symposium, I don’t see how this strategy is supposed to succeed. Unless there is widespread agreement on the meaning of the essentials – in which case there is no reason to remove them from the give-and-take of democratic politics – the deflection will be quixotic. The most it can accomplish is to shift the level of conflict from debate over the acceptability of essentials to debate over the meaning of those essentials. In the end, this is a difficulty that Michelman does not resolve (Michelman 2022, 133–35). Nor could he. As Michelman notes elsewhere, the distinction between ‘secondary acts of interpretation’ and ‘originary acts of legislation’ is unstable – and this fact alone, he suggests, is reason to preserve such questions for the people themselves (Ferrara and Michelman 2021, 30). ²⁰

Consider the question whether socio-economic rights should be included in the constitution. These present difficult practical issues, Michelman points out. A Rawlsian framer might reasonably decide that, even if socio-economic rights are to be included, implementation might be left to the ordinary legislative process.

[P]rinciples of justice addressed to fairness in distributions . . . are ill-suited to the justificatory purpose because, Rawls says, of inevitable reasonable uncertainty and debate over the policies and actions required for their satisfaction, as compared with the classic liberal, mainly negative, “basic rights and liberties,” about whose satisfaction at any moment, Rawls thought, “it is far easier to tell.” (Michelman 2022, 39)

But why are socio-economic rights opaque and basic liberties easier to make out? It might be because liberty claims can be decided on abstract principle. On this view, ‘a due regard for basic liberties requires no balancing of conflicting claims . . . ; liberalism weighs at nil any [liberty] claims from governments or their minions’ (Michelman 2022, 187). Michelman (2022, 188–90) rejects that view. Instead, he assumes that, though cast at ‘accommodating levels of abstraction’, these rights and liberties ‘carry a core of commonly agreed meanings’ that are ‘sufficient to render coherent and persuasive the claims of citizens to each other’ (Michelman 2022, 52). Does this work any better?

It cannot, and for two reasons. First, one of the best-documented findings in cognitive science is the phenomenon of prototype effects: category members that are reliably rated better examples of the category than others and that figure prominently in reasoning. For the most part, prototype effects arise because human categorization involves complex, gestalt conditions – that is, a confluence of factors whose sum is psychologically more fundamental than the individual parts (Winter 2001, 76–92). The ‘central cases’, then, are just those that match a constitutive categorization. They stand out because they fit some category or categories that the observer already holds. The experience of prototype effects creates the impression that there are clear, readily available meanings for most abstract concepts. But this experience of clarity – what I have elsewhere called ‘the illusion of transparency’ (Winter 2013, 115) – is deceptive. Meaning involves complex processes of intelligibility that draw on tacit knowledge of the relevant domain including context, purpose, social understandings and assumptions, models of behavior, etc. Under conditions of pluralism, differences in norms, values, and forms of life will structure different categories and produce different ‘central cases’ along with different principles of extension. ²¹

Second, basic rights and liberties are not different from socio-economic issues, neither in complexity nor in the trade-offs they pose. Many free speech claims are distributional: Restrictions on hate speech, regulation of corporate speech, or limitations on campaign expenditures pit the rights, interests, and relative wealth of private actors against one another (Post 2005, 149–51). The same is true of affirmative action or the application of the one-person-one-vote principle in the context of gerrymandering or malapportionment. In considering a liberty claim, a decision-maker must weigh multiple competing interests, regardless of whether those affected are parties to the case. Consider a challenge to a law requiring vaccination. The decision-maker must weigh the challenger’s interest in bodily integrity against the government’s interest in the public health, and that incorporates the bodily integrity claims of all those who might be exposed to contagion. It is a distributional claim that balances the challenger’s social utility against the social utility of others. (So, too, for the criminal law) There will be inevitable uncertainty both about these trade-offs and about the practical actions necessary to overall public well-being. ²² It is a fair point to say that only a democratic politics can answer these kinds of open-ended policy questions. ²³ But the same is true for all rights claims – including property claims under the Takings Clause, as Michelman (1967) has famously argued, or for the conflict between religious liberty or free speech claims, on one hand, and the claims for recognition and inclusion by LGBTQ citizens, on the other. ²⁴

Whatever levels of abstraction are chosen, a constitution cannot avoid the problems of pluralism, policy, and interpretation. Even if everyone in a polity has consented to the constitutional essentials, what they have consented to is a matter of interpretation and thus of current contestation. We may all agree in principle on the equality of all citizens. But, notoriously, the intension of that principle can be expressed in terms of formal or substantive equality, equality of opportunity or equality of result. ²⁵ So, too, equality before the law leaves open the question whether the law must treat everyone the same regardless of circumstance or whether (and to what extent) it may consider differences of situation, status, or ability. The latter understanding was the approach taken by nineteenth century courts, which used it to uphold racial segregation. In the modern affirmative action debate, the positions are reversed. Progressives argue that equality must take situation and context into account and that officials may consider race in allocating various social goods in order to redress past discrimination or to achieve diversity. Conservatives, on the other hand, maintain that equality requires that each person be treated as an individual without regard to race and its attendant social circumstances. ²⁶

On one hand, a text must always be interpreted. On Michelman’s own account, there is no way to do so without adverting to the background traditions and assumptions that underlie it. For a respectful interpreter, he says, the constitution ‘will be read against the backdrop of a political moral purpose – Dworkin’s lesson – that the judicial reader cannot simply find already in the text (because it indispensably informs the reading of the text) and so must of necessity bring to it’ (Michelman 2022, 131). On the other hand, some basic parts of the U.S. Constitution – the peaceful transfer of power, the rule of law and the independence of the justice system, sovereign immunity, ²⁷ the two-term limit prior to the Twenty-Second Amendment ²⁸ – are part of the basic public platform without being specified in the written text. Indeed, one crucial lesson of the Trump presidency was how very much of our constitutional system depends on such unspoken conventions.

It could not be otherwise. Consider what I call the ‘ashtray problem’. We can reflect on any concept and tease out its implications; we may, as with the concept of an ashtray, be able to specify its necessary and sufficient conditions. But we will always already be reflecting on our culture and history and discovering there the social deposit that shapes our thinking. ²⁹ An ashtray is always an ashtray. But it has no meaning outside a cultural practice of the consumption of tobacco (or similar substances) by smoking. In just this way, Michelman concedes that ‘any defense of “a constitutional scheme with judicial review” must advert to “historical circumstances and conditions of political culture”’ (Michelman 2022, 41 n.26). Repeatedly, he acknowledges that the situated social practices and attendant values that constitute our democratic nomos play an overtly regulative function. ³⁰

Illustrative is the question of ‘fixation’ and temporality. Early on, Michelman asserts that the procedural function of a constitution cannot work without ‘a text or text-like object, the terms of which are made visible in the public space where everyone can point to them and agree that, yes, that is the country’s constitution’ (Michelman 2022, 42–49). Indeed, he later insists that ‘the justificatory aim cannot itself get off the ground without an observable fixture in place already’ which, in turn, entails ‘an originalistic inner bound to any practice of constitutional interpretation’ (Michelman 2022, 134–35). But just as the only future we can apprehend is the one we have in the present imagined, the only past we can have is that constructed in the moment from our current perspectives with all their values, commitments, and pressures. An ‘originalist’ starting point for justification-by-constitution necessarily starts with the understandings, assumptions, and expectations of contemporary interpreters. ³¹

Justification-by-constitution needs a way to incorporate the changes in understandings, circumstances, or social values and still maintain fidelity to the perceived past. The problem of change over time is, on Michelman’s account, accommodated via a four-part ‘dialectical process of long-term adaptive modulation’ in which citizens bring to bear their current understandings of the historical tradition and its demands (Michelman 2022, 79). ‘A just constitution’, Michelman says, is ‘never fully realized’ but ‘a project to be carried out’ (Michelman 2022, 86). Exactly so. Our scripted, publicly ascertainable constitutional essentials, it turns out, are always in flux. ³² As we learned from Robert Cover, ‘law is always becoming’ (Cover 1975, 6, 123–29).

Just self-government

Abstraction is the sleight of hand that risks transmuting the gold of democratic self-governance into the lead of institutional dogma. ³³ We expect liberal constitutionalism to support the legitimacy of a democratic rule of law. But, with diplomatically ‘thin’ essentials and institutional arbiters, liberal constitutional practice all too often undermines it.

A traditional response to the anti-democratic implications of constitutional review is a set of legal doctrines that institutionalize the idea of judicial restraint. These include remands to political branches, abstention on nonjusticiable matters, and deferential ‘rational basis’ review. Such doctrines, Michelman explains, temper the overreach of judicial interpreters without running afoul of ‘a strong-form understanding of the supreme court’s’ role in enforcing the constitutional essentials (Michelman 2022, 169–70). Michelman offers as an example the Supreme Court’s decision in Rucho (2019).

Rucho held challenges to partisan-political gerrymanders nonjusticiable. A lot is at stake for the liberal constitutionalist. On one side of the equation, a partisan-political gerrymander allows a political ‘party, although overall outvoted, nevertheless to gain and keep control of the legislative body in question’ (Michelman 2022, 161 n.23). On the other, if the courts’ involvement in partisan-political squabbles undermines confidence in their neutrality, then the entire project of justification-by-constitution is in trouble. ³⁴

Rucho decided that, because of ‘a perceived impossibility’ of finding objective standards that would safely put the Court’s involvement in the highly politicized task of redistricting ‘above suspicion’, the issue must be left to ‘nonjudicial political agencies’ (Michelman 2022, 161). Although his subtle use of the term ‘perceived’ conveys skepticism, Michelman nevertheless appears to give Rucho a pass. By referring some issues to other branches of government or back to the people themselves, he says, doctrines of judicial restraint respond to the ‘standing dilemma’ of maintaining confidence in the court that is ‘internal to the political-liberal project of justification-by-constitution’.

Granting that some form of institutional settlement is needed here, the dilemma would abate with entrustment of that function or some share of it to some more openly political process, one whose perceived nonpartisanship would rest not (or not exclusively) on attribution of special skills of decryption of legal texts and precedents but rather (or also) on breadth, quality, and vigor of participatory input. (Michelman 2022, 162)

Note that this argument relies, once again, on the ongoing democratic debates of the living citizenry (the ‘openly political’ and ‘participatory’ process) to do the work of institutional settlement that was supposed to solve the problem of pluralism. But, of more immediate concern is that judicial abdication on this issue is doubly inappropriate. On one hand, it is an axiom of American constitutional thought that judicial review is most warranted when invoked to protect the integrity and fairness of the political process. ³⁵ On the other, the issue in Rucho is of the sort least suited to resolution through democratic politics. The point of a political gerrymander is to entrench the party in power by deliberately tilting the playing field. Its purpose and effect, in other words, are to undercut the quality, vigor, and effectiveness of the participatory process. For the Court to refer the problem of gerrymandering to the very political agencies, the state legislatures, that are its beneficiaries is both disingenuous and self-defeating. Nothing could more cynically undermine democracy.

To a strong democrat concerned with the organizational arrangements that maintain democratic legitimacy, ³⁶ Rucho looks like Public Enemy No. 1. The question is: Why doesn’t it look that way to the liberal constitutionalist? Is the right to vote not part of the package of basic rights and liberties? Is the bedrock fairness of the political process not essential to the liberal principle of legitimacy? Rucho should be a case where the project of legitimation by constitution supports intervention. To use it as a paradigmatic example of judicial restraint is puzzling.

These are rhetorical questions. I am sure that Michelman, too, thinks Rucho badly decided. I raise these questions, rather, to put in bas-relief the issue that has been the focus of this comment: the nature of constitutional law and its relation to our underlying democratic nomos.

For Michelman, the point on which the Rawlsian and the strong democrat disagree is ‘whether some of the norms they take as regulative for a democracy worth fighting for have to be, or ought to be norms of law, meaning within the keeping of some institutional setup outside the daily operations of democratic politics. . . .’ (Michelman 2022, 181). But Rucho undermines this characterization. On the gerrymandering question, our positions are reversed: It is the Rawlsian who tolerates judicial abstention in favor of the political process. And it is the strong democrat, typically skeptical of judicial review, who thinks that the gerrymandering issue (to paraphrase Learned Hand) importunately demands a judicial answer (Hand 1958, 15). How did this happen?

This apparent reversal is an artifact of a false distinction between law and social institutions (Michelman 2022, 177). It is (in a word he does not like) a reification. Indeed, he signals this hypostatization when he italicizes the word law in the above quote. But law is not some abstract entity, special and above. It is not, as Michelman once said, ‘an autonomous force’ that provides ‘an external untouchable rule of the game’ (Michelman 2022, 176–77). ³⁷ Law is just another institutional form – a performative practice – that does (or does not) reflect our democratic nomos.

My objection to Rucho isn’t that, in abstaining, the Court failed to do law. The Rucho Court did law when it applied the judge-made political question doctrine – a legal rule that no citizen in the history of the Republic ever voted on. ³⁸ Rather, my objection is that Rucho failed to engage with the implications of its ruling for our central democratic commitments. It feigned impossibility instead, scorning any consideration of realistic alternatives and punting the issue to an already broken political system. ³⁹

In the abstract, judicial restraint promotes democracy. Whether it does, in fact, depends on the circumstances. Consider the interaction between Rucho and the subsequent decision in Dobbs (2022). In overruling the precedents upholding abortion rights, ⁴⁰ the Dobbs majority soothingly explained that the Constitution ‘leaves the issue for the people and their elected representatives to resolve through the democratic process in the States or Congress’ (Dobbs 2022, 2305). But many of these states have heavily gerrymandered legislatures that are significantly more conservative than their voters, a substantial majority of whom (even in red states like Kansas) support reproductive freedoms. ⁴¹ In the abstract, the Court can pretend to be promoting democratic decision-making. But, as always, the devil is in the details. The full implications of Rucho emerge only when we recognize the many ways and the countless contexts in which an unrepresentative political process can no longer ‘be expected to bring about repeal of undesirable legislation’ (Carolene Products 1938, 152 n.4).

These cases underscore the folly of externalizing a text-like object and entrusting it to an unaccountable arbiter. ⁴² They illustrate the danger of abstraction that is indifferent to real-world practices and unmoored from our central democratic commitments.

Liberal constitutionalism founders on the paradox of its own metaphors. ⁴³ On Michelman’s view, we are always subject to law. ⁴⁴ But the fixed, external quality of law is a projection: It is the community, as Robert Cover astutely observes, ‘that posits a law external to itself’ (Cover 1983, 45). Law may be the object (from the Latin ob, ‘towards, against’, + iaciō, ‘I throw, hurl’). ⁴⁵ But we are always the subject (from the Latin sub, ‘under, at the foot of’, + iaciō). ⁴⁶ To be committed, as a strong democrat, to the rule of law as an essential element of ‘a regime of collective self-government among equals’ (Michelman 2022, 181) is to be committed to the idea that we make the rules by which we govern ourselves. It does not mean that one must buy into a reified set of ground rules that are ‘autonomous’ or ‘external’. And it hardly commits one to an institutional arbiter whose authority depends on the ‘attribution of special skills of decryption of legal texts’ (Michelman 2022: 162). ⁴⁷

Equal voice, equal power, and equal law are just self-government. They are by definition internal to – that is, constitutive of – the game. Without them, one would be doing something other than democracy. That does not make them ‘untouchable’. To the contrary, we are painfully aware that they are being manhandled nearly every day. But they nevertheless describe a minimalist constitution that fully legitimates itself in its performance.