A liberal, Robert Frost famously joked, is a man too broadminded to take his own side in a quarrel.
In the midst of assaults against liberalism at every level, in a venue dedicated to its revitalization after a long crisis, I think the usual thing to do with this joke would be simply to reject it. Liberalism, one expects to hear in response, is a “fighting creed,” and we would do well to remember it.
I think that’s all true, and yet I think there’s something important to Frost’s line. The creed for which liberals fight is one about which the joke will always ring at least a little bit true; a revitalized liberalism will and should remain one that has a little bit of ambivalence about taking its own side.
To see why, let’s consider a couple of ideas that are standardly contrasted with the idea of liberalism as a “fighting creed”: liberal neutrality and liberal norms.
Get Liberalism.org in your inbox.
Liberal Neutrality: Against Favoritism
Neutrality is among the least loved ideas that have often been connected to liberalism. Neutrality among religions, arguably the core example from which much else has followed, is uninspiring compared with substantive religious belief. It is widely thought to be bloodless and abstract, derided as a vision of what Harvard communitarian Michael Sandel famously caricatured as the "unencumbered self”—unencumbered by beliefs, commitments, loves, passions, projects.
Critics of liberalism despise it. They think it’s simultaneously false, because liberalism promises neutrality among various ways of life, while it also and insidiously makes alternative ways of life unsustainable; and unworthy, because real, vital, manly political actors agonistically commit to a substantive goal with their whole heart. But liberals themselves are often uneasy with neutrality: in good times, they call to surpass it for something more ambitious and inspiring; in bad times, they worry that it discourages necessary political action; and in both, they wonder whether it neglects questions of ethos and character that are necessary for the maintenance of freedom.
These complaints often misunderstand the type of neutrality at stake. Liberal neutrality is not personal apathy or indifference; its opposite is not commitment. Rather, liberal neutrality is impartiality—the impartiality of rules, of procedures, of freedoms. Its opposites are favoritism, nepotism, privilege, and personalism. Its standard cases, no less important for being familiar, are state neutrality as among religions, and judicial neutrality as among persons.
The latter doesn’t mean, and has never meant, neutrality between criminal and innocent action. And it doesn’t rest on the fantastic claim, never achieved in any society, that the effects of the law will fall equally on everyone. Rather, it means the demand that the same procedures and principles will apply regardless of who appears as a plaintiff or victim and who as a defendant. The ideal represented by justice being blindfolded is older than liberalism, though it traditionally meant impartiality as among members of a privileged class: the same rules applied regardless of which propertied citizen, or which man, or which nobleman, appeared in each of the seats.
The eventual advent of liberal governance didn’t invent the rule of impartial law, but it presupposed it, rested on it, and generalized it. In the important account given by Douglass North, John Joseph Wallis, and Barry Weingast in Violence and Social Orders, the emergence of what they call the “open access order” was the generalization and democratization of what had previously been organizational tools reserved to ruling-class elites. Organizing a business as a corporation no longer depended on royal favor, whether from personal connection, family status, political alignment, or plain bribery. As mercantilism gave way to liberal market capitalism, it became a simple, universally available bureaucratic procedure. As the elite political competition of eighteenth-century Westminster gave way to an inclusive liberal democracy, ordinary citizens likewise gained access to the franchise and to the ability to join political parties. The right to form and belong to a church followed the same path, as grudging toleration among Protestant sects became liberal religious liberty.
It’s noteworthy how much the current crisis involves backsliding on exactly these dimensions: from neutrality, impartiality, and impersonal governance to favoritism, partiality, and personalism. The global trend has been ably documented in Stephen Hanson and Jeffrey Kopstein’s book The Assault on the State, but it has accelerated since their book was published two years ago. The attack on liberal commercial markets has taken the form not only of protectionist nationalist tariffs and an increasing state ownership stake in large firms, but an astonishing degree of personal favor, enrichment, and spite on the part of political leaders. Dangerous purges of the military officer corps, law enforcement, and the previously independent expert civil service have quickly spread from Türkiye and Israel to the United States, turning executive authority and security into tools of personal power and spite.
The always-difficult problem is that the partisan elected leadership of the executive has authority over law enforcement, something I analyzed in various ways here, here, and here. The problem has now become a genuine crisis. It’s brought the abuse of both prosecutions and pardons for personal and political gain. The challenge to liberalism today is not just a matter of liberal political actors being defeated by conservative or socialist rivals within a stable, underlying, and ultimately liberal system, within which the winners would also work. It is now a matter of liberal orders and systems crumbling into the mercantilist and personalist “natural state.” As North, Wallis, and Weingast show, that state is the historical rule, to which liberalism has been a rare and brief exception.
The complaint that neutrality is false because nothing is ever really neutral is utterly beside the point here. Yes, the open access order, the rule of impersonal law, the professional civil service, and the open market are substantive political goods, difficult to attain or to maintain, requiring substantive commitment. Yes, they have substantive effects on the social order. But, yes, they are neutral in the relevant institutional sense: in their various ways, they all further impartiality among persons. I discussed this idea in greater depth on The Institute for Liberal Studies’ Curious Task podcast with Alex Aragona.
Playing Hardball: Norms and Sanctions
Probably the most famous book about the backsliding in the last decade, Steven Levitsky and Daniel Ziblatt’s How Democracies Die, foregrounded the idea of democratic norms and their erosion. They identified the central norms as mutual tolerance, different parties’ acceptance of each other as legitimate rivals, and forbearance, a willingness not to push to the limits of institutionally permissible power. In terms more familiar to public law and constitutional theory, forbearance is a willingness to refrain from what Mark Tushnet calls “constitutional hardball.”
Levitsky and Ziblatt say that these norms are a necessary but previously under-noticed complement to such formal institutions as the separation of powers. They argue that these norms have been gradually eroded in the United States, mainly from the right, since the 1980s. And they say that this is a feature of cases of democratic and constitutional backsliding around the world. Their treatment of democratic norm erosion has been widely influential, and it now organizes scholars’ treatment of such topics as whether and when members of the public believe that certain kinds of speech can legitimately be met with official censorship or unofficial violence.
Whether or not it’s directly inspired by Levitsky and Ziblatt’s work, we hear a similar kind of discussion about norms in public intellectual debates on the center-left and left about how to respond to the two Trump presidencies. It is a commonplace on the left that the Biden administration in general, and Merrick Garland’s leadership of the Department of Justice in particular, was characterized by too much deference to norms, too much forbearance, especially in the pace of prosecutions for the January 6, 2021 attacks, but also, for example, in the refusal to try to expand and pack the Supreme Court, or to abolish the filibuster. This line of argument shapes criticism of Democratic congressional leadership on an ongoing basis, charging them with trying to govern according to outdated norms of bipartisan mutual toleration.
More broadly, some people seem to believe that if bad things happen in politics, it must be because their own side didn’t fight hard enough. That easily becomes a belief that their own representatives decided not to fight hard enough, deliberately handicapping themselves with rules and restraints. The language of “norms” has become, for some critics, a shorthand for such half-hearted politics. This mindset looks at the contemporary crisis, and asks of liberals a version of the question Anton Chigurh asks in No Country For Old Men: If the norm you follow brought you to this, of what use was the norm?
Levitsky and Ziblatt’s history of the erosion of forbearance and toleration is valuable. The broader discourse that treats forbearance as the model of liberal norms—political, or liberal, or constitutional, or democratic, norms—is, however, actively misleading. It muddles together norms with niceness.
Norms are neither laws, though laws may variously enforce, specify, or ground norms, nor virtues, though norms may habituate persons into virtues. It helps organize one’s thinking about norms to remember that they are both normative and normal, and that each of those matters. A norm is a shared expectation, in both senses of “to expect”: not only the predictive “I expect rain today” but also the prescriptive “I expect better of you.” A norm describes an approved-of behavior that can also be widely relied on, and the internalization of the pattern of behavior. Moreover, the normative and the normal reinforce each other. The widespread pattern encourages internalization through the flawed but standard psychology of sociable beings: what is done around here is a primary source for thinking about what ought to be done. And the belief that the norm ought to be followed leads to penalizing its violation, which serves to reinforce the pattern.
Sanctioning need not be a matter of formal punishment. Think of Tushnet’s metaphor of “hardball.” When a pitcher pitches too aggressively to the inside of the strike zone, retaliation in kind by the opposing pitcher is a standard response. The second pitcher isn’t contributing to the erosion of baseball’s norms; they’re enforcing the norms. Tit-for-tat self-help has limits as a way to enforce norms, and in many settings we try to replace it with third-party judgment of one kind or another. But often such judgment is not available, and if we want a norm to remain normal, to remain an expectation on which we can rely in common, it will have to be supported by horizontal enforcement. Without some sense of legitimate sanctioning of violation—even if it is only social criticism—I think that the habit, custom, or pattern isn’t actually a norm.
Liberal Norms and Their Institutions
As with neutrality, so with norms: their relevance to liberal politics is institutional.
Institutions, Robert Goodin argues, are themselves sets of norms. They are, as Goodin put it, the relevant “organized patterns of socially constructed norms and roles, and socially prescribed behaviors expected of occupants of those roles, which are created and re-created over time.” This is true even for institutions that rest on legitimate coercive physical force; “legitimacy” is a normative word and an observed social fact.
To return to the analogy for a moment: the norm against inside pitches is a genuine norm within the game, but the sport itself is an institution, built out of nothing but norms. The rules of conduct constituting the game—nine innings, three outs, three strikes, infield fly rule, and so on—are all stable expectations. Their violations are either sanctioned by the umpire, or they render the game invalid. If a violent mob storms home plate after the final play and demands that the result be overturned, the normative structure of the game is compromised, and the people are doing something other than playing baseball.
A constitutional government, a market-governed economy, democratic elections, the rule of law, a professional and impartial civil service, a separation of church and state, the freedom of speech and the press: these are all institutions built out of norms. The evolution of liberal orders consisted of these and similar norms gradually becoming both normal and normative. Religious groups could safely refrain from the struggle for political power, increasingly secure in the knowledge that they weren’t at risk from some rival religious group seizing it and persecuting them. Defeated political parties could peacefully leave power, knowing that their rivals would do likewise when the time comes. Producers could concentrate on economic production rather than mercantilist rent-seeking or channeling wealth into the pockets of rulers. And so on.
During the long process of the consolidation of liberal orders, these norms all demonstrated virtuous patterns of self-reinforcement. The more stably they could be counted on, the more robust the belief in their appropriateness; the more widely shared the belief in their appropriateness, the stabler the expectation.
It’s worth noting that these are all norms for circumstances of pluralism and disagreement. Religious liberty was a response to the fact of serious religious conflict. Liberal democracy, understood as a system in which parties cede power after they lose votes, only makes sense given the factionalization of political life into two or more parties. To focus on the latter case, liberal democracy is not a system in which liberal political parties necessarily win every election. This is Frost’s paradox: to be a partisan liberal is to support the establishment or maintenance of a system in which your own side can regularly lose.
Norm Erosion and Norm Collapse
A norm that is an independently normatively valuable aspiration might still erode as an observed normal expectation. (Conversely, it might gradually take hold; the spiral can be vicious or virtuous.) Standard examples from the literature of norms that have this quality include norms against littering, against leaving dog waste on the ground uncollected, and against smoking in particular times or places. Even in a place with litter everywhere, we can still make sense of what a person is doing who declines to litter, or who makes it a point to clean up a little bit every time they go for a walk. They’re doing their part, even if no one else is. If they’re just one person, they’re probably exercising a virtue rather than following a norm, but they might be trying to set a norm: They aim to show by good example and to get others to think about what they’ve been doing. In more intermediate cases, there’s enough compliance to describe the norm as present but not having fully taken hold, or as having eroded from some previous time but not yet collapsed. Today there is still a norm against talking on cell phones during a movie, though it’s significantly eroded.
By contrast, a norm that is normative primarily because it is normal, a coordination convention, won’t look like this. There’s no such thing as an eroded norm about which side of the street to drive on. If even 10% of drivers stop observing the local convention, then no one can safely rely on it. We immediately collapse into a much worse equilibrium, each driver having to navigate around every other driver in both directions, all the time. Where the normative force depends on keeping to the normal pattern, we should be on the lookout for collapse rather than erosion.
Some norms aren’t as purely conventional as this, but they’re close enough that they similarly collapse rather than eroding. Waiting in an orderly queue is, in my view, like this. Unlike which side of the road we drive on, there’s a moral difference between the rule “everyone wait your turn” and “whoever has the sharpest elbows goes first.” But once 10% of the people have committed to the latter, the people who remain aren’t like the person individually picking up litter. They’re simply suckers, abiding by an imagined rule that is not the operative norm, and they’ll never get to the front. Everyone understands this and reacts accordingly; they adjust to the fact that this particular crowd isn’t lining up in an orderly way—and the results can be violent, even fatal.
I suggest that liberal institutional norms are very often of the sort that risk collapse rather than mere erosion.
If one party in a two-party system refuses to abide by the norm of peacefully accepting election losses, the democratic norm of selecting leaders peacefully through elections is at risk of collapse. If one party violates the norms of respecting a free and independent press, using the full power of the state, from the bully pulpit, to antitrust law, to harassment lawsuits, to bring the press to heel, that norm is also at risk of collapse. The press won’t reset to an independent posture when the other party takes back power; it will rationally respond to the threat of the initial party’s return.
The same logic holds across the range of personalist, favoritist, mercantilist, and other politicized attacks on the norms of neutrality. If one party purges the professional civil service, or the officer corps of the military, if it replaces the impartial rule of economic law with tariffs, subsidies, and antitrust permissions that depend on bribery and personal access, trust in those norms as an expected practice is shattered. They are not the sort of expectations that function in a four-years-on, four-years-off system.
This puts liberal political actors, movements, and parties in a difficult position. The open-access order’s emergence and stability relied on a political consensus about the positive-sum character of the peace and prosperity that liberal rules made possible. One party could take the lead, but it couldn’t unilaterally create the whole order. Under circumstances of norm crumbling and collapse, a party that plays by no-longer-operative rules just creates a permission structure for surrender to the other side. Even when the norm-abiding party is in power, the press, the civil service, big business, the military and police, and so on will have every reason to curry favor with its rival, on the expectation that the rival party will return to power and restart retribution and favoritism soon enough.
Does this mean that abandoning norms is the right strategy? No, that dodges the real difficulty. Liberal politics is, in large part, the politics of trying to instantiate liberal norms. Freedom of religion and speech, the rule of law, the impersonal open market, free and fair elections: those are neutral institutions across persons, but that kind of neutrality is the liberal’s fighting creed. Reversing the other side’s friends and enemies lists, changing around who gets treated with favoritist privilege and who with spiteful retribution—these aren’t liberal politics at all. These are the destruction of the order that liberal politics tries to create.
The difficulty is real. Without the possibility of sanctions for noncompliance, a norm becomes just an aspiration. But the institutions that can sanction violations within a constitutional order—courts, most obviously—are themselves creatures of constitutional norms. Without third-party enforcement, sanctions reduce to tit-for-tat self-help by the parties involved. But for norms of the sort we’re considering here, that amounts to a different kind of surrender for liberals: a surrender of the overall order at which they aim.
Norms, Not Niceness
Some political norms are a lot like the norm against inside pitches: there’s a ready tit-for-tat enforcement mechanism that leads straight to mutual forbearance. Others are more like the norm against an armed mob storming home plate; without third-party enforcement, the norm may vanish. If critics of liberal norms from the left sometimes forget the difference and treat all normative restraint as equally suspect, some avowed supporters of liberal norms blur the distinction for the opposite reason. They treat norms of forbearance and nice behavior within rule-governed structures as if they’re just the same as the rules that constitute the structures themselves. We are currently in the middle of a major round of tit-for-tat gerrymandering in the United States, at least some of whose participants would prefer a symmetrical scaling down of the tactic. Joseph Fishkin and David Pozen refer to this as anti-hardball, retaliatory action aimed at ending the escalation.
This is, I think, one of the sources of a general distaste for norms-talk in the contemporary crisis. The intra-elite comity of the sort that characterized bipartisan governance in the United States in the mid-twentieth century was a norm, too. Before we get nostalgic about it, we should recall that it was a norm that rested on the ugly fact of Jim Crow. White conservative southern Democrats could get along with white moderate Republicans nicely enough; the white Democratic solid south meant that Democratic dominance in Congress wasn’t ever really at risk, and the exclusion of Black voters meant that many divisive questions were kept off the agenda. (This has been widely studied; most recently, see Stephen Skorownek’s The Adaptability Paradox.) There’s something especially perverse about nostalgic elites whose commitment to such niceness constrains them from forthright criticism of the violations of fundamental, system-level norms. Here at last we find the actors who genuinely won’t take their own side in the political quarrel.
This is probably not unique to the United States. Elite esprit de corps is very real, even in the most democratic and inclusive open orders. Niceness across the aisle is psychologically easy. Confronting the task of trying to build or rebuild system-level norms when opponents refuse to cooperate is not. This is the challenge for liberal renewal: a willingness to take our own side robustly, while remembering that our own side is in large part the advocacy of norms that are neutral between the sides in such quarrels.
