On Friday, February 11, 2011, newly-elected Wisconsin Governor Scott Walker unveiled his proposal for the state's budget-repair process. Budget repair was a routine event at the midpoint of the fiscal biennium, and Walker's Republican colleagues enjoyed commanding majorities in both houses of the state legislature. Under the circumstances, one might have expected the governor's call for immediate passage to be uneventfully complied with. Earlier, Walker had briefed the president of the state Senate, Mike Ellis, on the bill's contents, which included sharp restrictions on public-sector collective bargaining. “My God, this is going to cause a firestorm,” Ellis told the governor.Footnote 1 He was right. Within a week, mass rallies at the Capitol had become a daily event; teacher protests had led to school cancellations across the state; and all fourteen Democratic state senators had departed for Illinois, blocking the quorum required to act on budget bills. Mobilization led to counter-mobilization, and the intensity of conflict spiraled upward. On March 9, still without Democrats, the state Senate separated the collective-bargaining and fiscal provisions and voted its approval of the former. By the time the state Supreme Court upheld the Senate's action—ruling 4–3 that a lower court had “usurped the legislative power” in applying the state's open-meetings law to lawmakersFootnote 2—petition drives had forced the scheduling of recall elections for six Republican and three Democratic senators (effectively new contests for each seat), and the six Democratic challengers faced primary opponents recruited by state Republicans.Footnote 3
Like the protestors in Madison, national commentators framed these events as much more than a hard-fought distributive conflict. In the Washington Post, Charles Krauthammer fumed that Democrats had hindered Walker “by every extra-parliamentary maneuver short of arson,”Footnote 4 while Paul Krugman of the New York Times implored “anyone who cares about retaining government of the people by the people [to] hope” Walker's bill would fail.Footnote 5 For all the intensity of feeling, however—and notwithstanding a raft of allegations—it has yet to be established that any Wisconsin public official violated constitutional, statutory, or chamber rules in the course of the budget-repair battle. So why did the situation explode? Wisconsin's budget breakdown serves to remind political scientists that the formal rules by which democracies resolve disagreement do not operate in a vacuum. Instead, they coexist with a framework of unwritten or informal rules that structure collective expectations about how disputes will be resolved. Violation of these expectations can lead to destabilizing conflict, as clashes over policy escalate to clashes over the political process itself. The unwritten rules of legislative procedure that were invoked in Wisconsin—what we will call its informal institutions—manage fundamental democratic tensions, enabling majorities to rule (“lawmakers should vote on bills even when they know they are going to lose”)Footnote 6 while granting a measure of influence to even badly outnumbered minorities (“proposals that are both novel and significant should not be enacted the day after they are announced, even if they enjoy majority support”). In Wisconsin, both sides used powers to which they were formally entitled in defiance of longstanding, informal expectations about the policy process. With the unwritten rules under strain, debate shifted from the merits of policy to the democratic legitimacy of actions and actors alike. As one state representative said of the straw-man recall candidacies, “There are no rules anymore.”Footnote 7
Given their role in stabilizing and legitimating democratic contestation, the unwritten rules of the political process should be of particular interest to political scientists. While research on the developing world has seen a recent growth of interest in informal institutions,Footnote 8 however, scholars of established democracies have preferred to focus on constitutions, statutes, and other formal rules. This preference is reinforced by the undeniable challenges of subjecting informal institutions to convincing empirical analysis. Where data on other subjects are plentiful and widely available, political scientists' disinclination to investigate informal institutions might be thought to reflect a reasonable allocation of disciplinary effort.
This article challenges that conclusion. We reject the widespread, if often implicit, assumption that informal institutions are politically significant mainly where formal institutions are new, underdeveloped, or dysfunctional. Informal institutions, we argue, demand political scientists' attention also where the formal rules of politics are established, elaborate, and seem to be running smoothly. This is so for both substantive and theoretical reasons. Substantively, a failure to notice informal institutions at work can leave us unable to explain important outcomes, such as the recent breakdown of orderly governance in Wisconsin. As events in Wisconsin suggest, the unwritten rules of politics shape both the process by which political conflicts are decided and the content of those decisions—the relative power of different groups (majorities and minorities; insiders and outsiders) and the distributive share each ultimately receives. Theoretically, attention to unwritten rules offers a new way to confront problems of broad importance in institutional analysis. We focus on two of these problems: how to theorize the coexistence of multiple, clashing institutions in one political domain, and how to account for institutional change without recourse to exogenous shocks. We address the first question in part by classifying informal institutions' functions in settings already dense with formal institutions: informal institutions complete, coordinate, and operate parallel to their formal counterparts. With respect to dynamics, we identify the formal/informal institutional interface as a site of tension productive of change in both written and unwritten rules.
We illustrate the application of this theoretical framework by analyzing four informal institutions in American politics. These are the two-term norm of presidential service (a completing institution), the norms governing extended debate in the U.S. Senate (a parallel institution), the unwritten rules of the presidential nomination process (coordinating institutions), and the public presidency (a norm with all three functions). The exceptional density and diversity of its formal institutions make the U.S. an instructive case: the written rules to which we refer include constitutional provisions, federal and state laws, political-party rules, and the internal rules of a legislative chamber. Neither alone nor in combination, however, can formal rules fully account for political activity in these domains. Behavior is also shaped by unwritten rules—that is, by informal institutions. We suggest not that Americanists have failed to notice this fact, but that they share with scholars of other developed democracies a tendency to look past the informal rules of politics where this can be done without apparent difficulty. Resisting this tendency, we argue, shines new light on recent trends in American politics and may help to strengthen cross-subfield lines of communication in the discipline. In this cross-disciplinary spirit, we begin by considering how institutions have characteristically been studied in different areas of political science.
Formal and Informal Institutions in Political Science
Political life is full of patterned behaviors; some prescribed by written rules, others not. Political scientists consider many (though not all) recurring, predictable behavior patterns to be the product of institutions, and definitions of “institution(s)” often admit both formal and informal types. Peter Hall and Rosemary Taylor, for example, characterize institutions as “formal or informal procedures, routines, norms and conventions embedded in the organizational structure of the polity.”Footnote 9 Douglass North, adopting the view of rational-choice rather than historical institutionalism, defines institutions as “humanly devised constraints that shape human interaction,” a society's “rules of the game.”Footnote 10 Nothing in North's definition specifies that these rules must be in writing.
If institutions can be formal or informal, how can we distinguish the two—or distinguish institutionalized behavior from other behavioral regularities? These are difficult questions, as a valuable recent discussion by Gretchen Helmke and Steven Levitsky in these very pages demonstrates.Footnote 11 We follow Helmke and Levitsky in defining informal institutions as “socially shared rules, usually unwritten, that are created, communicated, and enforced outside of officially sanctioned channels.”Footnote 12 In other words, informal institutions exist when shared expectations outside the official rules of the game structure political behavior. When behavior has no pattern, or when patterning derives from something other than a collective (“socially shared”) understanding of right behavior in a given setting (the “rules”), that behavior cannot be described as institutionalized. On the other hand, when behavior is governed by written rules and violations are punished through “officially sanctioned channels,” the behavior-governing institution in question is formal rather than informal.
Formal and informal institutions have attracted different amounts of attention in different polities. In the study of formal institutions, theories and findings have characteristically diffused from the developed to the developing world. Hypotheses about the effects of electoral laws on party systems, for example, were initially formulated for Western Europe and the Anglo-American democraciesFootnote 13 and later refined for broader application around the world.Footnote 14 In the study of informal institutions, this direction of diffusion is reversed, with scholars of developing countries taking the leading role. Describing a recent surge of interest in informal institutions, Helmke and Levitsky refer to “a growing body of research on Latin America, postcommunist Eurasia, Africa, and Asia,”Footnote 15 a list from which North America and Western Europe are notably absent. The characteristic roles informal institutions do play in research on established democracies have contributed to circumscribing scholars' interest. Two habits are especially important: first, a tendency to see informal-institutional analysis as having been superseded by other analytic tools; and second, an inclination to turn to informal institutions only when accounting for the failures of democratic governance.
As an example of the first tendency, recall that informal institutions were once central to accounts of legislative behavior in established democracies. In Donald Matthews' classic study of the “folkways” of the U.S. Senate, that chamber appeared as an informally-structured social “world” within which members' actions were shaped by unwritten “norms of conduct” like apprenticeship, specialization and reciprocity.Footnote 16 Later Congressional scholars, however, came to view Matthews' account as obsolete. Some argued that the Senate and its institutions had changed.Footnote 17 Others rejected the analysis on its own terms, preferring to view norms as only one among many outcomes of rational decisionmaking in legislative settings.Footnote 18 The apparent distinction between norm-driven and interest-driven behavior was thus an illusion of the naïve observer—a plausible but mistaken perception producing plausible but mistaken accounts of legislative behavior. The rationalist rather than the anthropological perspective dominates contemporary Congressional analysis.
The second habit is to focus on informal institutions only when accounting for pathological outcomes: political phenomena at odds with what democratic institutions ought to produce.Footnote 19 There is a striking intersection between the set of longstanding democracies in which the informal side of politics is studied most often and those where governance is widely viewed as defective in some way. Consider research on corruption in ItalyFootnote 20 or clientelism in Japan,Footnote 21 both countries noteworthy also for a tendency to one-party rule. A classic U.S. example is V.O. Key's Southern Politics. For Key, the political dominance of outnumbered black-belt whites was maintained by channeling conflict through “amorphous political factions within the Democratic party,” an informal politics he viewed as defective in itself: “ill-designed to meet the necessities of self-government.”Footnote 22 In work of this type, the implication is that “informality rules”Footnote 23 where formal rules have been reduced to an irrelevance, a façade. Nominally, formal institutions operate as elsewhere—elections are held, parties compete—but outcomes are determined behind the scenes, by entrenched but unrecorded norms and practices.
These two approaches have dominated the study of informal institutions in established democracies. One consequence is that informal rules are often understood as undermining, or standing in what Helmke and Levitsky call a “competing” relationship with, formal rules.Footnote 24 A second is a tendency to treat informal institutions as a residual category, a backstop source of explanation where others have failed. Even so, the turn to informal features of politics has paid off in accounting for many otherwise puzzling phenomena: why bureaucrats behave in ways not prescribed by written procedures or formal hierarchies;Footnote 25 why party leaders form coalitions counterindicated by legislative arithmetic and ideological proximity;Footnote 26 why identical laws support different political outcomes in different places;Footnote 27 and, in a literature reaching back to Matthews, how informal structures and de facto rules shape behavior in Congress.Footnote 28 Informal sources of political authority have been identified for actors ranging from U.S. presidentsFootnote 29 to bureaucratic agenciesFootnote 30 to supranational organizations.Footnote 31 This body of work, all by scholars of established democracies, runs counter to the tendencies outlined above: informal institutions are seen to be of contemporary as well as historical interest and to promote both functionalFootnote 32 and dysfunctionalFootnote 33 outcomes, as well as mixed or ambiguous ones.Footnote 34 However, these strands of awareness have not been effectively integrated across sub-specialties: informal institutions are referred to by different labels and theorized in different ways, and these literatures rarely come into contact with one another. One aim of this article is to encourage greater conceptual and theoretical synthesis in the study of informal institutions. To that end, we build on a growing theoretical literature, much but not all of it in the comparative subfield.
Theorizing Informal Institutions
The empirical study of informal practices is a challenging endeavor. As David Mayhew observes of the filibuster, in the study of informal institutions, “[t]heoretical puzzles and data difficulties abound.”Footnote 35 Like Mayhew, we aim to approach theoretical issues with the concrete problems of empirical analysts in mind. Our discussion centers on three questions: (1) What are informal institutions? (2) What do informal institutions do?, and (3) How do informal institutions contribute to institutional change? As we will show, these questions connect to broader aims of contemporary institutionalism: theorizing the intersection of clashing institutions and accounting endogenously for institutional change.
The Nature of Informal Institutions
Some behavioral regularities reflect obedience to formal rules. Many other recurring behaviors, even in politics, are neither prescribed nor enforced by official means. These informal patterns have many sources. Some are collective but uncoordinated responses to common stimuli, as when pedestrians raise umbrellas on a rainy day or reelection-minded legislators engage in “credit-claiming” and “position-taking.”Footnote 36 Some reflect coordination on a choice arbitrary in itself but enduring once established. Some follow from shared mental states: perhaps an idea, like Keynesianism; perhaps a shared, articulable expectation of appropriate behavior;Footnote 37 perhaps a cultural frame so deeply rooted as to be incapable of articulation.Footnote 38 Other possibilities include coercion by powerful groups and the operation of biological requisites of human existence. While it is widely agreed that behavior governed by informal institutions, or norms, constitutes one subset of these regularities, there is less agreement as to what belongs in that category. Two broad perspectives exist: that informally-institutionalized behaviors should be understood as equilibria, and that they should be understood as rules. Footnote 39
On the first view, most prominent in the rational-choice tradition, an informal institution is a behavioral regularity, of a particular kind. Institutions are understood as equilibria: simultaneous best responses for all players of a specified game. This leads to definitions like Robert Axelrod's, for whom “[a] norm exists in a given social setting to the extent that individuals usually act in a certain way and are often punished when seen not to be acting in this way.”Footnote 40 Axelrod's definition turns on two features: behavioral regularity and punishment of non-conformers. On the second point, some in this tradition define norms as self-enforcing,Footnote 41 while others note the possibility of endogenous enforcement.Footnote 42 Fundamentally, however, norms-as-equilibria are understood to persist because it is in the interests of actors to maintain them. What is less clear is that the concept “norm” has meaning independent of “equilibrium.”Footnote 43 With respect to Congressional norms, for example, Kenneth Shepsle and Barry Weingast write that “‘deference’ labels a behavioral regularity; it does not explain it. The theoretical question of interest is why that behavior is an equilibrium.”Footnote 44 Calling deference a “norm” is not wrong, exactly, but it does nothing to explain deferential behavior.
Contrast this with a contemporaneous view of Congressional norms from David Rohde. For Rohde, “the concept of norm implies shared expectations of desirable behavior and enforcement mechanisms for deviance.”Footnote 45 The second point recalls Axelrod's definition, but the first claim is different: for Rohde, “[b]ehavioral patterns alone do not imply that a norm exists.”Footnote 46 This definition exemplifies the view of informal institutions as rules. Rules are characterized by what Daniel Brinks calls “normativity” and “facticity”: they “state a standard of conduct,” and they are enforced.Footnote 47 Not every unwritten (or, indeed, written) rule is enforced vigorously, but the consistent absence of sanctions applied to observed violators would suffice to demonstrate the nonexistence of a rule.Footnote 48
Like Brinks, Rohde, and Helmke and Levitsky, we understand informal institutions as rules.Footnote 49 This approach bounds the set of behavioral regularities that qualify as informal institutions. Not every action that is widely avoided and predictably costly, though formally permissible, violates an unwritten rule. An umbrella-less pedestrian may get wet, and a legislator who eschews publicity-seeking may fail to be reelected, but neither endures a sanction imposed by an agent upholding some collective expectation of right conduct. Defining informal institutions as rules clarifies analysis by forcing researchers to specify what behavior is rule-driven. Treating “civicness” as an informal institution, for example, would require that civic engagement constitute a normative expectation sanctioned in the breach. This might suit some analysts' purposes, but others would probably prefer to treat civicness as a variable attribute of communities rather than as an informal rule. Or consider the study of organizations as “natural systems.”Footnote 50 It is not obvious that informal networks qualify as rules. But networks, and other features of the natural system, may be tied to unwritten rules: rules about information dissemination, about deference, about organizational priorities, and so forth. These rules are enforced by sanctions, positive and negative, that affect status, project success, and other valued outcomes. Defining informal institutions as rules, then, helps to constrain our subject matter, calling attention to potential objects of investigation while distinguishing unwritten rules from related concepts like ideas, cultures, networks, and equilibria.
Understanding informal institutions as rules points to concrete issues empirical researchers must clarify: the content and scope of a given rule; the nature of deviance; how and by whom deviance is observed and punished (and/or conformity rewarded).Footnote 51 Specifying that informal institutions are enforced avoids some difficulties of studying self-enforcing equilibria or socialized understandings from which deviation is unthinkable, and not requiring that informal rules be upheld by non-state actorsFootnote 52 opens our view to informal institutions that closely involve agents of the state, including clientelism, corruption, and Congressional norms. Still, defining informal institutions as rules does not eliminate the problem of observational equivalence that affects studies of informal practices generally. A definition that turns on expectations and enforcement raises the question of how these things can be observed in the world. Direct evidence on expectations, prioritized by some scholars,Footnote 53 may be unreliable or hard to come by.Footnote 54 An alternate strategy, focusing on enforcement,Footnote 55 leaves the question of how to proceed if no violations are observed or threatened. A third option, relying on behavioral evidence, is seemingly in tension with our definitional insistence that informally-institutionalized behavior reflect at least semi-conscious obedience to rules.
These obstacles do not make the task impossible, however. That evidence on behavior may be unconvincing or evidence on expectations unavailable does not mean these problems will always arise. In the case studies that follow, we aim to show that informal institutions can be studied in a variety of ways, each with its strengths and drawbacks. Some of our cases rely mainly on behavioral evidence, others mainly on expectations, and we will keep the question of deviance and its consequences continually in mind. While we agree that interviews and other first-person sources can be of particular value in studying informal institutions, we would not restrict informal-institutional analysis to those who rely on such data.
Functions of Informal Institutions
Resolving what informal institutions are brings us to our second question: what do informal institutions do? Do they all do the same thing, or can distinctions be made between them? Several taxonomic schemes have been suggested for informal institutions, with intended reference to the developing worldFootnote 56 or to emerging democracies.Footnote 57 We propose a three-part classification designed for understanding informal institutions' roles in established, densely-institutionalized democracies. Informal institutions can complete or fill gaps in formal institutions; operate parallel to formal institutions, jointly structuring behavior in some domain; or coordinate the operation of intersecting institutions or institutional “orders.” This is a division by role or function rather than by type: its categories are not exclusive, as we will show by considering one of our cases from all three angles. An overview of these functions, with examples drawn from our case studies, can be found in Table 1.
Informal Institutions and American Democracy
A first characteristic role of informal institutions is that of completing formal institutions where those contain gaps or ambiguities.Footnote 58 A formal institution may remain “incomplete”—its terms not fully specified—due to unresolved conflict over what the terms should be, or because it would have been prohibitively difficult to specify all contingencies at the time of its creation (here, “incomplete institutions” resemble “incomplete contracts”). Carol Mershon illustrates a completing informal institution in her analysis of party coalition-building in Italy.Footnote 59 Mershon shows that the formal rules governing coalition formation are supplemented by unwritten but equally binding rules about who may initiate or lead certain coalitions, which parties make acceptable partners, and so forth. The operation of this completing informal institution significantly reduces the indeterminacy of the written rules, making coalition politics more predictable than it would otherwise be.
Second, informal institutions can operate parallel to formal institutions, exerting joint but separable effects on behavior in a given domain. When written and unwritten rules work in parallel, this does not mean actors choose between them;Footnote 60 instead, behavior is governed simultaneously by formal and informal precepts. (The two need not be equally salient, however. As in the Wisconsin budget battle, unwritten rules often gain salience when bent or broken.)Footnote 61 The parallel operation of formal and informal rules is especially important where politics involves complex procedures, as in legislatures and bureaucracies. Consider the debate over whether Congressional caucuses, as informal institutions, “counterbalance” or “mirror” the “formal organizing institutions” in each chamber.Footnote 62 On the first view, caucuses strengthen legislators powerless under the formal rules;Footnote 63 on the second, caucuses duplicate formal structures' power imbalances.Footnote 64 On either view, however, (informal) caucus and (formal) committee and party structures are understood to operate in parallel: both shape the behavior and power of individual members, and either—as the very existence of the debate indicates—might operate differently without necessarily altering the other.
The third role of informal institutions is coordinating the operation of formal (and/or parallel informal) institutions that govern the same actors or activities. The institutions coordinated may be actually contradictory, as when an activity is governed by written rules that cannot all be obeyed at once, but most coordinating informal institutions resolve tensions that are less stark. Institutions operating in the same arena often reflect diverse origins and goals: as Karen Orren and Stephen Skowronek observe, “the institutions of a polity are not created or recreated all at once, in accordance with a single ordering principle; they are created instead at different times, in the light of different experiences, and often for quite contrary purposes.”Footnote 65 This complexity can result in clashes, collisions, or, less dramatically, the interweaving of institutions not clearly reconcilable on their own terms. Coordinating informal institutions operate within these settings, creating stable expectations where there would otherwise be conflict or uncertainty. One example would be informal organizational practices developed to cope with irreconcilable formal mandates, as when public agencies subject to conflicting statutory demands rely on unwritten rules to prioritize tasks or allocate resources. A related example would be the “adaptive” informal institutions Kellee Tsai discovers in China, where entrepreneurs and local officials evolved informal routes by which to reconcile capitalist markets with the formal institutions of party and state.Footnote 66 Note that it is not the clashing institutions themselves we describe as “coordinating”: rather, coordinating (or perhaps “mediating”) informal institutions are the unwritten rules by which political actors resolve, or at any rate contain, inter-institutional tensions and conflicts.
Like the earlier typologies proposed by Helmke and Levitsky and by Anna Grzymała-Busse, ours hinges on the relationship between informal and formal institutions—an especially appropriate focus in densely-institutionalized democracies.Footnote 67 In highlighting intersections between institutions, moreover, this conceptualization advances a broader aim of contemporary institutional analysis: moving beyond the study of institutions in isolation to understand how institutions coexist and co-determine consequential outcomes. Scholars of American political development, for example, emphasize that the U.S. polity contains a multiplicity of institutional frameworks, or “orders,” in which competing values, ideas, and logics collide.Footnote 68 The analytic problem becomes one of understanding the joint effects of overlapping and perhaps warring institutional wholes. We suggest that explicit attention to informal institutions can help scholars gain purchase on this problem, and we view our taxonomy as a tool with which to identify and differentiate informal institutions' functions in complex polities.
Informal Institutions and Institutional Change
As institutionalism developed, researchers broadened their focus from the effects of institutions to what Peter Hall has called “the second-order problem” of accounting for institutional change.Footnote 69 As Hall notes, explaining change in phenomena that interest scholars precisely for their persistence (enduring paths; stable equilibria) is logically as well as empirically difficult.Footnote 70 The logic of institutional stability favors explaining change by reference to exogenous shocks: “shocks” as events that disrupt established patterns (whether or not dramatic in themselves); “exogenous” in occurring for reasons external to the pattern disrupted.Footnote 71 With persistence and change attributed to different causes, institutional development appears discontinuous: periods of continuity “punctuated” by episodes of change. The limitations of this approach have prompted a growing interest in endogenous mechanisms of institutional change: dynamic processes arising within or among existing institutions.Footnote 72 Explaining institutional change endogenously is especially important in established democracies, where few institutions are truly created from scratch and institutional legacies have been shown to survive even dramatic exogenous shocks.Footnote 73
While we do not propose a complete theory of institutional change, endogenous or otherwise, we do suggest that informal-institutional analysis can contribute to understanding institutional dynamics. For one thing, attention to informal institutions may shed light on otherwise puzzling observations about institutional change. If behavior patterns are seen to change where formal rules have not, a change in unwritten rules should be considered as a possible explanation. Similarly, if behavior remains stable in the face of formal-institutional reform, analysts should explore the possibility of compensating change in informal institutions. Second, and more importantly, we suggest that the interface between formal and informal institutions is itself dynamic, and that written and unwritten rules can each promote change in the other. Previous work identifying the formal/informal interface as a site of change has tended to stress one direction of influence only (mostly formal → informal).Footnote 74 We suggest that causal effects operate in both directions (i.e., also informal → formal), and that both are amenable to empirical analysis.Footnote 75 Changes or gaps in formal institutions, or tensions between them, can stimulate political actors to create or revise informal institutions—in order to stabilize expectations, neutralize incipient conflicts, or settle points not resolved by the written rules. In the opposite direction, violations of or dissatisfaction with informal rules can motivate actors to redress the perceived problem by altering or creating formal institutions.
The mechanisms that give rise to endogenous change at the formal/informal institutional interface should vary with the institutions in question. Here again, our typology of informal-institutional roles helps to clarify analysis. Completing informal institutions mesh closely with formal institutions: if written rules change, the unwritten completion of those rules must often change as well. Reversing directions, a change in formal rules may follow when informal rules are challenged or violated, as in the case of presidential term limits. (Contrast Axelrod's prediction that the norms most likely to be formalized are the strongest ones.)Footnote 76 Parallel informal institutions, by contrast, are less tightly linked to their formal counterparts: either may change without necessarily producing change in the other. Still, the parallel operation of formal and informal institutions can create cross-pressures for change on either side. Dissatisfaction with the effects of unwritten rules may lead a coalition to mobilize for formal reform—perhaps with the goal of limiting or prohibiting an informally-sanctioned practice. On the other hand, parallel informal institutions may change as actors adjust their collective expectations to an altered framework of formal rules. This effect can either reinforce or thwart the intended impact of formal reform, depending on the power relationships involved. Coordinating informal institutions, finally, should undergo change as the institutions among which they mediate themselves evolve or (dis)appear. Our understanding of coordinating institutions suggests that the net effect on politics may often be stability, or at any rate greater stability than the tensions between coordinated institutions would suggest. Failure or breakdown of coordinating informal institutions, however, might then presage episodes of significant change, as with punctuated equilibria. In addition to indicating mechanisms of institutional change, finally, a focus on the formal/informal interface may help researchers identify coalitions likely to promote or resist change at a particular moment.
As an example of pressures for institutional change arising from the link between formal and informal institutions, consider the events with which we began. In Wisconsin, a formal recall process with a surmountable activation threshold had long coexisted with—in our terms, been completed by—an informal expectation that out-party activists would mobilize to cross that threshold only under exceptional, incumbent-specific circumstances. Faced with an unprecedented number of successful recall petitions, leaders of the state's Republican majority began to advocate raising the formal barrier to initiating recall elections.Footnote 77 Were they to succeed—or were the state Senate to pass new rules enforcing attendance on its members—these shifts would exemplify a characteristic mechanism of change in completing institutions: informal violations prompting a coalition to mobilize for change in formal rules. The Wisconsin case also illustrates what may be characteristic rhetoric surrounding such a move. Advocates of formal rules change presented themselves as reformers challenging anti-democratic practices, while their opponents saw an entrenched interest defending its position and privileges.
Analyzing Informal Institutions: Cases from American Politics
We have argued that informal-institutional analysis can untangle puzzles of stability and change in deeply-institutionalized polities. We turn now to a second task: showing that informal institutions can be subjected to empirical analysis in feasible and useful ways. Our illustrations are drawn from the established democracy in which formal-institutional analysis has shaped political science most profoundly: the United States of America. We consider four cases: the historical two-term norm of presidential service, the expectations surrounding legislative obstruction in the Senate, the unwritten rules of the presidential nomination process, and the norms of the public presidency.
These cases have several useful features. All involve substantively important outcomes, and three of the four are situated in large research literatures. (Political scientists have had much less to say about presidential tenure.) They are also diverse, touching both sides of major divisions in the American subfield: institutions and behavior; heavily quantitative and more qualitative literatures; areas influenced by rational-choice theory and by historical institutionalism. We aim to show that informal-institutional analysis has something to offer across all these divides, and that informal institutions can be approached with different types of data. Remembering the theoretical lessons above, we attempt throughout these mini-case analyses to specify the content of unwritten rules; to understand how informal institutions function; to observe whether and with what result rules have been violated; and to explore episodes of institutional change.
The Two-Term Presidency
Our first case is an unwritten rule that endured for a century and a half before being given formal, constitutional expression. The substance of the rule is simple: presidents shall not serve more than two terms. The two-term norm of presidential service exemplifies a completing informal institution, one that fills a gap or resolves an ambiguity in a formal rule. Here the formal rule is Article II, and the gap is its failure to specify any limit—or no limit—on presidential reeligibility. The framers bypassed this issue not because it was too complex or because they did not consider the possibility of a written provision, but because they were unable to resolve their differences over the contours of the presidency. At one extreme, Alexander Hamilton urged the possibility of lifetime service for presidents. Ranged against Hamilton were advocates of various more restrictive stipulations on term length and reelection.Footnote 78 While the term-length question was eventually settled, the issue of reeligibility was not, and the Constitution was ratified with no consensus as to the maximum appropriate duration of a president's career.
The two-term tradition that soon developed is usually attributed to George Washington's decision not to seek a third term in 1797. Washington does not seem to have intended his withdrawal to set a precedent,Footnote 79 but the second president to face the third-term question took a different view. In Thomas Jefferson's words:
If some termination to the services of the Chief Magistrate be not fixed by the Constitution, or supplied by practice, his office, nominally four years, will in fact become for life, and history shows how easily that disintegrates into an inheritance. Believing that a representative Government responsible at short periods is that which produces the greatest sum of happiness to mankind, I feel it a duty to do no act which shall essentially impair that principle, and I should unwillingly be the person who, disregarding sound precedent set by an illustrious predecessor, should furnish the first example of prolongation beyond the second term of office.
Footnote 80Jefferson's language expresses not political calculation or personal preference, but the sense of an informal expectation with normative force. He writes of a “duty” to maintain executive rotation, and he implies that Washington's “sound precedent” has given that duty definite form: at most two terms of presidential service. The rise of this expectation resolved the debate left open twenty years before, creating collective certainty on a point not addressed in Article II.
Violations of this rule are easily identified: a deviant is a two-term president who seeks (and certainly one who wins) a third or subsequent term. Among nineteenth-century presidents, only Grant's name was put in convention nomination for a third (nonconsecutive) term, and his bid for the nomination failed.Footnote 81 That Grant's run was thwarted in part by pronouncements in favor of the two-term limit by Pennsylvania state Republicans and the U.S. HouseFootnote 82 suggests an effort at preemptive enforcement, even in the strongly federal, decentralized party structures of the day. The first ex-president to contest an election for a would-be third term was Theodore Roosevelt in 1912. TR was defeated, and the circumstances were unusual in several ways: he had been elected only once before,Footnote 83 was not the incumbent, and ran on a third-party ticket, having lost the Republican nomination. Even so, the Democrats' 1912 platform called for a constitutional amendment barring presidential reelectionFootnote 84—a proposal that presaged the consequence of a later, successful violation of the two-term rule, though it did nothing to dissuade Woodrow Wilson from seeking reelection four years later.
How can we know that this pattern reflects an informal institution? Bruce Peabody and Scott Gant dismiss the idea of a “tradition,” arguing—in an important alternative explanation—that political circumstances consistently but unsystematically conspired against third-term candidacies.Footnote 85 One way to counter their objection would be with evidence on motivations, as from Jefferson above. Even observing outcomes alone, however, the persistence of the two-term pattern is remarkable. And Peabody and Gant are not wrong to consider politics: the fact that the two-term norm created regular succession opportunities for in- as well as out-party politicians gave ambitious would-be presidents strong reason to enforce the rule on others.
Also striking, and easier to explain on our argument than Peabody and Gant's, is that the one successful violation of the rule—by Franklin Roosevelt—was soon followed by its formalization in the Twenty-Second Amendment. (The economic and geopolitical crises of the timeFootnote 86 may also counter the view that a three-term presidency could have been politics as usual.) Roosevelt was comfortably reelected, and his opponent's efforts to campaign on the third-term issue had no apparent effect.Footnote 87 But the normative violation, soon doubled, did spark an effective counterreaction: a coalition mobilized to formalize the two-term limit, and a corresponding constitutional amendment was ratified in 1951. The coalition for change united two groups: New Deal opponents and liberals who favored the restriction for its own sake.Footnote 88 With the constitutionalization of the two-term rule, FDR's presidency became a historical anomaly.
The norm of two-term presidential tenure illustrates the operation of a completing informal institution. It also reveals characteristic ways such institutions emerge and change: here, an underspecified (because contested) formal institution prompted the creation of an unwritten rule, and the eventual violation of that rule motivated formal reform. Completion by informal means is an important device enabling ambiguous and lacuna-filled formal institutions, such as the U.S. Constitution, to function as regulators of complex political domains. The political questions resolved in this way can be weighty ones: from the Philadelphia convention to the debate over the Twenty-Second Amendment, participants understood the issue of presidential tenure to have broad implications for the nature and power of the presidency.Footnote 89 The possibility that similar gaps or ambiguities may be completed by different informal rules in different settings suggests interesting avenues for comparative research as well.
Senate Obstruction
Congressional procedure is governed by a framework of formal institutions: written rules that determine how proposals are considered. But the legislative process—and, by extension, the substance of public policy—is also shaped by informal institutions. These unwritten rules operate parallel to their formal counterparts, and legislative behavior cannot be satisfactorily explained by reference to either alone. Parallel informal institutions create distinctive paths to institutional change: legislators' dissatisfaction with informal rules can prompt efforts to reform formal procedures (and vice versa), but either type of rule can also change independent of trends in the other. We focus here on the Senate, and specifically on its rules governing the practice of minority obstruction.
Gregory Wawro and Eric Schickler describe the nineteenth-century Senate as almost entirely lacking in formal rules: senators' behavior was structured by “shared, stable procedural expectations” and associated norms—that is, by informal institutions. As the chamber and its workload grew, demand for formal rules increased.Footnote 90 One important change was the creation, in 1917, of Rule 22, which enabled a two-thirds supermajority to limit floor debate by voting to invoke “cloture.” In 1975, the majority needed for cloture was reduced to sixty (three-fifths), and procedural obstruction by unlimited debate (the “filibuster”) was made possible even with no speaker on the floor.Footnote 91 Neither Rule 22 nor later changes, however, did much to affect senators' use of extended debate: cloture reform and filibustering are at best loosely related.Footnote 92 To understand change over time in observed Senate procedure, we must refer to the chamber's unwritten rules.
Both before and after Rule 22, the Senate functioned essentially as a majoritarian body, and cloture votes were very rare. From 1917 to 1958, more than two cloture motions were filed in only two biennial Congresses,Footnote 93 and nine Congresses saw none at all. Major, controversial policies like the New Deal were approved by simple majorities.Footnote 94 Although the rules made it possible for a minority of senators to obstruct legislation, there was no assumption that filibusters would occur as a routine matter, and in fact they did not—except on civil-rights issues, whose distinctive treatment was itself an unwritten procedural rule. Today, however, a new and very different rule exists. In Gregory Koger's words, “the Senate transitioned from a majority-rule legislature with an occasional case of logorrhea to a sixty-vote Senate predicated on the ability of every senator to obstruct any bill at any time.”Footnote 95 Thus in January 2010, the Democrats' loss of their sixtieth Senate seat was greeted as a near-fatal blow to the policy ambitions of a party that still dominated the chamber 59–41. This replacement of a majoritarian informal rule with a supermajoritarian one has created a very different chamber. That the change took place without change in the written rules illustrates the power of parallel informal institutions.
Can the shift in the use of the cloture rule be explained in some other way? This again is the problem of distinguishing informal-institutional effects from other behavioral regularities, and we approach it here with quantitative evidence on behavior. Consider three alternative hypotheses. The first derives from Keith Krehbiel's theory of “pivotal” lawmakers.Footnote 96 The “filibuster pivot” hypothesis predicts fewer cloture motions when the filibuster pivot is a member of the majority party (that is, when the majority commands enough votes to invoke cloture) and more when this filibuster-proof majority is lost. A second, “new majority” hypothesis predicts a rise in cloture motions after party control changes hands. As the newly-installed majority moves to enact its agenda, the now-minority should seek to block change.Footnote 97 A third hypothesis derives from Sarah Binder and Steven Smith's argument that a cross-party “conservative coalition” dominated the Senate for decades before dissipating in the 1980s.Footnote 98 This suggests the “liberal Senate” hypothesis: cloture motions should have risen as liberal, policy-ambitious Democrats grew in numbers and status-quo senators obstructed change.
We can evaluate these hypotheses with the data in Figure 1, which shows counts of cloture motions filed and successful votes to invoke cloture in each Congress since the creation of Rule 22.
The filibuster pivot hypothesis refers to periods when one party holds a filibuster-proof majority. This has happened only twice: in the 89th Congress (1965–66) and for part of the recent 111th Congress (2009–10). The predicted drop in filings does not occur, nor do invocations of cloture decline. The 89th Congress saw slightly more cloture motions than the 88th and 90th (seven as against four and six), and the 111th Congress only very slightly fewer than the 110th (136 vs. 139) and 184% of the total for the 109th. The new majority hypothesis expects more cloture motions following shifts in the chamber majority. Between 1945 and 2010, party control of the Senate reversed nine times.Footnote 99 Among the succeeding Congresses, only one shows a noticeable increase in cloture motions: the 110th (2007–08).Footnote 100 The liberal Senate hypothesis, finally, predicts more cloture filings as the conservative coalition declines, a process that began with an influx of liberal Democrats in 1959.Footnote 101 Cloture motions did start to rise in the 87th Congress (1961–62), but the initial increase was very slow, and the upward trend continued and intensified long after the “liberal Senate” had become established fact.
The evidence for these hypotheses is mixed at best. Cloture trends confirm a fundamental shift in Senate practice, but not one explained by the logic of the filibuster pivot, the alternation or composition of chamber majorities, or changes in formal rules. Instead, senators have replaced one shared understanding of acceptable practice—one unwritten rule—with another: “cloture has gone from taboo to commonplace.”Footnote 102 This does not mean that structural or tactical variables are irrelevant. Observe the ratcheting shape of the trend: once a given level of use becomes acceptable, majorities and minorities may find cloture and filibustering, respectively, too valuable to relinquish. Another possibility is that rising polarization has strengthened minorities' capacity to force cloture votes and majorities' capacity to invoke cloture.Footnote 103 If this is true, shifts in majority control might come to have the effect predicted by the new majority hypothesis: perhaps the striking jump in cloture motions in 2007–08 can be explained in this way.
Congressional procedure is defined by both formal and informal institutions. The former are insufficient in themselves to explain behavior: witness the Senate's de facto replacement of majority with supermajority rule without formal procedural change. The written and unwritten rules on extended debate thus operate in parallel: each has an independent effect on outcomes, and each is compatible with a range of states (but not every possible state) of the other. This last point brings us to institutional change. Parallel formal and informal institutions may be more or less compatible, and tensions between the two can prompt endogenous institutional change. Many revisions to formal Congressional procedure have in part been reactions to the effects of unwritten rules,Footnote 104 as would be the case today if senators dissatisfied with supermajoritarianism rewrote the formal rules on obstruction. In the other direction, formal rules change may cause legislators to develop new informal procedures, either reinforcing or circumventing reform's intended effect. Whether or not endogenous mechanisms produce change in the supermajority Senate, it is certainly an important feature of U.S. politics today.
Presidential Nominations
The formal institutions of presidential nominee selection are a complex blend of party rules and state laws. The nomination context is also less bounded than those above: choice points are geographically and temporally dispersed, and participants are numerous and varied. Many patterned, predictable features of this decision process are not governed by written rules. Do they reflect the operation of informal institutions? This section explores the nomination process with an eye to distinguishing behavioral regularities that are informally institutionalized from ones that are not. We argue that some but by no means all informal aspects of presidential nomination take the form of rules that prompt sanctions when violated. Informal institutions are most apparent among elites: where actors are comparatively few and known to one another, rules can be communicated and sanctions made effective. We characterize the unwritten rules of the nomination process as coordinating institutions that stabilize and integrate its overlapping institutional orders.
Nomination contests now begin long before the first vote is cast. In an initial phase, the “invisible primary,” aspiring candidates gather resources and party elites survey the field for an electable, broadly acceptable, and (preferably) congenial nominee.Footnote 105 Candidate resource accumulation and elite coordination are informal but not, in themselves, obviously rule-driven processes. Certain features of the invisible primary, however, can be expressed as rules: for example, that candidates may not enter the race later than a certain point, and that they must at least seek the acquaintance (if not acquire the support) of a broad range of party elites. (A measure of vagueness in these rules is in the interest of elites to maintain.) Of course, one reason candidates do these things is to garner the resources available thereby. But if even a high-profile, potentially donor-rich candidate like Sarah Palin could not win the nomination without “playing by the rules,” as Jonathan Bernstein maintains,Footnote 106 this would be a strong indication that party elites recognize and can enforce unwritten rules for the invisible primary.
What of the “visible” primary, the series of state primaries and caucuses? Here, too, many informal processes are at work. Early results influence later ones, as initial winners accrue resources and media attention.Footnote 107 The sequential selection process shapes voter behavior through mechanisms like momentum and signaling.Footnote 108 These dynamics do not appear to reflect unwritten rules. Voters might in principle punish (vote against) candidates who violate informal rules—by ignoring particular states, say—but an empirical showing that voters are rule-driven would be tricky. What of party elites? We suggest, again loosely, that informal rules operate to govern candidate withdrawal, and that violators are plausible (what Barbara Norrander calls “office-seeking”)Footnote 109 candidates who overstay their welcome in the race. Office-seeking candidates face a dilemma: since they aspire to a political future, the censure of party elites is costly to them; since they (effectively by definition) have a body of support, unity-minded elites will want them out as soon as a presumptive nominee appears. These pressures are reflected in the language of both elites and candidates: elites urge perceived holdouts to step down for the good of the party,Footnote 110 and departing candidates speak of party unity and personal duty.Footnote 111 By contrast, candidates understood to have no chance of the nomination can stay in the race without attracting censure (or, indeed, much notice).Footnote 112
The willingness of recent presidential candidates, who have sometimes invested years in the race, to withdraw well before reaching any formal barrier to the nomination is striking. It is also a historical novelty.Footnote 113 The precise date on which a hitherto acceptable candidacy will begin to attract criticism is difficult to predict, in part because party elites have no reason to tip their hand. Still, certain dates have clearly become informal focal points at which party and media actors assess the field: the Iowa caucuses, the New Hampshire primaries, Super Tuesday. This informality has consequences. With no formal mechanism to force trailing candidates out, some may ignore informal pressure—as Hillary Clinton ignored a barrage of polite discouragement in 2008.Footnote 114 Since formal imposition of a pre-convention Clinton withdrawal would probably have shattered Democratic unity, however, party elites likely preferred the rules' informality even in the breach.
A final place to seek informal nomination institutions is in the sequential framework itself. The first-mover positions of New Hampshire (since 1952) and Iowa (since 1972) are rules of the game whose status was for many years informal.Footnote 115 Again, states are self-interested as well as rule-bound in frontloading the primary season. Super Tuesday, for example, originated in a southern-state Democratic drive for greater nomination clout,Footnote 116 but the event has become a kind of informal institution: a date to which states are free to move their primaries. That both parties are now attempting to formalize first-mover states' status (with sanctions on delegate apportionment and convention representation) testifies to the perceived value of what were once informal rules of nomination timing. Interestingly, these rules are in tension with the unwritten rules of candidate attrition. Given a firm drop-out date for candidates, states' incentive to schedule primaries before that date is strong; if the roster of early primaries is to be fixed and stable, candidates' do-or-die moment must remain uncertain.
What role do these informal institutions play? The nomination system is complex and historically unstable,Footnote 117 with new institutional arrangements layered over older ones created for very different purposes. In this context, informal institutions serve a coordinating role, reconciling mismatched institutional orders. Three moments of institutional innovation are especially important. The first, the party-building of the 1830s, concentrated power in the state and local party leaders who gathered every four years to nominate presidential candidates;Footnote 118 its legacies include federalism and the vestige of the national delegate convention. Progressive-era reformers, hoping to shift power from parties to voters, added a second layer with the introduction of the direct primary.Footnote 119 A third moment, launched in the Democratic party after 1968, led ultimately to the near-universalization of direct, binding presidential primaries.Footnote 120 The integration of these diverse institutional legacies—especially the first and last—has been accomplished in part by coordinating informal institutions. The unwritten rules of the invisible primary, for example, reconcile the democratizing impulse of the binding-primary era with a central role for the descendants of Jacksonian party elites (by, some argue, fatally undermining the former).Footnote 121 The drop-out norm for candidates fuses nomination by sequential primary with party leaders' interest in avoiding competitive contests and consensus-busting nominees.Footnote 122 Finally, informal rules about state scheduling integrate the sequential primary, party-elite influence, and the underlying federal structure of the nomination system.
The informal rules of the presidential nomination process cannot be specified as cleanly as those of Senate supermajoritarianism and the two-term presidency. Some rules are contested; others are vague or hidden—often intentionally so. But we would argue that they certainly exist, and that their coordinating role enables the smooth operation of a complex mechanism featuring many decisionmakers and few (and late) authoritative decision points. That nomination contests are now typically reduced to one plausible candidate long before any formal cutoff is due in no small part to informal institutions. The conflicting principles of party unity, state autonomy, and popular sovereignty might be irreconcilable by other means. At the same time, informal nomination institutions empower parties at the expense of voters, for all the democratizing impulse of the direct primary system. Unwritten rules thus temper the effect of past reform—and, insofar as their coordination remains effective, they may forestall future reform as well.
The Public Presidency
The study of rhetoric assumes almost by definition that some presidential influence is informal. Still, presidents might “go public,”Footnote 123 or address citizens at large, for any of several reasons. Going public could be a tool of policy influence, with presidents leveraging their sway over public opinion to move decisionmakers in Washington.Footnote 124 We suggest a second, not incompatible possibility: that the public presidency incorporates an informal institution, an unwritten rule dictating that presidents address the public on issues of major concern. Richard Neustadt famously wrote that a president is “expect[ed] … to do something about everything”;Footnote 125 the institution of the public presidency requires presidents to say something about everything. In this section, we make a case for the public presidency as informal institution and suggest that it can be seen as performing all three informal-institutional functions.
Going public was not always a presidential role. In the nineteenth century, it was considered illegitimate for presidents to communicate directly with citizens on policy matters. After the rise of the “rhetorical presidency,” however, presidents were not only permitted but expected to speak (rather than write) directly to the public (vs. Congress) on major policy questions.Footnote 126 How can we know that these patterns reflect informal institutions? One way is by identifying episodes of enforcement. Nineteenth-century presidents who violated the norm against public promotion of their ideas faced tangible sanctions: one of the impeachment articles against Andrew Johnson condemned him for “certain intemperate, inflammatory, and scandalous harangues.”Footnote 127 Violators of today's prescriptive (rather than proscriptive) rule face less dramatic consequences. Still, a president's decision not to speak publicly can exact a political cost, as George W. Bush found in the aftermath of Hurricane Katrina and Barack Obama when a press commentator called him “missing in action—unwilling, reluctant, or late to weigh in on the issue of the moment.”Footnote 128 The expectation that the president will “do something” about an issue is now often fulfilled precisely by speaking about it, and media actors are prominent enforcers of this expectation. As a result, contemporary presidents go public even when their political capital is depleted (Bush on the 2008 Emergency Economic Stabilization Act), when issues exceed their constitutional authority (Obama on Wisconsin budget repair), and when events are beyond executive authority of any kind (Obama on the Tohoku earthquake). In so doing, they conform to an unwritten rule about the responsibilities of the office they hold.
Viewed as an informal institution, the public presidency plays several roles. With respect to the office itself, unwritten rules on public speech perform a completing function. The Constitution does not specify whether presidents shall address the public—it is Congress that is to receive “Information to the State of the Union”—making the rhetorical function of the office one of many questions Article II leaves unresolved. Both prescriptive and proscriptive norms on presidential rhetoric complete this gap in the formal design. Jeffrey Tulis implies such a view when he calls the rhetorical presidency “an ambitious reinterpretation of the constitutional order”Footnote 129—that is, a replacement for the initial, nineteenth-century interpretation. The emergence of a norm requiring presidents to go public is ironic in light of founding concerns about demagogic leadership,Footnote 130 although the vesting of executive authority in a single individual is structurally conducive to a public role (as the founders also knew).
Analyzing the public presidency as a coordinating institution requires a wider view. The institutional order created by the founders entailed a limited federal government with power centered in Congress. Beginning in the Progressive era, new ideas emerged about the role of the presidency, the reach of the federal government, and the appropriateness of a direct connection between presidents and voters.Footnote 131 None of this displaced the original constitutional structure, however. Instead, the new expectation (what Victoria Farrar-Myers calls an “informal script”)Footnote 132 of a more active, president-centered governing process was layered on top of the institutional order of the founding. The public presidency functions as a means of coordinating these two incongruous orders—one mostly formal, the other largely informal. Speech-making and position-taking are ways for presidents to “do something about everything” despite the formal limitations of their office. The effect of this coordinating institution has been to produce stability where inter-institutional tension might otherwise have led to conflict and change.
Finally, the public presidency can be understood as operating parallel to the “institutional presidency,” the greatly expanded executive-branch organization (with accompanying formal and informal rules) that developed in and around the 1930s.Footnote 133 That these two presidencies may work in what we call parallel fashion, jointly determining the contours of the office, is implicit in a recent article by Jeffrey Cohen.Footnote 134 Cohen notes that the public and institutional presidencies emerged together, temporally speaking, and outlines several possible explanations for this fact. Of these, he stresses two: that the public presidency developed to enable presidents to gain control over policy, and that the institutional presidency emerged to enable presidents to build public support by producing policy.Footnote 135 If the two institutions operate in parallel, both explanations may be true—providing another example of endogenous change generated at the intersection of formal and informal structures. Over time, the public and institutional presidencies have mostly reinforced one another, but tensions can also emerge. Presidential rhetoric often appeals to change and rejection of the past, while the routines of the institutional presidency foster continuity. This tension can limit the transformative potential of rhetorically ambitious presidents,Footnote 136 but at other times it may catalyze actual change.
The public presidency embodies unwritten rules as well as political strategy. Chief executives use the “bully pulpit” in hopes of advancing a policy agenda, but they are also constrained to speak by a norm requiring presidents to address the public on important issues. The case of the public presidency illustrates that the completing, coordinating, and parallel functions of informal institutions are not exclusive. Adopting one or another view, moreover, suggests different avenues for empirical research. If the public presidency is understood to operate parallel to the formal structures of the White House, scholars might consider when and why the two institutions reinforce or clash with one another. The coordinating story emphasizes the tension between founding and modern institutional orders, suggesting research into how this inter-institutional mediation succeeds and when it may break down, producing conflict or institutional change. Approaching the public presidency as a completing institution, finally, highlights the gap between the role presidents have come to play in American politics and the resources afforded them by the Constitution. Because expectations for presidential action and presidents' capacity to act diverge more in some settings than others, it would be interesting to know whether the magnitude of that gap can predict presidents' reliance on rhetorical appeals.
Conclusion: Informal Institutions in Established Democracies
Informal institutions are the unwritten rules of political life: shared understandings created and enforced outside formal or legal channels. This article has explored informal institutions in four domains of U.S. politics: legislative procedure, sequential primaries, and presidential tenure and rhetoric. In each area, we find that informal rules can be identified with reasonable precision, observed in the world, and distinguished from other sources of patterned behavior, including strategic self-interest and the operation of formal rules. Although brief, these analyses suggest the potential for studying informal institutions in many areas of the American subfield and with many types of data: quantitative as well as qualitative evidence bearing on expectations, enforcement, and behavior. Framing the case studies, we introduced a new theoretical apparatus for the study of informal institutions in densely-institutionalized polities, distinguishing informal institutions' completing, coordinating, and parallel functions with respect to other institutions. We identify the formal/informal interface as a crucial site of endogenous institutional change, but also (as our cases indicate) of stability—showing that such analysis can approach persistence and change as interrelated rather than disconnected outcomes.
Why should scholars of American politics study informal institutions? The first and most obvious reason is that political actors in the United States do, in fact, alter their behavior in accordance with unwritten rules. It is for this reason that many research traditions implicitly acknowledge informal institutions' existence. We suggest that this focus can be made more self-conscious, and the exploration of informal institutions more concrete and precise, without sacrificing the gains Americanists have made from formal-institutional inquiry.
A second reason turns on the normative significance of unwritten rules. Informal institutions embody and reconcile core democratic concerns: citizen input and intra-party consensus in presidential nominations; bipartisan compromise and party policy advocacy in the Senate; executive authority and rotation in office, together with policy leadership and a duty to inform the public, for presidents. That informal institutions unite these concerns, however, need not mean that they do so in normatively satisfying ways, or in ways equally visible and accessible to all citizens. What is more, because major political conflicts are sometimes resolved by informal means, the effects of informal rules are distributive as well as uncertainty-reducing: political life is made not only smoother but also different from what it would otherwise have been. Informal institutions have strengthened party elites in the selection of presidential nominees, minority parties in the Senate, and presidents of both parties as against Congress—but not against their successors, guaranteed a turn in power by the two-term norm. While informal institutions are an important functional element in democratic polities, then, not all of their consequences are appealing. Unwritten rules have allowed party elites to blunt democratizing reform in presidential nominations and, in the Senate, enabled a safeguard for minority interests to metamorphose into a general supermajority requirement for legislative action.
We would also suggest that Americanists may have particular cause to study informal institutions now, for reasons rooted in the contemporary crisis of polarization in U.S. politics. Viewed in comparative perspective, the U.S. political system has unusually many formal veto points, but these are rarely exploited to full effect. The inbuilt tension between majority rule and minority veto is—as we argued at the outset—to an important extent managed by informal institutions. Rising polarization makes this tension more salient, and the interaction of polarization with unwritten rules becomes politically important. In several recent episodes, a decision by one or both parties to violate informal understandings launched or escalated a destabilizing spiral of conflict: the budget-repair dispute in Wisconsin was one example, and the national debt-ceiling conflict of mid-2011 another. In other cases, informal institutions help to sustain intensity spirals. This appears to be the case in the Senate, where the expectation that all legislation will be subject to a cloture vote allows senators to vote against cloture as a form of position-taking, making polarization and the supermajority norm mutually reinforcing. The replacement (in the Senate) or violation (in Wisconsin) of older norms promoting cross-party collaboration lessens the two parties' sense of shared interest in legislative outcomes. The transformation of informally-elided veto points into effective veto players may also have important consequences for political outcomes. Alfred Stepan and Juan Linz suggest this explanation for the United States' unusually acute socioeconomic inequalities, which are less effectively mitigated by public policy than those in other advanced democracies.Footnote 137
Finally, the study of informal institutions may promote greater intellectual exchange across sub-specialties of political science. Our discussion of presidential nominations, for example, used the theoretical apparatus of American political development (clashing institutional orders) to address a subject mainly studied as a topic in political behavior. We see similar possibilities for learning across subfields. Nothing we have said about the operation of informal institutions is unique to the United States. Completing informal institutions emerge in a wide range of contexts where formal rules underdetermine political behavior. Examples might include the practice of abstention on the UN Security Council and parliamentary regimes' unwritten rules of coalition formation (which party leaders almost certainly prefer not to put in writing). The political system of the United Kingdom arguably operates substantially on the basis of completing informal institutions. Coordinating informal institutions, in contrast, could be said to exist where formal institutions “overdetermine” behavior: where political activity is regulated by multiple, not-always-consistent institutional frameworks. We expect that all established polities are home to institutional intersections of this kind. As a transnational example, observe that the coordination of technocratic, state-centric, and democratic institutional principles in the EU is achieved in part through unwritten rules. Finally, parallel informal institutions should arise to co-define the authority and responsibilities of public officials—legislators, executives, and others—in other countries just as they do in the United States.
To date, however, the informal institutions investigated in this article have been studied in comparative isolation: by and for an audience of American-politics specialists, and largely without reference to comparable or contrasting practices in other democracies. Just as hypotheses about the operation of formal political institutions have traveled from the developed to the developing world, promising perspectives on informal institutions, we believe, should now be encouraged to move across subfields.Footnote 138 As our case studies illustrate, informal-institutional analysis that attends to a common set of questions—specifying the content and scope of informal rules, the means by which they are enforced, and the systemic functions they perform—makes it possible to see quite diverse phenomena as structurally comparable. Such an incorporation of analytic tools from the comparative subfield into American politics could help to remedy the deficits Paul Pierson identifies in contemporary Americanist scholarship: a lack of attention to macro configurations, to questions not easily addressed with quantitative data, and to problems that do not fit neatly into the customary organization of the subfield.Footnote 139 The greater conceptual clarity we have sought to provide for the study of informal institutions may also be of assistance to scholars engaged in cross-national comparison. Researchers now working on informal institutions in emerging democracies will gain from increased attention to those institutions' role where democracy is long-established. And scholars of established democracies, Americanists included, can gain by adapting for their own purposes an analytic focus now animating scholars of the developing world—the study of informal institutions.
