The Yale Law Journal

December 2014

Romanticizing Democracy, Political Fragmentation, and the Decline of American Government

abstract. This Feature was delivered originally as the 2013-14 Ralph Gregory Elliot Lecture at Yale Law School.

author. Sudler Family Professor of Constitutional Law, NYU School of Law. I am grateful for comments received there and at a presentation of these views at the NYU Faculty Workshop. I would also like to thank Hélène Landemore and Nick Stephanopoulos for comments. For excellent research assistance, I thank Nabil Ansari, as well as the NYU library reference team, particularly Gretchen Feltes. I am also indebted to discussions with former Senators and members of the House, all of whom requested anonymity.


For many years now I have been interested in developing more of an institutionalist and realist perspective on the dynamics of democracy and effective political power, particularly in the United States. By this I mean a focus on the systemic organization of political power and the ways that legal doctrines and frameworks, as well as institutional structures, determine the modes through which political power is effectively mobilized, organized, and encouraged or discouraged. This perspective emphasizes, among other elements, the dynamic processes through which winning coalitions are built or destroyed in the spheres of elections and governance. The mutually influential relationship between these spheres ultimately determines the ways in which our democratic institutions function or fail to function.

This focus on the organization, structure, and exercise of actual political power in elections and in governance is what, in my view, characterizes “the law of democracy”—a systematic field of study in law schools for only the last twenty years or so.1 To sharpen up this initial description, I would contrast the approach of the “law of democracy” to those approaches to constitutional law and theory that center on protecting and developing the dignity, or the autonomy, or the “personhood” of the individual, and ensuring the equal treatment of particularly vulnerable groups. These are the aspirations of Taking Rights Seriously, for example—the arresting book title that defines the approach of someone who has been much on my mind lately, my recently deceased colleague, Ronald Dworkin.2

Even more, however, I want to contrast my focus on the systemic organization of political power to rights-oriented approaches applied to democracy itself.3 By rights-oriented approaches, I mean approaches that focus on interpreting and elaborating in normative or doctrinal terms the general, broad, political values of democracy, such as participation, deliberation, political equality, and liberty, or the associated legal rights to political association, to free speech, to the vote, or to political equality. These rights-oriented approaches typically pay less attention to the structural or systemic consequences—the effects on the organization of political power—of concretely institutionalizing these abstract ideals in specific settings. Rights-oriented perspectives also often rest, implicitly, on a conception of democracy that envisions individual citizens as the central political actors. We can see these approaches in constitutional doctrine, in reformist advocacy about democracy, and in scholarship on democracy in political theory, philosophy, and law.4 My suggestion, however, is that these approaches can spawn, and have spawned, doctrines and policies that undermine the capacity of the democratic system as a whole to function effectively. Instead of this rights-based orientation, I want to encourage more focus on how political power gets mobilized, gets organized, and functions (or breaks down).5

In this Feature, adapted from a lecture I gave at Yale Law School in November 2013, I will illustrate this approach by addressing a problem on many of our minds, what my title calls “The Decline of American Government.” In making this statement, I mean to appeal to a broad consensus of such a decline. Therefore, I do not refer specifically to an inability to act in areas of partisan conflict in which one side has a substantive policy preference for the status quo (climate change policy, for example). Rather, I refer to arenas where there is broad consensual agreement that government must act, in some fashion, but where American government now seems incapable of doing so—or where government does act, but only after bringing the country or the world to the edge of a precipice: government shutdown, the regular dancing on the knife’s edge of the first U.S. government default, and the like. I do not want to suggest that American government is in some state of extreme crisis; American democracy has faced far more dramatic challenges before,6 and as democratic observers from de Tocqueville to today have recognized, democracy is rarely “as bad as it looks” at any particular moment.7 It is enough to recognize serious dysfunction even in only particular areas to motivate a search for deeper explanations, as well as directions for possible paths forward.

I. political fragmentation

I want to offer two main ideas about how to think about the decline of America’s governance capacity and effectiveness.

First, I want to suggest that we cannot understand how our democratic institutions are designed and how they function without recognizing that a uniquely American cultural sensibility and understanding of democracy—one that I view as excessively romantic, particularly in the forms it takes today—informs a good deal of the ways we design and reform our democratic institutions. This uniquely romantic conception of democracy has, I believe, perversely contributed to the decline of our formal political institutions. This will be one of my themes: the dangers of democratic romanticism.

Second, in diagnosing the causes of government’s limited capacity to function effectively, there is a widespread temptation to focus on how polarized the two dominant political parties have become (as well as on whether polarization is asymmetric between the two parties).8 Much of the commentary on polarization has focused on the difficulty of fitting America’s increasingly parliamentary-like political parties into the Constitution’s institutional architecture of a separated-powers system.9 The understandable concern that many have today is whether in times of divided government—but not only then, given the Senate filibuster rule, which remains in place on policy matters—the absence of a “majority government” will make it too difficult to generate the kind of concerted political action required for legislation.

If the concern about polarization is best understood as one about effective governance, then we should perhaps refine the concern, particularly for pragmatists searching for potentially productive directions of plausible reforms. To do so, we should identify the issue not as political polarization alone but as one of political fragmentation. By “fragmentation,” I mean the external diffusion of political power away from the political parties as a whole and the internal diffusion of power away from the party leadership to individual party members and officeholders. My claim is that, for pragmatic reformers, political fragmentation of the parties (most obviously visible, at the moment, on the Republican side, but latent on the Democratic side as well) is a more important focus of attention than polarization if we are to account for why the dynamics of partisan competition increasingly paralyze American government. The government shutdown and near financial default were not a simple product of party polarization; they reflected the inability of party leaders to bring along recalcitrant minority factions of their parties and individual members to make the deals that party leaders believed necessary. The problem is not that we have parliamentary-like parties. Rather, it might well be that our political parties are not parliamentary-like enough:party leaders are now unable to exert the kind of effective party leadership characteristic of parliamentary systems.

If this analysis is correct, stronger parties—or parties stronger in certain dimensions—ironically might be the most effective vehicle for enabling the compromises and deals necessary to enable more effective governance despite the partisan divide. I will offer a quick sketch of a few policy proposals designed to re-empower political party leaders in order to make government more functional. But the specific proposals are less important in themselves than as illustrations of a direction of reform that might enable more effective governance in the enduring context of highly polarized political parties.

II. democratic romanticism

Let me begin by impressing upon you the uniqueness of America’s practices and institutions of democracy, taken as a whole, compared to those of other mature, stable democracies.

Jacksonian-era reforms have bequeathed us the world’s only elected judges and prosecutors.10 Indeed, we elect more than 500,000 legislative and executive figures, vastlymore than any other country per capita (one elected official for every 485 persons): we elect insurance commissioners, drainage commissioners, hospital boards, community college boards, local school boards, and on and on.11 Furthermore, we lack independent institutions to oversee the election process, such as specialized electoral courts, independent boundary-drawing commissions, and independent agencies—institutions common in most democratic countries.12 This leaves partisan, elected, and mostly local officials in control of much of the regulation and administration of the electoral process, out of a perverse belief that doing so makes the process more democratically accountable to “us.”13

Our administrative state, in general, is far more subject to democratic control than those of other well-established democratic countries. Although there have been periods in which we embraced independent administrative agencies based on ideals of political independence and expertise, such as in the Progressive and New Deal Eras, the dominant and distinct characteristic of American administrative government has been the emphasis on political control (legislative or executive) over administrative agencies or what is often called “democratic accountability.”14 Indeed, the ever-increasing American skepticism of “expertise” and pressure for more and more “popular” or “democratic” control over our institutions makes it doubtful, in my view, that the political force could be marshaled today to create an independent central banking system, such as the Federal Reserve System created in 1913, if we were facing the issue for the first time now.

As another reflection of the degree of political control over public administration perceived to be necessary in the United States, there are roughly 1,300 positions in the federal government that require Senate confirmation, from the Supreme Court to the fifteen members of the National Council on Disability, not to mention the vast amount of time that administrators spend after appointment subject to the political pressures of myriad congressional committees before which they testify constantly.15 As another institutional example, our democratic culture produced an extraordinarily fragmented banking system for most of American history, from the 1830s until around the 1990s; this made American banking exceptionally unstable and prone to crises relative to the banking systems of some other democratic countries (averaging one crisis every decade).16 Democratic understandings and politics made our banking system uniquely subject to local, popular political control; our laws generated a highly disaggregated, decentralized system of tens of thousands of “unit” banks (individual local banks, with no branches) that were regulated overwhelmingly at the state level and thus politically controlled by coalitions of local bankers and agrarian populists. Indeed, the leading political history of banking systems in different countries characterizes the American banking system throughout the 1830-1990 period as “crippled by populism.”17

Even more to the point for my purposes now, Progressive Era reforms, such as the state-imposed requirement that political parties choose their nominees through primary elections, have long made our political parties more subject to “popular control” than in virtually any other democracy.18 We take for granted both that we vote for individual candidates, rather than for political parties, and that the parties must choose their candidates in primary elections, including for the most powerful elected office in the world. But primary elections are not the norm around the world—parties and their leadership choose their standard-bearers in many democracies.19

Indeed, our parties are unique in other ways that reflect our unusual understanding of popular sovereignty. Our parties have long been relatively “skeletal” organizations that do not require the regular payment of party dues, in contrast with political parties in most other countries, as well as most non-party organizations.20 To “join” a party in the United States is simply to check a box on a form or take a party ballot during a primary election. Patronage hiring and firing once played a role analogous to the role that membership dues in other countries play, but that, we have concluded, violates the First Amendment.21 In the absence of dues and the power of party leadership to choose the parties’ nominees, our parties have always been less tightly structured than those in European democracies. The discipline of party control is particularly firm in countries that use closed-list proportional representation electoral systems, in which voters can vote only for parties, not individual candidates. But weakened political parties do not empower “the people”; they empower the organized interests that are most able to take advantage of a system of political parties lacking sufficient organizational strength to countervail private forces. In at least twenty-three states we bypass formal institutional politics altogether through practices of direct democracy such as ballot initiatives, referenda, and recall tools that no other democracy uses to such an extent, especially since the revival of direct democracy in America that began in 1978 with the symbol of the “property tax revolt,” California’s Proposition 13.22

One of the best comparative accounts of the way in which the unique features of American democracy combine to affect both elections and governance remains Anthony King’s book, Running Scared: Why America’s Politicians Campaign Too Much and Govern Too Little.23 Using the concrete experiences of specific candidates and elected officials in the United States, Great Britain, and Canada, King identifies several features of the American democratic process that make American politicians “more vulnerable, more of the time, to the vicissitudes of electoral political politics than are the politicians of any other democratic country.”24 The unique features that combine to create this extreme vulnerability are the extremely short terms of office in the House; the use of primary elections in addition to general elections; the weakness of American political parties, which requires American candidates to be much more dependent on their own ability to raise money and get their message out; and the high costs of campaigns in the United States compared to those in several other democratic countries.25

The fact that American democracy exhibits these unique structures and features across so many different institutions in so many different domains is no accident. Underlying our institutions and practices is a singular democratic political culture that has always rested on a unique vision and understanding of the ideas of “popular sovereignty” and “self-government.” Indeed, I believe the very term “popular sovereignty” is invoked much more commonly in the United States than anywhere else. Put simply, I would say that American democratic culture has long had a distinctively individualistic way of understanding the “right” of self-government. This vision and the design of our political institutions have been mutually constitutive and reinforcing; as this unique understanding of popular sovereignty has led to institutional structures more subject to unusually direct popular control, the longstanding existence of these institutions has helped entrench and validate the cultural understandings. I will refer to the feature of American democratic culture embodied in the ideas and institutions that I have been describing as the “individualistic conception of democratic government.”

More specifically, our culture uniquely emphasizes—I would say, romanticizes—the role and purported power of individuals and direct “participation” in the dynamics and processes of “self”-government. This culture too often envisions an individualized form of political action, in which the key democratic elements are individual citizens, often pictured in splendid isolation, and a democratic politics that arises through spontaneous generation. This vision obscures the ways in which participation must be mobilized, organized, and aggregated to be effective; even worse, the pull of this vision often has led reformers and scholars to fail to appreciate the way in which “reforms” are likely to work in practice, given that the most effectively organized and mobilized actors will seize the advantage these reforms open up. As part of this romanticized picture of democracy, we uniquely distrust organized intermediate institutions standing between the citizen and government, such as political parties.

We can observe elements of this idealized image as far back as the Federalist Papers. Despite the brilliance and realist convictions of the Federalist Papers, these documents conceive of elections and government essentially in a kind of political vacuum. They offer no account of the critical role for intermediate political actors in mobilizing and organizing voters in elections (indeed, they conceived of elections as affairs of acclamation, not competitive political contests). Similarly, they do not provide an account of the need for organized, intermediary groups within elected government, such as caucuses and parties, to enable the concerted action necessary for government to function effectively.26 Like other eighteenth-century political thinkers, the Framers disdained political parties; recoiled when government soon divided into two distinct and warring Federalist and Republican camps; and viewed this division as a necessary temporary evil, not a permanent, legitimate feature of democracy.27 The worldview at the time of the Constitution’s framing encompassed citizens, elections, and government—but not the connective tissue of political parties, caucuses, and organizations that are so essential to organizing effective political power within the spheres of elections and governance. Of course, the eighteenth century’s vision of political representation was more elitist than ours, but its blindness to all of the critical intermediate organizations among citizens, elections, and government reflects a characteristically American way of thinking about democracy that has endured. We can see this in American foreign policy as well, in the naïve view that immediate elections will bestow legitimate and meaningful democracy on places emerging from non-democratic pasts, without regard to whether various underpinnings of democracy, such as a plurality of organized political groups competing for power, or a robust, independent press, have had a chance to develop.

The individualized conception of democratic government has pervasively shaped, and continues to shape, American democracy. We see this in institutional design, common critiques of democracy, and reformist efforts to “improve” American democracy. The conception is largely taken for granted, if recognized at all, let alone questioned. Since at least the Jacksonian era, the appeal to more “popular empowerment” or participation as the cure for political corruption has been a constant cultural and political theme in American democracy28—even as we struggle to correct for the dysfunctions that previous generations of reform in this direction have brought about.29 For example, in 1974 when Congress overturned the old seniority-based congressional committee system to dilute the power of committee chairs—at the time, conservative Southern Democrats—the result was the proliferation of committees and subcommittees. Yet some have argued that by undermining the power of committee chairs and diffusing power within Congress in this more “democratic” way, the net result has been to increase the power of private interest groups to block legislation by expanding further the number of veto points in the system, thereby diluting political power.30

Indeed, the central impulse behind many of our democratic reform efforts is not to criticize or challenge the individualist conception of democracy, but to insist on yet more “participation” and other ways of “empowering” individual citizens as the solution to our democratic disaffections. We require so many of our institutions to be chosen through elections, for example, on the view that “citizen” control will keep officials hewing closer to the common good, without any realistic assessment of how the electoral process actually works; with romanticized views of how much interest most citizens will take (or rather, fail to take) in voting for lower-level offices; and without regard for the degree to which organized private interests will be able to dominate in low turnout, low-salience elections. This approach is a longstanding one. For example, not only do we elect school boards in many parts of the United States, but Progressive Era policies urged that these (and other) local elections be held on a separate timetable from general elections, so that local decision making would be “more pure” and not entangled in broader political issues.31 Yet if turnout in school board elections is exceedingly low, it is even lower when these elections are held off-cycle; not surprisingly, the one interest that is always well represented in school board elections, no matter when they are held, is that of teachers, who have among the most direct stake in school board policies. Perhaps also not surprisingly, recent empirical work “is strikingly clear” in demonstrating that the lower the turnout in such elections, the more electoral and political influence teachers have—and the higher teacher salaries become as a result.32 Our culture seems to reel from one democratic dysfunction, to which the solution is more citizen empowerment, to another, in which we must face up to the perverse consequences of this prior solution, only to try yet another way to ensure more transparency and citizen control.

I want to push back a bit against that culture and the romantic vision of individualistic self-government animating it.

III. the causes of polarization

To begin to do that, I now turn to my analysis of why our political institutions have become so paralyzed in recent years.

It is well-known that our era of governance is constituted by what I have called “hyperpolarized political parties.”33 By all conventional measures, the parties in government are more polarized than at any time since the late nineteenth century.34 But keep in mind that partisan polarization is not necessarily bad, or all bad, from a broader democratic perspective. Political polarization, from my point of view, is a concern primarily insofar as it affects the capacity for governance. Others might be troubled with a political culture characterized by divisiveness, lack of civil disagreement, and the like, but my dominant concern is polarization’s consequences for effective governance. Indeed, polarization might well involve tragic conflicts between the domains of voting and governance, a much more general conflict in democratic practice than democratic theory has recognized. As responsible party government advocates have long argued, coherent and sharply differentiated political parties increase voter turnout, make the most salient cue in voting—the political party label—more meaningful, and through that cue enable voters to hold officeholders more meaningfully accountable.35 As a result, party polarization has distinct electoral benefits; it is not a matter of all cost and no benefit. We should therefore view partisan polarization as a significant problem only if and when its costs are substantial enough to outweigh these electoral benefits. Preventing government from taking effective action, even when broad agreement exists to the effect that government must act in some form, signals that the costs of polarization outweigh its benefits substantially enough to justify searching for measures that could mitigate these costs, including institutional design measures.

To understand what measures might be most effective—and to justify my argument that our search should move in a dramatically different direction than is typically suggested by those troubled by extreme partisan polarization—I need to begin by explaining the causes and suggested “cures” for our world of hyperpolarized political parties. What has caused the dramatic partisan polarization of our era? Polarization is not, in my view, a product of recent, or relatively contingent, forces or individual personalities.

I have argued that the hyperpolarization of today’s parties is overwhelmingly a product of long-term historical and structural forces.36 These forces were launched into motion with the Civil Rights Era of the 1960s, particularly the Voting Rights Act, as African Americans (and many poor whites) began the process of becoming full political participants.37 It is easy to forget that, from roughly the 1890s until the Civil Rights Era, the entire South was an artificially created one-party monopoly of the Democratic Party.38 The process of ending this unnatural political monopoly began in 1965, but the full effects of this change did not take place overnight; it took several decades of dynamic and mutually reinforcing processes for the Democratic Party in the South to move toward the left, for a robust and fully competitive Republican Party to rise, and for conservative whites to shift their party identification for Senate, House, state, and local elections to the Republican Party.39

Not until the 1990s, remarkably enough, do we see the kind of two-party political system in the South that the rest of the country had throughout the twentieth century.40 In my view, the racial redistricting regime of the Voting Rights Act (VRA) contributed to this process. The VRA took hold for the first time in the redistrictings of the 1990s as a result of the 1982 amendments to Section 2 of the VRA and, perhaps even more importantly, the Supreme Court’s 1986 Thornburg v. Gingles decision.41 The post-1990s redistricting regime shifted the political representation of the Democratic Party in the South towards its most liberal wing, dramatically reduced the number of officeholding moderate white Democrats in the South, and facilitated the rise of many more overwhelmingly conservative and Republican districts.42

Through this revolutionary set of historical changes, the two political parties, at both national and state levels, became “purified” into far more ideologically coherent entities. Voters now sort themselves into the two parties overwhelmingly, and correctly, by ideology, so that nearly all liberals are now Democrats, all conservatives now Republicans.43 This simply had not been the case for most of the past century.

If you accept my view on this, then it follows that the highly polarized partisan structure of our democratic politics should not be seen as aberrational. It should be understood as the “new normal.” Instead of being the product of contingent features of our present institutions or our present political moment, it is the result of deep and long-term historical processes. In other words, polarization should be accepted as a fact likely to be enduring for some time, not something that we can design away.

Nonetheless, a great deal of intellectual and reformist energy has been spent on the search for reformist solutions to extreme partisan polarization. This energy has been directed to restoring “the disappearing center” in American democracy.44 Given the recommended remedies for polarization that I describe below, it becomes necessary to explore briefly why certain solutions for polarization are likely to be unavailing and indeed, why such “fixes” might even be perverse, if the goal is to enable a more effective set of political institutions capable of overcoming current paralysis.

“Fixes” for polarization can be categorized into two forms. The first involves changes to the institutional structures of elections that will shift the mix of candidates and officeholders to empower a critical mass of more centrist officeholders who can bridge partisan divides. These institutional-design proposals include familiar ones that have been offered—open primaries; independent commissions to perform redistricting, perhaps with instructions to maximize competition; changes to internal legislative rules—and less familiar ones: eliminating laws banning “sore-loser” candidacies;45 moving to instant-runoff voting; or even more radically, abolishing primaries altogether and returning to a system of candidate selection by party leaders.

On the institutional front, the two fixes that have received the most attention are ending gerrymandering and opening up primary elections to a broader electorate than just party members. These changes might be desirable for many reasons, but in determining whether institutional-design changes in these areas are likely to make a meaningful contribution to reducing partisan polarization, we ought not be too sanguine about this prospect as more empirical evidence mounts.46 I continue to be more optimistic that changes to the structure of primary elections could make a difference, but there is little systematic empirical evidence to support this hope.47

The second category of reforms, on which I would like to focus more, seeks to reduce polarization in government by empowering “the people” more effectively. The idea is that greater citizen participation will be a solvent for political dysfunction and polarization. This idea is premised on the assumption that partisan polarization is not in us, but in our political parties; polarization in our formal politics is a corruption or distortion of the more moderate, centrist politics that we would have if only we could find ways to give “the people” more direct control or influence over elections and governance. The idea is part of a recurring wish or vision throughout American political history. But there are good reasons to distrust this idea and even to think that institutional efforts to reflect popular empowerment would make polarization worse, not better.

While earlier academic work suggested that “the public” was more centrist than those holding public office, more recent works reveal that polarization in government is not so obviously a distortion or corruption of the larger public’s less polarized views. Alan Abramowitz has shown that “politically engaged citizens” are just as polarized as the parties in government.48 Being “engaged” in this sense means little more than taking part in the most basic forms of democratic participation, such as: voting; trying to persuade a friend or neighbor to vote; displaying a bumper sticker or yard sign; giving money; or attending a campaign rally or meeting. Abramowitz’s findings therefore pose a serious challenge to the idea that more participation will translate into less polarization.49

Shanto Iyengar and his co-authors have found that partisans are far more uncomfortable today than in the past with their children marrying those who identify with the other party.50 And while citizens overall might not be as ideologically extreme as they are partisan, we are highly sorted along partisan terms today; 92% of Republicans are more conservative than the median Democrat, while 94% of Democrats are more liberal than the median Republican (twenty years ago, the figures were 64% and 70%, respectively).51 The percentage of those who are consistently liberal or conservative, rather than having a mix of such views, has doubled from 10% to 21% over the past two decades.52 As Marc Hetherington and others report, those who identify with one party express far more negative feelings about the other party than in the past; those of the opposite party to the President now largely report not trusting the government at all.53 A major recent study by the Pew Research Center finds that in 1994, only 17% of Republicans and 16% of Democrats had “very unfavorable” views of the opposite party, while today 43% of Republicans and 38% of Democrats hold such views.54 Other social scientists suggest that the public is even more extreme in its policy views than those in office or, at the least, that those whose views are categorized as “moderate” are actually ideologically polarized too.55 In addition, citizens, activists, and elected officeholders now see more issues in one-dimensional, partisan terms. As Carsey and Layman find: “The data are clear: across all three major domestic issue areas—social welfare, race, and culture—there has been a steady increase in the gap between Democratic and Republican citizens, elected officials and activists.”56 In state politics, we see a pattern similar to that in Congress. On average, state legislatures are becoming significantly more polarized.57

If political engagement correlates with increased polarization, as Abramowitz documents, then we should be skeptical about whether finding ways to increase popular participation will temper polarization. In addition, participation does not sprout up spontaneously, like mushrooms after a rain. Participation has to be energized, organized, mobilized, and channeled in effective directions—all of which requires the very organizations, and the partisans, that “citizen” participation is meant to bypass. Moreover, political engagement might not just involve individuals who self-select for partisanship, but might itself be an experience that generates polarization. Furthermore, despite all the cynicism about politics today, “Americans [now] are more interested in politics, better informed about public affairs, and more politically active than at any time during the past half century.”58 More and more of us are engaged in the ways that idealized democratic citizens are thought to be. And we are partisans. Cause and effect are difficult to disentangle here. But do you know many politically engaged people who are not partisans, outside of groups like the League of Women Voters, whose membership has dropped nearly in half since 1969, according to Putnam?59 Extremism in the name of moderation is no vice (that is certainly my own temperament), but it doesn’t raise a lot of money or draw a lot of volunteers. We should be wary of romanticizing a more engaged public as a vehicle that will save us from hyperpolarized partisan government.

Appealing to more “participation” as a cure for polarization thus reduces to a strange kind of hope that when the politically non-engaged become more engaged, they will not behave like those who are already politically engaged. They will pass untouched through the maw of the machinery of democracy but remain the same politically uninformed innocents as when they started. But their participation will have to be mobilized, organized, directed, and at least modestly informed. Will this not make them act in the same way as citizens who are already engaged?

Let me make this point concrete by turning to the specific, crucial issue of campaign financing. I show how certain proposals that focus on empowering more citizen participation are likely to have the unintended consequence of hindering effective governance.

I will state my preference at the outset: I favor a system of public financing, but not the kind of public financing centered on individual candidates that exists in the United States (in the few places we have it).60 Instead, I want to suggest a system of public financing in which more of the emphasis, and more of the flow of money, is oriented toward the political parties rather than individual candidates. I will return to this proposal shortly.

But to stay on the theme of empowering greater citizen participation, some proponents of public financing have suggested that campaign financing work not through the state, as in public financing around the world, but rather through individual vouchers provided to all of us.61 This is a distinctively American proposal, for it reflects, I believe, the peculiar and radically individualistic culture of American democracy, along with our characteristic distrust of more organized forms of political power.

Yet it turns out that individual donors are more ideologically extreme and more polarized than non-donors—as we’ve just discussed, the politically engaged are more polarized than the general public.62 Indeed, those who donate are more ideological even than “active partisans,” defined as those who identify with a political party and engage in more political activities than the mere act of voting.63 Even more to the point, individual campaign donors are also more ideologically extreme than most other donorsas well, such as PACs and the political parties.64 PACs tend to focus on moderate candidates, as well as incumbents; individual donors focus on more ideologically polarized candidates. In general, groups that give for access-oriented reasons tend to finance moderates and incumbents, while ideological donors favor challengers and more extreme candidates.65 Put another way, the most ideologically extreme money to campaigns comes from individual donors. Moreover, recent work concludes that the voting patterns of senators most closely track the policy preferences of their individual donors, rather than those of voters in the state or even co-partisans in the state—and that this pushes senators to the ideological poles.66 Democratic senators are more liberal, Republicans more conservative, than their voters, but these politicians are reflective of the views of their individual donor bases.

Furthermore, candidate campaigns have become dramatically more dependent on individual donors in recent decades than on all other sources combined, such as political parties and PACs, even as our candidates and parties have become more and more polarized. In other words, as our campaign finance system has become more democratized, our politics has become more polarized. In 1990, individual contributions to campaigns provided about 25% of a campaign’s money, and PACs provided about half; today, individuals are by far the largest source of direct money to campaigns (about 61% for Congress) and PAC contributions constitute less than 25%.67

Here is another fact to keep in mind in seeking to understand individuals, polarization, and money: a majority of individual contributions now come from out-of-state donors.68 Also not surprising is that out-of-state donors are the most ideologically extreme of all contributors. Consider the kind of individuals likely to give out-of-state money to the campaigns of Elizabeth Warren and Ted Cruz, as opposed to the more moderate senators or challengers about whom most out-of-staters probably know little to nothing in the first place.69 Are many individual voters around the country likely to send their money to Missouri for Claire McCaskill or to Tennessee for Lamar Alexander? Democratizing campaign contributions through vouchers might well, ironically, fuel the flames of political polarization, as compared to public financing systems funded in the more traditional way, through general revenues.

Voucher proponents might believe that the polarizing effects of individual donations will disappear once “all the people” are empowered to donate through vouchers. But this neglects the collective-action dynamics that influence all political activity. People have to become both motivated and engaged enough to choose to donate and to seek out information relevant to informed donations—just as they must to vote—and informing and motivating potential donors will take political organization and mobilization. Those who are most informed and motivated are likely to be partisans, and thus the groups most equipped to take advantage of these new political openings—as with other such openings—are also likely to be more partisan.

I say all this not to pick on voucher proposals in particular but to illustrate a larger point. Unless we attend to the ways in which political power is actually mobilized, organized, exercised, and marshaled, then policy proposals based on an individualistically driven vision of politics, or on non-grounded abstract democratic ideals such as “participation” or “equality,” can perversely contribute to undermining our institutional capacity to govern. If we want to adopt public financing in ways least likely to fuel partisan polarization, then more traditional forms of public financing through general revenues, rather than those based on individual donations, might be more appropriate.

Let me turn now to a different view: an institutionally and organizationally centered approach to the relationship between elections, governance, and effective political power.

IV. political power, political parties

In thinking about how to enable effective democratic action through our political institutions, we should focus less on individual citizens and turn instead to the current or possible organizational entities that have the most powerful incentives to aggregate the broadest array of interests into democratic politics—and to force compromise, negotiation, and accommodation between those interests. Organizational power inevitably exists in democracies; it cannot be wished away, and it is in fact crucial in order for democracy to be able to work at all. Of the various organizational entities that exist or that I can envision, the political parties, driven by the need to appeal to the widest electorate, remain the broadest aggregators of diverse interests.

This proposition might sound ironic, in light of how polarized the parties have become. But the electoral incentive means that it remains true. The overpowering need to put together coalitions broad enough to control one, two, or three of our national political institutions remains the single strongest unifying force capable of bringing together broad arrays of interests into two large coalitions—and, in doing so, inevitably forcing compromise among those interests. When African-American voters in the South were permitted to vote for three decades or so after the Civil War, there were effective office-holding interracial political coalitions, despite the era’s cultural attitudes about race.70 Electoral incentives and the desire to wield the tools of political power provide powerful motivations to compromise between groups in the pursuit of winning coalitions. In first-past-the-post election systems, the two dominant political parties serve as the principal vehicle for these types of compromise.

I see no other candidate on the horizon. Recently, proposals have emerged to form multi-candidate PACs that would raise and donate to “moderate” candidates.71 If such entities get off the ground, I am skeptical about how effective they will be. But of course, the multi-candidate organization par excellence that already exists is the political party. Indeed, the candidate contributions made by political party organizations tend to empower the forces in the center of the party; parties donate twice as much to candidates in the middle of the ideological spectrum as to those at the extremes.72 Party-based contributions to campaigns are a force for moderation compared to individual contributions.

However, this is where the problem of political fragmentation becomes acute. Parties, like all organizations, are complex entities composed of many, sometimes conflicting, components, including elected officials, organizational leaders, party voters, factions, and so on. Among these forces, it is the elected party leaders who have the strongest incentives to internalize national electoral incentives toward broad coalitions. The success of party leaders depends to a significant extent on making the party brand as broadly appealing as possible.

There, I suggest, lies the problem. Political fragmentation has drained partisan elected leaders of much of the power to control, unify, and discipline members of their own party. By “fragmentation,” I mean both the diffusion of the power in elections away from the formal campaigns and the political parties—and even more importantly, the diffusion of power in government away from the leadership of the major political parties to their more extreme factions. While some have characterized the parties today as “networks,”73 I believe “politically fragmented” better captures the structure.

Over recent election cycles, we have become well aware of the fragmentation reflected in the explosion of Super PACs, 527s, and 501(c) organizations that seek to influence elections and policy. Many of these organizations have much narrower ideological and policy interests than the parties as a whole. At the same time, party leaders also have less capacity to force party members to toe the party line. Members of the House and Senate are much better able to function as independent entrepreneurs and free agents. As Moisés Naím has documented across a wide array of public and private organizations, organizational “power” is breaking down in general.74

A specific representation of this phenomenon is the unprecedented power that senators in their first year in power have in relation to their party leaders and consequently over our politics. Republican Party leaders may have understood that shutting down the government and threatening to default would be destructive to the party’s interests (they did not permit the same mistake to be made twice). But today, they find it difficult to stop one or a few individual senators, or a minority faction, from doing just that. It is impossible to imagine even as powerful a figure as Lyndon Johnson playing the kind of role in his first years in the Senate that Ted Cruz has been able to play.

To those who think that this kind of political fragmentation is a problem only on one side, I disagree.75 The same structural forces are at work and the same kind of fragmentation lies latent in the Democratic Party; these divisions will become apparent under the right set of circumstances. The forces of economic populism—centered most directly for now on Elizabeth Warren, who already allegedly wielded an effective veto over her own President’s pick to chair the Federal Reserve76—do battle with the more centrist, establishment forces within the Democratic Party. For now, the presence of a sitting Democratic president exerts a sufficient unifying force to suppress these conflicts, but once this presence is removed, we may well see more overt political fragmentation within the Democratic Party.

If you accept my conclusion that intense polarization of the parties in government is likely to be an enduring fact for the foreseeable future, the question must then shift to the following issue: from where are sources of compromise and negotiation, deal-making, pragmatism, and the like most likely to emerge in such an overall polarized structure?77 Polarization and divided government make capacities and attitudes related to compromise more necessary—and, of course, more difficult. In my view, elected party leaders are the most likely sources of the kind of political compromise and pragmatism necessary to reverse the decline of American government.

In part, this is a numbers problem: negotiations between three to five leadership figures are easier to conduct than hydra-headed negotiations in which new factions or individuals pop up. In part this is because the trust in negotiations that is essential to deal-making is established by repeat players in ongoing relationships of regular deal-making. Second, my focus on elected party leaders stems from an empirical belief, reflected in academic studies, that party leaders in Congress tend to be ideological “middlemen” of their parties.78 They have stronger incentives to forge compromises both because their election requires appeal to broad constituencies within their parties and because they bear more personal responsibility and blame for the failure of “their” institution to function effectively.79

But party leaders can play this role only if they have the tools and the leverage to bring along their caucuses in the direction that these leaders believe best positions the party as a whole. Finally, then, let me explain why they have lost that leverage.

The problem is not that individual leaders are now “weak.” Part of the American tendency to individualize politics is to focus on personalities as the cause of political action or inaction. Personalities matter, but so does structure. Both George W. Bush and Barack Obama campaigned, and tried to govern for an initial period, in ways that reached across the aisle; both discovered that the larger structure of hyperpolarized parties in Congress made this aim exceedingly difficult.80 Furthermore, broader structural changes, including legal ones, have disarmed party leaders of the tools they previously had used to unify their members around deals that were thought to be in the best interest of the party as a whole. What institutional and structural changes might recapture some of the crucial capacities that enable effective partisan leadership and thereby also enable effective governance?

V. structural decline in the power of party leaders

Party leaders once had their greatest leverage over their members through the power of committee assignments. These assignments were valuable because they were the means to work on substantive issues a member cared about, ways to raise the member’s profile and stature, and ways to raise money for subsequent elections.

But two major changes have made committee assignments less meaningful when it comes to the ability to raise funds and enhance one’s public status and visibility, at least for those politicians who see themselves as upwardly mobile (that is, most of them): the communications revolution and the current system of campaign financing. Politically ambitious senators more and more now view the Senate as a quick pass over to a presidential campaign, particularly in light of President Obama’s success. Indeed, staying in the Senate for more than a brief period may be considered a liability to pursuit of higher office, since longer tenure means more need to take positions on divisive issues that will inevitably alienate some potential constituencies.

The first change is a cultural one we all recognize: the revolution in communications and information technology. According to Moisés Naím’s account, this revolution is the major force spawning the general unraveling of organizational authority and effectiveness across public and private sectors.81 Individual officeholders now have the capacity to reach large, intensely motivated audiences of potential voters and donors in ways that were simply not possible before; they are able to build a personal brand apart from the party label. Does Senator Ted Cruz, for example, spend more time on Twitter and television, including cable television, as well as televised speeches on the floor in the post-CSPAN era, than he does meeting with Republican Party leaders? What could Lyndon Johnson have done that would have been comparable? Party leaders do not control and cannot shut down these new channels of access to direct communication with voters and donors. At the same time that these channels enable individual officeholders to reach out, they also enable more widespread populist influence to reach in and factional interests within parties can be more easily mobilized. Of course, there is no way to unwind this communications revolution.

The second force behind the reduced leverage party leaders have over their members involves legal changes. Here I will focus only on the way we have changed election financing starting in the 1970s. We adopted the most aggressive regulatory structure in American history for controlling money in national elections in the early 1970s in the aftermath of Watergate. The system we created was a candidate-centered system of financing, in contrast to the party-centered systems used in much of Europe.82 The 1974 Federal Election Campaign Act Amendments83 imposed contribution caps and spending limits on campaigns in general, and they also treated political parties similarly to corporate and union PACs: party committees could give no more than $5,000 to candidates. In addition, individual contributions to the parties were capped at $25,000 a year.84 When the Supreme Court struck down the spending limits in this law in Buckley v. Valeo,85 the Court gave virtually no independent consideration to the Act’s regulation of political parties—either the restrictions on party donations to candidates or the caps on individual contributions to parties.

Then, in the early 2000s, we added on the second big change to our system of election financing: the McCain-Feingold campaign finance “reforms.” Before that moment, the political parties raised nearly half their money in what was called “soft money.”86 Without delving deeply into details, soft money entailed contributions to the political parties that were not subject to the caps in federal campaign-finance law.87 This soft money, which was used for party-building actions (including television ads, positive and negative, concerning specific candidates), was fully disclosed and transparent when Congress eliminated it. Some of these contributions were in huge amounts; around half of it came from individuals, the rest from corporations and unions.88

From the perspective of reformers, soft money was corrupting. The “purist” solution in McCain-Feingold was to ban the parties from receiving any soft money at all. From that point on, all money given to the parties became subject to contribution caps. The fact that Congress was willing to cut off the flow of soft money to the parties was itself a signal of the candidate-centered nature of our financing system and the reduced dependence of candidates (especially incumbents) on the parties for their electoral success.89 But the practical result now seems to have been to diminish the already-weakened political parties as a force in elections and to create incentives for this party “soft money” to flow to independent groups. Even “the ground game” in elections, the quintessential party electoral activity, is increasingly funded outside the parties.90

At the same moment that legislators became able to brand themselves and raise money independently of the parties, the parties were dramatically disempowered relative to other groups. Even worse, the “reforms” to campaign financing actively encouraged money to flow outside the parties to organizations that supported narrower, more sectarian causes. That is why first-year senators can now wield as much power within and over their parties as much more senior senators, including the party leadership.

VI. a party-based campaign finance system

Since we cannot undo the communications revolution, I want to suggest three proposals for legal change aimed at giving the political parties—and just as importantly, their elected leadership—more influence in elections and hence over how their members govern. Legal changes might not be necessary to re-empower party leadership; it is possible that organic processes, driven by national electoral incentives, will do so.91 But if legal changes turn out to be needed, I offer these three initial suggestions. The first two changes are modest, the third more dramatic. The specifics of any of these proposals are less important (and not developed here in any detail) than the general conceptual re-orientation I have in mind: to empower the political parties and their leadership so that the aggregative forces in democracy have as powerful a role as possible, and as much leverage as possible, in the democratic process.

These proposals focus on shifting our campaign finance policies in ways that would give a greater role to the political parties. As Ray LaRaja and Brian Schaffner have recently demonstrated, states that have more “party-centered” campaign finance laws tend to have less polarized legislatures than those that impose significant constraints on the amounts and means through which the political parties can support candidates.92 The mechanism through which this occurs, they conclude, is that the parties tend to use their financial resources to support moderate candidates more than other sources of campaign money; consistent with the incentive structures I described above, the elected officials who control party organizations internalize winning elections over ideological purity. The empirical evidence shows that parties, more than issue groups and other political committees, tend to concentrate their money on moderates and not on ideologues. As a result, LaRaja and Schaffner argue, states that give more freedom to political parties in the campaign-finance system end up with less polarized legislatures.93

The federal campaign finance system imposes caps on how much the political parties can directly contribute to the campaigns of their candidates.94 But the campaign finance rules also treat an actor’s campaign spending that is coordinated with a candidate’s campaign as equivalent to a direct contribution to that campaign. Moreover, while the rules are more generous in dollar terms for the parties,95 these rules similarly treat coordinated party spending beyond those amounts as prohibited contributions to the candidate’s campaign. Thus, parties can engage in only limited coordinated spending with their candidates.96 The fear is that party spending in coordination with its candidate would be a conduit for individuals to circumvent the contribution caps that exist on direct donations to the campaign and that bans on earmarking party contributions for specific candidates are not sufficient to address this concern.

The effect (and intent) is to use campaign finance law to try to build more of a wall between the political parties and their candidates when it comes to spending money on elections. In fact, the Federal Election Commission wanted to go even further and treat any money a party spent to support a candidate as a direct contribution to the candidate, which would mean this money would be subject to contribution caps. The Supreme Court put a stop to this effort by holding that political party spending that is independent is just as protected under the First Amendment as independent spending by any other entity.97 But we still live with the remaining constraints, which the Court endorsed, that impose limits on the ability of political parties to coordinate election spending with their candidates.98 Indeed, the Court has rejected any view that it should apply stricter scrutiny to limits on coordinated party spending than to that of any other entity.99

My first proposal, therefore—and it may sound startling—is to permit parties to work more directly together with their candidates and coordinate the party’s spending with campaigns. Contributors should continue not to be able to earmark contributions for specific candidates,100 and one can raise concerns about how effective those bans on earmarking might be, but the question of potential corruption should be seen in comparative terms: in a world in which individuals can contribute unlimited amounts to issue-advocacy Super PACs, including Super PACs dedicated to one specific candidate or issue,101 are we better off sharply limiting contributions to parties or their ability to engage in coordinated spending with candidates? Parties, after all, are constituted by numerous interests and many donors, including large donors; parties dilute the role of money by pooling so many interests and donors. This dilution is far from complete, of course, but again, it is probably better than the alternatives. Instead of treating cooperation between parties and candidates as potential vehicles through which individuals can corrupt candidates, we should recognize that, on balance, party coordinated spending at least has the virtue of linking parties and candidates more effectively. This link would help revive a more central role for the parties’ national campaign committees in their candidates’ success, and in turn give those in control of the parties more leverage over candidates.

My second modest proposal is to raise significantly the amounts of money that can be donated to political parties for election purposes. It is important to recognize, but complex to unravel, how the McCain-Feingold law’s ban on soft-money contributions affected the overall money available to parties and hence the role of the parties, in relation to other entities, in the democratic process. Let me offer just two quick facts to illustrate how this law’s ban on soft money immediately has affected election financing. McCain-Feingold, at least as much or more than Citizens United¸ accounts for the role of non-party entities in the way our elections are run today.

In the first election after the law was enacted, in 2004, the political parties appeared to be able effectively to replace the soft money they had lost through increased, successful efforts to raise more money from more individuals, in part because McCain-Feingold also raised the amount of money individuals could donate to the parties.102 But the law soon also encouraged a dramatic rise in spending by groups outside the party structure. From 2002, when the Act was adopted, until 2008—well before Citizens United was decided in January 2010—independent spending by non-party entities exploded, growing around 1122% in those 6 years (or 555% from 2000, the presidential election before the Act).103 In the 2012 elections, non-party spending grew only 207% from the 2008 election, even though the 2012 election was highly competitive to the very end.104 The money that had been going to the parties, and no longer could, simply flowed now to direct, independent spending by those who had formerly given to the parties. At the same time, spending by the political parties now does appear to have taken a significant hit.105

Keep this in mind the next time you hear Citizens United castigated as “the root of all evil” concerning money in politics. This view is wrong, for too many reasons to go into here, and Citizens United has become a too-convenient whipping post for those concerned about an excessive role for money in American elections.106 In fact, Citizens United has played a minor role in the recent explosion of non-party money, partly because the logic of Buckley itself made it inevitable that the First Amendment would prohibit caps on contributions to non-party entities that engaged only in independent election spending.107 Reforms like the McCain-Feingold soft-money ban created at least as strong an incentive for the rise of the centripetal and fragmenting forces in democracy as has Citizens United. In a world in which the potential pool of money to influence election outcomes becomes effectively unlimited, because Buckley v. Valeo makes any kind of spending limitations unconstitutional, restrictions on the flow of money to candidates and campaigns will inevitably lead that money to flow through other channels, as it did in floods well before Citizens United.108 For those who remain concerned about the flow of money into the political parties, we can debate appropriate limits on amounts and sources. But if we have limits on the amounts that can be donated to parties from appropriate sources, then those limits should be set at high levels to encourage a more effective role for parties in elections and hence in governance.

Both raising the caps on donations to parties and on party spending coordinated with its candidates do raise concerns that donors would be able to use the parties to corrupt those in office by making elected officials dependent upon large contributions to the parties—contributions that are then passed through to specific candidates who are aware of the ultimate source of the party’s spending or contributions. My final, more extreme proposal therefore takes to its natural conclusion the underlying idea of structuring the campaign finance system to support a larger role for the political parties, in a way that addresses this corruption concern. We could consider a shift to publicly financed elections, but with the important twistthat they be financed significantly through the political parties, rather than having the individual candidates be the exclusive or overwhelming recipient of the funds. In the limited experiences of public financing in the states, the money overwhelmingly flows through the candidates, not the parties—reflecting the typical individualist-based American conception of democracy.109

Public financing through the parties would most directly accomplish the aim of putting greater leverage into the hands of party leaders. It would also, perhaps not coincidentally, bring our system of financing elections closest to the most common system used in other well-established democracies.110 Of course, if we were to encourage or require money to flow primarily through the major political party organizations, it would become all the more important to focus on how the leadership of party organizations gets constituted—and, given the much greater significance that party organizations would have, the ways in which these party leaders are chosen would inevitably change (particularly for the party that does not control the presidency). Based on the views I have outlined here, perhaps the ongoing party organizations, the Democratic National Convention and Republican National Convention, would need to be put under the control of the elected party leadership in government, at least for determinations of how to use campaign funds. Elected national leaders of the party (a multi-member small group of such leaders might be appropriate) remain the actors most likely to internalize the incentives to make the party appealing to the widest constituency. But fixing the details is less important than generating discussion about this general direction for public financing schemes.

It is possible, of course, that organic developments might move political dynamics in this direction without formal policy change. The 2014 midterm election cycle, for example, has seen the emergence for Senate races of an exceptionally well-funded Super PAC on the Democratic side, The Senate Majority PAC, which is funded by individual billionaires making large contributions (the largest to date being $5.0 million) and labor unions.111 Much like the broad electoral aims of a political party, the purpose of this Super PAC is to preserve the Democratic majority in the Senate.112 The party has therefore spent large sums to attempt to preserve the seats of vulnerable but competitive incumbent Senators, which in turn means the most centrist Senators in the Democratic Party, who hail from purple or red states, such as Senator Pryor of Arkansas, Senator Hagan of North Carolina, and Senator Begich of Alaska.113

This PAC behaves much like a political party in the sense that its aim is not to support ideological purists, but in more pragmatic, electorally oriented terms, to support the party’s vulnerable candidates, regardless of specific ideology. Given the theoretical and empirical account that I provided earlier, it should come as no surprise that the organizers and leaders of this PAC have strong professional connections to Senate Majority Leader Harry M. Reid,114 or that President Obama has spoken at two of the PAC’s major fundraisers.115 Although Senator Reid does not formally control this PAC, one might comfortably speculate that if this PAC (as the biggest outside spender on the Democratic side) succeeds in enabling the election of Democratic Senators, Senator Reid might well end up with greater capacity to “persuade” those Senators to follow the leadership’s positions. If similar PACs closely tied to the leadership emerge on both sides of the aisle in both houses of Congress, party leaders might well end up with greater leverage over rank-and-file members dependent on this source of funding.

These large contributions to independent-spending “non-party” entities are flowing into this surrogate for the Democratic Party, rather than into organs of the party organization itself (such as the Democratic Senate Campaign Committee), because current law does not permit political parties to accept unlimited contributions that will be used only to engage in independent spending. But current litigation is challenging this restriction on First Amendment grounds, based on the argument that political parties, like other entities, should be able to accept unlimited contributions if they will be used only for the party’s constitutionally protected right to engage in independent spending.116 There is little doubt that if Democratic Party organizations could accept these unlimited, independent-expenditure contributions, then an entity like the Senate Majority PAC would disappear overnight, and all or nearly all of that money would flow to the appropriate party organization instead (more ideologically oriented Super PACs would continue to exist). Moreover, if this litigation succeeds, it would significantly reduce the importance of the issues I have raised about limits on party coordinated expenditures or caps on donations to the parties for money that will be used for contributions to campaigns, not independent spending. If the political parties were constitutionally entitled to receive unlimited contributions dedicated for use only for independent party spending on behalf of candidates, then we would likely see a significant reversal of the flow of money from Super PACs to the parties.

The Supreme Court’s recent decision in the McCutcheon case might already provide a gentle nudge in this direction.117 That decision left intact contribution caps on the amount an individual could donate to any particular candidate or to a political party, but invalidated caps on the total amount of money an individual (staying within these limits) could give to a group of candidates or party organizations. Before the decision, an individual could give no more than $123,200 in total to candidates and party organizations.118 The decision has triggered the formation of more joint fundraising committees; these enable a group of candidates to raise money collectively and accept a single check, which is then divided up legally among the candidates, and enable political party organizations, such as a national party organization and a number of state party organizations, to do the same thing.119 Despite claims about the additional torrent of money that McCutcheon would release through these vehicles, it remains unclear at this stage how much money these joint fundraising committees will be able to raise.120 But if party-based joint fundraising committees do turn out to be of considerable practical significance, then the effect of McCutcheon would likely be to cause more money to flow to the political parties, rather than to non-party organizations that had never been subject to similar aggregate contribution caps. If this occurs, McCutcheon would turn out to be the first Supreme Court decision in this entire field with the practical effect of creating incentives for money to flow to the parties rather than to non-party groups. Whatever else might be said about the decision, it would therefore encourage, to some modest extent, the forces inducing centralization of financing through the parties—emphasized in this Feature.121

To be sure, there is still reason to be concerned about the role in American elections of extremely large contributions or spending from single individuals or entities. But we only began in the 1970s to attempt to regulate the role of money in national elections in a comprehensive way.122 Moreover, as almost half a century of effort using that approach has shown, it is extremely difficult to limit the amount of money that flows into elections, as long as we continue to have a privately financed system and an understanding of the First Amendment that precludes limitations on election spending—a First Amendment constraint attributable to Buckley v. Valeo,123 not to any more recent decision. In light of this reality, the best policy we can achieve is probably to create incentives to encourage this money to be channeled in one direction rather than another. We should use these incentives to channel that money to flow through the political parties to a much greater degree than is currently the case.

VII. making deal-making possible

I have focused on campaign finance laws merely as one point of entry into my larger theme: the need to reinvigorate party leaders’ capacity to play a unifying leadership role. If we turn reform efforts in this direction, instead of the paths more often advocated, then other suggestions might start springing to mind.

For example, effective governance inevitably requires negotiation, particularly in our separated-powers system. But little in academic work on democracy, or even popular accounts of democracy, even addresses issues related to negotiation, such as the institutional environments or structural conditions that enable effective negotiations among political leaders. How could law and policy facilitate these structures and conditions?

Part of our romanticization of democracy has been reflected in an extreme emphasis on greater transparency as a solution to our democratic anxieties. In our culture, it is difficult to defend the need for secrecy in negotiations. But compare the environment in which successful international negotiations still work today (at least before WikiLeaks) to the ways in which our laws and culture of transparency have transformed the environment in which domestic negotiations over policy take place.

After the 1976 Government in the Sunshine Act required that congressional committee meetings to be public, surveys of senators soon concluded that these open meeting requirements were the largest single cause of a decline in the ability to negotiate and to make politically difficult tradeoffs.124 Today, we have the unfortunate Federal Advisory Committee Act,125 which extends these open-meeting requirements even to bodies that only provide advice to the federal government and ties these advisory groups in knots for little meaningful public benefit. If negotiations among leaders are a key to effective governance, particularly in polarized times, then we need a less moralistic, more realistic sense of the conditions under which negotiations effectively take place.

One structural condition for productive negotiation in theory and practice is likely to be the presence of long-term players who will interact over multiple negotiations. One-shot bargaining games are notoriously more prone to strategic withholding and manipulation of information, since there is no threat of future sanction in subsequent negotiations. In the political realm, this suggests that whatever the downsides to long-term incumbencies, one advantage that longer-serving members of Congress are likely to have is greater informational knowledge about what the other side values most and what it can afford to trade; which threats are realistic and which are bluffs; and the ability to trade off issues across time and policy spaces—an ability that can enable productive compromise.126 Once again, the presence of long-term players is a structural condition that tends to favor the role of party leaders, since they tend to have served for long tenures, along with other more senior members.

This is yet another reason why changes that empower recent arrivals to Congress might also make it more difficult to forge deals across partisan lines. To be sure, safe homogenous seats can yield long series of terms of office for candidates who can afford to appeal to more extreme poles of the spectrum without an electoral cost; once again, we face tradeoffs between democratic values. But the importance of repeat players to effective democratic negotiation in legislative bodies brings out a downside to another romantic but counterproductive populist “reform” of the democratic process: the movement for term-limits (especially short ones) for state legislators, in an effort to elect more “citizen legislators.”127 Yet term limits seem not to have any effect on the composition of those elected to office.128 The term limit effort was also designed to make legislators more accountable to the public. However, the effect of term limits is to “weaken the legislative branch relative to the executive;”129 to empower legislative staff, who can invest in long-term development of policymaking expertise; to boost interest groups, upon whom less experienced legislators become more dependent for information; and to force legislators to adopt shorter time horizons that are in tension with the longer-term, repeat interactions that make for effective political negotiation and problem solving.130

A second structural condition for effective negotiation across political divisions is, as suggested above, the ability for certain stages of the discussion and negotiation process to take place outside the public eye. Indeed, contrary to the popular emphasis on the pervasive importance of full transparency, studies of this issue now cause leading social science reports to issue such strong statements as that “the empirical evidence on the deliberative benefits of closed-door interactions seems incontrovertible.”131 Perhaps the reasons this is so are obvious, but they are nonetheless worth stating briefly, given the far greater emphasis the “democratic” benefits of transparency have received in recent decades. When the audience for a negotiation is public, the parties are encouraged to posture for their own constituents and, sometimes, to stand for principle by refusing to compromise. When negotiations take place in less public arenas, parties typically feel free to take greater risks in revealing their positions, the issues on which they have the most intense preferences, the issues on which they can give, and the benefits they must attain in return for any compromises they make. Similarly, negotiations work as part of packages of tradeoffs, but disclosing any one potential compromise in isolation, before the entire package of countervailing compromises has been agreed upon, can easily scuttle any potential deal. That is why, of course, one effective tactic for undermining negotiations is to leak the details of one potential dimension of compromise before the full range of provisions has been settled. Open negotiations can themselves foster polarization, which is why peace negotiations are frequently carried out in secret.132

The demand for greater transparency has been driven, of course, by genuine democratic concerns, including concerns regarding corrupt deals—ones that do not adequately take into account the full range of appropriate interests—or concerns that important affected interests will not be heard. One way to reframe the costs and benefits of transparency to democracy might therefore be to focus less on demanding full transparency of processes and more on asking for transparency of the reasons and purposes that explain and justify outcomes.133 The adoption of the Constitution provides one example: while the negotiations at the Constitutional Convention were held in secret, there was a robust, public ratification debate in which the justifications for various provisions, and the arguments against, were extensively tested in a prolonged open process.134 Of course, any decision to permit greater space for private democratic negotiation would itself be a decision that in most contexts would have to be made publicly and be publicly justified—though there are some contexts in which even the fact that negotiation is taking place might have to remain secret initially to have any chance of success.

Additionally, insulating the processes of negotiation from constant public monitoring to a greater degree would require policymakers to generate trust in the negotiating process itself. This might in turn necessitate public disclosure of the participants in the process, in order to ensure all relevant interests are represented, but without necessarily disclosing the detailed, step-by-step substantive proposals within the negotiation process itself. There are many controversial issues at stake, to be sure, in advocating greater space for less public policymaking negotiating spaces. However, we need to begin to take seriously the reality that full transparency can be in considerable tension with the prospects for productive negotiations and hence effective democratic governance in polarized times.

A third implication of moving away from the romanticized model of democratic governance is as follows. In a few short years we have learned that ending earmarks has eliminated one of the most direct benefits that party leadership could bestow upon recalcitrant members to generate their support on major legislation.135 A de-romanticized and less purist view of democracy might also have to accept that certain kinds of public side-payments—logrolling is itself an example, of course—are necessary to enable the compromise and negotiation required for government to function more rather than less effectively. Successful negotiation takes advantage of differential intensities of preferences; members of Congress who are moderately opposed or indifferent to legislation can have strong preferences for the concentrated benefits that public projects in their state or districts offer. Bans on bringing certain dimensions of policy into the negotiation dynamic can make tradeoffs and productive compromises more difficult.

VIII. less romanticized visions of democracy

Let me return to where I began. For many years now, private law scholarship has focused on the consequences of its rules for the dynamics of private power in contexts like market settings. My aim, and what I view as the aim for “the law of democracy,” is to do the same for public law, in the context of democratic elections and governance.

This approach recommends that we think in terms of measures that would encourage the forces of centralized authority within the political parties and discourage the forces of political fragmentation. Stronger parties are likely to remain the most effective vehicle for enabling the compromises and deals that are necessary in the face of what will be the ongoing polarization of the parties in government. Put another way, the problem is not that we have parliamentary-like parties; it is, I suggest, that our political parties are not parliamentary-like enough.

The obstacles to any changes along these lines will not merely be entrenched interests. In overcoming these obstacles, it will be just as necessary and important to confront head-on two powerful cultural trends that will generate resistance to publicly financed elections through the parties and other measures that aim at re-empowering political leaders.

The first is America’s exceptional and distinct ideology of “popular participation.” Any change in the democratic system that aims to empower political leaders will be cast in terms of Manichean conflict between “the people” and “the elites.” America’s cultural self-understanding of democracy has always invoked a rhetoric of “popular sovereignty” that is far more populist in meaning than in other Western democracies. But it is increasingly becoming clear, in our era at least, that the much greater participation enabled by the communication revolution breeds polarization as well as fragmentation. Instead of viewing a relentless expansion of participatory reforms as the cure for what ails democracy, we should start recognizing a perhaps-tragic tradeoff between the desire to make government more accountable, through measures like enhanced popular participation, and the capacity for government to function effectively. In the past, for example, I have supported matching private-public election financing systems, such as the system used in New York City and now being adopted elsewhere. But for the reasons I discussed earlier, I have become wary that these systems will only exacerbate polarization and fragmentation. Indeed, one recent study has found that in “clean-money” public financing systems, such as systems that match public dollars to private contributions, candidates’ positions move farther away from the ideological center of public policy preferences once the clean money system has been adopted.136 The mechanism involved, this study suggests, is candidates’ need to appeal to ideologically extreme individual donors to qualify for public funds. Other studies do not reach such conclusions,137 and it is too early with these systems to draw conclusions about whether individual donor-based public financing systems will contribute to polarization. But we should be alert to the possibility that they will, and we ought not take for granted that individual donor-based public financing systems will inevitably and automatically reflect the actual distribution of policy preferences among the general electorate.

Second, efforts to empower party leadership will run into America’s characteristic and unique distrust of political parties. Part of the culturally distinctive understanding of “popular sovereignty” in America has been a romantically individualist vision of democracy: a vision that sees organizational intermediaries between citizens and government, such as political parties, as a corruption of true democracy. Furthermore, if parties must be tolerated, then they must be put under the control of “the people” as much as possible; hence the Progressive Era anti-party creation of the mandatory primary election. Therefore, a robust ideological defense of political parties, as well as of party and political leadership, will have to be willingly and forthrightly undertaken in order to mobilize support for any set of practical measures that seek to re-empower party leadership.

If I am right that the problem is effective governance; that political fragmentation might be a more productive focal point for effective reform efforts than polarization per se; and that the right direction for fresh thought is how to re-empower political and party leaders, then it is also necessary to understand the deep sources of resistance that must be engaged as a prelude to any practical movement along this path. These sources lie in the distinctly American attachment to a romantic vision of democracy centered on the individual citizen, rather than on effective governance and the central role of organized political power, particularly the political parties, in determining how well a democratic system actually functions in delivering the appropriate level and forms of public goods.


American democracy has always rested on a balance between a mythology of “popular sovereignty” and the reality of what is needed to organize political and governing power effectively. The key to effective democracy might be cast in the following way: we need to sustain the appropriate elements of popular participation while maintaining a coherent and decisive enough structure of political leadership to enable effective governance.

We have to be careful not be seduced by an overly romantic and individualized conception of democracy that has a deeper resonance in American political culture and history than in any other nation. We should also be careful about invoking democratic values, such as political equality, freedom of association and speech, and participation, in overly idealized and abstract terms that fail to attend to the actual consequences of institutionalizing these values in particular ways on effective political power and governance. This is a particular risk for legal scholarship and advocacy, both of which tend to be based more on analysis and argument concerning values and principles than on empirical facts about the actual organization of effective political power.

I realize there will be no rousing ovation for any of this. Who cheers for centralizing more power in the political parties at a time when the parties are at their least appealing? Who cheers, worse yet, for a particularly elitist vision of the political parties, centered on empowering party leaders? People will not “go to the streets” in favor of political parties and party elites. All this runs counter to the DNA of America’s democratic sensibilities.

But that is part of my purpose: to challenge those sensibilities. In the midst of the declining governing capacity of the American democratic order, we ought to focus less on “participation” as the magical solution and more on the real dynamics of how to facilitate the organization of effective political power. I have tried, today, to give you a glimpse into this alternative, institutionalist approach to democracy and legal thought.