So far, we have discussed the competing traditions about what counts as democratic. However, while politicians steer by their hunches, and judges by concepts, traditions, and philosophies, political scientists examine the consequences of choice for the present and future of democratic governments. Their research invites us to ask whether the founders’ prescriptions stand up to the test of modern science or are challenged by it.
Before the twentieth century, elective government was so rare that it was almost impossible to study scientifically. England, France, and the United States spread democracy in the first half of the twentieth century, first as examples and then by control over Western Europe, Japan, and territories abroad. Elsewhere, democracy was occasional and unstable.
With increasingly robust data and methods, political scientists have investigated causal relationships, why some democracies broke down and became authoritarian, and why others successfully changed from monarchy, aristocracy, or dictatorship to democratic rule. Those studies give us an opportunity to address the future of American democracy.
What makes elective government survive or fail? Elective government is a minimal definition of democracy; but to avoid arguments about whether some definitions are too stringent or too political, we use a basic definition and let political scientists address the requirements of democracy. Some researchers organize their work around the degree of democracy, others around more elaborate definitions.1 Their work leads in the same general direction.2 This should not be surprising—compromising the foundations of democracy makes elective government vulnerable.
Rights are essential. Without basic freedoms, democracies shrivel and die. Sometimes freedoms are lost after dictators have taken power; sometimes the loss of freedom allows dictators to seize power.
Nazi inhumane perversions of law led to worldwide agreement on protections in the United Nations Universal Declaration of Human Rights. The U.S. Supreme Court’s defense of basic freedoms quickly became an international beacon. Core rights in the U.S. Bill of Rights defined ways government protects or abuses its population.
The Constitutional Convention listed guarantees against government subordinating, suppressing, or terrorizing its people, specifically blocking methods then used in England: an expansive definition of treason, denial of habeas corpus, and abuse of the criminal process to reach foreordained conclusions.3 Two years later James Madison proposed a Bill of Rights with protection to speak, publish, assemble, and petition, plus procedural protections to forestall the abuse of the criminal power of the state.4 He was explicit about the instrumental character of the latter.5 Later the Reconstruction Amendments filled out the promise of the Declaration of Independence that “all men are created equal,” now requiring equal treatment.6
The founders, not confident written rights would protect anyone, referred to them disparagingly as “parchment” guarantees, out of skepticism that fallible and sometimes malicious people would honor them when they had motives to do otherwise.7 The Constitution’s guarantees rest on the backbone of U.S. citizens, not the stiffness of the paper they are written on. Protections are weak when they seem unimportant to or for political opponents.
Rights would not have stopped the Nazis. Once in power they ignored or perverted existing rules. Dictators abuse the criminal process in order to put opponents away and stifle revolutionary movements, like the imprisonment of Nelson Mandela on Robin Island, Aung San Suu Kyi in Myanmar, mass arrests of dissidents in Iran, and too many others.
Bills of rights do not protect democracies from external threats, civil war, or internal deterioration; from misusing or squandering their armies,8 closing themselves off from the rest of the world,9 or destroying their industrial base.10 Bills of rights cannot protect democracies if authorities can sidestep them by claiming to counter other threats. As the ACLU puts it, “freedom cannot protect itself.”
The Bill of Rights has not stopped American governments from tracking antiwar demonstrators and questioning who they voted for, in the name of national security.11 The FBI’s well-known effort to crush Martin Luther King and cripple the civil rights movement was cut short by Presidents Kennedy and Johnson. It harassed Americans whose politics were unpopular with FBI Director J. Edgar Hoover. The participation of FBI agents in the KKK and other racist organizations was sometimes lethal. Frank Church, former Senate Intelligence Committee chair, said the resources of the National Security Agency (NSA) “could be turned around on the American people” leaving no privacy because it could “monitor everything.” Church added that if a dictator came to power, the NSA “could enable it to impose total tyranny, and there would be no way to fight back.”12 His point, in part, was that the ability of government agencies to gain business and personal information can be used to target or crush disfavored individuals. Recent revelations of NSA tapping and storing of phone and digital records underscores his point.
The use of drones to execute people abroad takes life without any defense, let alone due process. Restriction to the field of battle once offered noncombatants some protection, but the find-and-execute policy obliterates it. Used against Americans abroad, the policy short-circuits the Bill of Rights, and invites more targets at home to enemies abroad.13
Detention without due process is no longer merely foreign behavior, based on the 2012 National Defense Authorization Act. President Obama denied some implications in a signing statement but defended the law against a suit by well-known journalists. On September 12, 2012, District Court Judge Katherine Forrest, an Obama appointee, held that portion of the Act unconstitutional.14 The administration immediately requested that her order be delayed pending appeal. On July 17, 2013, the Second Circuit decided that the plaintiffs did not have standing, that is to say the legal right, to challenge the law.15
The Bill of Rights depends on the backbone of legislators and judges, the compliance of the authorities, and the means to protest. After 9/11, the administration rounded up Muslims who lived in the United States and refused to identify those in custody, for the benefit of relatives, friends, or lawyers. The administration blocked access to phones and moved detainees, convicted of nothing, to sites far from their families.16 Many Americans appear to have accepted those risks and to have been convinced that such risks are essential for “security,” even as the tools used by the administration threaten security from within.
Unfortunately America has had dictators. The slave-holding South before the Civil War treated slaves tyrannically and violently suppressed antislavery dissent.17 Open warfare in the post–Civil War South killed Republicans and removed an elected biracial government by coup d’état.18 Until the late twentieth century, roving bands of white enforcers, the Ku Klux Klan, and similar organizations enforced a code of total subservience throughout the former Confederacy, many border states, and sometimes well beyond. They infiltrated all levels of law enforcement.19 At the same time, big city machines controlled elections with a combination of intimidation, bribery, mayhem, the machinery of public administration, police, and the courts so that one had little choice but to cooperate.20
Abuse by law enforcement officials is more serious than most Americans realize. It includes using the power of office and weapons for illegal purposes, framing the innocent, and covering the officials’ own misbehavior. Serious official abuse is continuous,21 in the East,22 the West,23 rural areas,24 police labs,25 prosecutors’ offices26 and at the FBI.27 Weapons are used inappropriately,28 and people are injured or killed for reasons known only to the officers involved.29 Minor crimes like breach of the peace or interfering with a police officer are prosecuted to cover official abuse of people they believe are not respectful enough.30 And racial abuse is ubiquitous.31 Dealing with it has so far proved unsuccessful.32
Available statistics are notoriously difficult to interpret, but Kami Chavis Simmons described excessive force, perjury, and racial profiling in Los Angeles, New Orleans, Atlanta, Chicago, and New York, among others.33 Documented abuse runs the gamut from quotidian but collectively damaging abuse of minority communities34 to wonton shootings—“New Orleans police officers opened fire upon several citizens as they were crossing Danziger Bridge to flee . . . [Katrina’s] devastation.”35 Simmons reveals that “nearly 30% of the officers surveyed believed that ‘the use of excessive force is a serious problem facing the Department.’” And almost 4.6 percent of officers believed they were justified in physical punishing suspects with “a bad or uncooperative attitude.”36 In effect as Simmons, among many others, has noted, this is not merely a few “bad apples” in police departments, but a culture that protects the officers and defends their behavior.37
Investigations intended to stop flagrant abuse of power are recurrent, including the Wickersham38 and Mollen Commissions39 in New York, the Christopher Commission40 in Los Angeles, and congressional hearings,41 among others. The scholarly literature about the problem is lengthy, from Paul Chevigny’s pathbreaking study in the late 1960s42 to Michelle Alexander’s in 2012.43
Torture used to seem foreign. But America introduced waterboarding to the Philippines during the Spanish-American War; later it would torture Iraqi, Afghan, and other detainees for information—despite moral and practical reasons not to do it, and against the advice of experienced interrogators to avoid torture altogether because the “information” it yields is untrustworthy and because they have more effective ways to interrogate prisoners. The Bush administration refused to describe “detainees” from Afghanistan as prisoners of war precisely because that would have called for the application of specific legal rules governing their detention and treatment, effectively making it illegal to hold them in hot, open cages, hooded, blindfolded, masked, chained, or kneeling for extended periods—as well as incommunicado, and without access to attorneys. Prisoners were deprived of sleep and many were placed in bindings that compress and make limbs swell.
Americans are not angels. When the Bill of Rights and Reconstruction Amendments have not been enforced, Americans experience serious mistreatment.
Freedom cannot protect itself; democracy cannot survive without freedom.44 Political scientists describe countries as democratic if they protect core rights including the right to vote, and autocratic otherwise.45 Elections are insufficient if people can be otherwise controlled.46 Regardless of description, societies that abandon civil liberties to search for terrorists sink into autocratic rule. Jennifer Holmes documented the way military reign drove Peru and Uruguay into dictatorship, while Spain, defending itself against the Basque insurgency, restrained itself and remained democratic.47 Paul Wilkinson added postwar Germany, France, and Italy as countries that maintained both freedom and democratic government while fighting terrorists.48 British courts fought back at attempts to imprison people without charges, hearings, or trials.49 Without the restraint of courts and due process, petty tyrants are empowered to pillage and control their own people and governments become as debased as the guerillas; both groups sell out democracies.
Elections define the people’s choice if the people can vote. Keeping people out of the electorate gives some people the power to rule others. That might be described as colonialism, apartheid, or occupation.
Those in office are generally satisfied with prior voters; new voters threaten their power and they often try to prevent them from voting. Whatever magnifies the time, effort, and cost to vote tends to exclude the poor—costs for transportation; special forms of identification required by new voter ID laws; lost wages while they wait in lines, respond to spurious challenges at the polls, or go to court to seek their right to vote can be prohibitive for the poor but are minimal or nonexistent burdens for more economically secure voters. Other methods exclude immigrants, students, and those caught in or abused by the criminal justice process.50 This is not about democracy but about attempts to control the results.
Equally important, inclusive electoral systems are more likely to remain democratic than uninclusive ones. The smaller the group of people to whom the leadership owes its power, the fewer the leadership need to pay off; the fewer people a ruler needs to satisfy, the less he needs to do for the people at large, and the easier to satisfy his crucial supporters. That kind of society collapses into a kleptocracy. Stealing from the poor and giving to the favored few has been the method of tyrants in all generations.51 One way to accomplish this is by excluding people from the polls. Conversely, the larger the group of electors and the larger the group of people who can keep the leaders in power or put them out, the more the leadership is driven to adopt policies that serve the public, and not merely the friends of those at the top.
Inclusiveness is only partly about election rules. The world struggles for honest elections. Criticism of dishonest elections abroad is nonpartisan here. Honest elections are the sine qua non of democracy for judges and political scientists. Yet domestically clean elections have often been a casualty of politics.
Stuffed ballot boxes can take over democratic government. For much of our history, election results had the security of gold shipped on stagecoaches through narrow valleys in movie westerns. Armed mobs and state militia faced off at the Pennsylvania capital in 1838 over disputed election returns and determined the winners.52 Landowners and their overseers controlled the votes of many white tenant farmers in New York’s Hudson River Valley until the antirent wars of 1839–46 changed their relationship.53
Landowners, employers, and local leaders prevented blacks from voting in southern states well into the twentieth century and carefully picked those few who were allowed to cast ballots.54 The Civil Rights Act of 1957 authorized appointment of federal registrars to register black voters where discrimination had been rampant, and the 1965 Voting Rights Act gave the Justice Department new powers to secure the right to vote. Nevertheless, those in power continue to intimidate, disqualify, and block many blacks from voting.55
Large-scale, effective election chicanery starts at the top. Big city machines intimidated and bribed voters well into the twentieth century.56 Voters responded by demanding secret ballots and enclosed voting booths.57 Political bosses then sent people to watch whether voters spent enough time in the voting booths to split tickets instead of casting a straight party line vote, to assist voters inside the booths, or to rig the machines.58 Small-town registrars did not bother to count some ballots.59 Political bosses and their henchmen were well-practiced masters at stealing elections, miscounting votes, stuffing ballot boxes, and other chicanery.
In the late nineteenth century, mining and similar companies with isolated work forces in one-industry states gained all-encompassing economic power over employees, coerced their votes, and gained control over entire states like West Virginia and Montana.60
In the twentieth century, the New Deal safety net, which included Social Security and unemployment insurance, reduced the attractiveness of jobs for the legions of people required to intimidate, bribe, or fix ballots. And it rebalanced employer-employee economic power by substituting a system of labor negotiation for preexisting violence. Soldiers returning from World War II had less taste for dirty politics, more courage to resist, and the GI Bill to find a better life.61 We rely on those improvements at our peril; attacks on the safety net and growing economic insecurity make many people more vulnerable.
Seventeen states made it impossible to check election equipment to verify the 2012 count.62 Hurricane Sandy wrought havoc with the election, creating lines of voters, discouraging some, and making the wait impossible for those with inflexible obligations such as work or family care.63 Equally devastating, misallocation of polling places and equipment, photo identification laws, limits on early voting, electrical and other failures and decisions make it difficult, prohibitively expensive, or impossible for some people to vote, skewing elections.64 Voting machines misfire, whether by accident or design.65 Election administrators are politically selected, leading New York Mayor Bloomberg to object that party leaders “should not be . . . picking their buddies to supervise the basis of our citizenship.”66 New voting technologies bypass the need to employ large numbers of people for voter fraud by making the count itself insecure and unverifiable.67 People get used to abuse of election machinery.68
Each new weapon or strategy in the battle for honest elections is eventually met and defeated by another. There is no known unanswerable protection for elections and no substitute for vigilance. The United States has had relatively clean elections only since World War II in the North and the civil rights movement in the South. But the jockeying continues.
For most living Americans, discomfort with the 2000 presidential election tally was a new experience. The New York Times examination of the handling of absentee ballots made it clear that partisan differences in the acceptance of illegal ballots in Florida were sufficiently large to affect the results—George Bush’s election would have been confirmed if a recount examined only those counties which Al Gore requested, but Al Gore would have been elected if the entire state had been recounted.69 Others studied Florida’s systems for disqualifying and discouraging likely Democratic voters and reached the same conclusion.70 Questions have also been raised about the 2004 presidential election.71
If the ballot count can be gamed, then the voting process is ineffective. Had the secretaries of state been judges in battleground states like Florida and Ohio, they would have been disqualified by their stakes in the outcome from supervising the elections and affecting the result.72 The political culture and the risks of exposure are not strong enough to protect everyone’s right to vote. Skepticism about the accuracy of the count and doubt about the seriousness of the risk of trying to game it are corrosive.73
Malapportionment corrupts democracy. Delegates to the 1787 Constitutional Convention criticized British elections as having been corrupted by rotten boroughs, those boroughs with few voters.74 The delegates also understood that malapportionment had poisoned state representation. Settled western portions of the thirteen colonies had fewer representatives in state legislatures in proportion to population than the older seaboard communities. The seaboard was unwilling to readjust inland representation. At the Convention, delegates criticized the unfairness to western counties. They insisted on a census for allocating seats in the House of Representatives among the states, with the crucial exception of slavery.75
Predetermining power by apportioning representatives was crucial to the North-South compromise at the Convention. By eighteenth-century calculations, delegates expected southern states to control the House. Their calculations assumed that the potential population base in an agricultural society would be proportional to the size of the states. Southern states were much larger. It dawned on northern representatives that their protection against the slaveholding states was in the Senate where the tiny New England states would each have as many senators as Virginia, then the most populous state. So the distribution of seats in the two houses of Congress in the original Constitution confirmed both southern power in the House and northern power in the Senate.76 Of course, the eventual path of population growth did not work as expected. But the “Connecticut compromise” governed national power for decades, regardless of voters’ preferences.
Congress soon decreed that states could not elect representatives at large, a method in which state representatives are elected as a block, similar to how the electoral college works.77 Instead, each state must elect representatives by districts. That produced a vote more in keeping with the balance of forces within each state, although the shape of districts became vulnerable to partisan manipulation.
As population growth shifted from farming communities to cities, legislatures ignored the changes. The shift left rural districts with fewer people per representative than urban and suburban districts, and therefore, more representatives and greater clout in legislative halls than their voting population would have warranted. Representatives of farming communities continued to dominate the malapportioned Congress and state legislatures.
The Warren Court required “one person, one vote,” or, more precisely, that each legislative seat should represent the same number of people. If not, then people from favored regions would have more legislators than their numbers justified.78 The Warren Court’s reapportionment cases—from Baker v. Carr in 1962 to the final 1969 quartet at the end of Chief Justice Warren’s service on the Court—shifted considerable legislative power from rural to suburban parts of the United States on the basis of their relative population strength.79 None of us are, in Orwell’s famous words, “more equal than others.”80 The Burger Court, however, in its last year, 1986, came to the conclusion that two to one was sometimes equal enough.81
The Court is often described as undemocratic because the justices are not elected. But the Burger, Rehnquist, and Roberts Courts have been increasingly undemocratic in a different way—their decisions have been dismissive of malapportionment, gerrymandering, miscounting, and other ways to minimize the voting rights of qualified voters.82
Gerrymandering survived the reapportionment rules. The word comes from the name of Elbridge Gerry, member of the 1787 Constitutional Convention and governor of Massachusetts, when a strangely shaped district was created to increase his party’s representation.
The way lines are drawn can make a legislature safe for one party, regardless of voters’ preferences. Using census, statistical, and political data, line drawers waste opposition votes by “packing” or “stacking” opposing voters into a few districts and “cracking” the remaining pockets of opposing voters so they can be defeated easily—divide and conquer by drawing maps. Line drawers sometimes call legislators to tell them their careers are over; one called a prominent congressman and classmate of mine, whose district was carved up to enable the Hispanic community to elect a person of its choice.
Courts rarely intervene. Justices Clarence Thomas and Antonin Scalia long since denied there is any useful concept of democracy.83 Their Court decided it is permissible to draw lines for the purpose of protecting incumbent legislators’ seats.84 Strong opponents can be isolated. The Court upheld a 2004 redistricting plan that gave Republicans victories in twelve of nineteen Pennsylvania congressional districts even though the statewide vote for Congress substantially favored the Democrats.85
Incumbents are protected by designing districts with overwhelming support for them or their parties. In turn so-called safe districts encourage political extremism. Freed of competition, middle-of-the-road independent voters can be ignored. Victory means pandering to party activists; no district residents have the clout to make objections stick.
Legislators in competitive districts must appeal to centrist voters who hold the balance of power. Squeezing out competitive districts makes independent and centrist voters impotent, leaving the primary the only mechanism of voter choice. Primary elections, however, are contests for more partisan, ideological candidates. The primaries forced Romney closer to the right wing of his party. American voters now have more of what Barry Goldwater termed “choice; not an echo,” with intransigent Tea Party members facing ever more determined Democratic opponents. As moderates willing to compromise were forced out, legislative politics shifted from compromise to conflict, although the resulting Congress is deeply unpopular with American voters.
Gerrymandering is undemocratic—because incumbents pick their voters, protecting themselves against political accountability, and because it determines the partisan outcome of voters’ choices at the polls. Gerrymandering diminishes the incentive of elected officials to produce the kind of public goods that would broadly benefit the population.86
Proportional representation gives legislative seats to parties in proportion to votes and usually provides seats across the spectrum, from moderate to extreme. Such systems are sometimes thought to raise the political temperature because individual legislators have no incentive to moderate positions. Statistically, democracy may survive equally well with proportional representation as with the single-member districts, commonly used in the United States. However, extremism has led to the breakdown of democracy and played a role in the demise of democratic governments in countries like Italy, Germany, and Austria between the world wars.87 Although single-member districts are often more moderate, they become problematic when safe districts polarize the country.
The politicization of the districting process and the appeal to the courts to evaluate the lines is a peculiarly American problem. Other democratic countries with geographical districts like ours have used nonpolitical commissions to draw the lines.88 Nevertheless, courts could deal with gerrymandering. Political scientists have been using symmetry, also described as neutrality, to measure bias for decades. Symmetry requires treating both parties the same way so that a given percentage of the statewide vote should give either party approximately the same number of seats. (By contrast, packing and cracking of voters are reflected in deviations from the statewide average of expected party voters. Predictions are based on comparing expected deviations.) Symmetry does not guarantee what voters will do. Symmetrically designed districts, however, are not a stacked deck. A mathematical formula allows measurement that approximates the precision of the one person, one vote standard for reapportionment.89 Despite questions around the edges, the central concept is hard, clear, and powerful.90 Districting can be done in such a way that losing parties are highly unlikely to win the majority of the seats.
Symmetry is consistent with preferences for homogenous or for diverse district populations, but not with unequal treatment of supporters of different parties—what’s sauce for the goose is sauce for the gander. As Justice Kennedy explained in another gerrymandering case, there are fundamental rights involved.91 The parties have a constitutional right to equal treatment. Symmetry measures it with considerable precision, using the same data that the parties use to rig elections.
Partisan management of election outcomes is particularly virulent where it has the support of prosecutors and courts, both here and abroad.92 Democracy is only as secure as the people who work it.
There are two issues here: One is whether the separation of powers excuses us from worrying about threats to American democracy; the short answer is not much. A separate issue is how best to understand the value of separating the powers of the branches.
To much of the world, the American model is summed up by a powerful president and a powerful court. All democracies have legislatures so long as they last. The separation of powers among the three branches creates checks and balances.
The separation of powers was the French Baron de Montesquieu’s inaccurate description of British government, which many in America’s founding generation admired.93 Most European democracies have followed a parliamentary model where the ruling coalition, often led by a powerful executive, dominates all branches. The separation of powers is not a good description of the checks and balances of that system. But few European democracies seem in any danger of collapse into autocracy.94
American constitutional law is bathed in the light of the Federalist Papers.95 Written as an argument for the adoption of the Constitution in 1788, they embodied the best eighteenth-century thinking about democracy. American courts have assumed in turn that the success of the Constitution is sufficiently explained by its structure as explained in the Federalist Papers.96 Empirical science, however, tells a different story.
Gauging the effectiveness of the separation of powers suggests comparison between the failure of many parliamentary systems in Europe between the two world wars, which fell to the Nazis, with the presidential systems in Latin America since World War II, which have repeatedly fallen to dictatorship, often at the hands of their presidents.97 On average, parliamentary systems have lasted longer than presidential ones.98 Many other differences among those countries and regions may account for those outcomes, but they certainly cast doubt on the belief that our separation of powers has much to do with the stability of our democratic system.
Should we blame the model or the culture? Political scientists dispute whether presidential and parliamentary systems are an important cause of the short or long lives of democracies. Juan J. Linz, a Spanish political scientist who recently passed away after a long and distinguished tenure at Yale, was born shortly before Generalissimo Franco defeated the Spanish Republic and made himself dictator. Linz devoted his life to understanding why democracies rise and fall, and he believed presidential systems more likely to succumb to takeovers, with Latin-American countries his prime examples.99 G. Bingham Powell, an award-winning political scientist at the University of Rochester, discounts Linz’s observation about Latin America, saying the difference is essentially the result of economic factors.100 Either way, the system of checks and balances that go with presidential systems have not prevented coups d’état and takeovers by chief executives. Fear of executives is well justified. Democracies are likely to be taken over by their presidents and prime ministers.
Both presidential and parliamentary systems reflect and react to the distribution of wealth and power. Where political and economic power is concentrated around the executive branch and its supporters, the country will function as a dictatorship. Similarly, a powerful aristocracy may offer little protection for the rest of the population. Social and economic hierarchies prove more crucial than the choice of presidential or parliamentary systems.
Nevertheless, two centuries of tradition have shaped our practices and expectations and they matter here even if the separation of powers is not exportable for the purpose of controlling executives. They create, and were intended to create, checks and balances and to enable the departments to act with “energy” where appropriate. Some courts have treated the separation of powers as a method for protecting certain powers as independent privileges of one branch of government beyond the oversight of the others.101 Treating separated powers as independent tends to exceed historical understanding—Madison thought blended a better description of the powers than separated102—and treating separated powers as independent conflicts with the purpose of distributing power so that executives are not dictators.103
The Constitution’s federal structure is quite important to the way American democracy functions.104 The overwhelming subjects addressed by federalism are coexistence of people in different places, the effectiveness and fairness of their governing structures, and, in turn, the survival of democracy. A separate issue is the kind of federalism that matters. There is little or no evidence that the differences about which American courts have spilled so much ink matter.
Federalism typically comes about by means of a deal among the powerful, designed to protect their control over territory, people, or issues or to protect themselves against control by those they fear. Political scientists refer to constitutions that come about in that way as pacted constitutions. Federalism, then, may be an unavoidable compromise or a wise one—either an efficient way to get things done by decentralizing power or an inefficient duplication of resources. It may reflect tolerance for different cultures or it may be an invitation to intolerance by segregating people and maximizing the differences among them. If travel (sometimes called “exit”105) is less threatening than staying put, people may segregate themselves as they did in India and Pakistan and in many recent conflicts in response to racial, religious, and ethnic attacks. If there are opportunities that cut across federal units, people may be inclined to learn to work and live together. If different sections share problems they may be inclined to work together in solving them. How federalism turns out is determined by history—there may be no universally right way to divide populations, although there may be wrong ways that lead societies to threaten their members with mayhem.
With federalism designed to solve so many different problems in so many different ways, there is no single correct definition of federalism for political scientists. Germany is federal although the Länder (federal subunits somewhat like American states) are required to carry out national policy,106 and the United States is federal although it has a powerful central bureaucracy unlike Germany.107 National and local politics determine what is constructive, what diffuses power or conflict, and what hobbles government or threatens minorities.108
Yugoslavia functioned as a federal republic.109 It may not have been possible to create that state other than as a federal one.110 The various Yugoslavian states had a great deal of power.111 But Yugoslavia became a poster example of federalism gone awry.112 When international and economic conditions worsened in the 1980s, the federal government had little power to deal with its problems. So the politicians in the various Yugoslavian states blamed each other’s states for their predicament; and they sank their country into civil war.
That is one of the biggest issues for political scientists—federalism either relieves political stress or makes it much worse; and it either strengthens or weakens democratic government.113 As in Yugoslavia, the consequences can be disastrous.114 Local allegiances can breed resentment and distrust of the nation and eventually lead to dismemberment or civil war.115 Switzerland has held together,116 but Belgium,117 Canada,118 and other countries have gone through great internal turbulence related to their federal structure. Our own country nearly came apart when the eleven states of the Confederacy seceded from the Union. Political scientists point to whether federal borders follow or cut across ethnic or other fault lines in society.119 Borders that follow fault lines can exacerbate problems; crosscutting borders may calm them down.
Political scientists are somewhat divided on how best to make deeply divided societies democratically governable.120 The coincidence of ethnic and provincial boundaries may be too simple an explanation for strife and violence, given that we are learning that identities can be reshaped and, as in the birth of Pakistan, ethnic groups can be exchanged—to use a very sanitary word for a very nasty business.121 And the distinction between federal boundaries that do and do not cut across ethnic lines may be particularly unhelpful when the political bargains demanded as the price of a unified democratic state depend on boundaries that follow deeply felt distinctions.122
Federalism that enacts a political bargain may have to define political boundaries very sharply in order to engender trust. And as long as the society remains deeply divided, there may be no alternative to sharp boundaries among groups and between federal and national powers. It makes sense in such societies for courts to police such bargains. The French Constitutional Court, for example, was developed as part of a political bargain to protect the Gaullist Constitution.123
As democracy ages, however, issues that could have erupted and torn the state apart are likely to dissipate. Judicial supervision of the antique federal-state line becomes increasingly irrelevant. Moreover, outdated distinctions can throw monkey wrenches into the machinery of democracy. Slavery is one such outdated distinction, but it cast a strong shadow on the eighteenth century world in which the American constitutional language of federalism was written.124 Judicial “benign neglect” is not likely to undermine a two-century-old democracy, but judicial activism, using federalism to make policy, can do a great deal of damage. Precisely because federalism is a method of conflict resolution, its contribution changes with the conflicts and therefore is not permanently defined by stone tablets.
Federalism may be the midwife of emerging democracy.125 It may help to develop talent and inculcate democratic attitudes.126 It offers opportunities for participation in which people can learn and develop skills. And it offers people the opportunity to solve local problems as well as an escape valve for strong local feelings. Those advantages, however, depend on contemporary problems and not on ancient distinctions. It is best treated as a political problem, not a judicial one.127
In a mature democracy, a clear division between federal and state powers may be precisely the problem, encouraging extremists and aggravating conflicts where candidates think they can benefit from stirring the hornets nests,128 or where the national institutions become incapable of dealing with common problems because of jurisdictional issues.129 In those situations, political organs may police the proper boundaries better than courts.130 More permeable bargains may actually assist in defusing conflict by moving conflicts to arenas in which they can be handled more successfully. In such societies courts would be better advised to keep out.
Federalism may contribute more to a mature democracy if national and state powers overlap, keeping all political actors in check and preventing a slide toward autocracy.131 Ian Shapiro has described freedom as the “multiplication of dependent relationships” (in stark contrast to slavery which imposes dependence on a single other).132 Federalism can reduce freedom by simplifying government jurisdiction. Juan J. Linz133 and Robert A. Dahl134 describe decentralization as moving problems to smaller units that may be less able to handle them fairly.135 A more flexible federalism may alleviate some of those problems.
Madison and Hamilton described important ways in which the two levels would restrain each other.136 Describing the Constitution as partly national and partly federal,137 Madison argued for a blending, rather than a clear division, of powers. Similarly, in the context of the separation of powers among the branches he also argued that blending was the more effective guarantee.138 He told us that by blending powers we would enable the constituent parts of government to control each other.139 It was precisely the overlapping of powers that created the possibility of protecting liberty. Overlap between state and federal jurisdictions permits each to investigate the other, to compete for public support, and to provide a staging ground for opposition as the Virginia and Kentucky resolutions did in 1798 and 1799.140 Opponents of our Constitution wanted the kind of clear division of powers the Court has advanced. But those who wrote and supported it had a more sophisticated understanding of federalism.
Political scientists are also concerned by the impact of federalism on the quality of democratic government and on the extent to which government reflects and responds to the governed.141 Madison famously argued that a larger republic would be fairer;142 for example, it would be fairer to the slaves than the southern states were. Martha Minow described the same problem of fairness with respect to several modern conflicts.143 Majorities can be thwarted by federalism in the same way that gerrymandering can thwart them. State, county, and municipal boundaries stack and crack groups just as gerrymandering of election district lines do. Majorities can be defeated, submerged, and subordinated by empowered cultures or groups.144 There is no automatically appropriate federal division of the population that reflects democracy best.
Robert Dahl, a foremost student of democracy, has explained this problem at length. The majority principle and the principle of fairness may, or may not, require ever-greater inclusiveness so that the needs of the larger population are met.145 There is no principled point at which one can stop enlarging or dividing the borders.146
Beyond the fairness of geographic lines, clear division of powers between federal and state governments makes some problems unsolvable, thus frustrating democracy. This is true wherever the source of the problem extends beyond state borders and requires regulation of external entities, such corporations, businesses, or other states. Democracy can be frustrated when the state political machinery is too corrupt to clean itself up.
In short, for political scientists, there is no morally better level of government—which level of government is or will be fairer and better depends on the distribution of people and prejudices.
The Court’s explanations and its inferences of federal structure from the Constitution are virtually unrecognizable to political scientists. State sovereignty is a problem, not an explanation.147 Clarifying the lines of authority has a nice ring to it, but poses important trade-offs for democratic government. The Court’s clarification of authority eliminates much of the mutual restraint that provides checks and balances.
Different versions of federalism make different trade-offs among various measures of fairness. Local control can mean democracy to the anti-Federalists or “faction” and unfair treatment to Madison and Federalist supporters of the Constitution.148 Political scientists find that localities use power to keep “classes, races, ethnic groups, genders, and life-style groups in their places.”149 Against a standard of just government, federalism can be cited by both sides. In effect, the definition of federalism is political and contextual rather than a subject of universal, unchanging principles.
Federalism can thwart liberty; many groups around the world look to more cosmopolitan entities to protect them.150 It was that threat to liberty that led Congress to propose and the states to ratify the Fourteenth Amendment to change federal relations.151
Justice Blackmun agreed that federalism can protect personal liberty: “Federalism secures to citizens the liberties that derive from the diffusion of sovereign power.”152 Justice Blackmun, however, meant that federalism is valuable when and if it advances liberty, but not otherwise. He made the comment in a case where the petitioner was being denied any review, state or federal, of a conviction for murder. His point was that deference to the states does not automatically advance liberty. “Federalism,” as Blackmun put it, “has no inherent normative value.”153 In Lopez, the Court barred federal power to ban guns from schools.154 It prevented the federal government from protecting children from weapons in their schools. How that protects liberty is much less clear.
Was James Madison right when he told the Constitutional Convention in Philadelphia that a larger republic would be fairer than a smaller one?155 The government of the larger republic would have to consider different perspectives, and people in a larger republic would be less likely to gang up on small or weaker groups, African Americans included. So it would be wise to give central authorities the powers they need.
Or was Madison wrong? Political scientists find support for decentralized systems, and the data indicate that decentralization can sometimes increase stability.156 But scientists are appropriately agnostic on the fairness of federalism.157
Courts focus on definitions, and sometimes on objectives defined by their understanding of the founders. Political scientists, by contrast, have focused on the ambiguous relationship of federalism and democracy. Federalism often helps secure democracy but also deepens conflicts that tear it apart.
Political scientists’ theories of pluralism suggest that clarity of boundaries will also have an ambiguous relationship to democratization and the stability of democracy—possibly reducing stress and strengthening democracy when regions do not trust each other, but potentially weakening democracy by emphasizing regional competition and differences. In other words, whether the boundaries need to be sharp depends entirely on political circumstances. There is no reason to believe that the courts will prove themselves wiser than the politicians in discerning whether and where the tensions are in fact so sharp and what resolution would calm them. And there is good reason to believe that an aging conceptual federal-state boundary will make it harder to resolve contemporary disputes. The Court may have had it right, from an empirical perspective, in Garcia, when it left that issue to the political process.158
Federalism doctrine as the Rehnquist Court left it in 2005 at Chief Justice Rehnquist’s death paid careful attention to eighteenth-century feuds—indeed it “protected” states from exercises of national power that were sought by the vast majority of states.159 Despite all those issues, all branches of government can use federal structures for better or worse. If managed pragmatically, with an eye toward the context of needs and problems, federalism can contribute to democracy: It can be part of a compromise necessary so that people can agree to a democratic government; it can be handled as part of a system of checks and balances, limiting abusive behavior wherever it happens; it could be part of self-government, making it more efficient, effective or improving people’s satisfaction with their government; and as many political scientists address it, federalism could be constructed as a system for managing (that is, cooling and moderating) conflict.
Federalism is about statesmanship, not lines in the sand.
A healthy democracy depends on protecting its people from the kinds of abuse that can be used to break or intimidate them, protecting the procedures of democracy, from speaking and joining to voting, and including the whole population. Morally, the right to govern depends on self-government and therefore on universal and equal treatment and inclusion of the population. Pragmatically, that becomes an important protection and predictor of the health of democracy.
By contrast, whether government follows a presidential model with the separation of powers or a parliamentary system where executive and legislative powers appear more blended has no convincing relationship to the survival of democracy, except, and this may well be important, to the extent that they bring out the best in their peoples. Stability may be more important than the supposed virtues of a different system.
Federalism is a tool that can be as damaging as it can be helpful; it needs to be used and managed with intelligence.