A Radical Plan to Change American Politics

This kind of acquiescence in tradition--pious if you approve of it, mindless if you do not--might be justified if the U.S. Congress were an effective and popular institution. The consensus of its inmates and its constituents, however, is that Congress does not work. Widespread discontent with the political system could have been expected to produce proposals for reform of the legislative branch, and it has. Unfortunately, the most familiar proposals for reforming Congress would do so at the expense of democracy.

THREE BAD IDEAS

Bad idea No. 1 is term limitation. Why should any constituency be prevented from sending the same representative or senator back to Washington again and again? The direct election of U.S. senators, brought about by the Seventeenth Amendment, did not, as advertised, eliminate corporate influence on legislative politics. Only the very naive believe that the same promised result--the sidelining of "special interests" and the filling of the legislature with public-spirited citizen-legislators--will be produced by frequent rotation. In a capital city in which expertise is power, the frequent circulation of amateur legislators would only increase the relative influence of the permanent congressional staff, the federal bureaucracy and the entrenched Washington establishment of lobbyists and insiders.

The misguided panacea of the moment, term limitation joins two long-standing proposals for congressional reform, one associated with the left, the other with the right. For much of the twentieth century Progressives, New Dealers, and then left liberals, in the name of "responsible party government," have dreamed of reforming the United States to resemble an idealized Britain--bad idea No. 2. (Ironically, today constitutional reformers in Britain are seeking a new system with many "American" features, such as federalism, judicial review, and a written constitution.) Those who seek to reform Congress along parliamentary lines are not concerned about the separation of powers, an antiquated system they deride. Instead, they favor the "separation of parties." In their view, the American mass electorate should be given a clear choice between two sharply defined ideological parties; a party that wins a bare majority in the House, the Senate, and the presidential election should be given carte blanche to govern with as few impediments to the exercise of power as possible, rolling over not only the opposition party but also dissidents within the majority party itself.

If a parliamentary Congress was the dream of midcentury liberals when it appeared that the Democratic Party would control all three branches forever, the recent inability of Republicans to elect congressional majorities has created a constituency on the American right for bad idea No. 3--what can almost be described as presidential rule. Effectively abandoning any hope of ever recapturing Congress, many Republicans have favored strengthening the power of partisan Presidents to alter policy directly, through executive orders, rather than the old-fashioned way--by pressing for the repeal or the amendment of laws.

Since Richard Nixon began the present era of Republican presidential dominance, Republican claims for "inherent" executive prerogative have grown ever more extravagant (recycling the arguments of Democratic presidentialists in the era from Franklin Roosevelt to Lyndon Johnson). President George Bush has claimed vast powers in domestic as well as foreign policy. In addition to asserting that he had the power to wage war in the Persian Gulf without congressional approval, he has declared that he will not enforce provisions that he considers unconstitutional in laws that he has signed. Some conservatives even argue that Bush should exercise a supposed "inherent" line-item veto that has been read into the Constitution (where, somehow, it has eluded discovery since the days of George Washington, who observed, "From the nature of the Constitution, I must approve all the parts of a bill, or reject it, in toto").

Further magnification of discretionary White House power--even in the service of legitimate goals--runs the danger of creating a North American form of Latin American plebiscitary presidentialism, in which a demagogic Presidente circumvents a weak legislature and rules by fiat, in the name of a "mandate" from the national electorate.

Many of the problems of Congress that term limiters, parliamentary reformers, and presidential-power enthusiasts seek to address are genuine: swollen congressional staffs and perks, arrogance by entrenched committee chairmen, subservience to wealthy and well-organized lobbies, micromanagement of executive agencies by incompetent or venal legislators. These abuses cry out for remedies--but for republican remedies.

A GOOD IDEA

Other reforms of Congress can extend rather than restrict democracy in America. One is an overdue increase in the size of the House of Representatives.

In 1789 the first House of Representatives had sixty-five members, each representing about 30,000 inhabitants (and far fewer qualified voters). Anti-Federalist opponents of the new federal Constitution protested that districts of 30,000 were too large: congressmen would be far too remote from the concerns of their constituents (the constituencies in most state legislatures at the time were much smaller). One of the Federalist papers (No. 55) is devoted to justifying such enormous districts.

Today each of the 435 members of the House represents about twenty times as many voters as the first representatives did. Whereas the United States has, on average, one representative for roughly every 600,000 inhabitants, the ratio in Japan is one to 238,600, in Germany one to 120,000, in France one to 96,300, and in Britain one to 87,500. Obviously, the disparity is due largely to the differences in population between the United States and other Western countries, but the number of representatives also plays a role. The U.S. House of Representatives is small by Western democratic standards. Germany's newly revised Bundestag has 662 members, the British House of Commons 651 members, France's National Assembly 577 members, and Japan's lower house of the Diet 512 members.

There is nothing sacred about 435, the number at which the-House of Representatives has been frozen by law since 1929. In the intervening six decades the United States has more than doubled in population. A vote for a representative, therefore, counts less than half as much in 1992 as it did in 1929--and a twentieth as much as it did in 1792. The weight of a vote will continue to decline: fifty years from now the average congressional district may include 750,000 people. Already, Montana has gone to court to challenge the present apportionment system, which would give the huge state only one congressional seat, representing 800,000 people. What is more, at the rate that the U.S. population is growing, in the foreseeable future a fifth of the states may have only one representative--and a third only two or one.

Clearly, the membership of the House cannot be increased in direct proportion to the population. If the original 30,000-member constituencies had been retained, by today there would be more than 8,000 representatives. (The House would have to meet in a stadium.) Nevertheless, moderate increases in House size can help limit the dilution of democracy caused by population growth. If the experience of other Western democracies is any guide, a 500-member House would not be unmanageable, nor would a 600-member House. The House is not the Senate; it loses its special function as districts grow too large and representatives find themselves ever more remote from their constituencies. If there is an argument against moderate increases in House membership beyond 435, it must be more compelling than mere tradition. Sixty-three years is not time immemorial.

PROPORTIONAL REPRESENTATION

The comparatively small number of representatives in the popular branch of the U.S. legislature is not, however, the only respect in which the United States is less democratic than the countries of Western Europe.

Electoral systems can be divided into two fundamental varieties: the plurality, or "winner-take-all," method and the party-list method, with proportional representation (PR). Under the plurality system a representative is elected by a simple plurality (or in some cases a majority) of voters in a single-member district. In contrast, in PR systems the country is divided into multimember districts (in a small country such as Israel, the entire nation may be a single district). Several parties present lists of candidates within each multimember district; the electorate casts its votes for the parties, rather than the candidates; and then the seats are allocated among the parties, on the basis of the proportion they received of the total vote.

The United States has inherited the plurality method from Britain, which maintains it as well (Australia has broken with the British tradition in favor of more-modern methods). The Anglo-American method makes possible distortions of the democratic process which are simply impossible under PR. Imagine a country with a plurality system in which there are two major parties, X and Y. Those who vote for party X, even if they make up no more than 51 percent of the voters in each single-member district, may elect 100 percent of the representatives; those voting for party Y, although they constitute 49 percent of the population, MAY END UP WITH NO REPRESENTATIVES AT ALL. Now suppose that X and Y are joined by a third party, Z. If in every district X receives 40 percent of the vote, Y 38 percent, and Z 22 percent, X will control every seat in the legislature, even though 60 percent of the population voted for other parties.

These examples may seem extreme, but there are cases in which plurality systems have elected one party even though another received a greater number of votes. For example, in 1974 the British Conservative Party lost its majority of seats in the House of Commons, even though it received 300,000 more votes than the Labour Parry. In the latest British election John Major's Conservative Party retained a majority in the House of Commons, even though a majority of British voters cast votes for parties other than the Conservatives.

Similar distortions exist in the United States. In 1990 the Republican Party won 45 percent of the popular vote but was reduced to 38 percent of the seats in the House. The Democrats, with 53 percent of the popular vote, received 61 percent of the seats. As in Britain, a strong third-party challenge could permit a minority of American voters to elect a majority of members of Congress. Nothing remotely comparable to these distortions is possible under proportional representation.

Another advantage of PR is the way it makes gerrymandering difficult or impossible. In multimember districts every party or voting bloc will be represented more or less in proportion to its strength in the entire electorate, regardless of how the district lines are drawn. It is only in plurality systems, in which an area of several blocks may make the difference between losing everything and winning everything by a few percentage points, that there is a strong incentive to gerrymander.

Along with partisan gerrymandering, today's government-mandated racial gerrymandering could be eliminated by PR without curtailing the voting power of ethnic minorities. Federal courts have gone from striking down a "strangely irregular twenty-eight-sided" district drawn to prevent black voters from pooling their strength to requiring the creation of equally strange districts to encourage the election of black candidates. Under PR, blacks and Hispanics would find it much easier to elect candidates of their own ethnic group--if they chose. But they would not be maneuvered into such a choice by being electorally ghettoized in safe "minority" districts. Other ethnic minorities, who do not receive preferential gerrymandering, would benefit as well. For example, the federal courts have recognized the right of blacks to have districts redesigned to their benefit--but not the right of Hasidim. PR would eliminate the need for heavy-handed efforts that force some electoral minorities to waste their votes while artificially magnifying the weight of other minority votes. At the same time, PR would increase the power of all minorities--ethnic, religious, ideological, economic, blacks and Hasidim--to elect representatives to Congress, on whatever grounds they chose.

Proportional representation has an additional advantage, insofar as it permits the election of talented or distinguished persons who can get a minority of the vote in a district but who disdain to indulge in the vulgar exaggeration and false promises necessary to win a majority. Noting that "the highly cultivated members of the community" find it difficult to be elected under a winner-take-all system, John Stuart Mill wrote,

"Had a plan like Mr. Hare's [for proportional representation] by good fortune suggested itself to the enlightened and patriotic founders of the American Republic, the Federal and State Assemblies would have contained many of these distinguished men, and democracy would have been spared its greatest reproach and one of its most formidable evils."

PRACTICALITIES

What would it take to install proportional representation for members of the House of Representatives? Nothing more than an act of Congress. No constitutional amendment is necessary. Article I, Section 4 of the Constitution states, "The Times, Places and Manner of holding Elections for Senators and Representatives shall be prescribed in each State by the Legislature thereof; but the Congress may at any time by Law make or alter such Regulations, except as to the Places of chusing Senators." This gives Congress the power to preempt all state electoral laws and to mandate a system of multimember districts with proportional representation for electing members of the House of Representatives.

Because the idea of PR is to give electoral representation not only to parties or candidates that win by slight pluralities or majorities but also to second-, third-, and fourth-place winners, it requires multimember districts. The best way to adapt PR to American conditions would be to consolidate present contiguous congressional districts into multimember districts with some manageable number of members--seven, say, or nine. The larger the number elected from a multimember district, the smaller the number of voters needed to elect at least one representative to the House.

By promoting a close correspondence between groups of voters and groups represented in Congress, PR minimizes the waste of votes that occurs under our present system. For example, the same hypothetical voting population that, divided into nine single-member districts, now returns nine Democrats, each with a slight plurality in his district, might, if organized into a nine-member district with PR, send four Democrats, three Republicans, one Libertarian, and one Conservative to Congress--WITH THE SAME VOTES CAST. By the same token, Democrats who today waste their votes in largely Republican areas might be able to elect a minority of a consolidated multimember district's congressional delegation. Many voters who today are resigned to never electing a congressman of their party or their philosophy, simply because they happen to belong to permanent electoral minorities in their local communities, would suddenly be able to help elect one or more representatives, without changing either their residence or their views.

Multimember districts are nothing new in the United States. Combined with both plurality elections and PR, they have long been used at the state and municipal levels. For example, from 1870 to 1980 the cumulative vote within three-member districts (similar to PR) was used to elect legislators to the Illinois Assembly. As early as 1867 Senator Charles Buckalew proposed that cumulative voting for House elections be imposed by Congress on reconstructed southern states seeking readmission to the Union; following this experiment, either cumulative voting or proportional representation (which he thought too complex for the Americans of his day to comprehend!) could be adopted nationwide. That his proposed reform got nowhere was a tragedy not only for the black, white Republican, and Populist electoral minorities in the South, who were soon effectively disenfranchised by the dominant Democrats, but for the country as a whole, which might have adopted PR by the turn of the century, as many European democracies had.

What about the states that are so sparsely populated that they have only one congressman apiece? PR works only when districts elect multiple members--at least two or three, and preferably more than four or five. At present six states have such small populations that they have only one representative apiece (Alaska, Delaware, North Dakota, South Dakota, Vermont, and Wyoming). Perhaps single-member states would elect their representatives under a plurality system. Still, it seems unfair not to allow their citizens the benefits of PR. The fairest approach would be to increase the number of representatives to give these states more representatives.

These proposals may seem radical. They are. Established institutions need be accorded respect only to the degree that they have earned it. In this case it is difficult to argue with the political scientist Theodore Lowi that "nothing about the present American party system warrants the respect it receives."

Discontent is manifested not only by third-party challenges but also by split-ticket voting, a practice that expresses the inability of many Americans to identify completely with either national party. In 1988, for example, a quarter of the Ohio voters who voted for George Bush also voted for the liberal Democratic senator Howard Metzenbaum. Split-ticket voting like this has produced today's deadlocked government. The blame lies with a party system that does not allow the preferences of substantial numbers of Americans to be expressed.

Because of our peculiar electoral law, the U.S. government is divided between two parties. The American people are not.

THE CUMULATIVE VOTE AND CONGRESSIONAL LEADERSHIP

The principal behind PR--the representation of electoral minorities as minorities in the government--would be thwarted if it were not carried through to the selection of the leaders of the legislature. The instrument to carry it through is the cumulative vote.

The technique of cumulative voting was devised by corporate lawyers to ensure minority-shareholder representation on the board of directors of a corporation. The technique gives a voting bloc a choice, whether to distribute its votes among all the candidates who are to be elected to the board at the same time, or to pool its votes, in order to be sure of electing one or several candidates. The theory is that it is better to be represented by a minority on the board than not to be represented at all.

Cumulative voting could easily be applied to the selection of both the House and the Senate leadership. Under the present system for selecting the speaker of the House and the Senate majority leader, whatever party or faction controls a bare majority of each body can control the leadership and elect all the committee chairmen. Suppose, however, that each house of Congress were run by a leadership committee. Suppose further that each representative or senator did not have to cast one vote for each committee position but could cast all his or her votes for one candidate. Minorities within each house could then elect members to the leadership committees.

Why should a party that controls 51 percent of the seats in a legislature control 100 percent of the chairmanships of legislative committees? Why should 49 percent of the population go unrepresented in the congressional leadership, merely because they are 49 rather than 51 percent? Why should there be a revolution in committee leadership in a closely divided House if, say, half a dozen districts return representatives belonging to the former minority party?

Life is full of crude approximations, and the ancient convention that treats a 51 percent majority as a surrogate for the whole might be tolerable if there were not an alternative. But there is: the cumulative vote.

EXTREMISM--OR CONSENSUS?

Compared with proportional representation, the Anglo-American plurality method is so manifestly unfair that it must be justified in terms of other values. Most of the criticism of PR is directed at multiparty systems, which PR is thought--correctly--to promote. Plurality systems tend to encourage two parties to dominate, while PR systems generally reward the formation of small parties.

Some claim that the multiparty democracy encouraged by PR is "undemocratic," because coalition-building is undertaken in the legislature by minority parties, rather than within broad, would-be-majority parties before winner-take-all elections. Insofar as smaller parties must make more compromises than large parties in order to get their programs passed, PR is said to deny voters the right to choose between two clearly defined party platforms, one of which has the chance of becoming the platform of a stable legislative majority. This assumes, however, that there is no logrolling in two-party systems. It also assumes that most voters identify wholly with one or the other party. In reality, Americans, to the extent that they vote for parties rather than individual candidates, choose one party over the other because it is the lesser of two evils, not because they fully endorse the national party's positions on all issues. The addition of new parties would increase the chances that voters could find representatives who shared most or all of their values and concerns.

The charge that a multiparty system leads to governmental instability is untrue for a country like the United States, whose government is based on the separation of powers. In multiparty parliamentary regimes, because the Prime Minister and executive officers belong to the parties in the dominant coalition, the entire government may fall if a small party defects. In such a system a small extremist party may be able to exact major concessions on policy, as the price of its joining a coalition. Reportedly, in the spring of 1990, for example, a single Brooklyn rabbi, by virtue of his influence over a tiny religious party, prevented the formation of a legislative coalition in Israel's Knesset. This kind of instability would remain impossible in the United States, where the President would continue to be elected by the plurality system, and would serve for a fixed term no matter what multiparty coalitions formed in the House of Representatives. Israel, now on the way to adopting direct elections for the Prime Minister, is moving away from parliamentary instability toward a system like the one I advocate, combining the separation of powers with PR for elections to a popular legislative chamber.

NEW PARTIES

It is anyone's guess what the parties in a House elected by PR would be. (The Senate would continue to be elected the way it is now, and such effect as there would be on it would be indirect.) Although voters might continue to identify with the two major parties for a time, dissidents would soon learn that third-party votes were not wasted.

The largest party in the House might well be the Republicans. As it is, their party, with its homogeneous and stable group of core voters and centralized, disciplined organization, is far more like a European party than the Democratic Party is. If PR were adopted, the Republicans might lose their right wing to a new conservative party or parties, but the number of right-wing Republican voters is fairly small (as the Patrick Buchanan campaign unintentionally demonstrated, by adding only single-digit figures to large protest votes). The Republicans might compensate for their loss by becoming a more consistently classical-liberal party--pro-business, pro-choice--and attracting fiscally conservative social liberals who now identify with a Democrat like Paul Tsongas. Such a neoliberal Republican Party might hold steady at 35 to 40 percent of the House.

The Democratic Party, an incoherent coalition of smaller proto-parties, lobbies, interest groups, and machines, which are brought together only by the winner-take-all logic of our electoral system, would, however, probably disintegrate. The breakup of the Democratic Party as the result of PR would not mean that the power of today's Democratic voters would decline. On the contrary, the kind of moderate Democrats represented by the Democratic Leadership Council, freed from the electoral necessity of appeasing ethnic and liberal lobbies, might well prosper. Together with "Reagan Democrats" wooed back from the Republican presidential coalition, moderate Democrats in Congress might form a Populist Party equivalent to Christian Democratic parties in Europe. Slightly right of center on social issues and foreign policy, slightly left of center on middle-class benefits, such a party could be expected to draw substantial support from Northeastern ethnics, middle-and lower-middle-class whites and Hispanics in the South and West, and perhaps socially conservative blacks. It would be heavily Catholic. A Populist Party might be the nearest rival to the Republicans for the status of largest party in the House.

The United States, unlike Europe, probably would not have a strong Social Democratic Party, given the low level of unionization and the lack of a mainstream socialist intellectual tradition here. There might nevertheless be something that called itself the Social Democratic Party, representing unions, farmers, public-sector employees. Heavily black and Hispanic, such a party would favor protectionism and government subsidies to industry.

A small American Green Party would almost certainly arise from the decomposition of the Democrats. Appealing to New Age environmentalists, pacifists, feminists, and gay-rights activists, such a party might have trouble winning five percent of the vote in successive elections. So might other fringe parties that are easy to imagine: the Conservatives, a far right-fundamentalist alliance; the Multicultural Coalition, a coalition of ethnic-separatist parties; and anti-tax Libertarians.

Carrying this speculative exercise one step further, we might assign percentages of House membership to these hypothetical parties. Based on European experience and American political subcultures, the pattern might be as follows: Republicans (40 percent), Populists (30 percent), Social Democrats (15 percent), Greens (5 percent), Conservatives (5 percent), Multiculturalists and Libertarians (5 percent between them).

One question that immediately comes to mind is, Would a multiparty system exaggerate the power of small, extremist parties in the House? The danger that such parties might hold the key to House leadership would be eliminated by the leadership reforms proposed above. If the position of speaker of the House rotated among the members of a leadership committee elected by cumulative vote, moderate parties would be assured of exercising power in their turn and having substantial weight on the leadership committees, without having to make any compromises with extremists.

The chances that tiny fringe parties would win significant representation would be limited by the division of the country into multimember districts. The danger of extremism could also be guarded against by setting a threshold for participation in the legislature. Germany's requirement that parties get five percent of the nationwide vote, for example, has been enough to exclude both the neo-Nazi Republicans and the radical-left Greens. Anyone who argues for a higher threshold should explain just which 10 or 20 percent of his fellow citizens are so dangerous that they should be effectively disenfranchised.

It is extremely doubtful that PR could bring black-nationalist or white-supremacist parties to power in the House. Since blacks constitute only 12 percent of the population, and an even smaller proportion of the electorate, most black voters would have to support an extremist party for it to clear a five-percent threshold. It seems unlikely that they would, because more-moderate multiracial parties would more accurately reflect the economic concerns and values of most black voters. Even if a black-nationalist party passed the five-percent hurdle, it could be kept out of the House leadership. To elect one member to a five-member leadership committee, at least 17 percent of the House membership would be necessary. Black radicals would have to form coalitions with non-blacks.

What about white Klansmen and neo-Nazis? Right-wing nationalist parties have made gains in recent European elections, but these are protest votes by moderate voters who seek to change particular policies of centrist parties, not to replace those parties permanently with extremists. In the United States the support for Buchanan's extreme conservatism, even in the Republican Party, appears to have been in the single digits, and support for David Duke-style racism vanishingly small. This suggests that the only viable far-right parties in a multiparty House would be Pat Robertson-style Christian conservatives, not Aryan Nation skinheads.

Those who claim that PR was responsible for the electoral successes of the National Socialist party in Weimar Germany are mistaken, according to no less an authority than Hermann Goering. Enid Lakeman, a British advocate of PR, writes of Weimar Germany that "if the British electoral system had been in use [the Nazi party] might have been grossly over-represented....Goering said at his trial that under the British system that [1933] election would have given the Nazis every seat in the country, and he cannot have been far wrong." Hitler would have loved the Anglo-American electoral system.

For some, of course, a system that gives any influence to "extremists" like conservative Christians and social democrats is bad. Even if this intolerant position could be reconciled with a commitment to democracy in America, would PR be worse than the plurality system in this respect? Under our present system small but intensely motivated ideological minorities have the power to force moderate Republicans and moderate Democrats to placate them in order to get the small margin of voters sometimes necessary for election in winner-take-all races. In a multiparty system such intense minorities might gain direct representation in Congress but would lose their ability to dictate centrist-party platforms. To be sure, on particular measures large, centrist parties might need small-party help, but they would not need to ally themselves with small parties in order to win election in the first place. In the shifting coalitions formed to pass different bills, the larger, moderate parties would have the upper hand.

In addition to ideological extremism, today's two-party American system promotes "extremism of the center." This results from the need for the two national parties to exaggerate their differences in order to appeal to the crucial minority of white middle-class swing voters. These voters--Wallace-ites and Reagan Democrats--do not identify completely with either national party. Since 1968 they have elected primarily Republican Presidents who share their conservative cultural and foreign-policy values, and Democratic congressmen who protect their middle-class transfers (Social Security, Medicaid, student loans). The penchant of these "populists" for ticket-splitting not only exaggerates the indirect power of this group but also encourages Democrats and Republicans alike to pander to their worst instincts. The Democrats try to appeal to the populist swing vote with giveaways; the Republicans play on racial resentment and call for military strength. If populists had their own party in Congress, their leaders would presumably have to be more responsible and measured in their views than are swing voters whose prejudices and appetites are stoked and exploited by both of today's national parties. As long as there is ticket-splitting, extremism of this kind will be built into our present system--and will corrupt it. A case can be made, then, that PR is just as likely to reduce the political power of extremist minorities as to enhance it.

TRANSFORMING AMERICAN POLITICS

Of course, a little instability might help end the trench warfare in American politics. By multiplying the possibilities for coalitions, a PR-produced multiparty system could introduce movement and change where there is now only partisan maneuvering on the margins of deadlock. To use the parties in my example, a legislative coalition of Republicans, Libertarians, and Conservatives might be able to cap runaway spending and put through a bill enabling parents to choose the schools that their children will attend. Populists might be able to attract Social Democrats and Conservatives whose lower-middle-class constituents were dependent on federal transfer payments into an alliance on the issue of comprehensive health-care reform. Small-town Christian Conservatives and rural Greens might find themselves allied against environmentally disruptive development projects.

The formation of multiple parties could transform American politics in other ways. Although only members of the House of Representatives would be elected by PR, parties in the House would soon begin to run candidates for the Senate as well, so that the Senate, too, might be divided among several parties offering a choice, not an echo.

The effects on presidential elections could be even more interesting, because of the constitutional provision that if no presidential candidate wins a majority of the electoral-college vote, the choice will be left to the House of Representatives. "The presence of a third important party capable of obtaining seats in the House of Representatives and a few electoral votes would hardly throw every presidential election into the House of Representatives," Theodore Lowi has written. "Its presence WOULD, however, force each of the candidates for the nominations of the two major parties to look to the House of Representatives as the place where the real election MIGHT TAKE PLACE. This would transform the presidency because Congress would become the president's direct constituency." The possibility that a splitting of the presidential vote among several candidates would give the House the opportunity to choose might deter Presidents from demagogically "running against Congress," as have Jimmy Carter, Ronald Reagan, and George Bush.

A much more likely scenario than elections thrown into the House is the formation of multiparty alliances to elect presidential candidates. Not only might populists have the dominant party in the House, but also they might be the dominant partner in winning presidential alliances. No longer would populists be tempted to split their votes between Republican Presidents and Democratic Congresses. Divided government could come to an end.

Such a system might create incentives for interbranch cooperation, rather than confrontation. Since Presidents would tend to be elected by the same multiparty coalition that dominated voting in the House, they would be inclined to view major congressmen as allies, not enemies. At the same time, the multiplication of parties might increase presidential influence in Congress, because it would be difficult for centrist majorities in the House to entice Greens, radical Christians, and other small parties into the supermajorities needed to override presidential vetoes. Presidents who found it easier to organize legislative coalitions, in turn, would be less tempted to circumvent Congress and undermine constitutional democracy by bureaucratic policy changes in the name of "inherent executive power."

There might even be a parallel to the "era of good feelings" that, in hindsight, can be seen to have existed from 1936 to 1968. Contrast the accomplishments of that era--the winning of the Second World War, the Marshall Plan, NATO, GATT, the GI Bill, interstate highways, public education, the Civil Rights Act--with the dissension, deadlock, and deficits of the period from 1968 to the present. Contributing to the nonpartisan achievements of that earlier period might have been the fact that the United States had a de facto three-party system. From their rebellion against FDR's court-packing scheme until their downfall as the result of the civil-rights movement, conservative southern Democrats in Congress functioned in effect as a third party, voting almost as often with the Republicans as with liberal Democrats. Their power had evil roots, to be sure, in black disenfranchisement; recognizing this, one can also recognize that they played a moderating role in American politics, checking the excesses of left-wing New Dealers and isolationist Republicans alike. If something like the post-1968 two-party system had existed from the Depression until the 1960s, with a deep division between left-wing New Dealers and far-right Republicans combined with nearly constant control of the presidency and Congress by opposing parties, then America might have been far less stable and successful both abroad and at home.

The Founding Fathers did not think that the long development of democratic political forms in America and the world had come to an end with their work. Neither should we. America was once the laboratory of democracy. If the alliance of tradition and cynicism against democratic reform of Congress prevails, America may become democracy's museum instead.