America's Unwritten Constitution: The Precedents and Principles We Live By (67 page)

Most of these statutes might at first seem easy to evade. Formally, a clever
president is free to propose an appointee who is only nominally connected to (or independent of) a given party. In practice, however, opposition-party senators are often able to deter any sly evasions that presidents might envision.
¹⁷

The law of bipartisan replenishment of independent agencies is thus enforced through congressional party politics, which in turn are shaped by bipartisan rules and laws regulating the structure of Congress and the election of congresspersons. For a century and a half, Congress has been dominated by the same two major parties, Republicans and Democrats, who have alternated in power. In an intricate meshwork of statutes, house rules, and customary practices, this two-party system has threaded itself into durable formulas and folkways determining how many seats and staff-ers each party will have on each house committee, how party leaders will interact on committees and on the floor, and so on. These formulas and folkways both presuppose and help ensure that at any given moment, each house of Congress will consist of two main groups—one “majority” party and one “minority” party.
¹⁸

This basic dichotomy is visible in the very architecture of each house chamber, with members of one party traditionally sitting together on one side of the chamber and members of the other party likewise clustered on the other side, with an aisle literally and metaphorically separating the two groups. This remarkably steady, stable, stolid two-party system stands in sharp contrast to the kaleidoscopic arrangements one sees in many leading democracies around the globe, where three or more parties routinely win a significant number of seats in the national legislature, new parties arise with some frequency, and major parties occasionally collapse.

AMERICAN ELECTION LAW
has created conducive conditions for this entrenched Republican-Democrat duopoly. The cornerstone of this legal foundation is a simple rule in the United States Code that disaggregates each state into single-member congressional districts. For example, if a state is entitled to twelve seats in the House of Representatives, it must have twelve congressional districts, each of which picks one House member. The state may not divide itself into, say, three districts, each of which elects four House members. Nor may the state create a system in which all
twelve seats are filled in a single statewide election, with each voter allowed to cast one vote. Each district must have one and only one representative, and in this system it is hard for more than two parties in any district to thrive in long-term equilibrium.

Students of political science will recognize this empirical regularity as “Duverger’s Law.” The basic mechanisms driving the regularity identified by Professor Duverger are not hard to understand. When one and only one seat is up for grabs in a given congressional district, the victor must win a majority, or at least a plurality, within this district. Most sophisticated voters understand that a vote for a third-party candidate is usually a wasted vote, for several reasons.
¹⁹

First, a start-up party, by definition, has no track record of past victory—only the two established parties can claim such a track record—and the need to achieve a district-wide majority or plurality sets the bar of success quite high for the start-up entrant. Second, even a voter who sincerely prefers a third-party candidate should understand that a vote for this candidate, instead of a vote for the lesser-evil of the two major-party candidates, may increase the odds that the voter’s least favorite (greater-evil) major-party candidate will prevail.
^*
Third, even if these two factors do not initially sway a given voter, that voter should understand that these factors will likely sway other sophisticated voters, and that fact, in turn, provides an additional reason to think that a third-party vote would be a wasted or perverse vote. Thanks to the dynamic and reinforcing interplay of these three factors, the prophecy that the two established parties will continue to be the only realistic options becomes largely self-fulfilling—a stable equilibrium that can be disrupted only by a massive external jolt, the political equivalent of an asteroid strike.

Separation of powers and federalism further reinforce this equilibrium. Because America has one and only one president, and because this one-man
executive is elected independently of Congress, Duverger’s Law predicts that two and only two major presidential parties will survive in the long run. A similar dynamic ensures that within any given state two parties will routinely vie for gubernatorial control. The great majority of state legislatures also use single-member districting systems. The fact that virtually every congressional district has the
same
two parties today—Republicans and Democrats—is a product of certain electoral economies of scale and the ticket-system facilitated by the tight coordination of state and national elections.

In a multi-member districting system, by contrast, more parties may be able to thrive if friendly voting rules are in place. Imagine a district with four seats to be filled, and suppose further that each voter is given only one vote. In this system, any party with slightly over 20 percent of the vote is guaranteed to win a seat. (If Party X has more than 20 percent, it must necessarily be among the top four parties; if there were four other parties, each of which got more votes than Party X, the total votes would exceed 100 percent—a mathematical impossibility.) The victory threshold is thus much lower than a district-wide majority or plurality. In such a multi-member district, as many as five different parties could survive in long-term equilibrium, in a musical-chairs game with, say, three parties at 21 percent apiece and two slightly smaller parties credibly vying for the last remaining chair. More generally, any multi-member district with
n
seats and friendly voting rules can sustain, in the long run, as many as
n
+ 1 parties. Multi-member district systems and variants thereof are used throughout the world, and many national elections abroad feature robust multiparty competitions year after year.

American congressional elections, by contrast, narrow the playing field to two major parties. In any given district, there is one seat to be won in the musical-chairs game, and thus in the general election there are typically only two contestants—one Democrat and one Republican—plausibly vying for the seat. (We should recall here the critical winnowing role of primary elections and/or party caucuses, which enable the typical general election to become a focused contest between one Republican and one Democrat.)

The federal law that laid the foundation for this two-party congressional system was first enacted in 1842, pursuant to Article I, section 4,
which authorizes Congress to legislate rules governing the “manner” of congressional elections. According to this statute, House members were to be elected “by districts composed of contiguous territory equal in number to the number of Representatives to which said State may be entitled,
no one district electing more than one Representative
.” Over the ensuing century, Congress repeatedly revisited its election laws in connection with the decennial House reapportionment mandated by Article I, section 2. Most of the time, Congress reenacted the 1842 statute or some close cousin, but occasionally Congress allowed the single-member-district law to lapse, only to revive the law in a later reapportionment cycle. Since 1967, the single-member-district statute has been a fixed feature of the U.S. election code, a politically entrenched and politically entrenching provision cementing in place the current two-party system about as effectively and enduringly as any explicit constitutional text could ever hope to do.
²⁰

To be clear: There is no constitutional text that explicitly or implicitly requires a two-party system. Nor is there any constitutional text that explicitly or implicitly requires single-member districts, with or without the gloss of the past 170 years of actual government practice. Congress has the power to create single-member House districts, but not the duty to do so. Nothing in the 1842 law, or the 1967 law, or any of the districting laws in between invites us to read any specific constitutional clause in a manner that suggests that House members must be elected this way or that way. Rather, this long string of laws merely confirms that Congress can choose to require single-member House districts if it wants to.

But that is enough, for once single-member districts took root and solidified a two-party system in Congress, Congress lost almost all incentive to change the basic structure. Thus, this structure became deeply entrenched—a self-perpetuating element of America’s unwritten Constitution far harder to change than, say, the current size of the Supreme Court. To change the Court size, Congress need only pass an ordinary law, and it is actually imaginable that some future Congress might wish to do this—imaginable precisely because such a change in law would not necessarily harm congressional incumbents in any obvious way. To change the current two-party system, Congress likewise need only pass an ordinary law, but it is almost unimaginable that a future Congress might wish to do this—almost unimaginable because such a change would indeed harm
incumbents by opening the door to new parties that might threaten the existing duopoly enjoyed by the two major parties tied to these incumbents.
^*
If ever it makes sense to call something a “constitutional” rule even though this rule does not in fact appear in the written Constitution, America’s basic two-party system is such a thing.

DESPITE ALL THAT WE HAVE
seen, it cannot be said that the Constitution directly addresses political parties in a comprehensive fashion. Is this because, as some scholars have claimed, the document’s rules concerning elections and the political process—especially its provisions governing presidential politics and presidential authority—are the petrified fossils of an eighteenth-century world, wholly ill-fitting the political realities of modern America?

The evidence suggests otherwise. At the very moment that national parties arose, they began to integrate themselves into the Constitution in both text and deed. America’s modern presidency is not the product of eighteenth-century mistakes that later Americans have simply been unable to comprehend or correct. Although the presidency was originally designed for a nonpartisan figure—George Washington—the office was repeatedly redesigned, via many different amendments adopted over the course of many decades, to fit the rise of more partisan chief executives, including Thomas Jefferson, Abraham Lincoln, Franklin Roosevelt, and Lyndon Johnson. Most of the rules of presidential power are robust. These rules first worked without an entrenched two-party system and now work within such a system.

To put the point another way, virtually all states have created governorships that look amazingly like the presidency, and most states created these presidential look-alikes after the rise of America’s two-party system. Almost no state constitution comprehensively regulates political parties, even though many written state constitutions are quite detailed and relatively easy to amend.

All this evidence suggests that there is a different reason why political
parties receive rather spotty treatment in America’s fifty-one written constitutions, state and federal. The explanation, quite simply, is that it is far from clear what a more comprehensive constitutional regulatory framework should look like.

True, several advanced democracies across the oceans feature more detailed constitutional regulations of political parties, but few of these foreign constitutions provide models for easy American emulation. Most foreign regimes lack America’s special combination of an entrenched two-party system and an executive elected independently of the legislature. The fifty state constitutions are more obvious models for possible federal constitutional rules, precisely because all fifty-one constitutions, state and federal, share a great deal in common (including electorally separated branches and two major parties). But, to repeat, virtually no state constitution regulates political parties in dramatically different fashion than does the federal Constitution. Unless and until several state constitutions come along and demonstrate a better mousetrap for addressing American-style political parties, most Americans are unlikely to view the federal Constitution as defective in this regard.

There may well be deep wisdom in America’s piecemeal approach to political parties, which are, after all, multifarious, protean, and complex creatures. Consider just a few of the complications that could confound a proposal to regulate parties in a more truly systematic fashion: America’s two alternating governing parties are qualitatively different from its wide assortment of fringe parties. The boundaries between official political parties and broader social movements are porous. At any instant, official party membership lists will only imperfectly reflect real partisan allegiance and inclination. Sometimes a particular party may be tightly knit and ideologically pure; at other times, less so. A party that opens its primaries to independent voters will likely act differently than one that operates only closed primaries. Other matters of internal party governance—whether the party favors primaries or caucuses, whether it strongly privileges party chieftains or is more open to insurgents, whether it embraces plurality rule or proportional representation in picking delegates to its conventions, whether conventions in turn follow majority rule or supermajority rule in selecting party nominees—introduce additional
wrinkles. Parties often operate differently within Congress than within the executive branch.

In periods of “unitary” government when one party controls both Congress and the presidency, distinctive political possibilities and pathologies arise that are usually dormant in eras of “divided” government, which present their own classic routines and rhythms. Within any chamber of government, the ruling party often confronts opportunities and temptations not faced by the loyal opposition, and vice versa. Partisanship may play out differently in moments of extreme ideological polarization and high-stakes reform—as during the Jeffersonian revolution of the early 1800s or the Reconstruction of the 1860s—than in eras of relative quiescence, when party ideologies are blurred and patronage looms larger than principle.