The Quest for an Unproblematic Selection Process

On the surface, the prospect of a House contingency election might not strike people as a major concern. However, the Framers of the Constitution hoped the presidential selection process would serve several functions, from controlling the ambitions of candidates to encouraging the selection of able individuals (Ceaser 1979, 10–27). One of their concerns was that the selection process be unproblematic—able, in Ceaser's words, “to produce a winner without confusion or delay” (19). While it is certainly true that the selection process has not evolved as its designers intended, the goal of an unproblematic process is still valuable. In fact, some of the reforms of the presidential selection process are at least partially attributable to the desire of political leaders to forge a less problematic system. For example, the original dual vote mechanism created great strategic difficulties for electors, leading to the debacle that was the 1800 election, when Thomas Jefferson and Aaron Burr, members of the same party, tied in electoral votes, thus throwing the election into the House. By establishing distinct ballots, the Twelfth Amendment solved that problem. Similarly, the absence of a strong party system led to the rise of numerous personal factions and regional candidates in 1824, preventing any single candidate from winning the necessary majority of electoral votes, once again throwing the election into the House. The development of a strong two-party system—defended by Martin Van Buren in part because of the 1824 experience —served to limit the field of candidates, making a House contingency election unlikely (Ceaser 1979, 103–106, 143–146; McCormick 1982, 64–69, 82–88, 117–120; Busch 2001, 33–36).

Unlikely does not mean impossible. Whenever third party movements become powerful enough to claim electoral votes the nation flirts with House contingency elections, and the experiences of 1800 and 1824 should give pause to anyone who desires a clear and unproblematic process. Even supporters of the electoral vote system agree that the contingency election opens up the possibility of great mischief, and should be avoided (Best 1996, 13–14; Schumaker and Loomis 2002, 199–202). We can get an indication of what might occur in a contemporary House contingency election by examining my home state of Texas. In my hypothetical 2000 example, the Texas House delegation would have had to choose which candidate —Gore or Bush—to support with its single vote. In a state that voted for Bush over Gore 59-38%, one might think this would have been a proverbial no-brainer. However, in January 2001, Democrats controlled Texas's House delegation, 17-13. The cross-pressures are fascinating to contemplate. Texas Democrats might have opted to support Bush because he was their governor, or because he won the state popular vote. They might have opted to vote according to their district results, which would also have placed the state in Bush's column. On the other hand, Texas Democrats might have voted along strict party lines, basing their argument on the national will of the people, since Gore won more popular votes nationwide. In any case, Republicans would have had to convince three Democrats to vote for Bush in order for the state to align with its popular vote in a House contingency election, and there are numerous constitutional arguments Texas Democrats could have used to resist that ploy. No doubt similar dynamics would have been in play in other states. Raw political considerations would have been masked as arguments based on constitutional principle, and the result would have made the Clinton impeachment experience look like a tea party.

Judith Best argues that the magnifier effect of the current electoral vote system reduces the threat of a contingency election to “no real problem” (1996, 14), and it is quite true that the winnertake-all system makes it prohibitively difficult for third parties to throw the election into the House. Again, however, the even number of votes currently at stake makes a tie possible, a problem no magnifier effect can eliminate.

Political leaders understand the utility of using odd numbers in decisionmaking. For example, the Constitution calls for the vice president to cast tiebreaking votes in the Senate, by its very nature an even-numbered body. Constitutional convention delegates recognized the need for a tie-breaker in the Senate, and Alexander Hamilton specifically acknowledged the importance of a such a role in Federalist 68, pointing out the need for the vice president “to secure at all times the possibility of a definitive resolution of the body.” Supreme Court Justice Joseph Story, in his Commentaries on the Constitution of the United States, also argued that the absence of a tiebreaker would lead to “indecision and inconvenience … very prejudicial to the public interests” (Madison 1966, 596; Hamilton et al 1961, 383; Story 1987, 268–270). Although there is no comparable discussion in these sources about the precise number of legislators in the House, it is interesting to note that the lower chamber was originally set at 65 by the constitutional convention. In the 13 apportionments from 1790 until its size was fixed at 435, only three were set at an even number (Congressional Quarterly 2000, 892–899). The utility of avoiding the feuds, intrigues, and agitations feared by the Framers makes even mundane issues like odd and even numbers important considerations. “Definitive resolutions” are at least as important in presidential elections as they are in Senate votes.

The Political and Constitutional Problem

The only way to solve this problem is to do something about the constitutional dynamic created by the Twentythird Amendment. Here there are several options. The obvious solution is to make Washington, D.C., a state. The Senate would then have 102 members, the House would stay at 435, and the electoral vote total would drop to 537, making 269 the magic number for presidential candidates. The problem with this solution is that it is highly unlikely, politically and constitutionally. Politically, the two major parties are split on the issue of statehood for the Federal District. Recent Democratic Party platforms have supported statehood, while Republican Party platforms have opposed it (Harris 1995, 214–215). There is simply no political reason for the GOP to support statehood for a city that is so uniformly Democratic. District voters gave 85% of their support to Al Gore in the 2000 presidential race, a level that dwarfs even the next highest Democratic total (Ceaser and Busch 2001, 266–268). Statehood would immediately add two more liberal Democrats to a tightly split Senate, and reduce the representation in the House of a possibly Republican state to accommodate Washington's overwhelmingly Democratic population. This question of diluting state representation is not solely a Republican concern. The best opportunity for District statehood came in 1993, when a Democratic president, Bill Clinton, supported the creation of the state of New Columbia. While all but one Republican opposed the House measure supporting statehood, 40% of Democrats helped kill the effort (Harris 1995, 231–232; Meyers 1996, 5). Opposition to statehood owes as much to a nonpartisan suspicion that the city cannot manage itself as it does to partisan motivations.

Constitutionally, statehood for the District also has problems. Article One, Section Eight, of the Constitution allowed for the creation of a federal district through the cession of land from particular states, and it would seem that statehood for that federal district would require a constitutional amendment. Some scholars argue that, at the very least, the Twenty-third Amendment would have to be repealed, while others rest their argument on the original intent of the Framers, who desired a federal capital free of state interference (Best 1984, 64–77; Meyers 1996, 77–80; Harris 1995, 223–228). Whatever the relative merits of the constitutional arguments, there is sufficient partisan support for the arguments opposing statehood to make the matter moot. Even the argument that the District is bigger than many states—a common argument as recently as the early 1990s—is losing strength. The District of Columbia now ranks number 50 of 51 entities in population, beating only Wyoming. It was the only entity that actually lost population from 1990 to 2000. Its population has been declining since 1950, and if current trends continue it may move into last place by 2010.

A second option would be to grant to the District what is called nominal statehood. This proposal would give Washington, D.C., full voting representation in Congress, while retaining its special status as a federal district. Here, the electoral vote change would be identical to that described above for full statehood. Again, however, there are significant political and constitutional obstacles. Best argues that the proposal violates equal suffrage in the Senate, thus requiring unanimous ratification of any constitutional amendment (1984, 43–60). However, it seems unlikely that a constitutional amendment granting nominal statehood for the District could pass even the normal constitutional constraints, much less Best's higher standard. In 1978, Congress mustered up the necessary twothirds support in each chamber for a constitutional amendment granting nominal statehood for the District. The amendment then suffered a disaster at the state level. Only 16 of the necessary 38 states— all controlled by Democratic legislatures —ratified it in seven years. The same partisan and constitutional objections that manifested themselves 15 years later began here. The effort died in 1985 (Meyers 1996, 35, 85–86; Harris 1995, 206–207; Kyvig 1996, 420–423).

Some scholars and political leaders have proposed granting District residents the right to vote in Maryland congressional elections, while still maintaining the District's special status. While this plan draws at best a lukewarm response in some quarters, from the perspective of presidential elections it seems clear that such a measure would require repealing the Twenty-third Amendment, hardly a sure thing. Such a move would bring the total electoral vote count back to 535, with 268 being the magic number (Meyers 1996, 87–88; Harris 1995, 204–206; Raskin 1990, 438–440). Others argue that the majority of the District should be retroceded to Maryland, much as Virginia's portion of the District was in 1846. Scholars who oppose statehood on constitutional grounds tend to support retrocession, though even this move would require repealing of the Twenty-third Amendment. Otherwise, we might face the prospect of the few inhabitants of the federal enclave—the three thousand acres of monumental Washington—retaining three electoral votes (Best 1984, 77–83; Kyvig 1996, 424). In addition to the amendment barrier, there is apparently a great resistance on the part of Maryland to participating in such a scheme. State Republicans do not relish adding numerous Democratic voters to the state, and leaders from both parties share a reluctance to saddle Maryland with Washington's problems. In a 1990 survey of Maryland state legislators, an overwhelming percentage of respondents—82% of delegates and 92% of senators—said they would reject retrocession (Meyers 1996, 86–87; Harris 1995, 202–204; Kyvig 1996, 420; Schrag 1990, 319–320).

It will do no good to add other states to the Union, such as Puerto Rico. The Senate would remain an even-numbered body, and the House would stay at 435 after reapportionment. Making a territory a state without changing the status of the federal district would simply give us 540 electoral votes—still an even number. The problem is the Twentythird Amendment, and its addition of an odd number of electoral votes to an already-odd number of electoral votes. Only a constitutional amendment will change the status of the Federal District, and the necessary votes in Congress and the states simply do not exist.

The Legislative Solution

Considering the unlikelihood of what we might call the “Article V solutions” to this problem—those that require constitutional amendments—simple political prudence dictates another course of action. There is, in fact, a legislative solution that does not carry the political baggage of D.C. statehood, and might be sufficiently uncontroversial to get through Congress. The solution is to add one representative to the House, giving it a membership of 436. That would give us 539 electoral votes and no possibility of a tie in a two-person presidential election, barring faithless electors. The magic number would still be 270. Since it is federal law that limits membership in the House to 435, federal law can change that number. While this would have been an ideal move in a reapportioning year, it can still be easily accomplished before the 2004 presidential election. It seems likely that the new 436th House seat would go to Utah, which missed gaining a seat in the 2000 census reapportionment by just 856 people (Giroux 2001, 39).

The drawback to this plan is that, while we would now have an odd number of electoral votes, we would also have an even number of representatives in the House, and there is no outside officer to cast tie-breaking votes in the House as there is in the Senate. Clearly, the potential for deadlock in the House would exist. This downside is easily exaggerated, however. Senate membership tends to be a fairly stable constant, since state governors can quickly fill any vacancies. When Senators John Chafee and Paul Coverdell died in office, their seats were filled within a week. By contrast, membership in the House is often quite fluid. Vacancies occur due to death or resignation, sometimes remaining unfilled for many months, giving the chamber a fluctuating membership that, in most cases, is not as critical as similar instances in the smaller Senate.

To explore the issue of number in the House, I examined three years of Congressional Quarterly Weekly magazines, looking specifically at the “Datafile” box toward the beginning of each issue. This box lists the number of Republicans, Democrats, and independents in each chamber of Congress, as well as vacancies. It also explains each vacancy, and announces when the vacancies are filled. I looked at both sessions of the 106th Congress (1999 and 2000) and the first session of the 107th Congress (2001). During that three-year period, there were eight deaths and five resignations in the House of Representatives. Over the course of 158 weeks, 88 saw the House with an odd number of members. Seventy weeks saw the House with an even number. Only about one-third of the time did the House have its full complement of 435 members. Within each congressional session the fluctuation was more dramatic. In 1999, for example, the House was comprised of an even number of members twice as often as an odd number, 37 weeks to 17. In 2000, membership stayed at an odd number for 41 weeks, while in 2001 the House had an odd number for only 30 weeks. As of this writing, the House in 2002 has stayed at a constant 435—for the first time in four years. By contrast, in that same surveyed period the Senate slipped from its membership of one hundred for only two weeks. Someone may argue that the fluctuation in the House is no larger in percentage terms than that in the Senate. However, the issue is not about fluctuation in percentage terms—it is about decision-making. The Senate secures its ability to break ties with the vice president's vote. The House has no such mechanism, but its history indicates that the odd-or-even question is not nearly as important there as it is elsewhere.

So, we are faced with a choice: allow the House to retain its odd-numbered membership, which raises the specter of an Electoral College tie leading to a House contingency election, or eliminate the possibility of an Electoral College tie by slightly changing the membership in the House. Should the political dynamic someday favor a more permanent solution to the problem of representation for the District, Congress can always pass a law moving House membership back to 435. Some will no doubt argue that the once-a-century election ride we experienced in 2000 hardly warrants such a step. It is interesting to note that our current era of a closely divided electorate in which neither presidential candidate obtains a majority of the popular vote—for the third time now since 1992—looks not dissimilar to the late 1800s. At that time, the disputed election of 1876 began a series of five exceptionally close presidential elections. Those elections would never have been thrown into the House because of the odd number of electoral votes. We face a similar dynamic today, but without that numerical protection.