Law Review
Northwestern Law
Northwestern University Law Review : Colloquy : 2006 : Bennett

A Cautionary Tale:  By One Way of Counting the Democrats' Hold on the New House of Representatives Will Be Very Precarious

By Robert W. Bennett[*]

We all know that the Democrats gained a decisive House of Representatives majority in the 2006 election, so that their party will be comfortably in control of that body in the Congress that will convene in early January of 2007.  As of the time I am writing this, the Democrats have clearly won 231 of the 435 House seats for the new Congress, with five races still undecided.[1]  With a majority of that size, we can say with great assuredness that the Democrats will be able to elect the new Speaker and name Chairs for House committees.  That is the sense in which the Democrats will be in control—a very important sense indeed.  When they stick to their partisan guns, moreover, the Democrats will also be able easily to control final votes on legislation.[2]  

But even with party fidelity, the way the votes are to be taken on one matter would put the Democrats in control by the skin of their teeth.  The change in party affiliation—or temporary defection—of just one vote could deprive the Democrats of that bit of "control."  The matter I have in mind will never come before the House that will convene in January, but the precariousness of the Democrats' control in such a count should provide a warning sign for the future—for the Democrats, to be sure, but more importantly, for all of us.

The vote I have in mind is the House backup procedure for selection of the President of the United States.  The President is almost always chosen (as a formal matter) by a vote of what has come to be known as the "electoral college."  At the present time there are 538 allocated seats in the electoral college, which actually never meets as such, but rather votes in 51 separate smaller gatherings in the states and in the District of Columbia.  In order to prevail, a candidate for President must receive a majority of the "electors appointed."[3]

For a variety of reasons, however, the electoral college balloting may be indecisive, and in that event, the President is chosen by a vote in the House of Representatives.  The Constitution provides that the House vote "shall be taken by states, the representation from each state having one vote;  . . . and a majority of the states shall be necessary to a choice."[4]  In other words, what matters for House "control" over this vote is not a party's total number of representatives, but rather predominance in a majority of state delegations.

It now seems pretty clear that in the new House, the Democrats will control twenty-six delegations, a bare majority.[5]  The Republicans will have majorities in twenty-one delegations.[6]  Three delegations with even numbers of representatives (Arizona, Kansas, and Mississippi) will be in a partisan standoff.[7]

For several reasons, that Democratic majority is precarious, and dangerously so.  The Democratic margin is just one vote in eleven of those twenty-six delegations.[8]  The majority could thus easily evaporate tomorrow, even before the new House convenes.  For deaths and resignations from the House are not at all uncommon.  Over a recent stretch of fifteen years, for instance, there was an average of well over one death per year of a House member.  And in 2004, South Dakota's lone representative, a Republican, resigned and was replaced by a Democrat.  South Dakota will be one of the Democrats' twenty-six states in the new House, and it is just one example of a state that could shift to the Republicans should there be death or resignation[9]  of its representative.  Whether caused by death or resignation, moreover, House vacancies must be filled by special elections, which means that the seats can remain vacant for a good period of time.[10]  The Democrats could thus be deprived of their "majority" even before a replacement was seated, and even if they eventually retained the seat and the delegation.

The problem would be trivial if the lost Democratic majority was replaced by a Republican one, but that might not be the result.  Vacancies and partisan standoffs in delegations will often mean that no party will control the required majority of delegations when a presidential choice is required.  Thus if the Democrats lost just one of their twenty-six delegations, all other party affiliations of representatives remained steady, and all representatives remained faithful to their party in the vote, the result would be a stalemate in a presidential vote.

If a stalemate extended past inauguration day, January 20, the Twentieth Amendment instructs that "the Vice President elect shall act as President until a President shall have qualified."[11]  But if the electoral college vote had been indecisive for President, it likely would have been indecisive for Vice President as well, in which case that choice would have been relegated to the Senate.  The Democrats will also control a majority of the votes in the new Senate, but that depends on the votes of one pretty faithful independent, and another—Connecticut's Joseph Lieberman—whose vote may be not so dependable.[12]  Were Lieberman to abstain or vote with the Republicans, the outgoing Vice President, who serves as President of the Senate, would not be able to break the tie.  Normally he does have a vote when the Senate is "equally divided,"[13] but for the selection of a new Vice President a "majority of the whole number [of Senators] shall be necessary to a choice."[14]  If no new Vice President is chosen either, then a statutory fallback order of "acting" presidents kicks in.[15]

Unfortunately a House that was initially deadlocked but then able to choose a President would be almost as worrisome as one that was not.  For bargaining with the various representatives who could break the jam—in the Democrats' favor, or for that matter, in the Republicans'—would surely be ferocious.  Both the appearance and the reality of unseemly "deals" would be all too likely.  It is entirely possible that a President chosen in this way would have to weather an extended period in which the "legitimacy" of his claim on the office would be hotly disputed.

So how concerned should we be about the prospect of these unsettling scenarios?  That depends in good part on how likely it is that the electoral college balloting will be indecisive and the House will have to choose a President.  There are basically four reasons why there might be no required majority of appointed electors for any single candidate in the electoral college balloting:  a tie vote in an otherwise routine election, electoral votes split among three or more candidates, or abstention or disqualification of one or more electors.  In any given election none of these scenarios is likely to deprive one of the major party candidates of the required electoral college majority, but none can be dismissed as utterly remote.  The House had to choose the President in 1824,[16] and the Senate the Vice President in 1836.[17]  Those elections were long ago, but the 2000 and 2004 elections showed that at least a tie vote in the electoral college is a distinct possibility in today's political environment.

To be sure, even if the choice did reach the House, presidential selection might be both timely and relatively unproblematic.  One party might command a comfortable majority of the delegations, and that could well be the party that also had a majority in the more basic sense.  Or, even if no party controlled twenty-six or more delegations, a consensus might be forged so that any stalemate would prove to be temporary.[18]  The bad news, however, is that the present alignments in the Senate, and especially in the House, provide warning signs that indecisive votes or unseemly deals are entirely possible in those bodies, even if one party seems to be in "control."

In this respect—and indeed in others—there is great potential mischief in some of the obscure details of how we choose our President.  To many of the problems, there are solutions, but for the most part they are not self-executing.  It is my view that we would do well to pay attention now, before it is too late.[19]


* Nathaniel L. Nathanson Professor of Law, Northwestern University School of Law.

1., Elections 2006:  Balance of Power, (last visited Dec. 6, 2006) (link).

2. There are three supermajority requirements for the House made explicit in the Constitution.  A two-thirds vote is required for constitutional amendments, U.S. Const. art. V, to expel a member, id. at art. I, § 5, cl. 2, and to overcome a presidential veto, id. at art. I, § 7, cl. 2.  In addition, there is a debate about whether legislation could require supermajorities for subsequent legislation.  There are also some actions that require less than a majority.  As a matter of practice, moreover, the intricacies of House procedures leave a lot of control in leaders, rather than in majorities.  But the matter to which the text now turns is the only final action where the total number of members of the House voting for a measure is not decisive.

3. U.S. Const. amend. XII.

4. Id.

5., Elections 2006:  House of Representatives / Complete Results, (last visited Dec. 6, 2006) (link).

6. Id.

7. Id.

8. Id.  These delegations are Colorado (4-3); Illinois (10-9); Indiana (5-4); Iowa (3-2); New Jersey (7-6); North Carolina (7-6); North Dakota (1-0); South Dakota (1-0); Tennessee (5-4); Vermont (1-0); and West Virginia (2-1).  Id.

9. Or an expulsion.  See Discussion, supra, at note 2.

10. U.S. Const. art. I, § 2, cl. 4; see Robert W. Bennett, Taming the Electoral College 90-91 (2006).

11. U.S. Const. amend. XX, § 3.

12. Death or resignation (or expulsion) is, of course, possible for the Senate as well, but less likely than in the House, if for no other reason than that it is the smaller body.

13. U.S. Const. art. I, § 3, cl. 4.

14. U.S. Const. amend. XII.

15. See U.S. Const. amend. XX, § 3.  For a discussion of problems that lurk in vice presidential succession, and in the statutory order of succession, see Bennett, supra note 10, at 80-81.

16. David Leip, U.S. Election Atlas:  1824 Election Results, (last visited Dec. 6, 2006) (link).

17. David Leip, U.S. Election Atlas:  1836 Election Results, (last visited Dec. 6, 2006) (link).

18. Congress could reach a consensus in favor of the winner of the nationwide popular vote, for instance.

19. For possible solutions to a variety of electoral college problems, see generally Bennett, supra note 10, at 90-91, 115-18, 130-34, 144-60, 165-66.


Copyright 2006 Northwestern University

Cite As:  101 Nw. U. L. Rev. Colloquy 45 (2006),

Persistent URL:

(Download PDF)


(back to top)