Electoral rules and filibuster-proofing

Far more tense for election-night return-watchers than the presidential race is likely to be the Democratic Party’s quest for 60 seats in the Senate. Sixty seats would allow the caucus, when cohesive, to override a filibuster–that is, to prevent the minority’s veto of the majority’s mandate on policy changes and court appointments. ((And unlike past Republican Senate majorities, this Democratic majority actually is based on an electoral majority.)) There are two races where the electoral rules may be almost as decisive as the voting itself, given multicandidate contests. ((If the caucus has exactly 60, it will remain dependent on two independents who caucus with the big-D Democrats. One of those is Bernie Sanders, a reliable vote. The other is Joe Lieberman, less reliable, but more so than activist Democrats give him credit for. Even at 58 or 59, some filibusters could not be sustained, as one or more of the moderate Maine Republicans (among others) would be likely to work with the Democrats on an issue-specific basis.))

In Minnesota, as in most states, a Senate race is decided by plurality (first past the post). In addition to incumbent Republican Norm Coleman and Democrat Al Franken, there is a strong third candidate in the race, Dean Barkley of the Independence Party. ((This was Jesse the Body’s party, at least some of the time. In fact, Barkley previously served as an interim Senator, appointed by then-Governor Ventura.)) Barkley stands at just under 15% in the latest Pollster.com polling aggregate. His support has been slipping after flirting with the 20% mark in mid-October.

Franken leads in the polling aggregate, but precariously, 39.5–37.6. If Coleman pulls out a narrow 40% win, it will be literally legitimate (as in within the law), given the use of plurality rule. However, it would be a subpotimal outcome if an incumbent were reelected so narrowly. ((It would also be suboptimal if a challenger were so elected, but less so, inasmuch as an incumbent is the more known quantity, and if he can’t either obtain close to 50%+1 or win by a wide margin, he arguably did not “earn” a renewed mandate.)) But plurality rule may allow it to happen. We do not know which candidate would win a runoff if there were one, because pollsters do not ask about second choices.

Meanwhile, in Georgia, Democratic challenger Jim Martin has closed a once formidable gap on incumbent Republican Saxby Chambliss. The Pollster.com aggregate shows the race at 46.7–43.9 Chambliss. However, Georgia requires a Senator to be elected by a majority of votes cast. Thus if neither Chambliss nor Martin clears 50% in tonight’s count, there will be a runoff between the two on December 2. The main third-party candidate is Libertarian Allen Buckley, who is at 7% in the most recent poll shown at Pollster.com (Rasmussen). Other polls in recent weeks have shown him as low as 2% or, oddly enough, have not asked.

Even if Martin takes the lead and holds a plurality on election night, he has not been elected.

The last time Georgia had a US Senate runoff was in 1992, when Republican Paul Coverdell defeated Democrat Wyche Fowler. ((Subsequently, Democrats in the state legislature changed the rule to plurality, but in 2005 Republicans changed it back to majority runoff.)) In that election, Fowler won 49.23% in the first round to Coverdell’s 47.67%. In the runoff, on November 24, Coverdell won, 50.65% to 49.35%.

If Georgia and Minnesota used the Double Complement Rule that I often recommend, there would have been no need for a runoff in Georgia in 1992, but there would almost certainly need to be one this year in Minnesota. The Double Complement Rule allows a sub-majority win if, and only if, the second-place candidate’s shortfall from 50% is more than double that of the leading candidate. Fowler’s 0.77 percentage point shortfall in 1992 would have been “good enough”; the only way Coverdell could have defeated him would have been to convince Libertarians in the same, larger electorate–as we’ll see below, the inter-round turnout differential was substantial–to desert their party’s candidate and give him the plurality.

The approximate 40–38 result that we may see in Minnesota, on the other hand, would necessitate a runoff under the Double Complement Rule. In fact, the only way we would see a one-round result, if the leading candidate failed to break 40%, would be if the trailer fell below 30%–something that obviously would not happen without a doubling of Barkley’s support creating a close race for second place, rather than for the plurality.

As for this year’s Georgia race, we would have a squeaker on our hands: the current polling aggregate is just over the runoff-triggering margin by about half a percentage point.

In 1992 in Georgia it was likewise a Libertarian candidate whose support in the first round resulted in the need for a runoff. Did most of the Libertarian voters in 1992 vote Republican in the runoff, given that it was the Republican who won? More likely, they simply stayed home.

In fact, the turnout in the runoff in 1992 was very low–barely half what it had been in the first round. ((1,253,991 vs. 2,251,576.)) One might imagine it will be a bit higher this year–especially if the Democrats have won 59 seats after Tuesday night.

0 thoughts on “Electoral rules and filibuster-proofing

  1. Doesn’t your rationale for the double complement rule depend heavily on situations like Georgia in 1992 where supporters of the “spoiler” candidate sat out the second round, as well as the pernicious effects of unequal turnout? For example:

    45: A > B = C
    35: B > C > A
    20: C > B > A

    Under your proposal, A wins without a runoff, even though the supporters of B and C are clearly a majority coalition. B is the rightful winner by almost all standards of fairness (although you’d have to know the lower preferences of the A supporters to be sure about the Condorcet criterion).

    Granted, top-two is not guaranteed to elect B unless enough C supporters show up for the second round. But IRV/alternative vote would elect B.

    So my question is, are you suggesting that double complement is better than IRV/alternative vote? If so, what are the advantages? Thanks!


  2. Bob, your question is answered at the DCR thread. (Click the name of the DCR formula in the post, or as of this moment, the comment in question is just below yours on the comment feed at right.)


  3. 538’s final projection suggests Dems with only about a 15% chance of getting to 60, but a better than 50% chance of getting to 58.

    On Georgia: Nate assigns “a 50 percent chance to a Chambliss win outright, a 40 percent chance to a run-off, and a 10 percent chance to a Martin win outright.”

    And on Minnesota, conflicting late polls lead him to say that this “may be the race to watch today, possibly even exceeding the presidential contest.”


  4. Georgia will have a runoff, apparently!

    Regarding the runoff, for the second day in a row, 538 refers to it as a “special election.” Not to be too pedantic about it, but a special election is one called to fill a vacancy. A runoff is not a special election, but a mandated continuation of the general election.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.