The Value of Super Delegates

Super Delegates are getting a terrible rap these days. The 'slogan' arguments against them are good (undemocratic, they count more than my vote, they are tools of the elite) and it may be that if the slogan arguments are good that is all it will take these days to change something. But the underlying reality is much more complicated. The Democratic presidential nomination system has a lot of advantages, as well as some rough edges that super delegates smooth and that can’t be filed away by other means. In addition, there is a core fairness in rewarding the party members who do work that needs doing.

The roughness of the Democratic primary process results from the fact the Democratic primary process’ grounding principle is very different from the way we traditionally think about elections. The way we traditionally think about elections is that candidates are trying to get past the post, they are trying to get one more vote than their opponent to win. This approach makes sense and assures order. The presidency can really only be in one set of hands, a district can only go to one person. That is how our system works.

The Democratic party’s nominating process, however, doesn’t proceed in that fashion. So that it can more accurately reflect the views of all Democrats, the Democratic party uses a proportional system. If you get 45% of the vote in an individual content, you win 45% of the delegates. Proportionality is present in every single Democratic presidential nominating contest, which amounts to about five hundred separate contests. Each of the 57 different entities (50 states, Washington, D.C., five territories and Democrats Aboard) have a contest, as do regional districts that are mostly aligned with Congressional districts.

For the most part, this works great though it does have a wrinkle. With the exception of Democrats Abroad, every delegate who is selected has one vote in selecting the nominee. This means that in districts with an odd number of delegates a slight winner take all principal comes into effect. Winning by a vote means you snag the odd delegate. If a district has three delegates, winning by a vote gives you two delegates or 66% of all delegates. As one out of 4051 contests, this is not such a big deal. But if that particular result proves decisive, it would seem unfair in comparison to all the other contests.

In districts with an even number of delegates, such as four, a very large plurality is needed to get any advantage whatsoever. If you do reach that plurality but just barely that has a distorting effect as 62.5% of the vote becomes 75% of the delegates, and 62% of the vote becomes 50% of the delegates. This rounding approach likely does not matter because it does not inherently favor one candidate over another. Presumably rounding benefits one candidate in one contest and another in another. Still, a string of one-side plurality victories could produce an odd outcome in a very close race.

Besides the issue of rounding, two other factors make proportionality not entirely in keeping with one person, one vote. Caucuses typically have far fewer participants than primaries and attract more ideologically committed voters, thus potentially skewing the results toward political extremes. The difference this year between the Washington’s non-binding primary results (that favored Hillary Clinton) and the same state’s caucus (favoring Bernie Sanders) shows just how divergent caucus results can be.

A lesser known, but potentially as important wrinkle, is that the district and state level delegates are awarded to those units based on their Democratic performance. If a state or district votes more heavily for Democrats in a previous election, it receives more delegates to the next convention. The result is that African American districts, with their overwhelmingly Democratic performance, get more delegates than other areas. The most extreme example of this is in Ohio, where the Cleveland-based 11th district received 17 delegates to the Democratic National Convention, while all the districts currently represented by Republicans and with roughly similar overall populations received only four delegates each. Now this is not totally out of whack. Obama got 2.4 times as many votes in the 11th as he did in a typical Republican districts such as John Boehner’s former Ohio 8th. Still this inequity exists and is something that might be reconsidered in the future. (Doing what the Republicans do and granting every district the same number of delegates also has problems. For example, 1,223 Republican voters in the Bronx picked three delegates and 76,126 Republican voters in the rural area around Buffalo picked three delegates.

What these three points--performance bonuses, caucuses and rounding--demonstrate is that the Democratic nominating process has rough edges. They are probably worth it to maintain a proportionality that best captures Democrats preferences but they do mean the winner of the pledged delegates may not automatically be the person that the most Democrats want. And yet there is no easy way to fix this. Rounding is necessary unless you want lots of fractional delegates. Caucuses exist because some states don’t want to pay for primaries and because some states need the caucuses as organizing tools. The performance bonus is a feature not a bug that rewards people for voting for Democrats. (It is why Obama won in 2008.)

At the moment, none of this is a real problem because super delegates exist and can correct any glaring inequities. A nominee cannot be chosen simply on the backs of caucuses and rounding that produce an additional 10 delegates or so. If a candidate trailed by about a million votes, super delegates can step in and more accurately reflect the choice of the voters. This therefore is a great reason to keep super delegates. In their absence the party has to trust a process that can produce an unfair, unrepresentative outcome that is hard to fix. Even more so, if no one wins a majority of pledged delegates, the party will put an incredible amount of power in the hand of delegates pledged to third and fourth place finishers. This all creates the potential for chaos. Given that super delegates have never overturned the winner of the pledged delegates and that even the potential for such a result only exists in a close contest, this seems like a worthwhile stop-gap measure. (The Republicans, for example, would never have nominated Trump under the Democratic system.)

This is the mechanical case for super delegates and it is a strong one. It is important to remember that any changes in the rules would only matter in really close contests, otherwise the outcome would be the same.

There is an additional rationale for maintaining super delegates. A political party is not in fact a once-every-four-year enterprise. It is an everyday endeavor. Elections at all levels are held at least once a year, if not more. There are things that need to be done day in and day out. The people who are elected Democratic Super Delegates are the people who do this work. No secret cabal prevents someone who works hard for the party from becoming a super delegate as a result of that work. To a very large extent, people who regularly do the necessary work deserve a voice in who leads the party. Rather than ending super delegates, we should instead have robust contests over who can be elected a super delegate. By disempowering the people who do the work on a daily basis rather than those who simply show up once every four years, you make fewer people want to do that work.

It is also important to remember that in the current system a candidate who gets 58% of the pledged delegates wins without a single super delegate. A candidate who gets 53% of pledged delegates requires only a third of super delegates. It is not a system that actually lends itself to overthrowing voters wishes, particularly because overturning the will of the voters is usually not a winning strategy and super delegates want to win. The super delegates approach is mostly an in-case-of-emergency-brake provision and that is valuable.

A political party is quite like a real party, someone has to throw it. In a way super delegates throw the party. People who throw a party want the guests to have as good a time as possible. They might know something about the house or the neighborhood that the guests don’t. Giving them a say just makes sense. Super delegates function to make the road safer, they don’t change who wins the race. It is easy to fall into the trap of seeing them as an interference. It also may be that the optics of having super delegates have become too damaging. But their actual role in the process is valuable and should be maintained.