When I started building the LeanTossup Democratic Primary model for our site in the spring of 2019, I decided to add a pie chart to show the odds that no candidate would win a majority of pledged delegates (a Contested Convention). I asked twitter to vote in a poll on their preferred labels for the “Odds of a Contested Convention” graph; either “Yes/No” or “Everything is fine/Chaos”. Obviously, Twitter chose the “Everything is fine/Chaos” labels, mostly because it’s funnier, but also because it shows just how crazy that would be: legitimately nobody knows what would happen. Much to the delight of Twitter, the LeanTossup model has always projected a Contested Convention more likely than not, even when Biden was stronger in the National polls.
However, people were quick to point out it literally never happens. A contested convention is the dream of political analysts, as it allows them to publish nearly endless think pieces about what might happen, with the consuming public eagerly awaiting their every word. Even in Presidential races that haven’t been particularly close a contested convention has been speculated about, such as the 2016 Republican primary, where hundreds of secretly elected Cruz loyalists would deny Trump the nomination. We all know how that turned out – predictably flaming out. This year, however, it’s different.
Everything always happens one way until it doesn’t. Analysts banking (or building models) on past primaries that didn’t lead to a Contested Convention are working with a small and flawed sample size. Almost everything about this Primary so far seems unique. Never has a nominal front runner (Joe Biden) done so poorly out of the starting gate, with very little reprieve likely to come in Nevada. The closest analogy seems to be Mitt Romney in 2012, but at this point Romney won New Hampshire, after losing Iowa (but being declared the winner the night of) while Joe Biden’s fifth place in New Hampshire is somehow worse than his Iowa finish of fourth.
If there was ever a modern party that would have a contested convention, it’s the Democratic Party. The simple fact is that their delegate allocation rules seem almost perfectly designed to create a Contested Convention. On the Republican side there are some winner-take-all states, or some states that have a share of the delegates that are given only to the winner of each state or Congressional district (CD). This allows a candidate to build a lead by winning states and districts by a small margin, and not having to win an outright majority. However, those rules do not exist on the Democratic side. All delegates are given out proportionally to candidates above 15%.
In most Democratic Primaries, this usually creates a two-person race, as one (or multiple) lower polling candidates will always start falling below the threshold for delegates, forcing them to exit the race. However those same mathematical forces, that foster the creation of a two person race, can also very easily incentivise a three or four person race, if the percentages align. The polling we are currently seeing is almost exactly the ideal scenario to generate that multiple person race; candidates winning states with small margins, multiple candidates getting delegates in each state, nobody even close to 50% of the vote or delegates.
One might argue that this is similar to 2008. where Clinton was originally favoured, but Obama surged out of Iowa, and eventually won a long contest. The problem with that is, after Iowa, Obama surged into a strong National position, a position where he would always win roughly 50% of the delegates in each state. After an inconclusive Iowa result, and winning New Hampshire by a small amount, Bernie Sanders stands at a very unimposing 26% nationally per the LeanTossup Average, which is not likely to translate into a large delegate lead anytime soon.
The arguments against a contested convention come down to the past – but when have we had a billionaire pump endless amounts of money in the second wave of states, while skipping the first month of contests? When have we ever had 5 candidates above 10% of the vote this late in the contest? If you want to argue that someone – be it Sanders, Mike Bloomberg, or Pete Buttigieg – can rally the majority of delegates, that case is increasingly hard to make. The case for the likelihood of a contested convention has been made before, and it stands through two contests. The model currently has the chances of Chaos at 75%, and if Tom Steyer surges in Nevada as some of the December polling of the Silver State suggested, that number will rise further. The Democrats don’t have an easy way out of a messy Milwaukee convention – and time won’t save them.