Vinod B
Published in

Vinod B

Bernie Sanders Will Lose Because He Didn’t Do the Main Job of a Politician: Forecasting the Rest of the Primary

On March 3 (Super Tuesday), Biden outperformed expectations; he also outperformed expectations in most states on March 10, trouncing Sanders in Michigan, which was expected by some to be more competitive.

This was unimaginable just two weeks ago when Biden’s campaign appeared dead in the water and Sanders, on the contrary, was the clear front runner. Sanders, however, lost his momentum and frontrunner status because he failed to campaign like a skilled politician would. Now he stands little to no chance of winning the Democratic nomination because of the rules of the Democratic primary.

There are a total of 3,979 delegates pledged in the first ballot. To win a candidate must get a majority or 1,991 of these pledged delegates by performing well enough in the 57 state and territory contests (the primaries and caucuses).

If no candidate can amass the required number, the party will head into what is known as a brokered convention. In this scenario all pledged delegates are up for grabs as each is no longer bound to support their original, pledged candidate. In addition in this second ballot there are 771 superdelegates who can now cast their vote (composed of party insiders and current and past Democratic politicians) to whomever they like.

Now a candidate must win a majority of the new total 4,750 or at least 2,375 delegates. The process can continue through various rounds of additional voting if this doesn’t occur directly in the second ballot. Crucially, in a contested convention under the current process, any candidate could come out on top as there is no set process for how things turn out.

While the exact results are a bit unclear because of mail in ballots being counted in many states, here are the current and expected delegate counts by candidate and contest.

Biden will likely come out with around 20% more delegates than Sanders after all votes come in.

Looking ahead, there are still 2,115 delegates up for grabs in 31 contests. Here are the current expected delegate counts for each day of voting.

Based on current state and national polls, Biden is expected to get on average 2,251 delegates while Sanders is expected to get 1,552.

Biden then can be expected to win because he will likely get more delegates than required to clinch the nomination; however, to fully understand the range of outcomes, we need to simulate each primary contest.

To do that we simulate expected vote shares for each contest using current polling averages from a Dirichlet distribution, adding in random variation based on the historical accuracy of primary polls. We also account for things like polling bounces from winning contests and outperforming expectations based on realized vs. polling vote shares.

These are all crucial. Historically primary polling has not been super accurate. The average error is around 10%, which means the actual range of errors is quite large. Furthermore, primary contests are highly nonlinear and path dependent. Winning a contest or day of voting and outperforming expectations drives the media narrative, providing candidates a bounce in the polls. Consequently, performing poorly can lower a candidate’s support dramatically.

Across simulations, we find that Biden is 90% likely to win and Sanders’ has a 3% chance. In the rest of the simulations, neither candidate gets the required delegate totals to win the nomination, meaning there is a 7% chance for a brokered convention. While anything can happen in this scenario, a more realistic take is that the delegates when unbound will coalesce around Biden, the more popular candidate. Basically then, Sanders has little to no chance of winning.

Bernie has a strong core of support. He is quite popular among his base of voters, meaning that he has a high floor. Indeed, looking at 538 national polling averages, we see that his support hasn’t changed much since February 27, which was before South Carolina, and the start of Biden’s surge. It was roughly 29% then and is now 33%. Biden has just picked up proportionally more supporters due to the rest of the moderate field dropping out.

Bernie did nothing to expand this base of support though.

His campaign and supporters instead became divisive and attacked the Democratic party, making him a polarizing candidate. To govern effectively Sanders would have needed support of the Democratic establishment. He should have instead focused on his core issues and used that to expand support beyond his base: things like affordable healthcare for all, fighting corruption, and being well positioned to beat Trump, all of which matter a great deal to voters.

Moderating his stances a bit on immigration and social issues could have helped him expand his base of support into older voters as well, which currently favor Biden a lot and are highly likely to actually turnout and vote unlike Sanders’ current base of younger voters, which haven’t increased in turnout like he hoped.

The angle Sanders should have played up more was that Biden’s gaffe prone nature and past support for unpopular positions like the Iraq war and cutting Social Security would put him at a disadvantage in the general election. Sanders failed to break these points through to the media, which a better functioning campaign could have done.

Because primary contests are path dependent with bounces to both the winner and candidates who outperform expectations, this was a crucial mistake. Sanders’ campaign rarely had any good news coming out recently, making it difficult to turn stories into strong polling increases. As a result he got stuck in a negative feedback loop where negative coverage reduced his polling numbers, which in turn created more negative coverage.

On the contrary, Biden’s performance on Super Tuesday became the start of a positive feedback loop for him. His win was driven mostly by late deciders. For example, in Texas exit polls where Biden overperformed expectations to win, late deciding voters broke for him 45% to 15% compared to those deciding earlier which went for Sanders 33% to 29%.

Late breaking voters can dramatically swing results in each contest, the media narrative that emerges after voting, and therefore the relevant polling bounces for future contests. If Sanders could have changed the media narrative, he could have captured late breaking voters, turning his campaign around with positive momentum feeding on itself in a virtuous cycle.

Sanders also didn’t work the Democratic party to get endorsements. He should have done everything he could to get a strong endorsement from Elizabeth Warren in order to capture her voters and also work local party officials in each contest to get their endorsements.

Indeed, going back to South Carolina, Biden was able to capture many voters due to his endorsement by Jim Clyburn. This set off a chain reaction of events leading to the current state of the primary.

Sanders didn’t make the necessary changes to his campaign strategy that could have turned it around while Biden did. Whatever magic he pulled to get Klobuchar and Buttigieg to drop out before Super Tuesday, get key endorsements, and expand his base of support, was at least partially the result of savvy political maneuvering. Biden proved to be the more skilled politician, and that is why he will be the Democratic nominee.



Politics, Policy, Economics, & Data

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Vinod Bakthavachalam

I am interested in politics, economics, & policy. I work as a data scientist and am passionate about using technology to solve structural economic problems.