Who Will Win the Democratic Primary? Our New Prediction Model Says ...
WHY YOU SHOULD CARE
Because this Forecast is better than any polls.
By Nick Fouriezos
It’s easy to get lost in a sea of information about the likely outcome of a presidential campaign, from polls to stirring campaign trail reportage. As OZY sets out to bring you in-depth analysis of 2020 that you can’t get anywhere else, we wanted to distill the noise into solid numbers that will give you insight into what’s happening — and what comes next.
So OZY teamed up with Washington-based Republican technology and data firm 0ptimus and its sister company Decision Desk HQ, which reports and analyzes election results, to present “The Forecast.” Launching today, The Forecast is an exclusive prediction model that breaks down polls, demographics, fundraising figures, historical trends and more to determine who will be the next Democratic nominee for president. (More on the methodology here.)
The Forecast is built on the back of the OZY/0ptimus model that was so good that it came within one House seat and one Senate seat of precisely nailing the 2018 midterms. After crunching reams of data, the model runs tens of thousands of simulations to determine the likelihood of each result. “It’s like simulating the Super Bowl with Madden” the video game, says Scott Tranter, CEO and co-founder of 0ptimus, “except if you were re-running simulations after major plays, every injury and every quarter, giving us continuous fresh insights and predictions.”
So who have we got? After running the 0ptimus (Transformer-themed) computers red-hot all weekend to take in the latest results — Bumblebee and Ironhide accrued some serious overtime — our first Forecast finds …
Joe Biden is the most likely Democratic presidential nominee — but there’s a better than one-in-three chance of a contested convention.
After running 10,000 simulations of the Democratic race, we found Biden to have a 37 percent chance of winning the nomination. The second-most-likely nominee is Bernie Sanders (18 percent chance), followed by Elizabeth Warren (8 percent) and Pete Buttigieg (1 percent). But there’s a 36 percent chance no candidate gets a majority of delegates and this turns into a floor fight at the Democratic National Convention.
Biden was an early, if shaky, front-runner — but for the better part of a year, he’s held up. He has survived early attacks, mixed debate performances and the rise of other contenders (Warren in the fall, Buttigieg through the winter) to remain on top. And what may be most surprising is how Biden wins. Our model predicts essentially a three-way tie in Iowa, with Sanders projected as the narrow winner. It also has Sanders winning New Hampshire with a 10-point cushion, which could very well make Biden 0-2 in the crucial first voting states.
But Biden at this stage is projected to prevail handily in Nevada (+12) and South Carolina (+32). If he were to storm back and win the nomination despite not winning Iowa or New Hampshire, Biden would be the only Democrat other than Bill Clinton to do so since the modern primary system was born. “Biden is doing decently across the board on virtually all factors. He’s not really weak in any one, and he’s not blowing the doors off of any one,” Tranter says.
Looking under the hood, The Forecast sees Biden as a strong candidate because of his high national and state polls, as well as a surplus of news media attention (a key factor in helping Donald Trump rise to the top of the heap in 2016). And while Biden lags behind many other candidates in total amount of contributions and number of individual contributions — Sanders, Warren and Buttigieg all exceed him — the former vice president remains close enough in those arenas to still be the front-runner.
That seems to be the “story of the election,” notes Kyle Kondik, managing editor of Larry Sabato’s Crystal Ball at the University of Virginia. “Biden has been very durable throughout this process.” While that durability may be tested if Biden loses both early states, “it all depends on whether or not that has downstream consequences, particularly with the African American vote, and he loses his grip on South Carolina.” It’s certainly better for Biden if New Hampshire and Iowa are split, Kondik says, as our model suggests could happen.
One word of caution: The Forecast, and other models like it, are working with data available to us now … data that could change a lot depending on who wins the early primary states. “Once we get an idea of the strength of these candidates with certain demographics, the race gets easier to call,” Kondik says. “Probably the most important thing to remember is what we have on day one will change considerably over the next four weeks, and certainly after Super Tuesday,” Tranter says.
On that note, don’t count out Sanders, who has the second-highest chance of becoming the nominee, according to The Forecast. Not only is he in the running in Iowa and a clear front-runner in New Hampshire, but Sanders has also had a number of strong polls in Nevada, which votes third. His strength with Latinos, a key voting bloc, could see him gain ground on Biden — particularly if he ekes out an Iowa win.
Sanders remains the second-place person in our model, but the real wild card is the second place outcome. At this point there’s a whopping 36 percent chance that no candidate earns the 1,991 delegates needed to win the first ballot at July’s Democratic National Convention in Milwaukee. That outcome depends on several candidates hanging in through the spring, but Sanders has proven an ability to raise gobs of money from small donors, while Michael Bloomberg and Tom Steyer have the personal wealth to hang around as long as they like.
“With four viable traditional candidates, two self-funding billionaires and a very divided electorate, this year presents the best chance at a contested convention in a generation,” says 0ptimus data scientist Alex Alduncin.