Can Polls Be Predictive This Early? Yes, if Old Rules Still Apply

Spread the love

There are a mere 300-plus days until the Iowa caucuses. Believe it or not, this is a point when we typically could hope to know a lot about who might become the next Democratic presidential nominee.

Across both parties, candidates who have led in the polls over the first six months in the year before the primary have gone on to win more than half the time since the modern primary system developed in the early 1970s. Candidates who have led in early endorsements from elected officials have gone on to win more than half the time as well. It’s an impressive record, given how many candidates are usually in the running.

The traditional metrics suggest that there might be a clear front-runner for the 2020 nomination: Joe Biden. A decade ago, pundits might have wondered whether he was the “inevitable” nominee. Now it’s not clear whether the old metrics and rules of thumb still apply.

Presidential primaries have been getting strange lately, and it started well before Donald J. Trump defied every expectation to win the Republican nomination.

No candidate, Republican or Democrat, has received more than 55.2 percent of the popular vote during a presidential primary season without an incumbent president since John Kerry in 2004. Hillary Clinton got that 55.2 percent share in 2016, and she had the benefit of entering the race with stronger support in the polls than any modern candidate. But in six of the nine similar contests before 2008, that number was surpassed.

It’s not just that the front-runners have fared poorly in recent cycles. It’s that they face new and stronger kinds of challenges — underpinned by political and technological changes that aren’t going anywhere anytime soon.

Today there are media-driven boom-bust candidates, who come from nowhere to earn a shot at a lasting role in the race. The trend reflects the rise of cable news and the internet, which have changed the way voters consume information about the campaign. More candidates are drawn into the field, because they’ll have more chances to succeed.

One example is Pete Buttigieg, the 37-year-old mayor of South Bend, Ind. On Sunday night, he got an opportunity that hasn’t existed in every primary cycle: a nationally televised town hall on CNN. His performance was well received, and pundits seemed newly bullish on his chance of being a player in the race. Twitter was still abuzz the next morning. This kind of opportunity for early attention simply didn’t exist before cable news and the internet, particularly for long-shot candidates with few traditional qualifications and virtually no support in the polls.

Pete Buttigieg, the mayor of South Bend, Ind., is a Rhodes Scholar and a veteran of the war in Afghanistan. CreditJose Luis Magana/Associated Press

Technological change has also made it easier for factional candidates to thrive without elite support. There have always been factional candidates, like Pat Robertson or Jesse Jackson in the 1980s, but now they can raise the money to endure through the whole primary season.

The factional candidates also seem to benefit from larger bases of support, because the parties have become better sorted ideologically over the last 20 years.

Now the voting records of longtime establishment-backed politicians are out of sync with the increasingly ideologically consistent preferences and composition of their parties. This leaves them more vulnerable to the factional candidates who find it easier to thrive without establishment support.

Mr. Biden, who is expected to join the race soon, faces all of these challenges in 2020. He has a nearly 50-year record in Washington, with many votes that lag behind the modern Democratic Party’s increasingly progressive outlook. He faces a strong factional candidate in Bernie Sanders, who is well positioned to appeal to many of Mr. Biden’s natural skeptics. And beyond Mr. Sanders, he’ll face some charismatic but largely unknown candidates who will have far greater opportunities to break through than they would have a few decades ago.

Mr. Biden is still thought to have a significant chance to win the nomination. He just isn’t considered a clear front-runner, or even the front-runner at all. PredictWise, based on betting markets, has it essentially as a three-way tie between him, Mr. Sanders and Kamala Harris. Data-driven political analysts also seem skeptical of Mr. Biden; several make Ms. Harris the favorite.

Kamala Harris, a senator from California, campaigning recently in Myrtle Beach, S.C. She received an impressive boost in the polls after her announcement in January. CreditJason Lee/The Sun News, via Associated Press

These assessments might be right. But it is not consistent with the traditional measures, which make Mr. Biden a natural and perhaps a clear favorite for the nomination.

He has averaged 28 percent of the vote in national polls so far this cycle, with an average lead of about 10 points. If that holds through June (and it’s worth noting that Mr. Sanders has closed the gap since his announcement), you could conjure a case that it gives him at least a 40 percent chance of winning — if the old rules apply.

The historical track record for polls is pretty good. Poll leaders in the first half of the year with at least a 10-point advantage have gone on to win in seven of eight cases, and the leader of any size has gone on to win in eight of the 15 primaries without an incumbent president since 1976.

Recent cycles have defied expectations in many respects, but the winners typically polled well early in the cycle as well. You could consider President Trump an exception, but other than a very brief lead by Ben Carson, Mr. Trump led in the polling the whole way starting just a few weeks after he announced his presidential bid in mid-June.

Mr. Biden enters with high name recognition, which tends to diminish his chances relative to a candidate with equal poll numbers and less name recognition. But the name recognition penalty, historically, is not as substantial as one might expect — and that’s even if we don’t count Mr. Trump as a high-name-recognition candidate.

Perhaps Mr. Biden will be considered the front-runner if high-profile endorsements follow his announcement — another factor that has traditionally done a decent job of predicting the eventual nominee. It’s early to tell because he isn’t in the race yet, but as the former vice president, he is at the very least acceptable to the establishment.

Maybe the best early sign that 2020 will follow the last few cycles of presidential primaries is the large number of candidates who have entered the race, with a good chance that there will eventually be more candidates than in any other contest in the modern primary era.

Bernie Sanders is well positioned to appeal to many of Joe Biden’s natural skeptics.CreditSteven Senne/Associated Press

A larger field has tended to mean a greater chance of a nominee who wins with less than 50 percent of the vote. It has also tended to mean a greater chance of an upset. The crowd of candidates might in part be a reflection on Mr. Biden, who, unlike Mrs. Clinton, wasn’t able to clear most of the field of viable challengers.

Are the betting markets right that Mr. Biden is not the front-runner for the nomination? That conclusion could be justified based on serious reservations about his viability, but it is not a reasonable conclusion based on the data alone. If he goes on to win, a lot of analysis could look like a pretty serious overcorrection after 2016.

But recent primary battles suggest we know less than the longer historical record of polling implies. If so, it’s hard to know how much we really know at all.

Be the first to comment

Leave a Reply