Conventional wisdom is that presidential nominating conventions provide candidates with a boost among voters, a so-called “convention bounce.” Hence, some worry that presumptive Democratic nominee Joe Biden and his Democratic Party will be at a disadvantage if we ditch an in-person convention while Republicans recklessly proceed with theirs.
Conventional wisdom is wrong. Aside from missing out on some parties and needlessly burning cash, Democrats lose nothing by ditching the convention.
Let’s start with FiveThirtyEight’s poll aggregation from 2016 campaign:
“Aha,” you say, “there’s absolutely a bump there! I see it!” And technically you’re right—Trump’s “bump” was 1.6 points, and Hillary Clinton’s in the weeks after the convention was 2.9 points. That seems real, doesn’t it? Not really.
In situations like these, there’s one big factor to consider: response bias. In polling, the side that is most excited at the moment is more likely to answer a poll than those who are less enthused. That’s what oftentimes leads to wild swings in polling when, in reality, public opinion over something as polarizing as presidential preferences hardly ever moves. And fact is, we were pumped. First woman president, woooooo!
But that’s also why those bounces are short-lived. Eventually, whatever catalyzed one side to respond with greater intensity recedes into history, and we end up back at equilibrium. And that’s what happened in the 2016 race chart above as we hit the first debate. Equilibrium. The polling reset the race at exactly where it was before the conventions.
In fact, you want to know what’s crazy? The pre-conventions aggregate was Clinton 46.5%, Trump 44.9%, or a difference of 1.6 points. The final vote? It was 48.2% to 46.5%. The difference? Just 1.7 points. That’s how little public opinion actually changes in these campaigns!
But what about party building? Well, we were fortunately already tracking that at Civiqs. First up was the Republican Primary.
That’s a 3-month snapshot, from June 2, 2016, to September 2. You can see how steady this measure is, and yes, the Republican Party got a small bump, from 22-64 disapproval, to 26-62. The reason? Their convention united what was, at that point, a bitterly fragmented Republican Party. Here is the same chart, but only with Republicans:
Among Republicans, their party when from 53% favorable to 61%.
Democrats, on the other hand, saw zero change in the overall picture:
The Democratic convention was July 25-28 (unfortunately we forgot to flag it), and as you can see, there was zeeeeeero movement. Numbers among Democrats and independents were similarly unaffected. The Democratic Party, as much noise as the Sanders partisans made, was already unified at 81-11 favorable!
Today, both parties are now pretty united, at around that same 81-10 favorable marks. Trump and the Republican Party are one and the same. There are no other factions pulling it apart. So even if both conventions went ahead, full-steam, in an alternate universe with no pandemic, I’d bet on zero effect. Conventions are apparently great places to unify internally, but that’s something neither party critically requires.
Conventions are undoubtedly fun. They are celebrations of the values that drive the two parties, allow them to showcase up-and-coming talent (that’s where future president Barack Obama got his first national break), and it’s fun to drop the balloons.
But the world has moved to the digital realm—in organizing, in showcasing new talent (hi, Alexandria Ocasio-Cortez!), and in sharing with the nation the values and priorities that drive the presidential campaign.
It sucks that the world prohibits large gatherings, it really does. And it’ll suck that Democrats will likely have to cancel their convention while Republicans foolishly proceed (maskless, because it otherwise shows “weakness”). But will it hurt us politically?
In a normal world, there is no evidence that skipping a convention would cost us politically. And in a world where Donald Trump and his tightly bound party have killed hundreds of thousands of Americans, we can just assume that it’ll matter even less.
