Back in 1988, Gallup polling had Democrat Michael Dukakis up by 17 points over Republican George H. W. Bush in July of that year.
Bush went on to trounce Dukakis, who in the race’s final days was running such a listless and futile campaign that Saturday Night Live served up one of its all-time brutal presidential candidate skits the weekend before Election Day. The Dukakis experience helped fuel the myth of the post-political party convention poll bounce as being overinflated and irrelevant.
ALSO READ: 'Powder keg': Massive security presence on display in Chicago amid signs of trouble
What can we learn from this in 2024? Are national convention poll bounces ever anything more than temporary? Does it matter whether you have your convention first or second? Do Democrats get bigger bounces? And does having the bigger convention bounce benefit the candidate at the ballot box?
To answer these questions, analyzing the past helps inform our present.
For years, Americans looked to the Gallup polling firm for their “horse race” public opinion polls to determine which candidate was leading at a given point in time.
The American Presidency Project at the University of California at Santa Barbara has collected those numbers for 1964-2000 from Gallup Polling, as provided by Gerhard Peters. My analysis found that the average poll bounce for any presidential candidate, regardless of party, was 5.73 percentage points after their convention.
My calculations show that Democrats fared better with these post-nomination speech surveys, averaging a 6.18 percentage point bounce, as opposed to the Republicans who still find a mean jump of 5.27 percentage points after their conventions.
Recent cases show softer bounces.
More recent national convention surveys, aggregated by FiveThirtyEight, come from multiple polling firms. They run their comparison from the Monday when the convention begins to the Thursday after the convention ends. And they find that bounces just aren’t as big now — and probably weren’t so big back then.
My analysis of FiveThirtyEight’s data shows that during national conventions from 1968 to 1992, the average poll bounce was only about 2.71 percentage points. From 1996 to 2020, that bounce shrunk to a mere 1.64 percentage points. The average overall bounce was only 2.18 percentage points.
ALSO READ: Democrats to volunteers: 'Don't bother Jake Tapper.' But the CNN host has other ideas.
In digging through FiveThirtyEight’s numbers, I found that the Republican Party enjoyed a slightly bigger bounce — 2.36 percentage points to only a two-point gain for Democrats, on average. This is in large part caused by disastrous Democratic Party conventions in 1968, 1972 and 1984, where the candidates the Democrats nominated would get wiped out in the general election.
My research of FiveThirtyEight’s polling collecting reveals that it’s slightly better to conduct a national convention second, as the conventions staged later get a 2.36 percentage point bounce, on average, as opposed to the party that has its convention first, which can only expect a two-point shot in the arm.
While Gallup had Dukakis up 17 points on Bush overall in July 1988, the Massachusetts governor only got a seven point bounce from the convention, meaning he was already ahead of Bush before the convention in Atlanta began.
Republican ads and a disastrous second debate erased that Democratic advantage, showing that you still need more than a good convention to win.
In 2008, Obama didn’t get much of a bounce from Denver, because the U.S. senator from Illinois already had a big lead over his rival before arriving at Invesco Field to accept his party’s nomination. But that same year, McCain was trailing badly. He experienced a bigger jump because he selected Sarah Palin. That boost lasted at least until a disastrous Palin television interview several weeks before the election.
ALSO READ: Does hosting your political convention in Chicago equal victory? History has an answer
External forces can certainly have an effect on national political convention poll bounces.
An extreme example is the biggest post-convention jump of modern presidential history, when Democrats Bill Clinton and Al Gore surged in 1992. (Both the Gallup and FiveThirtyEight datasets confirm this.)
This occurred because independent presidential candidate H. Ross Perot temporarily suspended his campaign after the Democratic National Convention, which further amplified Clinton’s already sizeble convention boost.
Most think the 1992 Republican National Convention in Houston was a disaster, but President George H.W. Bush actually had the second highest jump — tied with Carter in 1976 — because he used the convention to help bring some disaffected conservatives back in the fold. But it didn’t help Bush win reelection, of course, as he couldn’t fully kick his image as a mushy moderate who struggled to hold firm on conservative principles.
Post-convention bounces clearly matter in one way: who gets a bigger jump in the polls after their convention ends is more likely to win on Election Day.
Kamala Harris fans pay close attention.
I found in Gallup’s polls that the candidate with the bigger bounce in public opinion won 63.6 percent of the time in November — taking seven of 11 elections, not counting one election where the two candidates had exactly the same post-convention poll bounce. For the FiveThirtyEight data, the candidate having the bigger post-convention bounce won seven elections, compared to four losses and three ties in post-convention bounces.
That means there’s additional pressure for Harris and running mate Tim Walz to have a successful convention, and not just rely on what’s been a very good start to their unexpected campaign, as Kyle Kondik with the Center For Politics and Sabato’s Crystal Ball, agrees.
And that’s why it’s also vexing to MAGA that Trump’s convention last month in Milwaukee ended with Trump delivering a rambling speech and selecting a running mate in J.D. Vance who many Republicans consider underwhelming.
Harris now has the convention stage. If she uses it well, history indicates it will bode well for her on November 5.
John A. Tures is a professor of political science at LaGrange College in LaGrange, Georgia. His views are his own. He can be reached at jtures@lagrange.edu. His “X” account is JohnTures2.
