Former vice president Joe Biden is now the presumptive Democratic nominee for president — after a primary involving months of campaigning, 28 Democratic candidates and over $1 billion in ads. Did any of that affect the result? Not much, our research suggests — but they weren’t meaningless. We’ll explain below.

Here’s how we did our research

More than 18 months ago, before the 2018 midterm election, we conducted a nationally representative survey of 1,046 likely voters, fielded for us by NORC at the University of Chicago and funded by Cornell’s Center for the Study of Inequality. One of the questions we asked was, “Now thinking ahead to the 2020 presidential election, which one person would you most like to see run for president on the Democratic ticket?” Our respondents could name anyone, allowing us to capture their most-preferred candidate. Among Democrats in our sample, Biden was the favorite by far, named by 22 percent. Biden’s closest competitor, Sen. Bernie Sanders (I-Vt.), was next, though mentioned half as often at 11 percent. Hillary Clinton came next with 8 percent, followed by Sen. Elizabeth Warren (Mass.) with just under 8 percent.

In other words, leaving out Clinton, who did not enter the primary race, the three candidates who’ve won the most delegates to date — Biden, Sanders and Warren — were also the most favored names 18 months ago, in that exact order.

Although Biden has been ahead in most polls for the past year, our results stand out for having come in 2018, before any primary campaigning; for having been the answers to an open-ended question, not picked from a list of names; and for exactly matching the order of the first-, second- and third-place candidates so far.

The figure below shows the 10 most popular names offered (16 percent of Democrats in the sample indicated they didn’t know or care and 7 percent skipped the question). As you can see from names like Hillary Clinton, Michelle and Barack Obama, and Oprah Winfrey, our respondents appear to have been naming the most prominent Democrats that came to mind.

But that is precisely the point. After 10 debates and over 1,000 campaign events, much of the final outcome is identical to what could be measured in survey data from a year and a half ago.

The primary campaigns do seem to have affected the outcome, at least a bit

But this doesn’t mean the primaries had no effect. First, notice that Sens. Cory Booker (N.J.) and Kamala D. Harris (Calif.) were the seventh- and eighth-most-popular names given — and the fourth- and fifth-most-popular out of those who entered the primary race. But those two were also among the first to drop out. That mismatch between their early popularity and their early exits from the race is consistent with claims that Democratic National Committee fundraising rules and the lack of racial diversity in early primary states disproportionately hurt minority candidates.

Our data also reveal how much money mattered in the Democratic primary. Not a single Democrat in our sample mentioned former New York mayor Mike Bloomberg. When we analyze all likely voters in our sample, including independents and Republicans, just three respondents (or 0.3 percent) mentioned Bloomberg. And yet in the end, he earned the fourth-highest delegate count. Given their early popularity, if either Booker or Harris had been able to spend more than $500 million, they might well have stayed in the race.

Finally, notice that none of our respondents named either business executive Andrew Yang or former South Bend, Ind., mayor Pete Buttigieg. That’s hardly surprising. Neither was a nationally known figure 18 months ago. And yet they were able to run national campaigns and stay on the debate stage for months — showing, perhaps surprisingly, that newcomers can still influence the primaries. Had Buttigieg, as well as Sen. Amy Klobuchar (Minn.) — the seventh-most-popular candidate name among Democrats in our data, stayed in the race through Super Tuesday, Biden might not have emerged as the presumptive nominee.

So did Buttigieg and Klobuchar help determine the election outcome? Or, given Biden’s popularity 18 months ago, was his nomination inevitable?

Peter K. Enns (@pete_enns) is an associate professor of government at Cornell University, executive director of the Roper Center for Public Opinion Research, and co-director of the Cornell Center for Social Sciences.

Jonathon P. Schuldt (@JonathonSchuldt) is an associate professor of communication at Cornell University, and a faculty affiliate at the Roper Center.

We thank Henry Manley for research support.