In the contest for the Democratic presidential nomination, front-runner Hillary Clinton holds only a modest lead over Bernie Sanders in pledged delegates from the first four states. But total delegate counts show Clinton with more than six times the delegates that Sanders has — a lead of over 400 delegates. Clinton enjoys this large lead because of her support from superdelegates, and Sanders’s supporters have, in turn, raised serious questions about the role that superdelegates play in the Democratic nomination contest. Last weekend, the New York Times editorialized that Democrats should clarify the role of superdelegates.

Superdelegates are elected officials and party leaders who serve as unpledged delegates to the Democratic National Convention. There are 712, or roughly 15 percent of the total delegates who vote to choose the presidential nominee. Democrats began to include these unpledged delegates through a series of reforms in the 1980s; these delegate slots guaranteed the party leadership some voice in a nomination process that had shifted heavily toward public participation in the 1970s.

The heated debate over superdelegates is not a new feature of the 2016 election —superdelegates came in for considerable scrutiny in the 2008 contest between Clinton and Barack Obama. As that long and close battle for the nomination dragged on, it appeared more likely that superdelegates could swing the nomination. In turn, the superdelegates felt growing pressure to line up with either Clinton or Obama, especially as the contest continued through the spring.

We examined the choices of congressional superdelegates in 2008 to understand who endorsed whom — and when. Based on our research, we think the concerns that superdelegates will “steal” the 2016 nomination from Sanders are overstated. Our results point to a main conclusion that is important for thinking about 2016: Superdelegates were responsive both to their Democratic constituents’ preferences and to the evolving dynamics of the national race.

We focused on about 270 superdelegates who were sitting Democratic members of the House or Senate. We tracked these members’ endorsements of Clinton or Obama over the course of the 2008 campaign, examining which factors increased or decreased the chance that a member would endorse either candidate on a given day. Constituency support for Obama depressed the chance of a Clinton endorsement, for example.

And as the Obama candidacy strengthened nationally (as measured by political markets data), congressional superdelegates became more likely to announce support for Obama. Other individualized political factors mattered, too. For instance, a member with political connections to Clinton (measured by contributions from Clinton’s PAC) was more likely to endorse Clinton and less likely to endorse Obama. The reverse was true for members with political connections to Obama.

Other researchers who have studied the 2008 case have reached broadly similar conclusions. Looking beyond the delegates who are electorally tied to congressional constituencies, other studies have found connections between state primary voting patterns and superdelegate choices as well as an effect for the trajectory of the national contest.

As in 2008, many superdelegates endorsed Clinton very early in the 2016 cycle. These superdelegates are likely to remain pledged to Clinton as long as the race remains competitive. Our data on congressional superdelegates in 2008 reveals that those who declared early stayed with Clinton until the end. Of the 74 congressional superdelegates who endorsed Clinton in 2007, all but three maintained their endorsement until June 5, 2008 (two days after Obama secured a majority of pledged delegates).

But the hopeful news for Sanders supporters is that those who endorsed later, during the 2008 primary season, were most likely to endorse the winner of their primary contest. And most of those who went against the outcome of their state were endorsing Obama, who had emerged as the national leader in pledged delegates.

Clinton supporters can take solace in the fact that Clinton now has about 150 more endorsements than she did at the same time in 2008. This support means that she will likely maintain a lead in the total delegate count for some time. As long as expectations remain strong that Clinton will be the nominee, superdelegates are more likely to endorse her unless Sanders wins their local contest by a solid margin.

Overall, we think that the 2008 case cautions against the prediction that superdelegates will play a decisive role in the Clinton-Sanders contest. The rules would, in theory, allow superdelegates to vote overwhelmingly for a candidate who had lost a close national contest for pledged delegates and give the nomination to that candidate. But, in such a scenario, the superdelegates themselves would likely be divided because of conflicting state and district-level results, and the tendency of later endorsers to follow the national trajectory may give an advantage to the candidate who is ahead in the pledged-delegate contest.

None of this detracts from the fact that the early superdelegate endorsers — like other party elites — can help to shape the nomination contest. As the process unfolds, however, superdelegates have strong incentives to follow public input.

Edward Hasecke (@ehasecke) is professor of political science at Wittenberg University.

Scott Meinke (@profmeinke) is associate professor of political science at Bucknell University.