Is a gender gap between women’s and men’s votes helping Hillary Clinton win in the Democratic presidential primaries and caucuses? Media reports on this have misconstrued exactly how the gender gap works. For instance, news outlets are generally reporting that Clinton won the women’s vote in Iowa and Nevada but lost it in New Hampshire.

But that doesn’t exactly answer the question. Let me explain.

Media outlets tend to report on whether most female voters cast ballots for Clinton. But that’s not the important point. Here’s the central question: Are women and men voting for Clinton at the same rates, or is she getting a larger proportion of the votes from one group or the other? That latter is the gender gap.

How big is the gender gap in the 2016 Democratic nomination race?

In the Iowa caucuses, Clinton won a majority of the women’s vote, capturing the support of 53 percent of women, according to the exit poll.

That’s not the gender gap; the gap is between how well she did among men versus how well she did among women. According to those polls, only 44 percent of male Democratic caucus-goers went for Clinton, giving her a nine-point gender gap. That nine points measures how much better she did among women than among men.

Even in New Hampshire, Clinton did better among women than men. There her gender gap was even larger: 12 percentage points. Clinton won the votes of 44 percent of women and only 32 percent of men. The gap was the same in the Nevada caucuses: 13 points, with 57 percent of female and 44 percent of male Democratic voters saying they chose Clinton.

In other words, Clinton’s 2016 gender gap so far is much like her gender gap in the 2008 presidential primaries, when nine percent (on average) more women voted for her than did men.

Did a gender gap occur in other Democratic primary contests?

Of course, male candidates can have a gender gap among their voters, as well.

In presidential primary voting from 1980 to 2000, my research finds a gender gap in about one-fourth of the Democratic primaries. The strongest came in in 2000, when in half the primary contests, women were five percentage points more likely than men to vote for Al Gore. That was quite different from the gap during Gore’s 1988 run; in that contest, more men than women voted for him 38 percent of the time. Perhaps women preferred Gore more when he was the establishment candidate in 2000 than when he was the Southern candidate in 1988. Or maybe men changed their minds about Gore.

We’ve seen gender gaps in other Democratic contests, as well. When Ted Kennedy challenged Jimmy Carter in 1980, more female than male voters favored him in one-third of the primaries. Jesse Jackson did better with female primary voters in 1984 and 1988; that’s because more African American women voted than did African American men. Finally, men preferred Jerry Brown’s populist message in 1992 more than women did, in 20 percent of the primaries.

Are there gender gaps in the Republican primaries?

Yes. In the 2016 Republican primaries, women and men voted relatively evenly for all the candidates in the Iowa and Nevada caucuses.

But in New Hampshire and South Carolina, more men than women preferred Donald Trump. In New Hampshire, Trump had a five-point advantage from men; 38 percent of male voters and only 33 percent of female voters chose him. In South Carolina, Trump had a seven-point advantage among men, gaining the votes of 36 percent of men but only 29 percent of women.

In the past, Republican primary voters were more likely to have a gender gap than were Democratic primary voters, with a gender gap in one-third of GOP primaries between 1980 and 2000. The biggest gap came in 1992: George H. W. Bush had an 11-point advantage from women, and Pat Buchanan had an 11-point advantage among men. Women also preferred George W. Bush more than did men in 43 percent of the 2000 primaries, while men were more likely to support John McCain. In 1988, Pat Robertson benefited from the fact that more women preferred a religious candidate.

Perhaps Republican women are more likely to prefer establishment candidates while Republican men are more likely to back outsider candidates. What’s more, women voting in these primaries are more ideologically moderate than their male counterparts.

Why is there a gender gap at all?

The gender gap in public opinion is not usually about women’s issues. Women and men generally have similar opinions on issues such as abortion and mandatory insurance coverage of contraceptives.

Different attitudes on compassion issues explain part of the gender gap. Women are slightly more likely than men to want the government to help people in need. The flip side of that difference comes on use-of-force issues; women are less likely than men to want their government to send in the military.

There’s a gender gap for party identification, as well. Thirty-seven percent of women are Democrats compared to 26 percent of men, for a gap of 11 points. That didn’t happen because more women became Democrats. Rather, since the 1960s men have become increasingly conservative and moved into the Republican Party.

How exactly will the gender gap break in the 2016 election? We shall soon find out.

Barbara Norrander is a professor in the School of Government and Public Policy at the University of Arizona and the author of “The Imperfect Primary: Oddities, Biases and Strengths in U.S. Presidential Nomination Politics.”