Skip to main content
ABC News
We Researched Hundreds Of Races. Here’s Who Democrats Are Nominating.

This story was produced in collaboration with ABC News and Ballotpedia.1 Research by Ballotpedia and Roey Hadar, Lee Harris, Adam Kelsey, Adia Robinson, Meena Venkataramanan and John Verhovek of ABC News.

You’ve probably already heard that 2018 is a new “Year of the Woman” in Democratic primaries. Women are being nominated in record numbers across the country. But that fact only tells you so much. To what extent is a candidate’s gender really affecting voters’ choices? And do other candidate characteristics make a difference?

Through Aug. 7, 811 people have appeared on the ballot this year in Democratic primaries for Senate, House and governor, not counting races featuring a Democratic incumbent.2 In partnership with ABC News and Ballotpedia, we collected data on every single one of those candidates. We’ve pored over campaign websites, news reports, state election websites, official campaign filings and more to figure out each candidate’s gender, race, sexual orientation, career background, policy positions and endorsements.3 The data gives us the most comprehensive look yet into whom Democratic primary voters are gravitating toward this year. For the next few weeks, we’ll be writing about our findings, including doing a similar exercise for Republican primaries. Today, we’re starting with something simple: the relationship between demographic and biographical factors and the share of votes candidates got in Democratic primaries.

Yes, it’s good to be a woman in a Democratic primary

According to our data, women have won 65 percent (90 of 138) of decided open Democratic primary races featuring at least one man and one woman. About 46 percent of all women who ran for office won the nomination — a stat we’ll call “win rate.” Men’s win rate has been just 23 percent (although part of their lower win rate is simply that more male candidates are competing against each other and, obviously, only one person can win each race). Women make up 48 percent (114 out of 2384) of the Democratic nominees in primaries that have been decided so far even though only 32 percent (263 of 811) of the candidates we analyzed were women. So, women are clearly having greater success than men.

In fact, all else being equal, being a woman has been worth an additional 10 percentage points over being a man in the open Democratic primaries we looked at.5 That’s one of the two biggest effects we found among all the variables we looked at. (The other variable with a similarly sized effect was having previously held elected office — more on this in a moment.)

Although women’s representation in Congress has almost doubled since 1992, the House and Senate combined have never been more than 20 percent women. But this year’s surge in the number of female candidates could change that, because studies have found that a major hurdle to women’s equal representation is that women are just less likely to run. But with women competing in 69 percent of open Democratic primaries this year, according to our data, that hurdle seems lower.

Women running in 2018 stand out from their male opponents in a few important ways. For instance, women this year are more likely than men to have previous experience as elected officials, especially if they’re running for higher-profile roles, like governor or senator. Fifty-six percent of the women who are running for governorships have previous experience as elected officials, compared with just 37 percent of men running for governor. In Senate races, the difference is even larger — 80 percent of women running for Senate have previously held elected office, compared with just 22 percent of men.

That squares with studies that find that women are less likely than men to see themselves as qualified for political office. Since previous experience as an elected official helps women to overcome the belief that they lack qualifications, it makes sense that the women running for a high-level office are more likely to have been elected before compared with men, who do not doubt their qualifications to the same degree. And because the women in our data set have more experience — which, as we mentioned, is an advantage in elections — that helps explain why they are more likely to win their races.

But women aren’t just winning because they have more experience. Even looking only at candidates with previous experience as elected officials, women are still outperforming men: 52 percent of previously elected women have won their primary races so far, compared with 40 percent of previously elected men.

Does race matter?

We might also expect race to be a significant variable in Democratic primaries. After all, we live in an era of “identity politics,” and Democrats are the preferred party of Hispanic, Asian and (especially) black voters. According to exit polls of the 2016 election, 45 percent of Hillary Clinton voters were nonwhite. So, do the Democratic Party’s nominees look the same as its voters?

Not quite. Out of 238 nominees so far, we identified 56 (or about 24 percent) who identified as nonwhite.6 However, just like with women, this shortage appears to be more the result of a pipeline problem than an inability to win — people of color accounted for just 25 percent (203 out of 811) of the total candidate pool.

Nevertheless, at least so far, being a person of color doesn’t actively help a candidate win a Democratic primary. Nonwhite candidates had a winning record of 28 percent in open Democratic primaries. Those whom we could not identify as nonwhite7 won at a rate of 31 percent. And, on average, being a person of color costs a candidate a few percentage points of vote share, but that relationship is not statistically significant.

How are LGBTQ candidates doing?

Only seven people currently serving in the House and Senate are out members of the LGBTQ community. But that number could more than double in 2019. Ten out of 34 LGBTQ candidates have won their primary contests so far, for a win rate of just over 29 percent. That’s essentially identical to the win rate among candidates whose sexual orientation was not mentioned on their websites, who won 30 percent of their primaries.

Experience helps

As we mentioned, political experience is typically an asset when you’re running for office, and so far the 2018 Democratic primaries are no exception. All else being equal, having held elected office before has given candidates a 12-point boost, on average, in vote share in open Democratic primaries. That’s the biggest such boost we found for any variable. Experienced politicians have an impressive 44 percent win rate; political novices clock in at 28 percent.

Part of the advantage of experience is self-explanatory, of course. Navigating voters, the media and rival campaigns takes some skills, and practice helps refine those skills, plus it means candidates have already built some of the connections and infrastructure that will help them run their campaigns. But there’s something else going on here too: Candidates with previous political experience are more likely to run in races where they have a greater chance at winning because they are more strategic about running for higher office than political amateurs. This is true for Democrats in 2018; experienced candidates ran in areas that were an average of 2 points more Republican than the country as a whole, according to FiveThirtyEight’s partisan lean calculation,8 compared with 15 points more Republican-leaning than the nation for inexperienced candidates.9

From the military to Congress

Being a military veteran is one of the most appealing attributes a person can have when being recruited to run for office. But in open Democratic primaries so far this year, veteran status doesn’t seem to make much of a difference in how a candidate performs. (We can’t say for sure whether it will matter in the general election.) Out of the 238 Democratic nominees in our data set, we identified 36 as veterans by looking for mentions of military service on their websites or in news reports. Candidates whom we identified as veterans won their primaries at a rate of 30 percent, compared with 31 percent for candidates for whom we couldn’t find a record of military affiliation.

Anything else?

We’re glad you asked:

  • In the age of Trump, we were interested in whether Democratic primary voters are going for millionaire business executives willing to spend their own money to get themselves elected. To find out, we used campaign-finance data to see whether the candidate had self-funded by loaning or donating at least $400,000 to his or her own campaign.10 Twelve out of 42 self-funders (29 percent) won their primaries, compared with 30 percent of non-self-funders.
  • We also looked at the success rate of candidates who had backgrounds in science, technology, engineering or mathematics — aka STEM fields. As FiveThirtyEight’s Maggie Koerth-Baker explored in depth, despite STEM professionals’ increased interest in politics and an emerging infrastructure to help them run for office, science isn’t a very big motivator for Democratic voters. In fact, it could be the opposite. Our analysis found that candidates we tagged as STEM professionals lost 3 points of vote share, on average, compared to those we did not associate with STEM (although this finding is not statistically significant). Only 32 STEM professionals (out of 142 who were running in the races that have thus far been called) have won Democratic primaries so far, giving them a below-average 23 percent win rate.
  • Finally, 34 alums of the Barack Obama administration or his campaigns have run for office so far this year, no doubt hoping to leverage their association with an ex-president who’s still popular among Democratic primary voters. Of those 34, 13 have won their primary races so far, for a win rate of 38 percent. In our estimation, being associated with Obama in this way was worth about 5 points of vote share in the average primary; however, the standard error is large here, so we can’t say for sure whether being an Obama alum really helps all that much.

Overall, Democratic primary voters in 2018 are jumping at the chance to nominate women and experienced politicians. Being an Obama alum seems to help too. No variable we looked at appeared to seriously harm a candidate’s chances in a primary (except maybe STEM training), although several — military service, race, LGBTQ status and self-funding — don’t appear to help. Here’s a summary of everything we found:

What are Democratic primary voters looking for?

How candidates with various demographic and biographical traits have fared in open Democratic primaries for Senate, House and governor in 2018

% of Nominees Win Rate all others’ Win Rate
Woman 48% 46% 23%
Past elected official 24 44 28
Nonwhite 24 28 31
Veteran 15 30 31
STEM background 13 23 33
Self-funder 5 29 30
Obama alum 5 38 30
LGBTQ 4 30 31

Win rates exclude candidates in races that have not been called. Open races are those that do not feature a Democratic incumbent.

Sources: Ballotpedia, secretaries of state, Associated Press, candidate websites, VoteSmart, Federal Election Commission, Obama Alumni Association, various news reports

But this is just the tip of the iceberg. Obviously, plenty of other questions factor into what kind of candidate pulls in primary voters — not the least of which is ideology. Stay tuned.

CORRECTION (Aug. 10, 2018, 3:27 p.m.): Due to a data-entry error, a previous version of the data used in this article incorrectly categorized David Trone, the Democratic nominee in Maryland’s 6th Congressional District. Trone should have been counted as a self-funding candidate without previous experience in elected office, not as a non-self-funding candidate who had previously held an elected office. The data has been corrected, and updates have been made to the corresponding sections of the text and table.


  1. Ballotpedia is an online encyclopedia of American politics and elections. Its goal is to preserve and expand knowledge about politics by providing objective information about local, state and federal politics and policies. Check out over a quarter of a million encyclopedic articles and analyses for free at

  2. We excluded races in which a Democratic incumbent ran because incumbents almost always win, which would heavily skew the analysis in their favor.

  3. You can find the data we used for this article, our criteria for how each variable was judged and all the details about how we collected it on our GitHub page.

  4. There are actually 244 races in our data set, but six were still unresolved at press time.

  5. We used a simple linear regression to estimate the impact of various candidate demographic characteristics on Democratic vote share. The estimated effects include an adjustment for the total number of candidates in a race and geographic location. The analysis excludes vote share totals from candidates running unopposed as well as from Washington state, as provisional and mail-in ballots there have not been fully counted as of Aug. 9.

  6. This count is most likely a lower bound, as we counted only candidates who explicitly identified themselves as a race other than white on their official campaign websites or for whom we could find a racial identification in news reports. Because candidates do not always mention their race or all use the same terms for race, our data was not complete or consistent enough to perform a more detailed analysis.

  7. So either white candidates or candidates whose race we couldn’t determine.

  8. The average difference between how a district voted in the past two presidential elections and how the country voted overall, with 2016 results weighted 75 percent and 2012 results weighted 25 percent.

  9. That pattern is likely also partly due to the fact that national party organizations — the Democratic Senatorial Campaign Committee, the Democratic Congressional Campaign Committee, etc. — will likely put more effort into recruiting quality candidates for more winnable races.

  10. Our source for this was federal campaign-finance data from the Federal Election Commission or, in the case of governors’ races, state campaign-finance data. We added together a candidate’s donations and loans to his or her own campaign to come up with a figure for his or her “total dollars self-funded.”

Meredith Conroy is an associate professor of political science at California State University, San Bernardino, and co-author of “Who Runs? The Masculine Advantage in Candidate Emergence.”

Mai Nguyen was previously FiveThirtyEight’s quantitative editor.

Nathaniel Rakich is a senior editor and senior elections analyst at FiveThirtyEight.