We Need to Talk About Elizabeth Warren
Inside the numbers — and the studies — that suggest why so many people feel bias or sexism influenced the results of her presidential campaign.
We need to talk about Elizabeth Warren.
The Massachusetts senator dropped out of the Presidential race last week, leaving two remaining candidates realistically vying for the party’s ticket. In a race for the Democratic nomination that began by featuring the most diverse pool of presidential candidates in American history, it ended with a field winnowed to two septuagenarian, white men.
Warren was among the frontrunners for the Democratic nomination up until voting began in Iowa in early February; through eighteen states thereafter, she failed to place higher than third in any state caucus or primary. Pundits have pointed to policy missteps or ideological miscalculations to explain how a candidate that was leading national polls as late as October of 2019 — and who was still polling within favorable probability to generate a valuable share of delegates from many states on the eve of Super Tuesday — failed to translate the numbers into tangible results.
Outside of her home in Cambridge last Thursday, Senator Warren was asked if sexism or gender bias played a role in her inability to garner the necessary support to win the nomination. She summed up what amounts to a trap question for women astutely. “If you say, ‘Yeah, there was sexism in this race,’ everyone says, ‘Whiner!’” Warren said. “If you say, ‘No, there was no sexism,’ about a bazillion women think, ‘What planet do you live on?’”
If you’ve been on the peripherals of the Democratic race, you may have been surprised at the visceral, fervent uproar that rippled across the country when Warren dropped out last week. You may bristle at the accusation that sexism or gender bias was at play. After all, didn’t we just elect a record number of women to the U.S. House of Representatives in 2018? Can’t we stop talking about overt gender discrimination or implicit bias?
Saying that bias is the sole reason Warren’s campaign fell short doesn’t respect the nuance and complexities of the subject; it’s this type of “all or nothing” polarization that artificially divides us on so many issues. But not confronting the very real ways in which bias impacted her run is problematic, too. Warren’s case is unique, because it didn’t look like some of the other campaigns led by female — and male — candidates that sputtered out. There wasn’t campaign mismanagement; it wasn’t as if she failed to attract significant voter interest (and fundraising), either.
Thousands — perhaps millions — of people just feel like something doesn’t add up.
And I think they have a point.
Before we go further, let’s start with the research. I’m going to quickly cite the key findings from three recent studies. You simply can’t tell this story without them.
The first one was released in the summer of 2019, by Ipsos. The study found that “defeating Donald Trump” was the top priority (82%) among Democrats and Independents in choosing a candidate for the 2020 election. It’s a finding that was corroborated often, and it suggested that the ambiguous notion of electability would, therefore, surpass ideology and policy for most voters this election cycle.
This is important. Women are often deemed less electable than men, largely due to a host of traditional gender stereotypes. And the Ipsos study went on to suggest precisely that: 74% of respondents said that they’d be comfortable voting for a woman, but only 33% believed their neighbors would be willing to vote for one. In an election to be determined by, well, electability, the voters’ perception that a woman wouldn’t be a viable choice for “their neighbors” is critical. More on that later.
Next, let’s look at a study recently released by MIT, the University of California at San Diego, and the University of Potsdam in Germany. The researchers found that people were reluctant to use the pronoun “she” when hypothetically describing the president of the United States. In this study, participants were asked to read a short passage about the next president. They pressed a button on a screen to reveal each word in the sentence. When they came across the word “she,” there was a “considerable disruption” (about one-third of a second) in reading time. This was found to be true in men and women. Again, considering how much of our biases live hidden below the surface, this is big: it suggests that the novelty of a female president is still a cognitive issue for many of us, even if we outwardly state otherwise.
Both of these studies suggest that implicit bias could’ve shaped voters’ perceptions of female candidates in the 2020 election. But bear with me. Let’s go one step further.
Just last week, a comprehensive UNDP Gender Social Norms Index study was released, which found that about 90% of people — men and women, worldwide — are biased against women. In the United States, nearly 40% of respondents thought that men made better leaders than women. Consequently, countries with higher accounts of gender bias featured a lower representation of female leaders in business and politics. Some of the insights from this study went further than bias; they demonstrated overt sexism. And while both genders exhibited bias and sexism, the men exhibited more.
(By the way, my cursory snapshots of these studies don’t do them justice. If you’re interested, I encourage you to check them out in full; I’ve provided the link to each study within this story).
Now: how do these findings translate to what happened?
Here’s what doesn’t add up about Elizabeth Warren’s campaign.
Before elections, there are really only two measurements we can take to determine public sentiment. One is polling. It stands to reason that an aggregate of well-grounded, scientifically-designed surveys can — at the least — give us a general sense of who the frontrunners in an election are. The second is campaign financing (well, unless you’re self-funded). Grassroots fundraising from small donors — while far from a perfect science — should, at worst, be able to give a similar sense of groundswell. And Warren was strong in both.
Consider this: in October of 2019, Elizabeth Warren was the frontrunner in many state polls; she was often first or second in national polls as well. Consequently, she was second in fundraising for that quarter ($24.6 million) to Senator Bernie Sanders ($25.3 million), with former South Bend Mayor Pete Buttigieg ($19.1 million) and former Vice President Joe Biden ($15.2 million) trailing behind them. Now, let me be clear: I’m not tying fundraising directly to polling. But these four candidates were the top four candidates in most polls; they were also the top four candidates in fundraising. At the least, you can reasonably suggest a correlation.
To further demonstrate that these two measurements are somewhat correlated, note that when Warren began to slip in the polls in the fourth quarter of 2019, her fundraising had started to dip as well. That quarter, her campaign fundraising tumbled about 15% to $21.2 million. Meanwhile in that quarter, Sanders had both a fundraising and polling surge, as did Buttigieg, who capped off the best fundraising quarter of his campaign ($24.7 million) with a rise in most national polls; he had even emerged as the frontrunner in Iowa state polls during this time.
But, oddly, any such correlation came to a screeching halt for one candidate when the voting actually began. Across the entire Democratic field, only the four candidates mentioned above raised more than $20 million in any quarter during the race. Three men. One woman. Of them, only one candidate — Warren — failed to finish first or second in any Democratic state caucus or primary.
Let’s be clear: this wasn’t merely because she peaked at the wrong time, as some pundits contended. Warren’s best month of fundraising came in February, days or weeks before voting would commence across sixteen states. Remember, she was polling within the 15% viability threshold in most state polls on the eve of Super Tuesday (and within the margin of error of leading some). Yet, she’d finish as high as third in only five states: including her home state of Massachusetts.
See, Warren’s story isn’t just one of a candidate who couldn’t generate enough popularity, or money, to stay afloat. That would make explaining this simpler. No: she hauled in some of the best fundraising totals of any candidate in the field. She had the second most individual donors of any candidate in the field. And it’s not just that she didn’t win the nomination; that would be one thing. But she didn’t even place as high as second in any state she ran in.
How can this be?
My hypothesis is that we can draw a line that explains the contrast between Warren’s fundraising, polling numbers, and voter turnout. She polled well early, before the election was getting real to voters (from the summer through October of 2019). This was still a time for ideology. For wonkiness. And her policies and experience truly resonated with potential voters. But, as the research suggests, electability was the prevalent issue in this election cycle, and since women are judged harsher than men, voters were going to be easily spooked. They were willing to sprint at the first sign of trouble.
So when Warren came under fire from critics for how she’d pay for her universal health care plan, her polling took an immediate hit. Granted, Bernie Sanders still hadn’t explained his plan — and when he finally did, it was widely maligned by critics — yet, his polling numbers didn’t budge. Electability was in the backs of voters’ minds. And we know that’s where bias lives.
This also explains why her fundraising climbed (astronomically) over the same period that her voter turnout stayed flat (or even declined) in February and March. Doesn’t it seem like the real-world representation of the Ipsos study? Perhaps many voters truly wanted Warren (another study found that she was the candidate that they’d want as president if they could “wave a magic wand,” and not have to worry about an election). But remember what the study found: voters were afraid that their neighbors wouldn’t vote for a woman. And thus, it became a self-fulfilling prophecy. Which explains why so many were willing to open their checkbooks, but not willing to turn out and vote. It’s kind of a warped example of the bystander effect.
Ultimately, this is still a hypothesis — based on inference. But check out the next set of numbers I turned up. These take it one step further.
On Super Tuesday, The New York Times conducted exit polls in twelve of the fourteen states (you can check them out here). I went back and looked at how many states each of the four major candidates had a higher percentage of men vote for them than women, and vice versa. For instance, in the twelve states the Times conducted exit polling in, Bernie Sanders had a higher percentage of men vote for him in all of them. Joe Biden had three states in which men voted for him at a higher percentage than women, and eight states in which women voted for him at a higher percentage than men (one state was evenly split). Mike Bloomberg? A higher percentage of women in six states, a higher percentage of men in five (one state was also split).
In the twelve states the Times conducted exit polls in, how many states had a higher percentage of men vote for Warren than women? Precisely none of them.
These numbers may fall short of scientific evidence. And yes, many studies demonstrate that bias against women exists in men and women. But remember that UNDP study? It found that men in the United States displayed more biased attitudes (and overt sexism) towards women than women did. And on Super Tuesday, every major candidate held a state that featured a higher percentage of men vote for them than women except the one leading female candidate?
Come on.
OK, I want to show you how subtle — often unconscious — gender bias actually plays out on the campaign trail. Let’s look at two examples that jumped out at me over the last few months.
In the South Carolina Democratic debate on February 25, Warren pressed former New York City Mayor Mike Bloomberg on his history with women in the workplace — specifically, the myriad sexual harassment claims that were brought against him and his company. She described a frequently reported incident from a 1998 lawsuit against Bloomberg, in which he was accused of telling a female employee to terminate her pregnancy.
Start this clip at 1:20:00. Warren recounts a personal story from her past related to pregnancy discrimination, and then references the aforementioned Bloomberg lawsuit by stating, “at least I didn’t have a boss who said to me, ‘Kill it,’ the way that Mayor Bloomberg is alleged to have said.”
Check out what happens at 1:23:25: CBS moderator Gayle King actually presses Warren on the issue, stating in reference to the allegation: “what evidence do you have of that?”
Now, I’d be all for fact-checking and source-citing during presidential debates. But, quite frankly, this never happens. It was the only time — all evening — a candidate was pushed by a moderator to cite his or her source. In fact, it was the only instance across more than ten debates in which a candidate was pushed by a moderator to provide evidence for their claim; lest we forget that this debate season has been filled to the brim with candidates attacking one another for allegations related to civil rights issues, social issues, and much more. And the many statements that have been made during these debates that were misleading, exaggerated, or flat-out untrue. But no others were challenged by the moderators.
Sexual harassment is one of the most under-reported crimes in the country, in part because women fear that they won’t be believed or that their claims will be minimized. And here, on a national debate stage, this fear was validated. An accusation was contested (in a dismissive tone, I might add), and by a female moderator, no less. To me, the moment read not only as a microcosm for the complexities and challenges facing women who bring allegations of sexual harassment forward. It was also a microcosm of Warren’s campaign, in which it often seemed that she was pushed to defend her policies or cite her sources far more often than her male counterparts.
OK, now let’s look at an example of where the female candidate herself projects these manifestations of gender bias back out, in a troubling feedback loop. Let’s go back to the December 2019 Democratic debate. Start this clip at 2:45:25. Being that it was held right before the holidays, the moderators asked the candidates to either “give a gift” to a fellow candidate, or “ask for forgiveness” from one.
At this point in the race, there were five men on the stage, and two women (Warren and Senator Amy Klobuchar). Check out what Warren says at 2:48:10: “I will ask for forgiveness,” she said. “I know that, sometimes, I get really worked up.”
To summarize the answers from every candidate: the five men on stage all “gave a gift” (in the form of themselves, or their policies, to their voters). And both women — Warren and Klobuchar — asked for forgiveness. For being too passionate, or too aggressive. It was living representation of the myriad studies that suggest a gender double-bind for women: take care, but be viewed as less competent. Take charge, and be viewed as less likable.
Democratic strategist Meredith Kelly summed it up well: “male candidates have the benefit of being judged by their potential for greatness. Women have to bring receipts.”
This is why we need to talk about Elizabeth Warren. It’s why we need to keep talking about Elizabeth Warren. Because for a candidate that was often viewed as the most competent; that was a frontrunner in polling and fundraising; but that failed to place second in even one state caucus or primary; you simply cannot tell the full story without assessing the role that gender played in shaping its ultimate narrative.
It’s why we must push ourselves to question not just our own biases — but to take stock of how a system was built around them. And it’s not just about raising awareness. That’s why bias training or implicit association tests cannot make meaningful impacts in reshaping our world views.
We must push further. We must demand structural change from the institutions, the organizations, and the leaders around us.
Because it’s not just about whether a woman wins or loses.
It’s about making sure that the next generation of women choose to show up to the fight.
