SZ: Let us imagine for a moment that Kamala Harris in November the election against Donald Trump Many commentators would argue that as a black woman she is sexism and racism would never have had a chance in US society anyway. What does the data say about this claim?
Dr. Thomas Schröder: In the event of a defeat, many would certainly argue this way. What I have investigated is the first impression that voters have of a politician with a migrant background. And this first impression is actually very positive. Women even have a bonus compared to men. In experiments, voters are significantly more likely to vote for a woman than for a man.
What about skin color?
On average, people in the US are just as likely to vote for a black American as for a white one. This also relates to first impressions.
How do these candidate experiments work to capture this first impression?
It is a survey technique in which respondents are shown a politician and then asked: “How likely are you to vote for this person?” Participants are also given further information about the politician: their political positions, their party, whether they have children and much more. For my analysis, I included all studies from around the world, with almost 70 percent coming from the USA. In total, there were more than 300,000 politician profiles. Then I aggregated all the data, re-analyzed it and combined it all into one large model. It is a really powerful technique. I did this for both origin or skin colour and gender. In the end, however, I did not publish my results for the influence of gender because another scientist was faster.
Oh, how annoying!
On the contrary, I was very happy that someone had beaten me to the gender issue.
Why is that?
At first, I didn’t trust my own results. This mountain of data suggested that voters were significantly more likely to vote for female politicians and just as likely to vote for politicians with a migrant background. When I presented all this data, many people didn’t believe me and I was skeptical too.
I can understand that.
So I ran my analyses again and always came to the same conclusion. I had done everything right. And then they published these other researchers found exactly the same results. I was therefore very happy about that.

The results of their meta-analysis are in stark contrast to our public debate. We keep telling ourselves a different story.
People often misunderstand statistics. I present average people’s reactions. It may be the case that the average reaction to women is positive, but at the same time there is a tiny, extremely sexist or racist minority that is loud and strongly heard. That just doesn’t mean that the majority of people think that way. To draw that conclusion would be a misunderstanding.
What are the consequences of this misunderstanding?
For example, that party decision-makers strategically discriminate against women or black people and do not dare to put them up as candidates, simply because they believe that they have less chance with voters – which is not the case, at least on average.
Why then is history so dominant that our societies are becoming increasingly racist?
On the one hand, this is due to the small minority mentioned above that is loud and racist. On the other hand, there is the negativity bias: bad news sells better and gets more attention.
Your study also shows that black voters in the US prefer to vote for black candidates. Other identity groups are also more likely to vote for candidates with whom they share their origins. Only white Americans do not show this pattern. If it were the other way around, would that probably be seen as evidence of racism?
Probably. That’s definitely a conclusion you could draw. On the other hand, I think it makes a difference whether the favored group is historically a minority or the dominant group.
White Americans are often described as the most racist voting bloc. Looking at data like this, one could argue that the opposite is true.
Yes, at least in terms of preference for their own group, which they do not systematically favor in the United States.
Have you also evaluated comparable data from Europe?
I looked at similar questions for the Netherlands, Germany and France. I could not find any significant preferences for female politicians. There was simply no gender effect, people prefer women just as much as men.
That would be the goal: content would count instead of identity characteristics.
Yes, exactly. In the Netherlands, Germany and France the picture is very similar in this respect. I also found no discrimination in terms of migrant background. I examined the most common migrant backgrounds and found no discrimination for any of them. Unlike in the USA, in Europe people with a migrant background significantly prefer politicians without a migrant background.
Why is that?
This could be because people with a migrant background are relatively new to Europe. It is only in the last 50 years that many have come to the Netherlands, Germany and France. One way to feel more positive is to favor the dominant outgroup. But it is difficult to say for sure. Religion does play an important role in Europe, however: in the Netherlands, the data suggested that Muslims were more likely to vote for Muslims. In Germany, this effect was moderate and in France it was absent.
And what are the reactions of the majority society towards Muslim candidates?
What I have also seen in experiments is that there is a very strong negative bias against Muslim politicians in the Netherlands, Germany and France. Perhaps it is more socially acceptable to judge someone based on their religion rather than the colour of their skin. But that is just speculation.
Let’s go back to the US. So the bottom line of your meta-analysis is that Kamala Harris has an advantage because of her identity?
Yes, I would definitely say that. But it could also be a disadvantage in some ways. There is a lot of literature on the fact that negative stereotypes about women and black people can be mobilized and activated over the course of a campaign, which may displace the positive first impression. So it could be that latent racism or hidden sexism that is not apparent at the beginning comes to the surface over the course of the campaign.
But one could argue that negative campaigns and stereotypes – with limits, of course – have always been part of election campaigns. Every politician does that. There are also many negative stereotypes about white men.
Right, yes.
So would you say that we as societies are much more progressive and better when it comes to racism and sexism than we tell ourselves?
I think we actually are. And I think a lot is happening and changing at the moment. On the one hand, it could be that we as a society are actually becoming less racist and sexist because we are becoming more aware of it. On the other hand, it could also be that we are becoming more aware of it and therefore learning to hide our racism and sexism better. That is an empirical question that we as researchers need to look into a little more closely.
Will Kamala Harris win the election?
Nobody really knows. And polls are extremely difficult, especially in the US where voter turnout is very important. It depends on who is going to vote. Are people who might be voting for the first time actually being picked up in these polls? It’s really hard to predict. But if I had to put my money on it, I’d say she’ll win.