• U.S.
  • Race

It Might Be Possible to Reduce Bias in Kids, New Research Suggests

14 minute read
Updated: | Originally published:

Once upon a time, not so long ago – but before the final seasons – millions of adults parsed the fine details of the televised struggle for power and influence between the Starks and Targaryens, the Lannisters, Baratheons and Martells, and, of course, their shared nemesis, the White Walkers, as if each Game of Thrones episode contained the only known key to the most noble and orderly form of social organization.

Something about the entirely fictional world created by novelist George R.R. Martin and his collaborators at HBO seemed to sharpen viewers’ ability to perceive earned and unearned privilege, inherited disadvantage, what law and tradition can do to sustain them. In recaps and fan forums and watercooler conversations, people often went on at length about all of that and what stops those at the top from caring about or even noting what’s happening to those at the bottom. 

In many ways, the ever more fractious debate happening across the country about the content of school curriculum, particularly what should be included in the teaching and learning of American history, is a debate about what happens if children are told the entire truth, a more flattering, even fictional version of history, or the spare facts with as little explanation as possible.

More From TIME

So, when I learned about a new study examining what, if any, effect identifying who created and who benefits from structural inequality has on how children ages 5 to 10 think about economic inequality and whether they support reducing it, I was intrigued. The study exposed 206 kids, most from the United States, to the differing fictional housing, job, and, shall we say, “life extras” two groups experienced in a made-up town using images and stories. The two groups were differentiated visually by styles and colors of clothing. The children were told the members of the two groups share a community but “are very different from one another.” Then, researchers measured how the children responded to the inequality depicted and the degree to which it was explained.

The study, published last week in the Proceedings of the National Academy of Sciences (PNAS) was conducted by the social psychologist and New York University doctoral candidate Rachel A. Leshin and the developmental psychologist Marjorie Rhodes, a professor at the same institution. It will give people on either side of America’s history wars some things to consider. What follows is a conversation I had with Leshin edited for clarity and length.

Let’s start with some basics about you. What do you tend to research? 

I describe myself as doing research at the intersection of social psychology, developmental psychology, and cognitive psychology. So I'm basically interested in how we as people from childhood represent the social world in our minds, what our representations of social concepts and social groups and social categories look like. A lot of my research actually looks at children's beliefs about gender categories.

Read More: What Happens When Kids Learn That Racism Can't Be Overcome

I'm really interested in how these ideas develop across the life span with kind of a focus on early childhood and what predicts variation. So some of my research looks at how things like parent ideology or the neighborhoods that children grew up in shape their representations of the world.

This study is also concerned with – I would call it a social concept – hierarchies and equality. And this is obviously looking at how specific messages shape their representations of the social world. 

How frequently does your work tend to deal with race or race and gender together? 

I have a few projects that look intersectionally at race and gender together. That's kind of surprisingly a novel research question. There's a kind of theoretical tradition – not a good one – in developmental psychology that assumes that children process social categories like race and gender in isolation from one another. And so there's a pretty stubborn tendency for studies, especially those that predate the last five or 10 years, to control for gender when studying race and control for race when studying gender. I have some projects that look at what happens when you don't. 

Basically I found that children do see the world intersectionally. So patterns of perception and stereotypes are differentiated when you look at Black women versus white women and Asian men versus Black men. Children are doing this from as early as 5. 


There's actually been a lot of different studies that have popped up in the last five or so years that have looked at explaining inequality structurally to children. But we felt that what could use more attention is this question about responsibility and creation of the structure. So we took a theme that clearly has resonance in the field right now and tried to see if we can, perhaps more robustly, shift attitudes and behaviors by adding this extra piece into the explanations, which to our knowledge hadn't been added in past structural interventions.

Could this be perhaps the missing puzzle piece?

Toogits (on the left in green) and the Flurps (right in blue).
Toogits (on the left in green) and the Flurps (right in blue).Courtesy of Rachel Leshin

Could you explain exactly what you all did in the experiment? It seems that the kids participating were given three stories. Is that right?

Yeah, absolutely. So I'll say that the majority of the paradigm [stories or conditions shared with the kids] was identical. The content of the explanation was the only thing that varied across the three conditions. But the setup of these two made-up groups – one with more resources and opportunities and one with fewer – that was all identical. So we said this “Toogit” lives in this house or this “Flurp” lives in this house because dot, dot, dot.

Then children in the “high-status” power condition were told that it was because of rules that Toogits – which is a high-status group – made up a long time ago when their grandparents were kids and that a long time ago Toogits had the power to make the rules for everyone. They made rules about what houses they can live in, what kinds of jobs they can get. And then we say Toogits made up rules that would be good for themselves and bad for others.

And then the “third-party” power explanation was actually quite similar. It was still a structural explanation about rules that were made up. But we said a long time ago, people who got to make the rules made a rule about where Toogits and Flurps could live, which types of houses they can live in, etc. There's some passive voice in there and we weren't implicating a member of either group. And then we said, “because this happened to their grandparents,” rather than “because of the rules that Toogits made up.” So all through the explanation we're referring to this third-party group quite passively and quite ambiguously.

Read More: Parents Should Have a Voice in Their Kids' Education But We've Gone Too Far

In the third condition, which was a control condition, we really didn't give any explanation. We just said it's been like this for a long time, since their grandparents were kids. For a long time now, Toogits have had these jobs and lived in these houses and Flurps have had these jobs and these houses. We just kind of reiterated the status inequality, but we didn’t provide any explanation, which we figure is a very popular strategy to use with young children. So that was a really important piece of the puzzle to include.

After that, we did some checks to make sure that children were understanding. And then, we had different measures of children's responses to the inequality. That included a measure of group bias, a measure of resource allocation – so whether they want to rectify the inequality by giving more to the low-status group, keep it as is by giving equally, or kind of perpetuate it – and then a few other measures.

What did you refer to these two groups as? What are they called?

They're truly odd, I know. So that high-status group we labeled the Toogits and the low-status group we label the Flurps.

But it's a common strategy in developmental psychology or psychology in general, when you're trying to understand a pretty basic process like how an explanation about a group shapes attitudes. The idea is to intentionally strip it of any real-world meaning so that the conclusions are based on this bare explanation rather than any prior knowledge or beliefs about specific real-world groups.

You're saying the group names and stories take kids into an imaginary space where you don’t bring into the experiment all the things that they know about the world?

Right, exactly. 

Got it. Now, when I read line one of the study, I thought, oh boy, there are people who would definitely disagree with this: “Children begin to participate in systems of inequality from a young age, demonstrating biases for high-status groups, and willingly accepting group disparities.” Many people have this notion of children as pure little hearts incapable of bias. Could you explain?

Absolutely. So there are these kind of pro-wealth or pro-high-status biases that you see from a pretty young age. Children are more likely to show more positive attitudes toward those with more resources, basically. And at the same time, children are also able to kind of link status with other social category dimensions. So children can make inferences, for example, about which racial group lives in which type of house from an early age. There's this early recognition of status, often defined materially, and then biases that emerge from that at quite a young age. So those would be attitude measures like, who do you like more, who do you want to play with, etc. 

And we find that – this is not research coming from our lab – children do accept the status quo pretty easily and sometimes perpetuate it. Sometimes they do rectify it. There are certainly different patterns that happen, but children often show biases for high-status others.

How early do you begin to see some of these patterns take shape and children?

Elementary-school age. So, kids in the range we studied. We had kids from 5 to 10, just turned 5 to almost 11.

The next line that stopped me was, “relative to those in the two other conditions, children who heard a structural explanation that cited the high-status group as the structure's creators responded to the inequality more adaptively: they showed lower levels of bias, they perceived the hierarchy as less fair and they allocated resources to the low-status group more often.” That means your conclusion is that structural explanations can be effective with children but only when they also implicate the high-status group. Please explain that to our readers. 

Yeah. The changes-to-the-structure piece, it's related to this rectification-of-inequality or allocation-of-resources issue. So that was the question, again pretty distilled for our child participants. 

We asked about the distribution of candies at a school. Inequality is couched in terms of material resources. 

We certainly think it's related, but I would say we didn't think about it as an effort to rectify the specific social structures that caused the inequality.

So, while children are able to understand candy inequality, and to grasp the way that structures can contribute to candy inequality, if the rule makers or people who caused the inequality and the fact that they benefit or benefited from the rule isn’t identified, then there’s no reduction in the bias children tend to have toward the lower-status groups and no willingness to redistribute candy resources. Is that right? 

Exactly. And then there's this third measure that we kind of captured within our responses to inequality, which is perception of the status hierarchy as unfair. So we asked, is it fair or unfair? And then: how unfair or fair is it? 

The most common distribution pattern across all three conditions for the resource-allocation question is equal allocation. A really common finding in literature is that kids have the fairness bias, they tend to allocate fairly. The difference is [this study found] many more kids chose to allocate preferentially to the low-status group in the high-status power condition. Most kids do think it's unfair in the control condition where you just describe this group has more and don’t explain why or what that means. So it's a relative difference.

Then why might it be important not to be nebulous or passive when talking about or explaining inequality to children? 

So, we didn't test this mechanism directly here, but our idea is that it has something to do with children's tendency to assume what is is what should be. This is sometimes called “is to ought” or a “descriptive to prescriptive.” Our thought was that in just describing a structural inequality, and chalking it up to rules that were made, or a kind of ambiguous creator, children might then make the leap that this is how it is and so it's OK. In some ways. 

And most kids did say it was unfair, so maybe it's not that it's OK, but it’s more OK than it might otherwise be. So the idea here is that assigning these selfish or self-interested motives to the high-status group might be a way to disrupt that intuitive link that children have. 

Is there an age at which most people move away from or outgrow assuming what is is also what is right?

I'm not sure I have the exact answer to that.  But it's certainly the case that at least we assume that adults don't or don't frequently engage in that type of thinking. And what's interesting, I think, is that a lot of the research with adults on structural explanations finds positive outcomes of structural explanations without implicating a high-status entity in particular. So it almost seems to be the case that specific component may be particularly relevant for children because of the “is ought” link that exists for children. 

This is probably way too in the weeds for your readers. 

Well, maybe not. Right now we're having this national debate about how children should be taught about history. If we are a country that continues to move in the direction of extracting out the full explanation for the inequality that is abundantly clear in our country, that children are capable of spotting, what does your research say about the likely effects of that decision? 

Yeah, so that's a great question. In some ways, a tricky question.  

I think the link here, which will be really important for future research, is this question of whether children respond to these explanations in a similar manner when the explanations pertain to a real-world social category like race. It's certainly a possibility. And that's clearly part of our framework for thinking about this. It can also be applied to categories other than race–gender groups, religious groups, what have you. 

I think it's an open question if that leap can be made. For a real-world social group, the kids would potentially be a part of one of them or have prior beliefs or knowledge about them. But I think that this suggests explanations could be beneficial in shifting children's attitudes and thinking about inequality.

I wonder if there is anything that you would say would be critical for adults to understand about their own conversations about inequality with children outside a school setting? 

I really think that this research suggests, potentially, that identifying and being specific about the groups responsible for creating structures of inequality is important and could have the power to shift children's attitudes and beliefs about inequality, broadly, about lower-status groups, however that's defined. 

Explanations are really important in shaping children's attitudes, especially if the goal is facilitating beliefs that are more egalitarian in nature. That is, this dimension that connotes who is behind the structures, who's responsible for creating them, and even that there is someone who is responsible for creating them. This study suggests that could be an important dimension of these conversations.

Correction, August 28

The original version of this story misstated the name of the researcher interviewed and her job title. It is Rachel, not Rebecca, and she is a doctoral candidate, not a professor.

More Must-Reads from TIME

Contact us at letters@time.com