White People’s Friendships are Racially Limited

While the average Black American has eight White friends, and Latin people in the USA have an average of 19 White friends, the typical White American only has one friend each from Black, Latin, Asian, mixed race, and other racial minorities.

Drawing on a nationally representative dateset (the 2013 American Values Survey, N=2,317 people), research by the Public Religion Research Institute finds that finds 91% of White people’s friends are White. On average, the rest of their friends include only 1 Black person, 1 Latin person, 1 Asian person, 1 mixed race person, 1 other race person, and the rest are unsure. In contrast, 83% of Black people’s friends are Black, and 64% of Latin people’s friends are Latin, and both Black and Latin people have more racial variability amongst the rest of their friendship networks.

Looking deeper at close friendships where people discuss important matters, 75% of White people have “entirely White social networks without any minority presence.”

Researcher Robert Jones and colleagues argue the lack of diversity amongst White social networks has a negative impact on civil society. White people lack a personal connection to Black history and culture because they don’t engage adequate information on these issues. As a result, they do not fully recognise the marginalisation that Black people experience. As such, Jones argues, White people are not “socially positioned” to understand the significance of events at Ferguson and other racial justice and civil disputes.

These data are backed up by a recent Pew Research study which finds that only 37% of White Americans think Michael Brown’s death and subsequent events in Ferguson raise important issues about race, in contrast to 80% of Black people who do see the connection to racism.

In sociology, we use the concept of homophily to explain the structure of social networks. This word literally means “love of the same.” In sociology, this concept measures how little people mix outside their groups, and the consequences of this lack of intermingling. On the outside, friendship groups seem diverse because we tend to think of relationships in reference to individuals. For example: I have a friend who is outgoing and likes horror films and manga comics; I have another friend who’s quiet and likes 70s rock music; another friend likes going to the art gallery and reading Margaret Atwood books; and so on. The fact is, that most people tend to know people similar to themselves where it really counts: along racial and socio-economic lines.

Friendships are not just a source of interpersonal support. Friendships connect us to a world of opportunities: access to new sources of information, a connection to jobs, an inroad to help us navigate social institutions (such as education), an entry point to political power, and other valuable social and economic assets (in sociology, we call this social capital).

Research shows that lack of diversity in White people’s friendship circles has a societal impact, in that it stagnates social change. White people hold the balance of social power, whether they like to admit to it or not. People of colour find ways to connect with White people, but the reverse is not true of the majority of White people. Friendships don’t just “naturally” happen; they aren’t even the strict outcome of personalities or personal preferences. Social relationships are one clear way in which power, resources and the status quo are maintained.

Learn More

Study: Daniel Cox, Juhem Navarro-Rivera, Robert P. Jones (2014) Race, Religion, and Political Affiliation of Americansā€™ Core Social Networks. Washington: Public Religion Research Institute.

Graphic: Washington Post: http://buff.ly/XUw3oz

Pew study: http://buff.ly/XUw3oA

On homophili:

A classic study: http://buff.ly/XUw5wS

Broader impact on social organisation: http://buff.ly/XUw5wT

Race and impact on elite structures in society: http://buff.ly/XUw5wU

6 thoughts on “White People’s Friendships are Racially Limited

  1. Interesting. Thinking of reposting with the header, “I am your one black friend.” šŸ˜‰

    As a side note, Washington Post, I’m guessing the rounding somehow throws all the numbers of on the chart. It claims 100 friends for both groups, but the totals are 99 and 101.


  2. That’s the spirit, Zoey ColeĀ it is sad and funny when you think about it. Unfortunately, it throws off the whole, “I’m not racist I have [one] Black friends!” argument, since three quarters of White Americans don’t even have any Black friends. It also throws into question the faux argument about “post-racial” societies where race is supposed to be a non-issue.

    Yes the figures are simplified for visualisation.


  3. What’s the null hypothesis?

    In a country where there is a larger fraction of white people, if people were friends with 100 others, then most of them would have more white friends. But that would also be true for black people.

    So the biases seem to be in both directions. Only a real quantitative analysis using e.g. the Kullback-Leibler divergence with the random friends distribution would tell us which ethnicity is further from random.

    What you’re really showing so far is that there is preferential attachment when making groups of friends and that may well be due to simple geographic distribution…


  4. HiĀ Romain BrasseletĀ Read the original link to the study where this comes from and then the other links I’ve provided. Yes, there are more White people in America (63% are White non Hispanic/Latino), but that does account for the lack of racial intermixing by White people. As Pew Research data show, White people self-segregate – that is, White people do not mix with other races on average on purpose (http://goo.gl/UzEbhL). Again, read the studies I’ve linked to before dismissing data you don’t like. And by the way – there’s no such thing as “simple geographic distribution.” Racial segregation is a result of racist practices. I’ve written about this here (http://goo.gl/W1uN9u).


  5. The Other Sociologist sorry I don’t have time right now to read everything, but quickly browsing I haven’t seen any clear-cut data analysis. But I admit I may have missed it.

    My question is whether the distribution of people explains the lack of interracial mixing by Black people. Could it be that Black people also self-segregate? I’m just wondering here. And I try to see to what extent the data says something about that. You see?

    BTW, I very much like this data, I have no problem with it. I just want to make sure the interpretation is solid and not biased. No ideology involved. Pure data analysis.

    Finally, I agree that segregation is also the result of racist practices, but not only and they might only be mild, see for example the Schelling model. I’m just wondering, once there is segregation, to what extent it reinforces unmixed social networks. In any city (segregated or not) people are more likely to be friend with neighbours, so in a segregated city, people are more likely to be friends with people of the same ethnicity. Now, finding which feature is the origin of the others is a hard problem.


  6. Romain BrasseletĀ The data are not biased otherwise I would not post this, but perceptions of data can be biased, such as when people refuse to accept data they don’t like because it conflicts with their personal politics or worldview. Racism describes a system of oppression that is reinforced in many ways. Residential segregation is one way this plays out. It is White class interests, not Black self-segregation, that keeps this system in place.Ā I’ve addressed all of this and the other issues you raise in the links provided, including a critique of the Schelling model. If you’re serious about wanting to learn more on this topic, you’ll read the links. šŸ˜‰


Comments are closed.