Race and the Internet

Let's talk about race.

Well, this'll be good.

The Internet is an arena where we are disconnected from defining physical characteristics. Things like race, gender, and other personal traits are only a part of your online identity if you want them to be. But of course, race still matters online despite being invisible.

How has the Internet impacted racial identity? How does race emerge on the Internet? As someone who couldn't possibly do the subject adequate justice, I will try to explore these questions.

First, let's employ some statistics. What is the racial breakdown of the Internet?

The U.S. census provides us with information on general population demographics, as well as the proportion of people in each demographic that use the Internet. The census data is from 2010, and the online demographic data is from 2011, so there will not be a perfect way to compare the two sources of information. That said, this data can still give us a rough idea of the digital landscape.

Of course, not everyone we see on the Internet is going to be American, which is an assumption that we're making with this data. We would expect far more Asians on our graph if we include the hundreds of millions of Chinese users in mainland China. However, as we've mentioned before in my previous post on China, mainland Chinese internet users and American internet users rarely interact with one another.

When Americans talk about the Internet, they are generally talking about a subset of the Web that they are capable of accessing and understanding. This means that the Internet to most Americans is just the part of the Internet written in English (just as the Internet to most of China is the part of the Internet written in Chinese). For the sake of brevity, I will be using the term "Internet" to refer to the English Internet in this blog post.

The census data doesn't directly tell us the proportion of people actually on the Internet. That's where we'll need to employ some math. If you want to peruse my calculations, you can follow this link. The result of those calculations is the following chart:

This is the census of the Internet. This is presumably the picture you would get if the Internet were to conduct a census on itself.

The thing that should leap out to you is that the Internet is mostly white people. In fact, the Internet over-represents white Americans in general - America's actual population of white people is 63% of the total population. Asians also get over-represented online, jumping from 5.1 to about 5.9% - though, of course, there are far fewer Asians overall. These numbers are not so kind to blacks and Hispanics, however - the black population goes from 13.1 to 10.4%, and the Hispanic population goes from 16.9 to 13.7%.

I emphasize that this graph should be thought of as the American Internet's national census. In fact, limiting our Internet census to American users makes the Internet seem less white. If we want to talk about the "English Internet", then that will include the UK, Canada, Australia, New Zealand, and other countries that have a lot of English speaking users. The countries I listed are all very white as well, and would probably contribute to the white majority.

Admittedly, what you don't see in this data is the extent of what is called the "digital divide". This refers to how a lower proportion of blacks and Hispanics use the Internet compared to their white and Asian counterparts. For example, according to that 2011 census data, more than 40% of black people reported not having a home Internet connection, while less than 25% of white people reported the same. Granted, this divide has been narrowing over time, especially with the advent of mobile Internet, but it still exists. Reports on the digital divide have been done before, as it carries pressing implications for marketing strategies and social inequalities.

But notice what happens when we start looking at the American Internet population, rather than at the American population that uses the Internet. Although about 40% of black people reported not having a home Internet connection, it would only take a 3% increase in online black presence for the online demographics to numerically represent real life. The digital divide is a pressing concern in discussions of getting more people to use the Internet, but actual online representation isn't too off the mark from real life.

So how do different demographics spend their time on the Internet?

Alexa doesn't offer information on race for free, so instead we'll turn to Quantcast. Quantcast compares a website's demographics to the demographics of the (American) Internet. For example, a website with a demographic value of 120 for white people means that a random visitor on that website is 1.2 times more likely to be white than a random Internet user. Quantcast's most comprehensive visibility is in the United States, and their averages primarily reflect American Internet habits. This is good, because it should match up with our use of the American census.

I fished around Quantcast's website listings for a bit, cherry-picked some websites that have been pertinent to previous blog posts, and corrected values based on our calculated Internet population. The results? Well...

There's a lot going on here, so I'll explain it bit by bit.

Everything under "Raw Representation" is data straight from Quantcast. As explained above, each racial group is given a number that describes how the website population represents the general Internet population. Like Alexa, this is a very relative figure. The bars graph gives you a visual aid for how over-represented or under-represented each ethnic group is.

Everything under "Population" is calculated. I took each Raw Representation number and multiplied it by the proportion of the racial group of the Internet. This gives us a very rough look at the racial breakdown of each website. You'll notice that the percentage values don't quite add up to 100 - they usually hover between 95 to 110. Again, not ideal, but for our purposes, this is close enough to 100 to give us a general idea of the racial environment.

Some of these data say "internet average" while others say "US average". This is a quirk of Quantcast. The ones that say "US average" indicate websites that actively work with Quantcast to verify their user statistics. The ones that say "internet average" haven't contacted Quantcast at all. In this case, Quantcast is forced to estimate the demographic information based on third-party sources from around the web (hence, they use an average derived from Internet statistics to determine representation levels). The "internet average" data points won't be as good as the "US average" data, but again, they'll give us a general idea.

The "Updated" column represents the last time Quantcast updated their statistics on each website. Quantcast doesn't actually list its "internet average" data points anymore, so I had to fish around Internet Archive to find archives of Quantcast pages back when they did post internet averages. The ones with November 2013 dates are often "US average" data points, so their information is of course better. But again, I'm comparing 2012-2013 Quantcast data to 2010 census data and 2011 web data. To parrot what I've been saying, this isn't ideal, but it's giving us a general idea.

So are you noticing any trends yet?

White people have a strong presence in every popular realm of the Internet. Even when we account for the comically large range of error in my data (this data's usually within 5 percentage points of adding up to 100%, but a few here and there go all the way up to 110%), white people remain the majority of the audience in most of these websites. Even when they're relatively under-represented, Caucasian numbers very rarely go below 50% of the population.

This phenomenon goes both ways. Despite Tumblr's over-representation of Asians, Asians don't even break 10% of the total Tumblr population. YouTube, by representation, is very favored by non-whites, but by absolute numbers, white people still constitute a YouTube majority.

Going forward, it will be important to look at both the relative representation values as well as the absolute proportion values. OkCupid skews mostly white people, Hispanics really like DeviantArt, and the Bleacher Report (a sports website) approximately brings white people and black people together in the average proportions.

Some websites don't have any statistical data out there. Pinterest has nothing listed for demographic information, and doesn't even have any old demographic records in Web Archive. 4chan, Cracked, Buzzfeed, and several other websites are quantified with Quantcast, but have deliberately requested Quantcast to keep their demographics information private.

Still, let's work with what we've got. How do news venues stack up in terms of demographics?

These are more progressive websites, ranging from traditional news like NBC News to vaguely-liberal video content like Upworthy. We do see some greater representation of Asians in the Guardian, but there is a striking lack of minority readership. You'd certainly expect this from conservative news (and surprise surprise, conservative news sites also lack minority readership), but progressive websites give us an expectation of seeking broader representation. Granted, progressive websites do boast more diversity than conservative sites, but you can hardly call them "diverse".

The unfortunate truth is that what is "progressive" can mean many, many things. The Left is hilariously splintered, to the point where progressive movements aren't necessarily always going to be friendly to minorities. We see this with trans-exclusionary radical feminists, for example. Perhaps minority groups generally don't feel like they're being spoken for in these websites. Perhaps some perspectives are missing from these progressive outlets that would draw more varied viewership in.

Where can we find more diverse news demographics?

The fact that Infowars seems to draw greater numbers of blacks and Hispanics than actual news sites should be worrisome. It could be a symptom of not having enough representation on actual news sites, and InfoWars happens to be one of the most high-profile "alt news" websites there is.

We should bear in mind that Infowars isn't a real news site, while these other website examples are in fact legitimate. However, with the exception of Telemundo, these other websites aren't very prominent in broader public consciousness. NGB and Global Post are both websites that focus on international news. The Atlanta Journal-Constitution does a good job in appealing to black people, but this makes sense considering that the AJC is the only major daily newspaper in the Atlanta area, and Atlanta is widely regarded as a center of black culture. As someone not from around Atlanta, I'd never heard of the AJC before.

That last one - the Grio - is a news website with content specifically targeting a black American audience. There are in fact a number of these around the Internet, though they tend to come in the form of an extra limb to a pre-existing web site. The Grio is an extension of MSNBC. Huffington Post offers Black Voices. Washington Post has The Root. These websites do succeed in drawing greater black viewership, but the fact that they're offshoots of larger news sites means that they aren't necessarily run by the people they're supposed to represent. It's also a bit strange that these websites are so partitioned from their main websites -  if you actually look at The Root, you don't see mention of the Washington Post until the very bottom.

There are also a number of other, smaller websites that aim for a racially-minded perspective. ColorLines is a fantastic resource for news analysis from a more diverse roster of journalists. Racialicious is a blog that focuses on racial issues and pop culture. Still Furious and Brave is a group blog run by sociologists and activists that often features solid racial commentary. Mixed Race Studies is an academically-minded blog that focuses on biracial and multiracial individuals. These smaller websites often have the advantage of being completely run by people who directly represent their audience.

Even in news outlets like the Grio, you see a strong white viewership, but this is a case where it's a net positive. This is perhaps the great victory of these news sites - not only do they have great racial representation, but there are white people actively reading about racial perspectives. Again, many things can fall under the progressive header, and it is heartwarming to know that there are news sources that can bring together a diverse audience. This would be a good place to have some values on actual website traffic by number of users, so that we could compare absolute numbers of visitors of websites, but unfortunately my analysis did not extend that far.

I worked up some other data too, which you can find if you follow this link into the footnotes (they're more for fun than for real discussion, though they're certainly interesting). It is important that we establish some data on online hubs before we move on, though.

So, we've strongly established that there's a clear white majority on the Internet and on many popular websites. How does this manifest itself in the way we interact online?

An interesting thing about the online setting is that it's possible to interact with multitudes of people without ever knowing anything about their age, gender, or race. On the Internet, we are truly colorblind. Unfortunately, being colorblind is not necessarily a good thing. If you're in an environment with very little diversity but you can't see that there is very little diversity, then how can you be aware that your environment lacks diversity?

One can argue that the colorblind quality of the Internet removes any first impressions that might come with race. This is true, but it also enables users to project qualities about themselves onto other people online. In some ways this can be humanizing - but in other ways, it erases important differences in personal experiences. And yes, being unaware of differences in personal experiences has been a problem since before the Internet, but now even the visual cues for differences in perspective is gone.

Conversations about racial issues tend to pop up in online hubs like 4chan and Reddit. Given these demographics data, it's possible that the contributing perspectives on these discussions are going to be mostly white people. That means that any misconceptions about race that are common to white people are going to be present in these conversations with very few perspectives offering dissent or contradiction. Nobody is necessarily aware that they're just talking to other white people about race, so it becomes harder to call out the gaps in contributing perspectives. In this way, misunderstandings about racial issues continue to fester.

Over the years, these misunderstandings can build up. The online echo chamber can make these misunderstandings seem like a consensus on reality. The result of this is a gradual formation of a racist environment in these online hubs. This toxic racial environment is capable of forming anywhere on the Internet. This echo chamber effect makes these hubs vulnerable to any bigot with an agenda online.

And yes, of course bigots have found their safe havens online. The website Stormfront is explicitly devoted to white nationalism. It is considered to be the most visited white nationalist website on the Internet, having been around since 1995. They'll occasionally rear their ugly head during major public events, but for the most part they're in the shadows of the Internet, seeking to proselytize as many as they can to their cause. On the Internet, Stormfront and other white supremacist websites can thrive. Since nobody really wants to deal with them, their websites continue to only accrue like-minded bigots. Hell, it wasn't until mid-2013 that Reddit finally started regulating its more unsavory racist corners, taking some responsibility for the content on its website.

Granted, there is some self-awareness on the Internet.  A UCLA girl making a racist video about Asians was quickly mocked and shunned among online goers. Reddit is continuously mocked for its blatantly racist corners (among other things). If the online racism is sufficiently overt, then you can expect it to get derided on the Internet.

You also have online racial humor focused around white people pointing out their own whiteness. The blog Stuff White People Like was an exercise in white self-effacement, pointing out things that white people of a certain demographic like to do and say. Black People Love Us was a lampooning of white attitudes towards black people (though, looking through its fan mail, not every viewer seemed to be in on the joke).

However, there's also a gray area of ironic racism online. If you were to look through the long list of advice animals in existence (not restricted to animals, but often represent general archetypes conserved through social interactions), you'll find a few centered around race. Successful black man opens with a negative African American stereotype and then follows up with a statement that makes the previous stereotype seem like a misinterpretation. Ordinary Muslim Man does the same thing with a picture of a Middle Eastern Muslim man. Successful White Man is similar, but instead opens with a positive stereotype and follows up with a negative statement.

Sure, these memes can be funny (hell, I find some of them funny), but we have to admit that they're only funny if you're in on the joke. What's the joke? That these individuals, because of their race, fit the (often negative) stereotype of the first line, but oh - surprise - they don't. This kind of content is certainly not as big a deal as the blatantly racist content that we mentioned earlier, but that doesn't mean that it isn't problematic. These memes can only exist in a context where these stereotypes are wrongly believed by some group of people. What's worse, they can end up inadvertently communicating these stereotypes to new audiences.

Then you have the meme Irrational Black Woman, which seems to skip any pretense of irony. The meme does nothing but put down black women. The fact that it's a black woman might be more important here - these image macros have a precedence for showing sexism. Yet her race is specified in the title of the meme. High Expectations Asian Father is another meme that seems to skip the conveyance of ironic racism, going straight for the Asian "model minority" stereotype.

In the mid-90s following the Mississippi River flooding, a photographer took some very intense photos of a black homeless man. These photos eventually made their way to the Internet, and presumably because the black man in the photo looked so intense, someone plastered the phrase  "You Gonna Get Raped" onto one of the pictures. The picture quickly became a meme. Allegedly, by the time it had gone viral, the homeless man had cleaned up his act...and then lost his job over the photo.

Racist memes and pictures can come in many forms. Some are less subtle than others, but then again, that's how racism seems to go in general. Racism can be as subtle as moving from one social network to another because of the crowd you see online. Misogyny can be equally subtle, but since women constitute about half of the online population, they are more capable of representing themselves. Not only is the non-white audience so amazingly scarce online, it's also splintered into smaller racial groups.

Despite minority groups being in the, uh, minority on most websites, they can get their voices heard on social media platforms. For example, if you look at the data, you'll notice that Twitter's got a rather prominent representation of black people. This has allowed for the emergence of what's known as Black Twitter.

Obviously, Black Twitter is a subculture in multiple respects. It is not a monolithic representation of black people, nor is it a monolithic representation of Twitter, nor is every participant in Black Twitter a representative of everyone else on Black Twitter. It is, however, an interesting corner of the Internet that seems to have similar cohesion as other online hubs. Every now and then, members of black Twitter create waves in the greater online sphere by speaking on the issues that face them.

When word of Paula Deen's blatant racism went public, word quickly spread on Twitter under the #paulasbestdishes hashtag. Paula Deen ultimately lost several endorsements from the whole affair, with several former employees coming forward to talk about her offensive behavior. Granted, those former employees might have come forward without the involvement of Twitter, but the wave of online presence must certainly have helped.

Then of course, there was the Geroge Zimmerman trial that (eventually) followed the death of Trayvon Martin. The Internet has had an ongoing role in the developments of the case. When Martin's parents had turned to Change.org to focus attention on their son's murder, word of the case quickly spread to Twitter. When former talk show host Geraldo Rivera suggested that Trayvon was to blame because he was wearing a hoodie, Twitter retaliated. Finally, on April 11th of 2012 - nearly two months after the shooting - George Zimmerman was arrested. When George Zimmerman was acquitted in his 2013, Twitter flared up in anger. Throughout the whole process, Twitter was one prong in the multi-faceted push for activism.

Even feminism for women of color has gotten use out of Twitter. In August of 2013, self-proclaimed male feminist Hugo Schwyzer had a public online meltdown, admitting to taking advantage of several women during his career and calling out several women of color. Now, Schwyzer already had a bad reputation, especially among women of color. But when this Twitter meltdown happened, Schwyzer somehow continued to find support from white women within the feminist movement. One woman of color noticed this, ranted to Twitter about it, and the #SolidarityIsForWhiteWomen hashtag was born. White feminists unaware of their own whiteness have since become a running joke among women of color.

The #SolidarityIsForWhiteWomen tag was doubly important because it brought discussion about race and gender into a more public sphere. Women of color have been in the unfortunate position of being routinely marginalized in feminist movements as well as marginalized in racial movements. The concept of being a part of multiple marginalized groups but not finding much solidarity within either community is explored in the study of intersectionality. Twitter was a great platform to bring some intersectional discussion to light.

Although Twitter is restricted by a short character limit, Twitter's hashtag infrastructure allows for rapid sharing of themes and ideas. Tumblr has similar functionality, though its freer-form blogging format allows for more substantial content to be shared. Both Twitter and Tumblr have important roles in online social justice. The hashtag infrastructure has enabled a minority with a righteous message to find a voice.

But of course, racial representation isn't just about social justice. There's also the rabbit hole of media representation.

Asian Americans are in an interesting position online, in that while they have a smaller population in America compared to other minority groups, they are proportionally represented - if not over-represented - on the Internet. There's also been an unfortunate tendency to exclude an Asian perspective from discussion on racism, despite Asians having to face fetishization and model minority-related stereotypes. The Internet has been a great platform for Asians to assert themselves as individuals beyond the traditional stereotypes.

Perhaps this has been most visible on YouTube, where one of the most prominent site-wide YouTubers is run by Ryan Higa, a Japanese American from Hawaii. In addition to his incredibly successful YouTube channel, Higa and other prominent Asian YouTubers founded YOMYOMF, a YouTube network that focuses on Asian-American pop culture. Unaffiliated with the YOMYOMF group of Asian YouTubers is fellow YouTuber Freddie Wong, whose works have evolved over time to include full-length webseries and a sprawling network of YouTube shows.

When we talk about media and the big names that run it, we know that we're talking about a very established industry that has not always been quick to shed problematic representations of minority groups. YouTube, however, falls outside of that establishment, and several people at the helm of new media creation are minority groups. The number of Asians on the platform could mean a brand new ownership of what it means to be an Asian American, by Asian Americans.

YouTube, of course, is a wide and varied place. In addition to the prominent Asian Americans that I listed, you can also find very popular Hispanic channels and YouTube personalities. You can also find several great black YouTubers. Beyond YouTube, there's Twitter's own video service, Vine, which allows Twitter users to make videos about six or seven seconds in length. Looking through some Vines, it's clear that the platform also has a pretty diverse array of contributors that make some hilarious video shorts.

But...at this point, I think I'm drifting pretty far from what I'm qualified to talk about. As far as my own background goes, I'm a Middle Eastern American guy, so my experiences only really extend into that realm. That's not to say that there aren't some online quirks to be had on that subject - usually when middle easterners are brought up online, it's usually about the Arab Spring or some kind of racism. But, as far as what it means to be Middle Eastern in America, I think that I'm basically a white guy - hell, the government thinks so too. Honestly, before 9/11, I don't think being Middle Eastern in America even really meant anything to most people. That kind of goes to show you how arbitrary racial distinctions can be.

Ultimately, my musings on how some people from other racial groups use the Internet come from an under-informed outside perspective. If you want to hear real discussion on black culture and the Internet, talk to black people online. Same for Asians. Same for Hispanics. Same for Native Americans (but, uh...I hear the digital divide impacts them pretty hard).

The universal point here is that the Internet has empowered the individual to define themselves in however way they like. Race's default setting on the Internet is "invisible". In some ways that's empowering. In some ways it lets sinister ideas come to light. It takes deeper examination to see how our cultural backgrounds translate into the online sphere. I think it shows promise to let racial dialogue find a broader audience and empower our society to understand - not merely tolerate - one another.

In the age of the Internet, race is a new beast with old problems. I'm not the first to write about this, and I won't be the last.



Footnotes

How I calculated the general Internet demographics:
Here's an example calculation. Say I wanted to figure out how many people on the Internet are white. That's like asking the following question: If I were to pick a random person from the Internet, what are the chances that the person I choose will be a white person? In terms of probability, we'd write that like this:

P(You're white | You're on the Internet)

In other words, this represents the probability that you're a white American given that you're on the Internet. Now, if you look at Figure 2 of our census data here, we are given the following information:

P(You're on the Internet | You're white) = 0.762

Ah, it's close to what we were looking for, but not quite what we want. In the census data, they asked every white person about their Internet activity, so we already knew that these were white people being asked. This data represents the probability that you're on the Internet given that you're a white American. It's flipped from what we want!

This is where we can use something called Bayes' Law to proceed with getting the info that we want (and yes, the last time I mentioned Bayes' Law was when I was telling you about a certain online cult, but let's not let math get ruined by crazies who don't know what they're talking about).

Let's remember that all of our information is confined to American demographics here. We are interested in white Americans and American Internet users. World data certainly exists for the Internet, but American census data is limited to Americans. Plus, let's be honest, I think the demographic proportions would be a bit of a landslide if we included the statistics of countries like China anyway.

The probability that you're a white American is also in census data, and the probability that you're an American Internet user is in Figure 1 of that previously mentioned census document. The calculation becomes easy: The probability that you're white given that you're on the Internet is about 0.67, or 67%.

We repeat this for all of our demographics (except for "Other" - Other just ends up being 1 minus the rest), and we get the chart in the blog post.

If you got to this footnote from the link within the blog post and would like to return to the place you were at in the article,follow this link here to go back to the Internet census chart.

Some other fun website statistics:
How do different music websites stack up in terms of demographics?

Okay. What about photos?

How about other social networks/online dating sites?

How about some general gaming/anime/nerdy stuff?

If you reached this footnote from clicking the link within the blog post, follow this link here to return to analysis of individual website demographics.