In 2001, Dna88 joined Stormfront, describing himself as a “good looking, racially aware” thirty-year-old Internet developer living in “Jew York City.” In the next four months, he wrote more than two hundred posts, like “Jewish Crimes Against Humanity” and “Jewish Blood Money,” and directed people to a website, jewwatch.com, which claims to be a “scholarly library” on “Zionist criminality.”
Stormfront members complain about minorities’ speaking different languages and committing crimes. But what I found most interesting were the complaints about competition in the dating market.
A man calling himself William Lyon Mackenzie King, after a former prime minister of Canada who once suggested that “Canada should remain a white man’s country,” wrote in 2003 that he struggled to “contain” his “rage” after seeing a white woman “carrying around her half black ugly mongrel niglet.” In her profile, Whitepride26, a forty-one-year-old student in Los Angeles, says, “I dislike blacks, Latinos, and sometimes Asians, especially when men find them more attractive” than “a white female.”
Certain political developments play a role. The day that saw the biggest single increase in membership in Stormfront’s history, by far, was November 5, 2008, the day after Barack Obama was elected president. There was, however, no increased interest in Stormfront during Donald Trump’s candidacy and only a small rise immediately after he won. Trump rode a wave of white nationalism. There is no evidence here that he created a wave of white nationalism.
Obama’s election led to a surge in the white nationalist movement. Trump’s election seems to be a response to that.
One thing that does not seem to matter: economics. There was no relationship between monthly membership registration and a state’s unemployment rate. States disproportionately affected by the Great Recession saw no comparative increase in Google searches for Stormfront.
But perhaps what was most interesting—and surprising—were some of the topics of conversation Stormfront members have. They are similar to those my friends and I talk about. Maybe it was my own naïveté, but I would have imagined white nationalists inhabiting a different universe from that of my friends and me. Instead they have long threads praising Game of Thrones and discussing the comparative merits of online dating sites, like PlentyOfFish and OkCupid.
And the key fact that shows that Stormfront users are inhabiting similar universes as people like me and my friends: the popularity of the New York Times among Stormfront users. It isn’t just VikingMaiden88 hanging around the Times site. The site is popular among many of its members. In fact, when you compare Stormfront users to people who visit the Yahoo News site, it turns out that the Stormfront crowd is twice as likely to visit nytimes.com.
Members of a hate site perusing the oh-so-liberal nytimes.com? How could this possibly be? If a substantial number of Stormfront members get their news from nytimes.com, it means our conventional wisdom about white nationalists is wrong. It also means our conventional wisdom about how the internet works is wrong.
THE TRUTH ABOUT THE INTERNET
The internet, most everybody agrees, is driving Americans apart, causing most people to hole up in sites geared toward people like them. Here’s how Cass Sunstein of Harvard Law School described the situation: “Our communications market is rapidly moving [toward a situation where] people restrict themselves to their own points of view—liberals watching and reading mostly or only liberals; moderates, moderates; conservatives, conservatives; Neo-Nazis, Neo-Nazis.”
This view makes sense. After all, the internet gives us a virtually unlimited number of options from which we can consume the news. I can read whatever I want. You can read whatever you want. VikingMaiden88 can read whatever she wants. And people, if left to their own devices, tend to seek out viewpoints that confirm what they believe. Thus, surely, the internet must be creating extreme political segregation.
There is one problem with this standard view. The data tells us that it is simply not true.
The evidence against this piece of conventional wisdom comes from a 2011 study by Matt Gentzkow and Jesse Shapiro, two economists whose work we discussed earlier.
Gentzkow and Shapiro collected data on the browsing behavior of a large sample of Americans. Their dataset also included the ideology—self-reported—of their subjects: whether people considered themselves more liberal or conservative. They used this data to measure the political segregation on the internet.
How? They performed an interesting thought experiment.
Suppose you randomly sampled two Americans who happen to both be visiting the same news website. What is the probability one of them will be liberal and the other conservative? How frequently, in other words, do liberals and conservatives “meet” on news sites?
To think about this further, suppose liberals and conservatives on the internet never got their online news from the same place. In other words, liberals exclusively visited liberal websites, conservatives exclusively conservative ones. If this were the case, the chances that two Americans on a given news site have opposing political views would be 0 percent. The internet would be perfectly segregated. Liberals and conservatives would never mix.
Suppose, in contrast, that liberals and conservatives did not differ at all in how they got their news. In other words, a liberal and a conservative were equally likely to visit any particular news site. If this were the case, the chances that two Americans on a given news website have opposing political views would be roughly 50 percent. The internet would be perfectly desegregated. Liberals and conservatives would perfectly mix.
So what does the data tell us? In the United States, according to Gentzkow and Shapiro, the chances that two people visiting the same news site have different political views is about 45 percent. In other words, the internet is far closer to perfect desegregation than perfect segregation. Liberals and conservatives are “meeting” each other on the web all the time.
What really puts the lack of segregation on the internet in perspective is comparing it to segregation in other parts of our lives. Gentzkow and Shapiro could repeat their analysis for various offline interactions. What are the chances that two family members have different political views? Two neighbors? Two colleagues? Two friends?
Using data from the General Social Survey, Gentzkow and Shapiro found that all these numbers were lower than the chances that two people on the same news website have different politics.
PROBABILITY THAT SOMEONE YOU MEET HAS OPPOSING POLITICAL VIEWS
On a News Website
45.2%
Coworker
41.6%
Offline Neighbor
40.3
Family Member
37%
Friend
34.7%
In other words, you are more likely to come across someone with opposing views online than you are offline.
Why isn’t the internet more segregated? There are two factors that limit political segregation on the internet.
First, somewhat surprisingly, the internet news industry is dominated by a few massive sites. We usually think of the internet as appealing to the fringes. Indeed, there are sites for everybody, no matter your viewpoints. There are landing spots for pro-gun and anti-gun crusaders, cigar rights and dollar coin activists, anarchists and white nationalists. But these sites together account for a small fraction of the internet’s news traffic. In fact, in 2009, four sites—Yahoo News, AOL News, msnbc.com, and cnn.com—collected more than half of news views. Yahoo News remains the most popular news site among Americans, with close to 90 million unique monthly visitors—or some 600 times Stormfront’s audience. Mass media sites like Yahoo News appeal to a broad, politically diverse audience.
The second reason the internet isn’t all that segregated is that many people with strong political opinions visit sites of the opposite viewpoint, if only to get angry and argue. Political junkies do not limit themselves only to sites geared toward them. Someone who visits thinkprogress.org and moveon.org—two
extremely liberal sites—is more likely than the average internet user to visit foxnews.com, a right-leaning site. Someone who visits rushlimbaugh.com or glennbeck.com—two extremely conservative sites—is more likely than the average internet user to visit nytimes.com, a more liberal site.
Gentzkow and Shapiro’s study was based on data from 2004–09, relatively early in the history of the internet. Might the internet have grown more compartmentalized since then? Have social media and, in particular, Facebook altered their conclusion? Clearly, if our friends tend to share our political views, the rise of social media should mean a rise of echo chambers. Right?
Again, the story is not so simple. While it is true that people’s friends on Facebook are more likely than not to share their political views, a team of data scientists—Eytan Bakshy, Solomon Messing, and Lada Adamic—have found that a surprising amount of the information people get on Facebook comes from people with opposing views.
How can this be? Don’t our friends tend to share our political views? Indeed, they do. But there is one crucial reason that Facebook may lead to a more diverse political discussion than offline socializing. People, on average, have substantially more friends on Facebook than they do offline. And these weak ties facilitated by Facebook are more likely to be people with opposite political views.
In other words, Facebook exposes us to weak social connections—the high school acquaintance, the crazy third cousin, the friend of the friend of the friend you sort of, kind of, maybe know. These are people you might never go bowling with or to a barbecue with. You might not invite them over to a dinner party. But you do Facebook friend them. And you do see their links to articles with views you might have never otherwise considered.
In sum, the internet actually brings people of different political views together. The average liberal may spend her morning with her liberal husband and liberal kids; her afternoon with her liberal coworkers; her commute surrounded by liberal bumper stickers; her evening with her liberal yoga classmates. When she comes home and peruses a few conservative comments on cnn.com or gets a Facebook link from her Republican high school acquaintance, this may be her highest conservative exposure of the day.
I probably never encounter white nationalists in my favorite coffee shop in Brooklyn. But VikingMaiden88 and I both frequent the New York Times site.
THE TRUTH ABOUT CHILD ABUSE AND ABORTION
The internet can give us insights into not just disturbing attitudes but also disturbing behaviors. Indeed, Google data may be effective at alerting us to crises that are missed by all the usual sources. People, after all, turn to Google when they are in trouble.
Consider child abuse during the Great Recession.
When this major economic downturn started in late 2007, many experts were naturally worried about the effect it might have on children. After all, many parents would be stressed and depressed, and these are major risk factors for maltreatment. Child abuse might skyrocket.
Then the official data came in, and it seemed that the worry was unfounded. Child protective service agencies reported that they were getting fewer cases of abuse. Further, these drops were largest in states that were hardest hit by the recession. “The doom-and-gloom predictions haven’t come true,” Richard Gelles, a child welfare expert at the University of Pennsylvania, told the Associated Press in 2011. Yes, as counterintuitive as it may have seemed, child abuse seemed to have plummeted during the recession.
But did child abuse really drop with so many adults out of work and extremely distressed? I had trouble believing this. So I turned to Google data.
It turns out, some kids make some tragic, and heart-wrenching, searches on Google—such as “my mom beat me” or “my dad hit me.” And these searches present a different—and agonizing—picture of what happened during this time. The number of searches like this shot up during the Great Recession, closely tracking the unemployment rate.
Here’s what I think happened: it was the reporting of child abuse cases that declined, not the child abuse itself. After all, it is estimated that only a small percentage of child abuse cases are reported to authorities anyway. And during a recession, many of the people who tend to report child abuse cases (teachers and police officers, for example) and handle cases (child protective service workers) are more likely to be overworked or out of work.
There were many stories during the economic downturn of people trying to report potential cases facing long wait times and giving up.
Indeed, there is more evidence, this time not from Google, that child abuse actually rose during the recession. When a child dies due to abuse or neglect it has to be reported. Such deaths, although rare, did rise in states that were hardest hit by the recession.
And there is some evidence from Google that more people were suspecting abuse in hard-hit areas. Controlling for pre-recession rates and national trends, states that had comparatively suffered the most had increased search rates for child abuse and neglect. For every percentage point increase in the unemployment rate, there was an associated 3 percent increase in the search rate for “child abuse” or “child neglect.” Presumably, most of these people never successfully reported the abuse, as these states had the biggest drops in the reporting.
Searches by suffering kids increase. The rate of child deaths spike. Searches by people suspecting abuse go up in hard-hit states. But reporting of cases goes down. A recession seems to cause more kids to tell Google that their parents are hitting or beating them and more people to suspect that they see abuse. But the overworked agencies are able to handle fewer cases.
I think it’s safe to say that the Great Recession did make child abuse worse, although the traditional measures did not show it.
Anytime I suspect people may be suffering off the books now, I turn to Google data. One of the potential benefits of this new data, and knowing how to interpret it, is the possibility of helping vulnerable people who might otherwise go overlooked by authorities.
So when the Supreme Court was recently looking into the effects of laws making it more difficult to get an abortion, I turned to the query data. I suspected women affected by this legislation might look into off-the-books ways to terminate a pregnancy. They did. And these searches were highest in states that had passed laws restricting abortions.
The search data here is both useful and troubling.
In 2015, in the United States, there were more than 700,000 Google searches looking into self-induced abortions. By comparison, there were some 3.4 million searches for abortion clinics that year. This suggests that a significant percentage of women considering an abortion have contemplated doing it themselves.
Women searched, about 160,000 times, for ways of getting abortion pills through unofficial channels—“buy abortion pills online” and “free abortion pills.” They asked Google about abortion by herbs like parsley or by vitamin C. There were some 4,000 searches looking for directions on coat hanger abortions, including about 1,300 for the exact phrase “how to do a coat hanger abortion.” There were also a few hundred looking into abortion through bleaching one’s uterus and punching one’s stomach.
What drives interest in self-induced abortion? The geography and timing of the Google searches point to a likely culprit: when it’s hard to get an official abortion, women look into off-the-books approaches.
Search rates for self-induced abortion were fairly steady from 2004 through 2007. They began to rise in late 2008, coinciding with the financial crisis and the recession that followed. They took a big leap in 2011, jumping 40 percent. The Guttmacher Institute, a reproductive rights organization, singles out 2011 as the beginning of the country’s recent crackdown on abortion; ninety-two state provisions that restrict access to abortion were enacted. Looking by comparison at Canada, which has not seen a crackdown on reproductive rights, there was no comparable increase in searches for self-induced abortions during this time.
The state with the highest rate of Google searches for self-induced abortions is Mississi
ppi, a state with roughly three million people and, now, just one abortion clinic. Eight of the ten states with the highest search rates for self-induced abortions are considered by the Guttmacher Institute to be hostile or very hostile to abortion. None of the ten states with the lowest search rates for self-induced abortion are in either category.
Of course, we cannot know from Google searches how many women successfully give themselves abortions, but evidence suggests that a significant number may. One way to illuminate this is to compare abortion and birth data.
In 2011, the last year with complete state-level abortion data, women living in states with few abortion clinics had many fewer legal abortions.
Compare the ten states with the most abortion clinics per capita (a list that includes New York and California) to the ten states with the fewest abortion clinics per capita (a list that includes Mississippi and Oklahoma). Women living in states with the fewest abortion clinics had 54 percent fewer legal abortions—a difference of eleven abortions for every thousand women between the ages of fifteen and forty-four. Women living in states with the fewest abortion clinics also had more live births. However, the difference was not enough to make up for the lower number of abortions. There were six more live births for every thousand women of childbearing age.
In other words, there appear to have been some missing pregnancies in parts of the country where it was hardest to get an abortion. The official sources don’t tell us what happened to those five missing births for each thousand women in states where it is hard to get an abortion.
Google provides some pretty good clues.
We can’t blindly trust government data. The government may tell us that child abuse or abortion has fallen and politicians may celebrate this achievement. But the results we think we’re seeing may be an artifact of flaws in the methods of data collection. The truth may be different—and, sometimes, far darker.
Everybody Lies Page 12