My research uses Twitter to understand the dynamics of protest and state violence, which also means I think a lot about media coverage.  I have a few thoughts on recent events.

TL;DR: There is very little looting going on across the United States.  There are lots and lots of peaceful protests and protesters.  You would not know that if your news comes from newspapers or television.

It is well known that newspapers prefer to focus on events when they are violent; see, for example, the work of Daniel J. Myers and Beth Caniglia.  ‘If it bleeds it leads’ and all that.  The same is true for television.  Both media sell engagement, not Truth, and negativity drives eyeballs in traditional news just as much as it does on social media.  The problem, of course, is that the sensational text and images become reified as the version of events that happened; we forget they are the version of events that sell the most advertising revenue.

This media bias means that when bad stuff happens, you are more likely to think the bad stuff is the event, not a small part of it.  For example, on May 30, a few dozen looters, possibly organized, descended on downtown Santa Monica while a much larger peaceful protest occurred a few blocks away.  I know this because I read the Los Angeles Times’ story of that event very closely.  If I only read the headline – “Hundreds arrested in Santa Monica amid widespread looting” – or looked at the article’s images, I would think the protest was only looting.  News coverage is dominated by this action, not by the much greater number of peaceful people nearby or peaceful events happening all over Los Angeles.  For example, the next day, my wife and I passed a peaceful protest of perhaps 50 people in El Segundo.  A couple of police cars monitored from a distance and passersby honked in support, but I could find no news coverage of it.

News media are especially biased when protests are anti-status quo, and it is hard to be much more anti status-quo than protesting police brutality or the failed justice system.  This framing is especially pernicious because bystanders really dislike protester violence.  Building on others’ work, coauthors and I have shown that perceived protester violence correlates with subsequent protests that are about 50% smaller than they otherwise would be; see Figure 4a of the paper at that link.  In other words, media focus on a very small part of the event because it drives advertising revenue, and that focusing may cause bystanders to not join a protest.  Media’s biased coverage has real effects.

A more accurate portrait of what happens at protests is actually possible by relying on social media.  My loyal readers will know that Josh Flagg is my favorite agent on Million Dollar Listing: Los Angeles, and he recently posted a story on Instagram from the Beverly Hills protest on Saturday.  The crowds were much smaller and peaceful than you would think from reading newspapers.  Here are other examples of non-violence that you most likely will not see in newspapers or on television:

• Flint, MI police march with protesters.
• More people are cleaning up after looters than probably looted.
• A peaceful protest in Van Nuys.
• That El Segundo protest I saw did have a couple of tweets about it.

What is especially distressing about the links I just shared is that all but the last come from journalists.   The stories those tweets document usually do not make the leap from Twitter to the nightly news or daily newspaper.  In other words, news on social media looks much different than news on television or in newspapers.

The difference is especially stark if you compare event records generated from images shared on Twitter to what you would infer from newspapers.  In another working paper I have, my coauthors and I show that newspapers report more events of repression (police presence or violence) than events of protest, which is hard to believe since police do not repress empty space.  The figure below shows the distribution of records of repression (left) and protest (right) in Venezuela in 2015.  The dotted line is events recorded from images shared on Twitter, and the other lines correspond to three leading news datasets.

There are two takeaways from this figure.  First, not only do social media record more protests than repression, news record much less protest than repression.  News likes violence.  Second, Twitter records many more events of both types than news outlets do.  In other words, you learn more about the world from following social media, and the world looks less violent.

What is the takeaway from my semi-coherent rambling?  Please discount what you read in newspapers or see on television.  The vast majority of people are peaceful.  The vast majority of events are peaceful.  The majority of coverage ignores that.  You can get more diverse, and therefore more accurate, perspectives on protests by looking at social media.