How Left-Wing Is Facebook? [With Final Update]

Facebook came under attack a few months ago for selecting liberal themes and news stories and downplaying conservative ones in its “trending” news section. Mark Zuckerberg responded by hosting a meeting with influential conservatives in which he was conciliatory and promised to do better. But my own experience during the past week has caused me to wonder what is going on at Facebook.

In two weeks, the think tank that I run will put on a lunch forum on the minimum wage. There will be three speakers: a liberal Minnesota state senator who is an enthusiastic advocate for a $15 minimum wage, an economist who will explain the reasons why a minimum wage in excess of the prevailing cost of entry-level labor is a bad idea, and a representative of Minnesota’s hospitality industry who will talk about the practical consequences of instituting a high minimum wage in a city like Minneapolis. This is the invitation we have sent out for the event:


As part of my ceaseless efforts to market our events, I did a post on Facebook and “boosted” it to the tune of $7. The Facebook post might have been the most inoffensive thing I have ever written:

Screen Shot 2016-09-04 at 1.57.01 PM

Imagine how stunned I was when I was told by Facebook that they had refused to boost my post because it violated their guidelines:

Screen Shot 2016-09-06 at 1.05.16 PM

So our ad apparently was not approved because it “contain[ed] profanity, harassment, or references to your audience’s personal characteristics (such as gender, race, age or name).” This was the complete text of my post:

Are voters catching on to the minimum wage? This Star Tribune poll suggests they may be. But please come to our lunch forum on September 20 to hear all points of view!

How racist can you get?

I thought this must be a simple mistake. Facebook has an appeal procedure when ads are declined, which I followed. I clicked the “appeal” box and wrote that this obviously was an error, as no one could possibly think that our ad violated any guidelines.

That was six days ago, and Facebook has done nothing to reverse its spiking of our ad. Which causes me to reflect.

My first assumption was that some stupid computer algorithm had misinterpreted words in my post and thought they somehow referred to race, gender, etc. (in some way, evidently, different from the countless references to race, gender, and so on, not to mention people’s names, that appear in millions of Facebook posts). But return to the language of my post. There is no way a computer algorithm could have interpreted it as being offensive in any way:

Are voters catching on to the minimum wage? This Star Tribune poll suggests they may be. But please come to our lunch forum on September 20 to hear all points of view!

So that makes me think that disapproving our ad must have been a human decision–a decision by a Facebook employee who didn’t want anyone hearing the arguments on the minimum wage, even where multiple points of view are represented. What other possibility is there?

And in six days, Facebook has taken no action on my obviously meritorious appeal. In many instances, inaction for six days would destroy whatever value the post/ad might have had. That will be the case here, if Facebook’s inaction continues for another week or so.

Are Facebook employees refusing, in some instances, to “boost” posts that include non-leftist points of view? I would hardly have thought it possible, in part because we have successfully boosted a number of posts that obviously had a conservative slant. But this experience leaves me wondering what other explanation there could be for Facebook’s spiking of our ad. Maybe this is one more situation where paranoia is justified.

UPDATE: Several people at Facebook read this post and investigated. I got a call this morning from a woman whom I knew slightly when she was working on a Republican presidential campaign several years ago. She told me that the ad’s disapproval was the result of a “false positive,” i.e., Facebook’s computer algorithm apparently mistook something in the ad for racist or otherwise offensive content. She promised to let me know if and when they figure out what triggered the disapproval. She also said that my appeal should not have sat in the “appeal queue” for six days, and they will look into why that happened. It was a very cordial conversation.

I also got an email from Facebook saying that the ad has been approved. This was done manually by Facebook employees who inquired after seeing this post.

I appreciate the prompt reaction from Facebook and will do a follow-up post if I get a further report about the reason for the ad’s disapproval. I continue to suspect that a rogue employee at Facebook disapproved the ad, on the ground that there just isn’t any language in the ad that could have triggered a negative computer response. But perhaps I am mistaken; maybe another explanation will emerge. If so, I will report it.

AND FINALLY: My Facebook contact called again and told me the mystery was solved. Apparently it was Facebook’s policy against “financial apps” that caused the ad not to be approved. The key language, she said, is the beginning of the post to which the Facebook post linked:

The minimum wage is a classic feel-good measure: who wouldn’t want relatively low-paid workers to earn a little more money? No one. In my view, the minimum…

Apparently the computer confused the words “earn a little more money” with the sort of spam comments that crop all across the internet: “earn $5,000 a month working from home,” or whatever.

As for the six days without a response to my appeal, my source told me that there was a technical glitch that caused a number of appeals, including mine, not to go into the queue.

Is that explanation accurate? I don’t know. It seems bizarre that any post that includes the words “earn more money” can’t be boosted. Mine can hardly have been the first to use those words. But I am confident that the woman I have been talking to believes it is the true explanation, and I am happy to leave it there. I appreciate Facebook’s responsiveness.

Notice: All comments are subject to moderation. Our comments are intended to be a forum for civil discourse bearing on the subject under discussion. Commenters who stray beyond the bounds of civility or employ what we deem gratuitous vulgarity in a comment — including, but not limited to, “s***,” “f***,” “a*******,” or one of their many variants — will be banned without further notice in the sole discretion of the site moderator.