Moveon.org’s MO Research puts racially charged questions to Facebook voters

About a year ago, I found myself talking to a bot. The bot was from MO Research, an app created by MoveOn.org.

The bot asked me a few questions about a political issue (I probably clicked on one of their ads for the bot to start chatting with me).

In any event, it asked me one question. And then every few months asks me more questions, like “Would you support or oppose a national health plan, sometimes called Medicare-for-all, in which all Americans would get their insurance from a single government plan?”

Then, over the weekend, I got a survey which was a bit shocking.

I was asked several questions, which options such as “Strongly Oppose”, “Oppose”, “Agree”, etc.

And here are some of the questions:

How much do you agree with this statement: The best way to solve the country’s illegal immigration problem is to make conditions so difficult for illegal immigrants that they return to their home country on their own.

How much do you agree with this statement: We should provide a path to citizenship for immigrants who are living in the U.S. illegally.

How much do you agree with this statement: Most Muslim people living in the United States are more prone to violence than other people.

How much do you agree with this statement: Most Muslim people living in the United States are more prone to violence than other people.

How much do you agree with this statement: Islam is fundamentally incompatible with U.S. national culture and values.

How much do you agree with this statement: Jewish individuals are more loyal to Israel than to this country.

How much do you agree with this statement: Jewish individuals have too much power in the business world.

The questions were shocking, to say the least.

Going to MO Research’s Facebook page (which is linked from the MO Research website), there is an explosion of anger over this survey.

And apparently, I received one set of surveys. There was another troubling survey sent out as well (I’m taking as screen shot from one of the negative reviews on the MO Research website):

Moveon’s MO with MO Research

Why would MoveOn do such surveys? Simple: they are attempting to emulate Cambridge Analytica.

Cambridge Analytica used data from Facebook to craft political ads. The core of Cambridge’s methodology was surveys, and then mining the survey taker’s friends to develop profiles.

MoveOn is attempting to do the same thing. According to an article in Wired, DeVries (MoveOn’s director of analytics) and his colleagues created a Facebook app called MO Research and targeted ads at people, asking them to answer survey questions like, “Should Congress pass stricter gun control legislation?” or “Do you approve or disapprove of recent NFL player protests?” Some 400,000 people answered an average of five questions each via Facebook Messenger. They also answer questions about things like their hometown, gender, and age. That allows MO Research to build demographic profiles of people and match them to their voter file records.

They place ads asking a controversial question, and then the bot follows-up.

This is not small amounts of money, either. MoveOn has spent millions. From a letter the organization sent to Facebook in 2018:

Finally, as a result of our experience as one of the largest political advertisers on Facebook in 2018 (MoveOn Political Action and our subsidiary MO Research together spent $5.5 million on Facebook ads from May to present), we have identified potential inconsistencies in the application of your advertising standards, including potential differences along lines of race and identity, that warrant investigation.

And, from OpenSecrets, confirmation that millions has been spent.

But why?

We can speculate as to MO Research’s intentions with this survey, but they’ve answered the question themselves: Moveon is attempting to sway the 2020 elections, just like the Steve Bannon did with Cambridge Analytica back in 2016.

They are building profiles – probably something like OCEAN models – and adding additional data to model behavior.

And they certainly have proven that they can spend the money.

Update: Paul Roberts independently contacted MoveOn and has confirmed that this poll was, in fact, launched by the MO Resaearch division of MoveOn — for all of you who are convinced this was a hack, it wasn’t. It was a MoveOn project.

Leave a Reply