Facebook’s data scientists used anonymized data to determine that “Just as certain genetic mutations can be advantageous in specific environments, meme mutations can be propagated differentially if the variant matches the subpopulation’s beliefs or culture.”

Take this meme:

“No one should die because they cannot afford health care, and no one should go broke because they get sick. If you agree, post this as your status for the rest of the day”.

In September 2009, 470,000 Facebook users posted this exact phrase as a status update. But a total of 1.14 million status updates containing 121,605 variants of the meme were spawned, such as “No one should be frozen in carbonite because they can’t pay Jabba The Hut”. Why? Because humans help bend memes to better fit their audience.

In the chart below you can see how people of different political leanings adapted the meme to fit their own views, and likely the views of people they’re friends with. As Facebook’s data scientists explain, “the original variant in support of Affordable Care Act (aka Obamacare) was propagated primarily by liberals, while those mentioning government and taxes slanted conservative. Sci-fi variants were slightly liberal, alcohol-related ones slightly conservative”. That matches theories by Dawkins and Malcom Gladwell.