You've heard about the filter bubble idea, whereby social algorithms show you want you want, and you end up getting information about the world from mostly agreeable people and sources.
The idea is that, instead of a filter bubble effect where information comes from only like-minded people, the “information gerrymandering” is when information within a filter bubble comes from people inside the filter bubble who also get information from outside the filter bubble.
When you have two filter bubbles — say, politically left and right filter bubbles — the filter bubble that is more “open minded” (has more information coming from members who are aware of information from outside the filter bubble) has an electoral disadvantage.
In other words, keeping party members ignorant of opposing political viewpoints is a winning strategy for shameless political operatives.
According to the research:
“The researchers’ analysis revealed that information gerrymandering could easily produce biases of 20%. In other words, a group that was evenly split into two parties could nonetheless arrive at 60-40 decision due solely to information gerrymandering.”