Marquette University’s Charles Franklin had an interesting way of visualizing partisan poll differences in a tweet earlier this week. He took data from a Reuters-Ipsos poll on President Trump’s immigration ban — one of several showing varying results — and imagined rooms full of people from three groups of people who’d given responses.

A room full of Americans in general would look like the square at left. A room full of Democrats, though, would look like the square in the middle — a lot more homogeneous. Likewise for the square at the right.

The question of Trump’s executive order is by no means the only example of a stark partisan split. On Friday, CBS News released a poll that showed partisan splits on a number of issues. Following Franklin’s lead, we made this interactive graphic to show what that partisanship looks like.

Now, your response to this might be that you’re almost never in a room with just Democrats or just Republicans. And that’s fair. But it’s important to remember that this split overlaps with — and is reinforced by — the fact that many of us live in places where people share our politics.

This can be hard to visualize, but a project from the Los Angeles Times makes it easier. They compiled a huge, gorgeous map of the 2016 results in every precinct and every county across the state. More importantly, they made the data behind it public.

We know that county data can be misleading. A county could have split its vote between Trump and Hillary Clinton last year, we can theorize, because the big city in its center backed Clinton but the suburbs and rural area surrounding it went for Trump. At the precinct level, the results offer a more fine-grained look at where people live — and who they live near. Precincts usually operate at the scale of dozens or hundreds of people, not thousands or tens of thousands.

Consider Los Angeles County.

(The Washington Post)

The scale from left to right is how overwhelmingly Clinton dominated the vote in a precinct; the vertical is how many precincts had that vote percentage. Notice that most of the precincts voted for Clinton. In about two-thirds of the precincts in the state, about two-thirds of the people voted for Clinton.

At the other end is Lassen County, which supported Trump by a wider margin than any other in the state.

(The Washington Post)

There, people often live in places where other people also voted for Trump.

The county that was closest in margin in the state was Stanislaus County. We might think that the margin there was a function of heavily pro-Trump and heavily pro-Clinton regions that balanced out.

It wasn’t.

(The Washington Post)

Stanislaus looks more like the boxes at left in our interactive than the more partisan ones.

The lesson is a simple one. Attitudes on policies often now come down to attitudes on partisanship — and how we think in that regard often overlaps with how live.

But, we are reminded: not always.