All politics are local, in the broad sense that it is voters in a congressional district or state who weigh in on whether a member of Congress continues to serve in that position. In the sense that voters vote on local issues, that’s also largely true, although in a moment with heightened partisanship and nationalized media attention to political issues, it’s less true than it once was.

Now that the Senate Republican leadership has finally released the health-care bill that it spent several weeks drafting behind closed doors, we’ll see an interesting test of the boundaries of how and where the Republican push for overhauling the Affordable Care Act butts heads with what voters in key states actually want to see.

We can visualize the Senate on a sort of grid, laid out in two dimensions. The first, the horizontal axis, is how the state a senator represents voted in the 2016 election. States to the right of the center line were more supportive of Donald Trump. States to the left, more supportive of Hillary Clinton.

On the vertical axis, how partisan each senator is in his or her actual votes. These data come from UCLA’s Voteview, which analyzes the voting records of each member of Congress. A negative score is a more liberal legislator; a positive score, more conservative.

The current senate, then, looks like this.

Notice that we’ve color-coded each person. Those with darker-colored dots are up for reelection in 2018.

There are two groups of senators worth paying close attention to here.

The Republicans

Note this little cluster. Included there are Sens. Jeff Flake (R-Ariz.) and Dean Heller (R-Nev.) — both of whom are up for reelection next year. Heller is already on record as being skeptical about the bill. So, too is Susan Collins (R-Maine), the second-farthest-left dot.

Our whip count (that is, our list of how senators plan to vote) includes a number of senators who aren’t in this block. But these are the Republican senators who represent states that were least supportive of Trump (among Republican senators) and who, therefore, may feel more pressure to buck their party.

It’s important to note, too, that vertical axis. Imagine a diagonal line running from the bottom left to top right. It’s not an apples-to-apples comparison, but the further above that line a senator might be, the more that senator may be more conservative than his or her state. The further below that line, the more he or she may be much more liberal. So a senator who is perceived as being too conservative or too liberal may seize upon a high-profile piece of legislation to prove a point.

The Democrats

We need to pay attention to some Democratic senators, too.

Of the 11 Democratic senators representing states that voted for Trump last year, 10 are up for reelection next year — during a midterm election of the sort that, in recent years, has been unfriendlier territory for Democrats. Several of those Democrats represent states that only barely backed Trump. But six senators represent states that backed Trump by at least 8 points; two senators represent states that backed him by more than 30 points.

Will Sens. Heidi Heitkamp (D-N.D.) and Joe Manchin III (D-W.Va.) — those two senators — hold firm with their party despite representing states that preferred the president who is calling for Obamacare to be overhauled?

Luckily for the Democrats, the legislation is broadly unpopular, making this an imperfect test of the primacy of national party loyalty over local concerns. West Virginia voters seem as though they’d be a lot less likely to punish Manchin for bucking the president they picked to pass a bill they don’t particularly like than Nevada voters would be to punish Heller for backing a bill they hate for a party they oppose.