The Washington PostDemocracy Dies in Darkness

How unrepresentative is Florida’s proposed new congressional map?

Florida Gov. Ron DeSantis (R). (Octavio Jones/Reuters)
Placeholder while article actions load

One of the interesting aspects of Florida Gov. Ron DeSantis’s (R) deeply partisan approach to his position is how far that deviates from the state’s actual politics. He won election in 2018 by less than half of a percentage point; President Donald Trump won there in 2020 by just over three percentage points. For more than two decades, Florida has been the quintessential swing state, but DeSantis has increasingly governed it as though he won by double digits.

Elections have consequences, as they say, and DeSantis may very well win reelection by a far wider margin this November. But it does make the new congressional maps that DeSantis is offering stand out. One analysis indicates that it sets up a Republican advantage in 20 of the state’s 28 districts — meaning that a state in which Trump won 52 percent of the two-party vote would allot 71 percent of its House seats to Trump’s party.

And yet, in the context of how other states have redrawn their own boundaries, this Florida map stands out only as deeply partisan, not as unusually deeply partisan.

Sign up for How To Read This Chart, a weekly data newsletter from Philip Bump

To evaluate this, we can look at the same metric across states that have completed their redistricting. Here we’re using FiveThirtyEight’s analysis of redistricting efforts including the estimated partisan lean of each new district (which we’ll come back to).

One might envision an ideal wherein a state would send representatives to the U.S. House in proportion to its voter preference. A state that’s 60 percent Republican, for example, might ensure that six of its 10 representatives are Republican. There are a thousand asterisks that litter that idea, of course, from questions about how you determine a state’s partisan composition — presidential vote? registration? — to whether states should be in the business of trying to skew partisan lean in districts anyway. But let’s just assume that as our goal and consider where states land.

Comparing Trump’s vote percentage in 2020 to the distribution of House seats in each state, a 1-to-1 ratio would fall on the diagonal dotted line, below. But as you can see, almost no states do, instead falling along a (unshown) axis that gives Republicans no House seats in states where he got 30 or 40 percent of the two-party vote to states where he got 60 to 70 percent of the vote and give Republicans all of their House seats.

The farther from the diagonal line, the bigger the gap between a state’s 2020 vote and its House-seat distribution. A lot of this is unavoidable, of course, since a lot of states have only a handful of seats to allocate. Yeah, Trump got nearly a third of the two-party vote in Vermont in 2020, but Vermont can’t give Republicans a third of its one House seat. Same goes for Alaska: President Biden got about 45 percent of the two-party vote, but Democrats get no part of the state’s House seat.

Then there are the iffier examples of partisan shutouts, like Iowa or Connecticut. The former has four House districts and the latter five but, despite neither Trump nor Biden earning 100 percent of the vote in 2020, each drew congressional boundaries that are expected to favor only one party.

For our purposes, let’s compare Florida not to those outliers but to states such as California, Texas and New York. Those are the four most-populous states in the country, and each offers a disproportionately larger share of its House seats to one party. Florida has a 19-point gap between the party split on its proposed map and its 2020 vote. That’s the same as California. In Texas, the gap is 10 points (meaning that Republicans get about 10 percentage points more House seats than Trump did a percentage of the two-party vote). In New York, it’s wider still: nearly 23 points. That map may not survive judicial review.

This is not the only measure of partisanship of course. If we compare FiveThirtyEight’s estimates for the partisanship of each district to the Cook Political Report’s state-level estimates (a metric called the partisan voting index or PVI), you can see how districts are weighted for or against parties.

I’ll note first that the metrics here are slightly different; North Dakota’s PVI for its congressional district is, of course, the same as its state metric, as there’s only one district in North Dakota. But we can still see interesting decisions being made.

In Nevada, for example, the districts were drawn so that there would be three districts that have a slight Democratic advantage and one that has a wider advantage for the GOP. This has been a subject of some hand-wringing on the left, since the lines could have been drawn to offer at least one safer Democratic district. In a heavily pro-Republican election year — like, potentially, this one — all four seats could go to the GOP.

In Louisiana, something different is demonstrated: one heavily Democratic district and a number of less-safe-but-still-safe Republican ones. This is a more common way to weight House districts, jamming the minority party’s votes into a small number of districts to allow a more subtle advantage for your own party more broadly. That pattern shows up a lot above, as in Indiana and South Carolina.

As you know, the gap between partisanship and House-seat allocation isn’t solely a function of trying to divvy up two seats equitably. 2020 presidential vote is correlated to how state legislatures look, too, and more Republican or Democratic states have the ability to draw more lopsided maps to favor their own party in states where legislators draw those lines.

One last aspect of Florida’s proposed map that bears noting. Matthew Isbell, the analyst who tallied the 20-to-8 split mentioned above, also estimates that the new map reduces the number of seats drawn to give an advantage to Black candidates from four to two, with some heavily Black areas split into multiple districts to dilute their voting power. This is a reminder of the other question of how power is distributed that’s at play in redrawing district lines: the extent to which non-majority groups are sidelined. Alabama’s lines were challenged in court on the grounds that they reduced the political power of Black residents, but the Supreme Court let the map stand. You can see the balance of its districts above.

Again, this imbalance in representation is to some extent a natural effect of divvying up the power of 330 million Americans among 435 people. It is also a function of partisan jockeying that often overlaps with racial power struggles.

So Florida’s proposed map isn’t unusual. But that doesn’t make it fair.