on representation, redistricting, and "gerrymandering" (with 2024 election data)

“Gerrymandering” is always a potential topic with our electoral system, but sometimes it moves to the front burner of the proverbial stove. We’re in one of those moments now– as the Texas GOP stirred the pot last month with its work to “redistrict” and Democrat leaders in a few states responded with angry rhetoric and promises to “fight back”. 

Let’s start with defining and explaining the relevant terms. Redistricting is the practice of redrawing the geographical lines that define a political district, given changes in national and state populations. These evolutions in population are measured at the beginning of every decade through the efforts of the U.S. Census Bureau. 

Let’s consider the 435 seats in the U.S. House of Representatives. If a state’s population grows/shrinks (enough) with respect to the national population, a state will gain/lose a seat (or more). In such cases, a state must redraw its district lines, accommodating more/fewer seats for the same geographical area. But even if a state’s number of seats remains the same, lines are likely redrawn as population shifts within the state– from births and deaths, immigration and emigration. 

In 25 states, these lines are redrawn by the state legislature (often approved by the governor). In the other 19 multi-district states, an independent or bipartisan group (to limit political shenanigans). And the process can be regulated by the judiciary. (Six states have one district, so they “don’t count”.) 

Gerrymandering is the practice of drawing these lines for political purposes. (The word was created by merging a salamander with the name of the Massachusetts governor who did this in 1812, Elbridge Gerry.) We would hope (and expect) that an independent process would avoid this. For example, a GOP-controlled state government might look to gerrymander its state population by putting most of the Democrat voters in a few districts. 

But there are some inherent tensions in such efforts. Let’s use some numbers to see how this works. If there are 600 GOP’ers and 400 Democrats in a state with ten districts, you could put the 600 GOP’ers in six districts and the 400 Democrats in four districts, creating ten utterly “safe seats” which lack political diversity. You could also put 60 GOP’ers and 40 Democrats in each district, creating ten GOP seats that would be relatively safe, but much more diverse and potentially competitive. As a political party draws advantageous district lines, they have to wrestle with the tradeoff between more seats on average with greater political risk and fewer seats that are much safer.

With the exception of political partisans who naively imagine that their party leaders are far more virtuous than those in the other major political party: it will be little or no surprise to learn that significant gerrymandering is bipartisan and not all that pervasive at the state-wide level. For one thing, the overall vote percentage (49.8% GOP vs. 47.2% Dem in 2024) implies a slight partisan gerrymandering advantage to the Dems, given the GOP’s far slighter seat advantage in Congress. 

With the institutional structures in play, we might expect more trouble when lines are drawn by partisan legislatures, but this is still regulated by the courts. And even so, we wouldn’t expect politicians to passionately aim for another seat or two, while increasing the likelihood of defeat in more competitive seats and looking nakedly partisan. (Politicians and bureaucrats usually look to reduce risk, not increase it.) 

One way to perceive gerrymandering is to look at the political map. When districts look highly irregular, it gives the impression that gerrymandering was the strategic product of “intelligent design”. For two terrific examples, check out the 13th and 17th districts in Illinois.

But what do the numbers say? Assessing representation by party is easy (number of seats), but what about vote percentages? The most obvious candidate would be the overall vote percentage in each state’s House races. But this can be distorted by single-candidate races, where one party receives 100%; the other receives 0%; and there are usually fewer voters. (For better/worse, I calculated average vote percentages across all districts in a state.) Given this concern, another reasonable proxy could be the percentage received by the presidential candidate in that year. (The two statistics turn out to be highly correlated, but I’ll use both here.) 

I’ll provide all of the data below, but I only see a handful of cases where the numbers reflect the potential for gerrymandering. Observations of note: 

-There are six single-district states where shenanigans are impossible. 

-Of the nine states with two or three districts, three raise a red flag (that favors Democrats): Maine with a 7-12 percentage point Democrat voter advantage and both seats (despite an “independent process”); New Hampshire with a 3-6% Democrat advantage and both seats; and New Mexico with a 6-10 Democrat advantage and all three seats. 

-Of larger states: Wisconsin advantages the GOP (1-3 percentage point voter advantage but 6-2 on seats) and Illinois strongly advantages the Democrats (8-11% advantage but a whopping 14-3 on seats). Despite “independent” processes, Colorado advantages the GOP (11-15% Democrat advantage but only 4-4 on seats) and New Jersey benefits the Democrats (only 6-9% advantages, but 9-3 on seats). 

-In the Deep South, efforts to “positively” gerrymander for racial considerations may have helped with African-American representation but hurt Democrats overall. Arkansas has an independent process and a 30-36% GOP advantage, so its four GOP seats are no surprise. In the four more rural states (LA, MS, AL, and SC), the GOP’s 24-35% advantage translates into a reasonable 18-6 seat advantage. But in the two more cosmopolitan states (GA and NC), the GOP’s 3-8% advantage translates into an impressive 19-9 seat advantage. 

Source: http://schansblog.blogspot.com/2025/09/on-representation-redistricting-and.html