Last week, a North Carolina state court panel threw out the current state legislative district map and ordered N.C’s General Assembly to enact a new one by Sept. 18. This article explains the substance of the panel’s decision and anticipates what it may mean for future district lines. The answer: Some change should be expected, but perhaps not the qualitative redrawing that many might assume.
The panel’s decision pertained specifically to the districting map used for North Carolina’s state legislature, rather than for its congressional delegation. Nevertheless, analysts speculate that the case could establish a precedent for future challenges to congressional district maps as well. The method of restraining partisan gerrymandering is significant because it relies on an interpretation of North Carolina’s state constitution. In other words, the decision implicitly concedes that the U.S. Constitution does not forbid the drawing of district lines to partisan advantage, as my recent study documented and which the Supreme Court recently reaffirmed. The panel decision instead emphasizes, in multiple places, that its conclusion is based on interpreting the North Carolina state constitution’s wording as distinct from that of the U.S. Constitution.
Challenges to partisan gerrymandering through federal courts were always doomed to failure because the U.S. Constitution tacitly permits, rather than forbids, a certain amount of the practice. The Constitution’s authors were well familiar with gerrymandering when they granted state legislatures the explicit power to draw congressional districts in manners of their own choosing, while also reserving for Congress the power to “make or alter” states’ election processes. Multiple federal apportionment acts enacted over the decades explicitly authorized electing members of Congress via geographical districts, a method that naturally creates an asymmetric impact on political parties’ fortunes whenever voting factions are heterogeneously distributed throughout a state, as they usually are.
The U.S. Constitution is indifferent to the welfare of political parties; what the larger structure of federal apportionment law seeks to ensure is that members of the same legislative constituency live relatively near one another. This is the foundational reason why Congress is chosen by geographic district, with each district electing one member, rather than through a system that allocates seats to political parties according to their total support within each state. It is also why many state constitutions, as well as several previous federal apportionment laws, have required that congressional districts be reasonably compact rather than irregularly shaped. Accordingly, the constitutionally appropriate way to limit gerrymandering, at least at the federal level, is by Congress’s limiting the allowable irregularity of district shapes rather than by courts mandating that legislative power be redistributed among political parties in any particular way.
The North Carolina panel opined that state elections could not be truly free, that freedom of speech and assembly was burdened, and that equal protection of the laws was undermined, if gerrymandering purposely gave one political party significant advantage over another. Although the panel was careful to couch these findings as interpretations of their state’s constitution, it is clear from reading their opinion that the panel’s reasoning and conclusions were fundamentally different from those the U.S. Supreme Court has employed throughout recent history.
In ordering the North Carolina General Assembly to redraw the districts, the panel specified that the redrawn districts must be contiguous and compact, follow local political boundaries, contain equal population, and not reflect partisan considerations. There are a few additional wrinkles. One is that the Assembly may seek when redrawing the lines to avoid pitting incumbents against each other within the same district. Another is that the redrawn districts must comply with the Voting Rights Act, in that they must allow African-American voters the opportunity to elect candidates of their choice.
Following the panel’s various specifications is a complex (and interesting) challenge for the General Assembly. Of particular note is the panel’s requirement that partisan considerations must not influence the drawing of the map. This, of course, means that the Assembly may not gerrymander the map for deliberate partisan advantage. However, this doesn’t suggest a map must emerge that treats the two major political parties equally. To the contrary, properly read, the panel’s instructions to avoid “partisan considerations” equally preclude drawing a map to deliberately achieve proportional representation of opposing parties, just as surely as it would preclude drawing a map to maximize partisan advantage. In fact, if the map is genuinely drawn without reference to partisan considerations, it probably will not deliver proportional representation across parties.
The statement above may come as a surprise to non-mathematicians, for we humans have terribly flawed instincts for statistics. If you show a typical person two rows of digits and ask which of them was randomly generated, most people tend to assume that a row in which each digit appears with roughly comparable frequency, which avoids certain digits being repeated several times in a short stretch, and also avoids appearances of recognizable patterns, is the randomly-generated one. But true randomness actually means that, sometimes, one digit will appear a few times in a row, or that patterns will emerge that appear too ordered to be coincidental. Similarly, if you draw a map without recourse to political considerations, it will only result in proportional power for opposing political parties under certain highly unlikely arrangements of voters. In the real world, it’s far more likely that a heterogenous distribution of voters across a state will cause one party to fare better than the other in statewide legislative representation.
These principles become much easier to grasp when using the 538 site’s handy tool for modeling different approaches to mapping legislative districts (note: although the recent decision pertained to state legislative districts, similar principles apply for the congressional districts analyzed by 538). For example, 538 anticipates that, as currently drawn, the district lines would over the long term result in the following tilt in North Carolina’s 13-member congressional delegation, on average:
Let’s pretend for a moment that the districts were drawn to partisan perfection, for example, to maximize Republican advantage. 538’s map reflecting that objective shows some pretty weirdly-shaped districts, resulting in average long-term strengths of:
The analysis above supports the contention that N.C.’s current districts were drawn to heavily favor Republicans – after all, 9.4 out of a possible 9.5 is about as lopsided an advantage as one can produce. 538 finds alternatively that Democrat-favoring redistricting would look more like this and have the following average breakdown.
Of course, the recent panel decision says the Assembly is not allowed to adopt either of those approaches. It can neither deliberately favor Republicans nor Democrats. But neither is it allowed to do what is depicted here, which is to create a map explicitly designed to deliver each party a congressional delegation proportional to its statewide strength. 538 projects that on average over time, such a map would have the following balance:
Now, remember that one of the charges from the panel is that the map must comply with the Voting Rights Act in creating some majority-minority districts. 538 also shows a map that would maximize majority-minority districts in North Carolina. A few things are striking about this particular map. One is that it is extremely gerrymandered, with a ridiculously improbably thin district stretching from Charlotte to Greensboro. Another is that even such extreme gerrymandering would create only three majority-minority districts – just one more than the two existing under the current map. Two majority-minority districts are thus likely the largest number any practicable map will be able to produce. A third observation is that the creation of majority-minority districts helps Republicans on balance, at least within North Carolina’s congressional delegation. The average expected party strength with this map is:
If, instead, you merely sought to give the districts as regular shapes as possible, while maintaining equal population in each, 538 finds you would get a map that looks like this. The balance of party representation under this map, on average, would be:
The panel’s instructions, however, require the Assembly to follow county lines where possible and to avoid splitting smaller municipalities and precincts. Maximizing compactness while following rules much like these would, according to 538, create a map that looks more like this. The average composition of the delegation under this approach would be:
Of all the approaches analyzed by 538, this last approach is closest to the one recommended in my study. My study actually does not suggest that mapmakers need to strive for maximum compactness; it only suggests a limit on the irregularity of shape that federal law should allow, without attempting to dictate whether mapmakers purge all political considerations from their minds. The N.C. state panel is far more restrictive in its instructions. Given that the panel explicitly forbade political considerations, following its dictates could well result in a map resembling this last one generated by 538.
In presenting these possibilities, I would stress that I have absolutely no idea what North Carolina’s General Assembly will do. The above exercise simply works from the principles laid out by the state panel and assumes that legislators appropriately attempt to maximize the compactness of districts while avoiding partisan considerations altogether.
What can we learn from this exercise? The main lesson is that even when political considerations are forbidden, we should not expect a well-drawn map to result in proportional representation of North Carolina’s opposing political parties. To the contrary, a map that produces proportional representation would more likely, and counterintuitively, indicate that political factors were considered.
If good principles of maximum compactness and respect for local political boundaries are applied to a redrawing of North Carolina’s congressional maps, without any consideration of political effects, it's likely that they will affect the delegation’s composition by only a couple of seats, at most. As long as we retain our system of representation based on geographic districts, we should not expect or demand that it produces symmetric treatment of opposing political parties, even in the absence of all partisan gerrymandering.
Charles Blahous is the J. Fish and Lillian F. Smith Chair and Senior Research Strategist at the Mercatus Center, a visiting fellow with the Hoover Institution, and a contributor to E21. He recently served as a public trustee for Social Security and Medicare.
Interested in real economic insights? Want to stay ahead of the competition? Each weekday morning, e21 delivers a short email that includes e21 exclusive commentaries and the latest market news and updates from Washington. Sign up for the e21 Morning eBrief.
Photo: Tasos Katopodis/Getty Images