The “Minimal Overlap” Solution to Gerrymandered Injustice

David Brin
8 min readSep 23, 2017


Gerrymandering has reached a point of such outrageous blatancy that it seems likely the US Supreme Court will have another look, soon. This New Yorker article dissects the problem, describing some new insights from logic and mathematics that might help the Court better to understand a foul practice that has warped and partly-stolen American democracy. A lot has changed since Justice Kennedy provided the deciding “we can’t see a way to do anything about it” vote, roughly a decade ago.

As I describe elsewhere, voters in many “blue states” have rebelled against their own Democratic politicians, ending gerrymandering via ballot measures. Hence, with a few dismal exceptions like Maryland and Illinois, this cheat has become ever-more associated with the Republican Party.

Ideally, solutions should come from negotiated legislation. When power abuse is generated by legislatures themselves, courts must step in. Hence, aware that losing this battle may end their lock on power, attorneys of the right argue that no alternative is intrinsically fair — including “impartial commissions.” Moreover courts are reluctant to interfere with state sovereignty.

Why did Justice Kennedy opt for the status quo, last time? Even in the face of blatant injustice, judges like to have two things:

* A simple, unambiguous metric that proves actionable harm. This has been provided recently by an elegant standard of “voter efficiency.”

* At least one clean and simple remedy they can point to as an example.

The first requirement has been powerfully met. But for the Court to articulate a workable standard limiting gerrymandering, what’s needed is a fallback solution that is inarguably better than the present state of affairs, that can be ordered if a state proves unable to devise a fair and impartial redistricting process on its own. To resolve Justice Kennedy’s dilemma, I will propose a solution so simple that it can be expressed in three sentences.

Here are those three sentences:


1. With allowances for contorted state borders, like Maryland’s panhandle, the districts that are drawn for State Assembly, State Senate and Congress shall meet a compactness standard, not falling below a reasonably generous area-to-perimeter ratio limit set by the court.

2. On advice from a non-partisan and unbiased commission, the State Legislature may assign boundaries to the districts of the State Assembly however they see fit.

3. Once those State Assembly boundaries are set, the drawing of boundaries for State Senate and Congressional districts will be computer-generated with the core provision that they must have MINIMAL OVERLAP with each other and with the State Assembly districts, sharing as few voters as practically possible.

There you have it. Three sentences. I’ve offered this suggestion for a decade and I promise that (alas) you’ll find it nowhere else. But what does it mean?

It means that the State Legislature may if they choose ignore the ‘neutral commission’ and connive, jigger or gerrymander districts for one house — the State Assembly — limited by some basic rule of compactness. But provision #3 ensures that the districts for State Senate and Congress will be utterly different. The more carefully the legislature’s majority partisans gerry-rig one house, the less effective will be their efforts in the other two.

The chief aim of gerrymander-cheating — to achieve government dominance by the most rabid of hyper-partisans — will be devastated and then grow weaker, over time.

== Illustrating the Minimal Overlap concept ==

For some reason, the notion of minimal overlap seems obvious to some people, while others find it difficult to grasp. So let’s try using illustrations.

Sentence/provision #1 takes care of the worst, egregious cases, illustrated in our first figure.

Fig1: A large fraction of gerrymander travesties would be eliminated by a compactness rule, setting upper limits to perimeter-area ratios. This limit can be fairly generous, since the rest of the solution happens through minimal overlap.

In figure #2 we present a strawman set of six State Assembly districts that are (for the sake of simplicity) highly compact.

Fig. 2 Simplified boundaries of a State Assembly, under compactness rule, but somewhat gerrymandered.

Let’s assume that the state legislature has, under rule #2, but limited by the compactness rule #1, arranged these assembly districts to maximize gerrymander benefits for the majority party.

Now, in our third illustration, let’s overlay districts for State Senate. These are required — under the court-ordered remedy of MINIMAL OVERLAP to be computer-optimized so that each senate district shares as little territory and as few voters as possible with any one assembly district.

Assuming the compactness rule is enforced, and that Senate districts are truly drawn according to provision #3, then Minimal Overlap — also called “anti-nesting” — means that the political character of the Senate will not be warped by gerrymandering. Citizens who were disenfranchised before will likely get attention and an effective vote, in at least one chamber.

The districts for Congress, presumably larger, will nevertheless be kept off-kilter from the gerried State Assembly districts. The party in power will thus only get to have one chamber warped by self-serving, partisan political cheating.

Moreover, even if this method has flaws, it is a clear limiting case that deprives the courts of any “we see no clear remedy” excuse. For all its faults, Minimal Overlap is palliative, equitable and enforceable. It also gives a nod to state sovereignty and legislature privilege, by allowing the legislature to continue complete, discretionary control over one chamber, while the other two are set by a neutral computer reacting to their assembly boundaries.

== Arguments against Minimal Overlap ==

One objection that opponents to such a solution will assert is that voters should be represented by “communities of interest.” For example, one of the commonly used excuses for gerrymandering is that contorted arrangements are necessary in order to ensure that minority populations get some representatives who are of their ethnic persuasion.

There are two, decisive answers:

(a) The “communities of interest” argument is served by having one of three chambers divided that way. So long as those communities of interest are firmly ensconced and represented in one chamber, there is no inherent need for duplication. This is an original merit of bicameral legislatures.

In fact, there are strong arguments in favor of voters facing different coalition needs, in different houses. Why should their Assembly, State Senate and Congressional delegates be clones of each other? Apportioned one way — say in the Assembly — the community of interest might map onto national political parties, or else be optimized for ethnic representation. But mapped orthogonally in another house, entirely different matters of community interest — based on geography, markets, or some other basis — might come to the fore. State Senators will discuss different priorities at their town hall meetings than Assembly members, to the benefit of political problem-solving.

Anyway, a state senator who must negotiate among multiple constituencies and interests will be a busier one, and possibly one who achieves a lot more to break down our divisions.

(b) This method is a fallback, intended to persuade the Supreme Court that gerrymandering can be solved intrinsically, in a simple fashion that is inherently more fair than the present, biased-partisan cheating. And what could be simpler than three sentences?

Under the Minimal Overlap method, voters who now feel completely disenfranchised in all ways will thereupon very likely see their position improved. They will gain a chance that at least one of their three representatives will be someone who heeds their concerns. That is an improvement and a palliation of harm, and one that is far from arbitrary.

Voters thus would be guaranteed some relief from a conspiratorial injustice, in a fashion that is simple to execute. States may opt for some other method to eliminate the injustice. Many already have. But this method provides a backstop ensuring that the worst, most pervasive effects of gerrymandering will end.

== Implications of Minimal Overlap ==

Notice one “judo” aspect of this approach — that it allows hyper-partisans to have their way — somewhat — for a while, in one house. This might lessen resistance to reform by the most fundamentally powerful entities in American political life, state assembly members. It also splits away the self-interest of State Senators, reducing their motivation for hyper-partisanship — which is a desirable outcome in its own right. Why should Assembly members and Senators connive together? Vive la difference!

Moreover, as State Senate and Congressional delegations become more moderate and less partisan, they will then tend to pressure the State Assembly to damp down its own cheating and partisanship.

The Court should also be made aware of the effect that impartial redistricting has had in many blue states and a few purples. While California remains dominated by the Democratic Party, impartial redistricting and other reforms (e.g. non-party primaries) have resulted in less bitterness between parties, not more. Less acrimony. Even in districts that wind up heavily Democratic or Republican, voters who are members of the minority party now feel more listened-to than before.

Earlier I mentioned that Illinois and Maryland and few other Democrat-dominated holdouts still outrageously gerrymander. Former President Barack Obama and former U.S. Attorney General Eric Holder have specifically targeted these states, arm-twisting state legislators to end gerrymandering. When those Democratic Party holdouts comply, this horrifically blatant cheat and crime will be seen as an odious offense perpetrated primarily by just one party against the citizens of this great nation.

Nevertheless, the best solution will come from the Supreme Court, whose past reluctance must be met with a web of logic that allows no escape or wriggle room for Justices Roberts, Alito and especially Kennedy, erasing their earlier excuses for inaction. Minimal Overlap can serve as a example of a backstop remedy that’s simple, fair, and undeniably better than the outrageous status quo.


David Brin is a scientist, tech speaker/consultant, and author. His novel about our survival in the near future is Existence. A film by Kevin Costner was based on The Postman. His 16 novels, including NY Times Bestsellers and Hugo Award winners, have been translated into more than twenty languages. Earth, foreshadowed global warming, cyberwarfare and the world wide web.

Dr. Brin serves on the external advisory board of NASA’s Innovative and Advanced Concepts program (NIAC). David appears frequently on shows such as Nova and The Universe and Life After People, speaking about science and future trends. He has keynoted scores of major events hosted by the likes of IBM, GE, Google and the Institute for Ethics in Emerging Technologies.

His non-fiction book — The Transparent Society: Will Technology Make Us Choose Between Freedom and Privacy? — won the Freedom of Speech Award of the American Library Association. (Website: )


A Summer School for Mathematicians Fed Up with Gerrymandering, By Dawn Chan, New Yorker 8/2017.

Gerrymandering American Democracy: More Fragile Than We Think, by David Brin, Ph.D., 2006.

The Supreme Court case that could shift how Americans vote rests on a simple math equation, Lola Fadulu, Quartz. 2017



David Brin

Author, scientist, public speaker. My books include The Transparent Society, The Postman, Earth, Existence, and Startide Rising.