25 Comments

Legislators usually redraw districts to ensure that as many incumbents as possible are re-elected. This reduces conflict. Over time districts evolve into safe districts.

Expand full comment

It's also something that legislators of the two parties can agree on. A Republican legislator who shifts Democratic voters out of his district and into an adjacent Democratic district is benefiting both himself and the adjacent Democratic legislator. On the other hand there is a conflict between the legislator's interest in being safe and his party's interest in getting as many congressional seats as they can.

Expand full comment

Definitely. It produces an effect which critics _call_ gerrymandering, but isn’t, strictly speaking.

Expand full comment

More fundamentally, there's a conflict between the legislator's interest in being safe and the public's interest in having a meaningful say in choosing their own representatives :-)

Expand full comment

This makes sense to me.

Expand full comment

You mention the UK and the anomaly in terms of percentage of MPs in the House of Commons compared with the vote share and this is accurately identified as a consequence of the FPTP voting system. What particularly demonstrates the foolishness of the Economist article is the fact that constituencies in the UK are either not gerrymandered at all, or, if so, to an insignificant degree. Since the Economist is based in the UK, the failure to understand the potentially false assumption of gerrymandering in the state of Tennessee seems particularly egregious.

Expand full comment

I have sent the Economist an email pointing out the errors, including the fact that the UK districts are not gerrymandered.

Expand full comment

Correct. But we don't know if the author of that article is from the UK. If he is from a state with proportional representation his mistake is a little more understandable.

Expand full comment

My (admittedly) crackpot "proportional representation" scheme that I haven't seen discussed (although commenter "Chartertopia" comes close): "Any candidate for the US House of Representatives who receives greater than 1% of the popular vote in the general election shall be entitled to a vote in the House equal to the fraction of the vote he or she receives."

(They also receive a fractional salary.)

I envision some advantages:

As long as their candidate got above that 1% threshold, people would have someone in office they thought of as "their representative", decreasing political alienation.

Conversely, the elected representatives would have a greater incentive to pay attention to (i.e., actually represent) the people who voted for them.

Citizens residing in overwhelmingly "blue" or "red" districts are probably marginally discouraged from voting under the current system. Why bother, when the outcome is foreordained? Under this proposal, they'd have more incentive to get to the voting booth. Maybe even more of an incentive to get informed on issues of interest.

Gerrymandering becomes much less of an issue (and my guess it would be negligible), since just about everyone gets "represented".

Expand full comment

That's roughly my old proposal. I add the ability of each voter to transfer his vote to a different representative at any time. Instead of your 1% limit I propose that any representative with at least (say) 1/450 th of the total vote gets a seat in Congress, needed to participate in debates, committees, introduce bills, but not needed to cast votes. Representatives with less than that can pool their vote count to get a shared seat, agreeing among themselves on who occupies it when.

Expand full comment

I did think of electing every candidate who got more than 10% or some other fraction of the vote. It would not work if legislatures had to meet physically in one building; you'd have to allow for far too many legislators to be practical. Even my 4 per district would turn the current House into 1740 winners. A 1% threshold could run up to 10,000 representatives. My 1740 would work only because the Chartertopia government is so powerless that legislators don't have to meet in person or do much.

I also thought of using percentages of the vote instead of absolute numbers. But then you do have to standardize district absolute sizes. A legislator with 50% of a 10,000 voter district should not have the same vote as a 50% winner in a 20,000 vote district.

My Chartertopia legislators also don't get paid. If legislators are expected to movie to DC and be full time, you have to pay them all the same, otherwise only the rich winners could afford to work for even half pay, and the 1% winners might as well not be paid.

One of the problems with the current 535 member House is too many legislators looking for something to do and some way to make a big splash with the public. It's partly why they have so many committees and subcommittees, so only a few (30 being a few, ha ha) legislators actually have to argue and bicker. Especially when they all insist on their 15 minutes of asking questions and speechifying, only a few legislators have any real power. It's also why they limit floor amendments and so many floor votes now are yea or nay on what the committee passed and the floor debates are so useless.

It would be better to have 1/10 as many legislators as now instead of 10 times as many. Even 100 Senators have too many idle hands looking for too many ways to draw attention to themselves.

When everyone is represented, nobody has a say.

Expand full comment

This seems to be a mathematical problem that politicians do not want to be solved by a computer. The problem to be solved is relatively simple: a set number of representatives has to be elected in a state, so the number of inhabitants per district to be approximated is known. Presumably (for organizational reasons), you wouldn’t want to break up the smallest administrative entity (like a municipality), so these entities have to be distributed over the districts. Add an additional rational condition like minimal length of borders of the districts and voilà, you have a problem that can be better solved by a computer than by a commission of political operatives.

Expand full comment

You're talking about how to draw districts "rationally", presumably in order to elect a representative from each district. That can indeed be done by a computer, but you have to tell the computer what qualities to optimize for, which is far from obvious.

Besides, any system with one representative elected from each district is already doomed to failure: as David points out, a party or voting bloc with a 51% majority in every district can elect 100% of the representatives, leaving the other 49% of the voters with no representation of their views at all. Not good for democracy.

Of course, a party or voting bloc perfectly equally distributed across districts isn't realistic. In a more realistic scenario, a party or voting bloc with a 51% majority statewide will generally get _at least_ 51% of the representatives, but usually more. A party with a 60% majority statewide can usually elect 70-80% of the representatives, enough to override vetoes, change the voting rules so it never loses its majority, etc.

Single-seat districts amplify majorities, unless you use politically-informed gerrymandering to ensure proportional representation, which usually means creating a lot of safely-blue and safely-red districts where there is no meaningful general election and all the action happens in primaries. And that's not good for democracy either: politicians who face only a primary challenge, not a general election, tend to be more extremist and less willing to compromise -- indeed, they often put on a big show of refusing to compromise on anything.

Expand full comment

> No court has interpreted the Voters Rights Act to justify forced redistricting to make the number of congressional representatives of each party proportional to the number of votes cast for that party. If one ever does, the Libertarian Party, with .42% of the vote in 2024, should sue for its two congressmen.

Maybe they should preemptively

Expand full comment

Much of the difficulty associated with redistricting with any majority-minority districts is that you basically must start with a district that is at least 60% Black. This makes making other (required, hypothetically) compact, contiguous districts with less than 1% population differneces really hard. Even then you end up with Tennessee having a "Black district" represented by very non-Black Crazy Steve Cohen.

It also happens in Michigan. "Create" that first Black district and then the other districts have to be hammered to fit.

We could be better off if SCOTUS mwe the requirements more specific. What the hell is "compact" anyway? There are mathematical ways to do that, but no one who wants lots of Black districts ever wants it.

20 or so years ago North Carolina had even worse districting than LA. One district basically ran along both sides of an Interstate, includinglots of small Black majority small cities and towns. It got tossed.

Someday someone may successfully sue and wn under the 14th Amendment that majorty-minorty districts are unconstitutional. Black Republicans have proven they can win even in districts where there isn't a Black majority.

Expand full comment

I wonder if Asians or Hispanics, in a state where they were a significant minority, could sue for districts where they were a majority, and win. Or whites, in a state with enough Blacks, Hispanics and Asians to make whites a minority.

The whole system seems crazy to me, an attempt to evade the fact that, in a winner take all electoral system, the electoral minority in a single winner contest doesn't get represented.

Now I'm imagining a doctrine under which the presidential candidate who gets a quarter of the votes gets to be president for the last year of the four year term.

Expand full comment
2dEdited

Hawaii is already that state and generally the gerrymandering is mostly just a tug between how to keep Filipinos and Micronesians out of Federal office while keeping the Japanese and Jews in while also ensuring the city of Honolulu is in charge of both districts even though it's a single municipality because the local politicians and populace live in abject terror the citizenry in the rest of the state have a say on anything; kind of like NYC and NY state but amplified. Basically through which specific Honolulu neighborhood the district line is drawn changes to ensure the city always has enough people in the other district to remain in control of it.

Also the electoral minority is overwhelming represented in Hawaii, of the 4 Feds you have one White, one Jew, and two Japs in a state that is overwhelming Hispanic (Filipino) and Chinese and I honestly don't think either have EVER had a Federal representative.

I think the thing missing here, and maybe my memory is bad, is the only reason this keeps coming up in the South is because they are under special redistricting rules that were Federally imposed post Jim Crow.

Expand full comment

It's entirely possible (I don't know, offhand, the lower bound) for a canidate to win 25% of the popular vote and still win the Electoral vote. It would be weird, but consider if Trump had been barred the ballot in the Pacific Coast states, Illinois, and NY. He would have finished probably 7,000,000 votes behind Harris and still won the election. Given an H Ross Perot strong challenge, and it would be worse, even if he still won the same states.

In 1860 Lincoln won with just less than 40%. in a 4 candidate race. So . . .

I have argued that we should develope a program that considers, in this order, the parameters of redistricting:

1) <1% difference in total population

2) Contiguous

3) Compact

4) avoiding splitting current governmental boundaries (i.e. counties, then townshiips, then incorporated local governments, then streets and roads, etc.

Use that to have computers draw up the 10-12 sets of new districts with the lowest total amount of boundary lines in miles.

Then have the legislature have to choose from among those possibilities. No voter characteristics of any kind allowed.

Expand full comment

To win the electoral vote in a two candidate presidential election you need 50%+1 votes in 50%+1 states. That is ignoring the fact that the ratio of electoral votes to population is higher in smaller states. Allowing for that means you can do it with a little less, so a bit under 25%. Third party candidates also push down the percentage required for a plurality.

So yes, you can win with 25% of the votes, provided they are distributed properly, with the other candidate getting 100% of the vote in states he wins, your candidate getting a bare plurality in states he wins.

Expand full comment

Why does a district need to be geographically based at all? It made sense 200 years ago that you would vote for one of your neighbors, whom you might actually know and do business with, but today the number of voters per seat is much larger, and people know their professional colleagues and their socioeconomic/educational cohort better than they know their neighbors.

We can easily produce districts that are equal in population, without considering "voter characteristics of any kind", by using at the last few digits of your Social Security number to determine your "district". Absolutely fair, absolutely objective, no political considerations at all -- and it would have the problem David describes, that a party with 51% popular support can win approximately 100% of the districts.

Perhaps instead we should draw districts by income: if there are ten seats to fill, give one to the 90th-percentile-and-up income people, one to the 80th-90th, and so on. That would at least ensure that different economic interests have a decent chance to be represented. (Of course, income isn't everything...)

Expand full comment

"Two thirds of the voters getting eight congressional seats out of nine would be anomalous in a proportional representation system, but US congressional elections are first past the post." - I am not sure how you are missing the point here. If less than 2/3 of the voters are Republlican. In a fair election should get 5 or 6 out of 9, not 8. And part of the probleem is that the election is "winner take all" but another part is the genrrymandering. Both things contribute to the problem and changing either one or both would help make elections fairer.

Btw, this is an interesting idea on how districts could be drawn fairly. https://theconversation.com/how-politicians-can-draw-fairer-election-districts-the-same-way-parents-make-kids-fairly-split-a-piece-of-cake-222859

Expand full comment

You are missing the point. Given first past the post it doesn't require gerrymandering for the 2/3 majority to get 8/9ths of the seats. If districts are constructed with no attempt to favor the majority they still get most or all of the seats unless the minority happens to be geographically concentrated. In the limiting case of no geographical concentration the majority gets all of the seats whatever the electoral map.

So why do you expect a fair election to give the majority six seats?

Expand full comment

The obvious solution is to elect all representatives within a state at-large, proportionally. But then they wouldn't represent districts, just states, and that's what Senators are supposed to do, isn't it?

So the next step is add a third chamber and revert the Senate to election by state governments, not people, so you have chambers representing the state government, the state people,and the district people.

Or do as in my Chartertopia, where each district elects the top three winners and they cast as many votes in the legislature as they won in their election; I call it proxying. But bill passage requires 2/3 majority by both proxies and head count. It eliminates the requirement for redistricting after every census, with the assumption that votes cast are roughly proportional to population.

Every voter can also submit a name, one of which is chosen at random as the amateur, who proxies all the remaining election votes. Of course the volunteer has to accept; many might not be interested in interrupting their life to live in DC for a full term and then having to resume life back home. The idea is that while a truly random voter is likely to have voted for one of the three winners, these volunteered names are more likely to not support career politicians and be more disruptive and annoying to the professionals who actually campaigned for office.

I also like how much harder it is to predict legislative votes by both politicians and pundits.

Expand full comment

The thinking and the law combine two mistakes:

The first, of course, is that proportional representation is a natural norm. Seems to me no judge has studied the actual performance of proportional representation with its coalition governments. To get majorities, parties form coalitions with those who just lost the last election. Never mind Weimar. Postwar that was the Socialists in Italy. In Germany it's the gloriously misnamed Free Democratic Party, even the Greens. Voters can't throw the bums out [Popper].

The second, of course, is that political preferences correlate highly with observed characteristics like skin color. Why not use income, or age, or religion, or ... ? There are endless possibilities! US judges and lawmakers are clearly referring to the Irish norm! :-) But even that didn't last forever.

Expand full comment

“But I cannot be too hard on the mathematical incompetence of the author of the Economist article…”

But you should. Or if not at the author, then on the editor, or the assignment editor.

The chances that they would have written this piece had the political parties been reversed is *maybe* as high as 0.01%.

MA 9 Congresspeople are *all* Democrats, but only ~65% of voters voted Democrat. IL has 14 of 17 Dem house members (82%), yet voted less than 60% Dem. WA Congresscritters are 80% Dem, yet voters were less than 60% Dem.

Shockingly - shockingly! - none of these were the subject of, or even mentioned in, the story.

The Economist used to be broadly centrist, even if its slant was left of center on a couple of issues and surely technocratic and usually in favor of government interventions

Now it is across-the-board left-leaning like almost the entirety of the MSM, and I can no longer read it.

Expand full comment