Hacker News Reader

That's likely due to the Voting Rights act which requires what they call majority-minority districts. It prohibits redistricting to disempower racial minorities. It's another complication in the whole gerrymandering debate that makes it a lot harder to solve, because it requires districts that will look gerrymandered in a lot of cases.

See, the Russell conjugation is that it demands gerrymandering, to keep Democrat strongholds blue. I'm really only speaking to the 10% of you who can think about this stuff without feeling those delicious squirts of tribal belonging, not the 90% of people who ignored what I said so they could yawp.

I think a lack of fairness is a inherent problem with districting first past the post elections.

We could solve the problem by moving to a proportional system with much larger or no districts. The more proportional representatives you have per district, the less value you get for gerrymandering it as it gets harder to ignore voters.

Proportional representation goes a long way of increasing fairness. Societies are not split cleanly by geographic lines. You can have a lot more diverse ruling class by coalescing single person electorates into at least 30 electorates per district.

But what would also be needed is to increase the total numbers of representatives by a whole factor at least. USA is one of the least represented democracies in the world (only India is worse). Even China has more representatives per capita.

This also extends to local elections where cities as big as Seattle or San Francisco only have 9 and 11 members in their respective city counsels. There is no way to represent such big and diverse cities with this few seats. In comparison Oslo has 59 seats in their city council.

You don't need much larger districts to get a much better distribution, depending on how you allocate the remaining seats. Somewhat larger districts would arguably also improve representation for people, on the assumption that if there is a local representative from a party that supports your viewpoints you're more likely to be listened to.

To the first part, while Norway have pretty large districts, the method used to allocate levelling seats to achieve reasonable proportionality (it's not maximally proportional, as there is a lower threshold of 4% to be eligible for levelling seats) maintains a regional link by sorting the parties in descending order of wasted votes (votes that failed to get them a direct mandate) nationally. Then you pick the party with the most wasted votes, then you pick the district with a free levelling seats where that party had the most "unused" wasted votes, and allocate it. You then remove the requisite number of wasted votes, and repeat the process until all the levelling seats are allocated. This means the representatives are all drawn from the regional lists, and are the closest possible match.

The higher the ratio of direct mandates to levelling seats are, the more likely the levelling seats will go to a party that matches the preferences of the population of that region, but even with one mandate and one levelling seat, most places will elect someone via the direct mandate that is withing normal expectations, and so will rarely be any worse off.

Another lesson to draw from Norway is that part of the problem of the US and UK systems is that the local link largely matters because of confusing political and largely apolitical local matters. In Norway, a lot of the issues people in the UK at least write to their MPs about are things you in Norway would write to the County Governor about (there are only 10 of them, for a population of 5m, scale accordingly). The role of County Governor is one often given to former politicians (e.g. about half of the current ones are former cabinet ministers, almost all are former MPs), who knows the system and have connections to get things done, but it's not party political, and they have a civil service staff to assist. When I moved to the UK the notion that writing to an MP was a thing was bizarre - it felt to me like an abject failure of government if you have to write to the legislative about non-political local stuff. You can certainly talk to them in Norway to. When I highschool I just called up parliament for a school project to talk to an MP for my region. But it feels like there is much less need to do so, because you have other avenues and if it's a political issue, the proportionality means it's more likely there's an MP advocating for your view already.

As with software, whatever the real goal is should be decided and then approached directly. Any effect that is brought about as an accidental side effect of an unrelated architectural decision should be regraded as a flaw in the system.

If we want better local representation, we can strengthen local government, or create governments smaller than a city. We can eliminate geographic districts, as units that elect representatives to some higher level of government, but we can still have excellent local government.

If we want cultural representation, we can create a new branch of government, that allows people to identify as a particular culture, and then vote for members of that culture.

If we want religious representation, we can create a new branch of government, that allows people to identify as a particular religion, and then vote for members of that religion.

If we want racial representation, we can create a new branch of government, that allows people to identify as a particular race, and then vote for members of that race.

If we want language representation, we can create a new branch of government, that allows people to identify as speaking a particular language, and then vote for members who also claim to speak that language.

Again, the same rule we apply to software architecture should apply here. Whatever kind of representation we regard as important, we should approach the matter directly, and not try to achieve the effect accidentally, as the side effect of some other architectural decision.

Indeed, many of the problems we see in our political system are exactly because too many goals are being overloaded on too few institutions. If this was a legacy software app, and I was brought in as a consultant to clean it up, I would immediately suggest that a dramatic increase in both encapsulation and polymorphism was needed, and could only be achieved by introducing new branches of government.

To a limited extent, I implied as much when I wrote about the division of labor and the need for highly specialized committees:

https://demodexio.substack.com/p/why-have-politics-in-the-us...

How do these multiple branches of government set policy for the population as a whole?

That's an open question and I think everyone would answer it differently. In an earlier essay I suggested that we live in a complex world with a sophisticated division of labor, and therefore all real power should move to the specialized committees. In that model, the legislatures simply appoint people to the committees. If we had 5 legislatures and they each appointed one person per year to a committee, for a term of 5 years, then all the committees would have 25 people. The committees would set policy and pass the laws. I spoke of specialized committees here:

https://demodexio.substack.com/p/why-have-politics-in-the-us...

The purity of that argument is appealing, but the reality is that the underlying neighborhoods themselves are, literally, gerrymandered into these weird shapes, and trying to decouple districting policy from those early-20th century decisions essentially ratifies the injustice. "This district looks stupid" is not a good enough reason to eliminate Latino representation from Chicago, which (I think?) has more Latino residents than Phoenix and Dallas put together.

Drastically increasing the number of representatives sounds like a good solution.

Another approach is to incorporate that neighborhood as its own independent city. I think we should think about these issues creatively, while sticking with the idea whatever the goal, we should approach it directly, rather than achieving the effect accidentally, as a side effect of some other design decision.

Your proposal is to literally expel from government the people who inhabit a minority neighborhood?

Setting entirely aside the questions of morality, you also lose economies of scale by doing this, and there are already hierarchies that enable local governance as well as local representation in regional governance.

I'm not saying there aren't ways to improve the system, but what tptacek is trying to say is that "this shit is complicated and all about trade-offs, and you can't always get everything you want". In some ways it's the definition of a wicked problem.

It's not about "creative problem-solving towards some goal" it's about balancing competing goals from competing stakeholders. The status quo has stakeholders who benefit and will not like change, and will not like the characterization of the status quo as "an accident" or "side effects of some other design decision".

I think I was clear. My proposal is to have no geographic sub-districts to any polity. Everyone in a city should be free to vote for all candidates for city council, everyone in a region should be free to vote for all candidates for the regional assembly, everyone in a country should be free to vote for all candidates for the National Assembly. But if you really wanted to rescue some unique neighborhood, my suggestion is that you do so directly, without trying to achieve the effect as the accidental side effect of an unrelated architectural decision.

This sentence has no coherent meaning:

"It's not about "creative problem-solving towards some goal" it's about balancing competing goals from competing stakeholders. "

There is no meaningful contrast between the two phrases.

> I think I was clear. My proposal is to have no geographic sub-districts to any polity

You said to have them incorporate their own city. This seems different?

> There is no meaningful contrast between the two phrases.

Sure there is: in these cases, there is no single well-defined all-encompassing goal, and it may not be possible to define one.

"You said to have them incorporate their own city. This seems different?"

Everyone in a polity (a town or region or nation or perhaps special district) should be free to vote for every candidate running for that polity. The polity should have no geographic sub-divisions. If, as was mentioned above, a neighborhood in Chicago has a special culture, then:

its residents have their city concerns met when they participate in city elections,

and they have their regional concerns met when they participate in regional elections,

and they have their national concerns met when they participate in national elections.

But if you decided that their culture was so special, and deserved so much special protection, that the above was not good enough, then my suggestion is that the neighborhood should incorporate as its own city. Instead of being a neighborhood in Chicago, it should be an independent city with its own elections.

As I said, if you really wanted to rescue some unique neighborhood, my suggestion is that you do so directly, without trying to achieve the effect as the accidental side effect of an unrelated architectural decision.

"Sure there is: in these cases, there is no single well-defined all-encompassing goal, and it may not be possible to define one."

I was referring to Thomas's goals. What does Thomas want to do? What is he trying to achieve? Whatever his goal is, he should approach that directly, without relying on accidental side effects of unrelated architectural decisions.

Ok, I’ll bite: his goal is for every stakeholder to feel listened to, for their lives to get better, for no one to have an undue advantage, and for people to generally have what they need and want to the extent possible.

What’s the next step, then?

Thomas wrote:

"But what's really happening there is that there are in fact very well-defined Latino neighborhoods in Chicago, and the earmuffs capture a bunch of them neatly: Pilsen, Little Village, Cicero, Belmont-Cragin. If you know Chicago, you know these places, and you also know what the boundary between, say, Belmont-Cragin and North Austin is like; however artificial it looks on a map, it is a real border. That these communities are where they are is also not purely happenstance: a lot of very unfortunate social engineering took place in the early-mid-20th century to put those neighborhoods (and all the other neighborhoods) where they are now."

Assuming his goal is to protect the unique culture of that neighborhood, two options are:

1. the neighborhood already has its unique culture protected because the people in that neighborhood help elect the city government

2. the unique culture of the neighborhood is not sufficiently protected by the city government, therefore the neighborhood should incorporate and so become its own independent city

Again, whatever Thomas's goal is, I'd suggest that he achieve his goal directly, rather than trying to achieve the result as the accidental side effect of other architectural decisions. Gerrymandering is a very weak way of trying to defend the culture of a neighborhood. If his goal is to protect the culture of the neighborhood, then he should think of an architecture that would do so directly.

You are presenting option 2 as though it "achieves his goal directly" but it also comes with significant costs which you completely elide.

No option is clean.

Why should Thomas do something direct if something indirect achieves his goal, if a little bit less, but with far fewer harms?

Direct is not always better than indirect, especially in domains where any direct action has myriad indirect effects too.

What do cities have to do with congressional boundaries? Districts span cities.

My proposal is to have no geographic sub-districts to any polity. Everyone in a city should be free to vote for all candidates for city council, everyone in a region should be free to vote for all candidates for the regional assembly, everyone in a country should be free to vote for all candidates for the National Assembly. But if you really wanted to rescue some unique neighborhood, my suggestion is that you do so directly, without trying to achieve the effect as the accidental side effect of an unrelated architectural decision. The neighborhood with the unique culture has voters who will express their interests in city, regional, and national elections, and that should be enough, but if you felt strongly that this unique neighborhood deserves additional protections, then my suggestion is that you pursue that goal directly. Incorporating that neighborhood as its own independent city would be a way of achieving that goal.

About this:

"What do cities have to do with congressional boundaries? Districts span cities."

My point is that districts should not exist.