Hacker News Reader

I'm not really a fan of proportional representation and don't really have a dog in this race, but I'll point out something -

People have historically underestimated "tail risks" to institutions and social situations that they take for granted. I had an international relations professor once who was in the Situation Room on August 19, 1991; he worked for the Pentagon, and they had just finished their 5-year plan about likely force build-ups and military risks in Eastern Europe. Nobody had envisioned that there would be no Soviet Union, no Eastern Bloc, and no Cold War. It just didn't fit into their possible calculus; the Soviet Union had existed for 69 years, the Cold War for 46, and that was the entire adult lifespan of most of the people in the room.

I believe there's a non-negligible chance that within my lifetime, the United States of America will cease to exist. And furthermore, there's a non-negligible chance that the concept of the nation-state as a societal organizing principle may also cease to exist. People forget that the modern nation-state as we think of it - a strong central government with a monopoly on military power, public infrastructure, and a unified ideological worldview - is only about 150 years old. It actually post-dates the U.S. Constitution; many of the individual amendments (eg. the 2nd and the 10th) are remnants of a time where primary power resided within the States, and we literally were the "United States of America", not the "United States of America".

If this does come to pass, it's critically important that there be robust academic discussion of alternatives, and particularly alternatives to those aspects of the current system that are most problematic. Otherwise, we risk falling back on systems (eg. corporate feudalism) that provide safety & security but don't even bother to give lip-service to freedom or democracy, which I think would be a huge shame. We would not have an America without the ideas of the Enlightenment, which were largely developed in response to rising contradictions within Europe between the official institutions and the actual economic & social power on the continent.

150 years ? Seriously ? A strong democratic state maybe but since the peace of Westphalia in 15xx you had strong states with centralized authorities with the monopoly of power, infrastructure and world view.

The seeds of the idea of nationalism date to the Peace of Westphalia, but few of us would actually think of the Westphalian "states" as nations in the modern sense.

I'd date the modern nation-state to:

The American Civil War. (1865). There's a reason why this is called "The War Between the States" in the South - before the Civil War, it wasn't clear whether the U.S. was a loose confederation that any state could secede from at will, or a unified political body that would fight to maintain its integrity.

The unification of Germany (1871). This is geographically the same area as the Westphalian states, but people think of "Germany" as the nation, not Munster/Brandenburg/Westphalia/Saxony/etc.

The unification of Italy (1871). Again, we don't think of the Papal States as a nation, we think of Italy as one.

The Meiji restoration in Japan (1868), where decentralized power under the shogunate was centralized under the emperor.

Russification under Alexander II of Russia (1859-1882), which included both the emancipation of the serfs and a systematic campaign to promote the Russian language & culture within territories governed by the Russian empire.

Could you explain to me what would make Britain or France not a state far before that time?

nation state.

The issue is that territories often shifted hands, and what territories happened to be ruled by the same people often boiled down to which monarch happened to inherit or conquer which areas.

There simply was no clear unified concept of a nation of people.

The rise of the national states came in part as a result of Romantic Nationalism (or Nationalromanticism) [1], and the wave of revolutions that swept Europe in particular as a reaction against the above:

For the first time there was the rise of the sentiment that certain areas belonged to certain states because their people belonged together due to common language or culture, rather than because the territories were owned by a specific monarch, or because of a given convenient political association.

This coincided by wholesale creation of cultural mythos. E.g. Norway had not existed as an independent kingdom since 1380, and had only existed as a unified kingdom for a few hundred years at that time. But with the rise of Romantic Nationalism, authors and artists and politicians contributed to creating a mythos of a Viking era Norway and Norwegian people that had a common destiny.

Norway was nothing special in that respect - it is one example of dozens in Europe alone where nations were spun out of mythology or ideas about a shared history or destiny that largely boiled down to finding convenient political or language boundary and creating a mythos that fit them.

[1] https://en.wikipedia.org/wiki/Romantic_nationalism

Peace of Westphalia is 1648, not 15xx. And the Peace of Westphalia wasn't exactly a centralizing influence--in confirming that the Holy Roman Emperor could not interfere in the matters of its constituent states, it repudiated the attempt of the HRE to centralize.

Strictly speaking, the Roman Empire was a surprisingly centralized state for its size--the provincial governors spent a lot of time asking the Emperor for guidance. (That was a large part of the reason why it fell--it was just too big to be centrally governed, and the administration couldn't really effect the necessary decentralization). Decentralization was the norm from post-classical times up until the development of gunpowder, which allowed central governments to start to exercise a monopoly of force over their subordinates.

> Given all that, talking about proportional representation is a complete waste of time, and it gets in the way of having constructive conversation about this subject. It is constitutionally impossible to disband the House of Representatives and switch to proportional representation

You are confusing "proportional representation" (an outcome which many systems approach) with Party List Proportional (an election method which is designed to achieve proportional representation.)

There is no Constitutional barrier to many PR systems (e.g., STV in multimember districts) in the US, though there are some structural limits to the degree of proportionality that could be achieved (you'd still have to apportion by population and have each state as a separate pool.)

There are some federal statutory limits mandating single-member districts, but those are much easier to reform than Constitutional barriers.

> Every time I see a thread here I really do look forward to reading thoughts that are actually about gerrymandering and solving gerrymandering

Distortions of the kind labelled gerrymandering are fundamental to the exclusive use of independent single member districts (and exacerbated if you use plurality or majority-runoff elections in those districts.)

This is an inescapable fact. You can't complain that you want to hear solutions to gerrymandering and simultaneously exclude all the things that actually address gerrymandering from the acceptable solution space. Well, you can, but you are making irreconcilably conflicting demands when you do so.

>Distortions of the kind labelled gerrymandering are fundamental to the exclusive use of independent single member districts (and exacerbated if you use plurality or majority-runoff elections in those districts.)

Not so much. In Australia we have single member districts but not really gerrymandering because our districts are set by an independent non-political body using an algorithm and some common sense. Look at an electoral map of Australia and you'll see that our electorates are reasonably sensible shapes, they haven't been set up to favour one party over another.

our districts are set by an independent non-political body

Heh, yeah...we used to have those in the US, too. Is there anything in AUS law that prevents infiltration by members more political than the body's charter anticipated?

There's always gerrymandering as a concept: If you build everything by algorithm, you are making the problem far smaller, but ultimately the decision of which algorithm to use is never going to be as good as proportional representation. Whether one only looks at population density and geography, or one adds ethnicities or party affiliations, there is no such thing as an objective algorithm that will not lead to biased results. The results will be more fair under any definition than the gerrymandering in, say, North Carolina, but they are still not going to be wonderful.

Given how the US political landscape is, it's not difficult to build a generic system that produces results that will favor the same party in all states. Compactness will, in general, be closer to favoring Republicans, while something that aimed for even population density in each district would favor democrats.

If I had to choose between any algorithm and what we have now, I'd pick the algorithm, but it's still going to be a worse output than someone actively redistricting aiming for maximally competitive elections or a proportional system.

I think nothing in the US Constitution prohibits states from electing their members to the federal House of Representatives by a proportional representation method. For example, California has 53 representatives, and these could be elected not in 53 separate districts, but instead in a single state-wide election.

Making this change would require action by a state's legislature, and likely changes to federal law as well, but no federal constitutional amendment is needed.

I think it's highly germane to discuss alternative election systems in a discussion on fighting gerrymandering.

If people are concerned about the geographic link that too can be solved.

E.g. Norway uses proportionality within regions, and counts votes per party list and then keeps track of how many "wasted" votes each party has.

Then a small fixed number of seats are allocated in descending order of remaining wasted votes (each allocation removes an equivalent number of votes from that party-lists pool) from the region where that party had the most wasted votes.

There's also a 5% threshold in the national vote to be able to receive seats from that pool.

It both gets relatively close to proportional, still makes it possible for very popular independent candidates to get elected directly, and the seats used to even out are also linked to regions. The areas are just a bit larger.

Even 2-3 seats per district plus a few seats used to even out representation would give dramatic improvements in proportionality.

Correct, not constitutional, but by federal statute.

This was done variously by allowing and disallowing multi member districts, after reapportionment via a decenial census.

Multi-member districts were un-regulated in the early 1800s, then disallowed in the 1842 statute, but had a number of states with multi-member districts were in the subsequent election allowed to have their multi-member-district representatives seated. Then allowed again by silence in the statutes in 1850, then disallowed by statute again in 1862, and generally not allowed in subsequent reapportionment statutes through 1911, allowed by absence of regulation in 1929 (via Supreme Court ruling in 1932, Wood v. Broom, that the provisions of each apportionment act affected only the apportionment for which they were written).

Then disallowed most recently in 1967 by statute (as part of a civil rights concerns, after the 1965 Voting Rights Act, that southern states might resort to winner-take-all at-large elections to eliminate voting strength of recently-enfranchised blacks in the South).

See: The History of Single Member Districts for Congress: Seeking Fair Representation Before Full Representation - by Tory Mast

http://archive.fairvote.org/?page=526

Thing is, "fixing gerrymandering" sounds a lot like fixing the wrong problem - instead of devising complicated clever plans for fixing it, engineers will feel compelled to adress what they perceive as the "real" underlying problem.

Whether doing so is realistic or not is another question, but as a non-american, I can totally relate to this attitude - the whole problem just seems bizarre to me. But don't get me wrong, I am neither judging, nor do I think that "just switch to proportional representation" is realistic for the US without a revolution, or at least some miraculous bipartisan effort.

It's quite possible for state legislatures on the US without involvement of the parties-in-government, in states where citizen initiatives are a thing.

National reforms are often led by the states.

Gerrymandering is considered a bad thing, but one might reasonably ask why we ought to care about geographical subdivisions of voters at all, why not just allow a pure majority vote within a broader administrative unit.

This is, incidentally, the argument behind eliminating the Electoral College.

But the fascinating thing about the EC is that it's basically a custom "consensus algorithm" that is intended to prevent large disparities between the political power of states with very large vs very small populations.

This has the effect of increasing the political power of voters in remote regions, and decreasing the political power of voters in states with large populations. In effect, what we consider "consensus" in the EC system is different than what we'd consider consensus in a purely majority-vote system.

Congressional districting is an attempt to do the same thing at the state level... Consensus requires more than just appealing to the largest numerical majority. Interests that correlate with geography get slightly more political power than they would in a purely majority-vote system.

As we know from cryptocurrencies, changing the consensus algorithm is a hot topic, and there is always a group of winners and a group of losers.

Gerrymandering creates voting blocs that make it more likely that the bloc will elect a candidate that strongly supports the bloc's interests, but it also reduces the impact of the bloc's views on elections held in nearby blocs.

How this plays out in any specific scenario is largely the result of historical accident, but I'd offer the argument that allowing states to change the consensus algorithm for their own voters is not necessarily bad for anyone outside of that state.

I support the Electoral College because I think that without it, the issues that Americans considered at the national level would be very strongly dominated by the concerns of the most populous urban centers, which would lead to even more fractiousness between urban areas and "flyover" areas.

So states that tinker with their own consensus algorithms are potentially helping (but likely hindering) the ability to achieve stable consensus. This is a bad thing, but not something that should be prevented by Federal law. We should let them make their own mistakes and let more enlightened states (if any exist) lead by example.

> I support the Electoral College because I think that without it, the issues that Americans considered at the national level would be very strongly dominated by the concerns of the most populous urban centers, which would lead to even more fractiousness between urban areas and "flyover" areas.

and with it the majority of voters get overridden because of those flyover states. I don't like this argument because you are arguing for exactly the opposite as the better way.

Anyway, I still think the easiest way is to remove the cap on the house or raise it a lot. You still get the electoral college benefit you want but the disparity in power isn't so high.

I don't really like that though. Personally I think a popular vote would just result in more moderate policies because you need to at least split the cities, and then the rural areas are still valuable to put you over the top.

> and with it the majority of voters get overridden because of those flyover states. I don't like this argument because you are arguing for exactly the opposite as the better way.

Urban voters have the benefit of living in a densely populated area. There's something to be said for safety in numbers.

> Personally I think a popular vote would just result in more moderate policies because you need to at least split the cities, and then the rural areas are still valuable to put you over the top.

Campaigns would visit fewer areas than they do now. Voter fraud could become more easier to commit. Swing states would disappear.

Popular voting sounds great until you realize those who live away from larger voting blocs would get the short end of the stick.

> Voter fraud could become more easier to commit

Why?

> Swing states would disappear.

Thank fucking god for that.

> Thank fucking god for that.

Do you want swing states to no longer have votes that matter? Or do you wish that the kinds of issues that result in swing states would no longer exist?

> Do you want swing states to no longer have votes that matter

I don't want them to be the only ones that end up mattering.

> the only ones that end up mattering

Even in a popular vote system there would still be small issues that ended up mattering (seemingly too much) they would just not be as likely to be from geographical areas with low populations.

There are issues with broad support, and other issues with narrow support. In order to win, politicians must cater to both. The narrow, tie-breaker issues in a popular vote system would also (quite likely be just as stupid as they are today).

>the issues that Americans considered at the national level would be very strongly dominated by the concerns of the most populous urban centers, which would lead to even more fractiousness between urban areas and "flyover" areas.

This is a common sentiment. I'm curious about what you think about the following question:

Let's say we order cities from most populous to least populous. How many cities would we need to add together to get to roughly 50% of the population? Bonus points for putting down a guess before doing the math and checking.

I guessed about a 100 and that's a little low. Half the US lives in the top 146 biggest counties.

http://www.businessinsider.com/half-of-the-united-states-liv...

There are a few places where the split into counties obscures the contiguity of the population, such as greater NYC (the five boroughs, part of NJ, etc.) and greater SF (includes Oakland, Berkeley, etc., down to San Jose). The same is true in the greater DC area and greater LA area.

Once you merge those into virtual counties, the number of counties is a lot lower.

And, of course, there are places where the county's reach exceeds the city's. For example, the city of Milwaukee is deep blue, but Milwaukee County elects Sheriff Clarke.

> How many cities would we need to add together to get to roughly 50% of the population?

While this is an interesting question, it's not the one that we should ask:

Instead, what is the lowest number of flights a candidate needs to take to visit a majority of the voters' nearby area? If the candidate can drive for under three hours to arrive at another area, the trip does not qualify for a flight.

See these links for the map visuals supporting this line of reasoning:

http://www.arbesman.net/blog/2011/02/02/the-city-states-of-a...

and

https://www.nytimes.com/2016/04/17/opinion/sunday/a-new-map-...

A popular vote system decreases the number of flights substantially compared to the EC system, and would make much of the US utterly irrelevant in a national election. The EC system was created to prevent this from happening.

Also, flights are a proxy for overall campaign spending. A popular vote system would reward spending on the most populous visits and (compared to the EC system) penalize visits to the least populous areas.

Consent of the governed is important. When the winning side is disenfranchised, repeatedly, it undermines the legitimacy of the democracy.

>Given all that, talking about proportional representation is a complete waste of time, and it gets in the way of having constructive conversation about this subject. It is constitutionally impossible to disband the House of Representatives and switch to proportional representation.

I'm curious, what is the Constitutional argument against proportional representation within states? A complete proportional representation system that spans the states is obviously unconstitutional, but I'm not sure that a proportional representation system within states is. Can you expand on this?

There is no Constitutional limit. There is, however, a statutory limit (2 USC §2c) [0].

[0] https://www.law.cornell.edu/uscode/text/2/2c

See: The History of Single Member Districts for Congress: Seeking Fair Representation Before Full Representation - by Tory Mast

http://archive.fairvote.org/?page=526

It is not strictly impossible to abolish the House of Representatives. We have the amendment process. Incredibly unlikely, yes. If you weren't exaggerating, I would legitimately love to hear your logic, it sounds quite interesting to me.

The reason people love to talk about proportional representation is because it is a comparatively easy and elegant method to solve the problem. Legislating math is hard. Folks won't understand it.

People feel their vote doesn't matter. I say this as someone whose district just became a swing one out of nowhere.

Then start a thread that is actually about proportional representation, and I promise I won't come in and tell everyone that what we really need is a better gerrymandering algorithm.

Really important point you make at the end there: Proportional representation has to be made to be understandable to the layman or it will backfire. The electorate has to understand that it's more fair or else they'll reject it. Sometimes the rejection won't be felt for a long time, either. One could say the current political situation is the result of laypeople rejecting the system long ago (39% of the eligible voters didn't vote in 2016).

For all intents and purposes it's impossible, especially if you consider exactly who would be ratifying such an amendment anyway. This is the GP's point: it's unproductive to talk about this topic because there is no remotely plausible way for such an event to occur.

> It is constitutionally impossible to disband the House of Representatives and switch to proportional representation

the constitution dictates how representatives are apportioned and sets out the rules for how election laws are handled. The districting rules are not in the constitution. So why would it require an amendment? You wouldn't dissolve the house, you would just change the apportionment as needed (people/rep) and pass the appropriate laws.

It is constitutionally impossible to disband the House of Representatives and switch to proportional representation.

There are multiple kinds of proportional representation. You're focusing on one kind called a party list system, but it's perfectly possible to have PR within a constituency with a transferable vote.

Instead why not introduce Approval Vote ballots in every congressional district? BOOM solves most of the major problems.

It can be done on a local level and in each such district you will see people hate each other less, third parties get a real chance etc.

Google APPROVAL VOTE.

To put it in terms people here understand: suppose you have a big legacy code base written in C. It works fine but it's not perfect. And every time a bug comes up, there's a bunch of guys who say "hey, let's just rewrite the whole thing in Scala!"

No, we're not rewriting the whole code base on a production system just because you think Scala is neat. The existing code is fundamentally fine, let's focus on this problem today.

But what if the existing code is terrible?

(I still agree that "rewrite in Scala!" is a bad answer, but some kind of extensive refactoring or rewrite may be called for.)

> It is constitutionally impossible

It is merely a piece of paper, containing stuff that was the pinnacle of political thought centuries ago. It is not the will of god almighty. It is merely the operating system of this country's laws, and we upgrade OSs all the time - when they're too old, start glitching, and crash your apps all day long.

Stop worshiping a piece of paper with some old words on it.