How many people per Congressional District is the ideal quantity?

[America Needs a 13-6 Progressive SCOTUS]

Lets see what people think.

I am personally unsure.

Log of what values people are posting as their personal preference (highest on top, lowest on bottom):

500,000 (singletxguyforfun)
350,000 (trajan)

1 total!

I want like 20 responses, and then I will discard the 5 highest and the 5 lowest responses, and calculate the average among the other 10, and then I want to try to draw maps with nonpartisan design based on that value, and I would encourage other posters to do the same.

[Bidenworth2020]

i like the wyoming model

[Strudelcutie4427]

500k

[America Needs a 13-6 Progressive SCOTUS]

i like the wyoming model

Way too arbitrary. The wyoming rule theoretically allows the possibility that the number of seats increases while the national population decreases, or that the number of seats decreases while the national population increases, as long as the population trends of the smallest state are more pronounced than the population trends of the country. I wish people would not talk about it as if it were some sort of way to improve the system.

[Xeuma]

Whatever the minimum is so that each state deserves their representatives (i.e., currently, the average district size is 700k but WY-AL represents only 500k, so Wyoming doesn't deserve that representative).

EDIT: I just realized that this is about the population size of each district, not the number of representatives. Obviously, for the above to work, we would be using the Wyoming rule. However, I've been tinkering around, and I really like the results of a 995-member house, which would make each district 350k, so that's my vote.

[Starpaul20]

I'm a fan of Wyoming rule, even if it's arbitrary.

[Hatman 🍁]

I also like the Wyoming rule. I'm sure it can be tinkered to deal with #KavanaghForPrison's criticisms.

[Indy Texas]

Those calling for a much lower district population should remember that if you start amping up the size of the House in the direction of 1,000, you're going to change the way it works.

With that many people, any single member is going to be more or less powerless. That's going to impact the type of people who run, and could end up with a New Hampshire General Court situation where the place is full of total goofballs because the barriers to running are so low in smaller constituencies.

[Hatman 🍁]

Those calling for a much lower district population should remember that if you start amping up the size of the House in the direction of 1,000, you're going to change the way it works.

With that many people, any single member is going to be more or less powerless. That's going to impact the type of people who run, and could end up with a New Hampshire General Court situation where the place is full of total goofballs because the barriers to running are so low in smaller constituencies.

No one is suggesting a congressional district should have like 10K people. Plenty of places in the world have smaller districts (e.g. Canada/UK/Australia where they are ~100k), and probably elect fewer 'goofballs' than the US congress.

[Orser67]

I'll be contrarian and say 1 million. I want a smaller House, even if it means that the smallest states are overrepresented.

[Xeuma]

Here are rough approximations for the suggestions thus far. I'll update this post periodically.

1 million:


500,000:


350,000:

[President Punxsutawney Phil]

https://uselectionatlas.org/FORUM/index.php?topic=297768.0
650 member house

https://uselectionatlas.org/FORUM/index.php?topic=289116.0
1 million people per seat

[Boss_Rahm]

I support the cube root rule, so the size of Congress and the size of the average district would grow with the country's population. Right now that would mean 691 seats, with an average of 476K per district.

[DC Al Fine]

Those calling for a much lower district population should remember that if you start amping up the size of the House in the direction of 1,000, you're going to change the way it works.

With that many people, any single member is going to be more or less powerless. That's going to impact the type of people who run, and could end up with a New Hampshire General Court situation where the place is full of total goofballs because the barriers to running are so low in smaller constituencies.

No one is suggesting a congressional district should have like 10K people. Plenty of places in the world have smaller districts (e.g. Canada/UK/Australia where they are ~100k), and probably elect fewer 'goofballs' than the US congress.

Our parliamentary system allows the parties to have better "bozo filters" though. American parties wouldn't do anything like our background checks.

[jimrtex]

I support the cube root rule, so the size of Congress and the size of the average district would grow with the country's population. Right now that would mean 691 seats, with an average of 476K per district.

Under the cube root rule, the size of each district is the square of the number of districts.

USPOP^2/3

[The Free North]

The New Hampshire system is impractical on a national level for logistical reasons, but representative/3k residents sounds good.

[pops]

100,000 people

[morgieb]

I like 500k, personally.

America is a bit too big to have the ideal amount of people per CD though which hurts.

[Co-Chair Bagel23]

75k

[MAINEiac4434]

100,000 people

[Lachi]

350k imo

[politicallefty]

I support the cube root rule, so the size of Congress and the size of the average district would grow with the country's population. Right now that would mean 691 seats, with an average of 476K per district.

That's what I like as well. It strikes a reasonable balance between a large population and an adequately-sized legislature.

[muon2]

If the concern is to make the small population of WY fairly represented, one way to determine the number is to return to the original Jefferson method of apportionment used in the early US. In that method you divide each state's population by a series of integers from 1 to the maximum number of seats and place them in a table with one state per row and one divisor in each column. Then to apportion N seats, look at the N largest numbers in the table and assign that number of seats to each state based on how many in each row are on the list of largest numbers.

In this case I'll let the number of seats vary until WY gets one seat, then stop and see how many seats are assigned. The equivalent process is to divide each state's population by a number d such that the quotient is larger than the WY population, but would be smaller if you divided by d+1. For instance in 2010 WY has a population of 586.3 K and AL has a population of 4,803.0 K. If I divide the population of AL by 8 it is 600.4 K and by 9 is 533.7 K, so AL would get 8 seats here. I can repeat that for each state and add them up to get the size of Congress such that WY is the threshold to get a seat.

In 2010 the House would have 523 seats.
In 2020 my projections are that the House would have 540 seats.

The seats could be left apportioned according to the Jefferson model (great for large states) or they could be reapportioned from that number based on the current Huntington-Hill method (best for smaller states).

[MarkD]

If the concern is to make the small population of WY fairly represented, one way to determine the number is to return to the original Jefferson method of apportionment used in the early US. In that method you divide each state's population by a series of integers from 1 to the maximum number of seats and place them in a table with one state per row and one divisor in each column. Then to apportion N seats, look at the N largest numbers in the table and assign that number of seats to each state based on how many in each row are on the list of largest numbers.

In this case I'll let the number of seats vary until WY gets one seat, then stop and see how many seats are assigned. The equivalent process is to divide each state's population by a number d such that the quotient is larger than the WY population, but would be smaller if you divided by d+1. For instance in 2010 WY has a population of 586.3 K and AL has a population of 4,803.0 K. If I divide the population of AL by 8 it is 600.4 K and by 9 is 533.7 K, so AL would get 8 seats here. I can repeat that for each state and add them up to get the size of Congress such that WY is the threshold to get a seat.

In 2010 the House would have 523 seats.
In 2020 my projections are that the House would have 540 seats.

The seats could be left apportioned according to the Jefferson model (great for large states) or they could be reapportioned from that number based on the current Huntington-Hill method (best for smaller states).

I like the second method the best, especially since I can't make head nor tail out of the first one. 540 seats is not too many (notwithstanding the fact that Capitol Hill will need a little bit bigger House chamber, and there will need to be more office space near the Cannon, Longworth, and Rayburn buildings).

[muon2]

If the concern is to make the small population of WY fairly represented, one way to determine the number is to return to the original Jefferson method of apportionment used in the early US. In that method you divide each state's population by a series of integers from 1 to the maximum number of seats and place them in a table with one state per row and one divisor in each column. Then to apportion N seats, look at the N largest numbers in the table and assign that number of seats to each state based on how many in each row are on the list of largest numbers.

In this case I'll let the number of seats vary until WY gets one seat, then stop and see how many seats are assigned. The equivalent process is to divide each state's population by a number d such that the quotient is larger than the WY population, but would be smaller if you divided by d+1. For instance in 2010 WY has a population of 586.3 K and AL has a population of 4,803.0 K. If I divide the population of AL by 8 it is 600.4 K and by 9 is 533.7 K, so AL would get 8 seats here. I can repeat that for each state and add them up to get the size of Congress such that WY is the threshold to get a seat.

In 2010 the House would have 523 seats.
In 2020 my projections are that the House would have 540 seats.

The seats could be left apportioned according to the Jefferson model (great for large states) or they could be reapportioned from that number based on the current Huntington-Hill method (best for smaller states).

I like the second method the best, especially since I can't make head nor tail out of the first one. 540 seats is not too many (notwithstanding the fact that Capitol Hill will need a little bit bigger House chamber, and there will need to be more office space near the Cannon, Longworth, and Rayburn buildings).

Actually the Jefferson method is one of the most mathematically simple, so I apologize for such a cursory description. By its more common name, D'Hondt, it's used by dozens of countries to award seats in multiparty proportional elections.

BTW in a visit I made to the US House in 2001, the capitol architect said that the chamber could accommodate about 600 without significant remodeling. Note that the House and Senate fit comfortably together there for the SOTU.