Varying multi-member districts vs. one person-one vote

[Mr. Matt]

In both houses of the Vermont legislature and the New Hampshire House (and probably others as well), there are many districts that vary the number of people that can win per district, e.g., the 11-member district of Hillsborough-37 (consisting of Hudson and Pelham, NH) while there are plenty of single-member districts throughout. I'm trying to wrap my head how a person in Hudson, NH gets to vote for 11 people while the Dixville Notch voters only get to choose one house member and not run afoul of one person-one vote idea or the Reynolds vs. Sims decision.

[America Needs a 13-6 Progressive SCOTUS]

In both houses of the Vermont legislature and the New Hampshire House (and probably others as well), there are many districts that vary the number of people that can win per district, e.g., the 11-member district of Hillsborough-37 (consisting of Hudson and Pelham, NH) while there are plenty of single-member districts throughout. I'm trying to wrap my head how a person in Hudson, NH gets to vote for 11 people while the Dixville Notch voters only get to choose one house member and not run afoul of one person-one vote idea or the Reynolds vs. Sims decision.

The number of members per District is scaled proportionally to the amount of people in the District.

Take the West Virginia House of Delegates for Example.

There are Districts with total numbers of members varying from 1-7 (or less than 7, I cannot remember which). Basically, as an example, the 3 member Districts have 3 times as many people as the 1 member Districts do. Its basically similar to how the US House would work if states used winner takes all at large instead of Districts.

[Kevinstat]

#KavanaughForPrison's reply was good, but I'd also point out that New Hampshire has "floterial" districts, so that is one town has 1.3333/400 of New Hamphire's population, and a neighboring town has 2.6667/400 of the state's population, the first town will constitute a single-member district, the second town will constitute a multi-member district, and the two towns together will constitute a floterial district.  Since 2/3 of the voters in the floterial district live in the larger town, that town (or that town's non-floterial district) could be thought of as having 2_2/3 (~2.6667) Representatives and the smaller town could be thought of as having 1_1/3 (~1.3333) Representatives.

Of course, the math usually doesn't work out so well.  If Town A has 1.65 quotas and Town B has 1.35 quotas, and each Town is given one Represenatative and the two towns share a third, then Town A (or the Town A "component district") will have effectively 1 + (1.65/(1.65 + 1.35)) = 1 + (1.65/3) = 1.55 Representatives overall while Town B (or the Town B "component district") will have effectively 1 + (1.35/(1.65 + 1.35)) = 1 + (1.35/3) = 1.45 Representatives overall.  Expressing that in terms familiar to those who are used to thinking in terms of deviation with single-member districts, the Town A component district's deviation is 1.65/1.55 - 1 or +6.45%, while the Town B component district's deviation is 1.35/1.45 - 1 or -6.90%.  One might feel like Town A is better off as it is a bit more likely to get the third representative, but strictly mathematically, using the so-called component method* to measure deviation, Town A is underrepresented while Town B is overrepresented.  That realization in New Hampshire resulted in some towns (like Pelham) remaining in large multi-member districts dominated by a larger town (Hudson in Pelham's case) even though they had been promised a district of their own as a result of the 2006 state constitutional amendment authorising the return of floterial districts.  There was also some Republican hackishness that contributed to that, like interpreting the 2006 constitutional language strictly in some ways and not in others that resulted in discarding what might have been some more sensible options using floterial districts.

*See the footnote on page 1765 of https://www.bu.edu/bulawreview/files/2019/01/FOSTER.pdf .

[Kevinstat]

Take the West Virginia House of Delegates for Example.

There are Districts with total numbers of members varying from 1-7 (or less than 7, I cannot remember which). Basically, as an example, the 3 member Districts have 3 times as many people as the 1 member Districts do. Its basically similar to how the US House would work if states used winner takes all at large instead of Districts.

West Virginia, interestingly enough, will be moving to entirely single-member districts for its House of Delegates beginning with the 2022 elections (after redistricting following the 2020 census).  See http://www.ncsl.org/blog/2018/04/10/west-virginia-moves-to-single-member-districts.aspx .

According to that article, the range of members per district is currently 1-5, but it might have been 1-7 in the past.

[Kevinstat]

A table of states with multi-member districts in either legislative chamber (which is wrong in regards to South Dakota, as the two Senate districts marked as having "2 posts" for the House of Representatives (with the implication that both posts still cover the entire districts) are actually each subdivided into two single-member House districts to facilitate the election of Native American Representatives): https://ballotpedia.org/State_legislative_chambers_that_use_multi-member_districts

I also think the "partial absention" mentioned in elections for the Arizona House of Represenatives does not reflect anything "exotic" about the election method itself, but just voter habit.  If I were on the ballotpedia staff I would have argued against mentioning that in the table.  Similar partial absentions may exists in other chambers with districts where voters can vote for several members.

Postscipt: I just re-read that article.  I serously doubt that any (let alone all) of the chambers listed as electing their members by the "Bloc" method actually require voters to vote for as many candidates as they are allowed to (or not have their votes for the candidates they do vote for count).  It's probably just that the abstention existing in Arizona has been more publically acknowledged.

[jimrtex]

In both houses of the Vermont legislature and the New Hampshire House (and probably others as well), there are many districts that vary the number of people that can win per district, e.g., the 11-member district of Hillsborough-37 (consisting of Hudson and Pelham, NH) while there are plenty of single-member districts throughout. I'm trying to wrap my head how a person in Hudson, NH gets to vote for 11 people while the Dixville Notch voters only get to choose one house member and not run afoul of one person-one vote idea or the Reynolds vs. Sims decision.

The Alabama Constitution apportioned (apportions, it has not been updated since the Reynolds v Sims decision) one representative to each of the 67 counties, and apportions the remaining 39 based on population. Though the constitution required an apportionment after every census, it had never been done since the 1901 Constitution was enacted. Even it the House would have been reapportioned, the 39 extra seats were insufficient to equalize representation.

It would be like if the US House had 75 representatives.

The apportionment would be CA 7, TX 5, NY 4, FL 4, IL 3. PA 2. O 2. MI 2, GA 2, NC 2, NJ 2, VA 2, WA 1, ....

A majority of the House could be elected from states with 32.4% of the population, and Washington and Wyoming would each have a single representative, even though Washington has almost 12 times as many people.

There would be insufficient representatives to provide any degree of proportionality. The problem was not that Jefferson County did not have representatives elected by district, but that too few representatives were apportioned to the county.

The rubric "one man, one vote" is misleading. If my district has 20,000 voters and yours has 100,000 voters, we each have "one vote" and assuming that we are both men.

The formulation is that our votes will have roughly equal weight. If the 100,000 voters can elect 5 representatives, and the 20,000 can elect one representative this would be true.

The 'Reynolds v Sims' decision itself suggests that multi-member districts might be used to achieve to comply with "one man, one vote" (equal protection).

Multi-member districts were later ruled to be unconstitutional in certain cases. If there were a minority neighborhood in Hudson or Pelham, it might be able to elect a Democratic representative, but with 11 at-large representatives they would be outvoted by (white) Republicans.

If the 11 representatives were elected by district it is quite possible that the Republicans could still elect 11 members. They had a pretty healthy plurality, about 58% in Hudson, and 60% in Pelham. If the Democrat did elect someone, it would be because of some oddball occurrence. If one of the R's was caught in a scandalous situation, voters might split their vote among the 11 D candidates, none of which would get that many more votes. But in a single-member district, there is only one D to vote for.