Currently in the US Senate, each state receives equal representation, no matter its size.
Some have proposed modifying this to give larger states more representation.
One idea would be give each state 2 sets of the following groups of senators:
- Divide the state’s population by the population of the average state, rounding down to the nearest integer, or if zero,
- A minimum of one senator
In other words, small states like Wyoming would still have a single senator elected in two out of every 3 biannual cycles, [2 total senators] for a 6 year term each, but larger states like California would get Floor (40 million * 50 / 330 million) = 6 senators in two out of every 3 cycles, for a total of 12 senators.
In practice, only the 4 largest states (CA, FL, TX, NY) would get more than 2 senators total, with current populations.
This means that while smaller states would still need a good single winner method, larger states should have some form of PR.
If, instead, you rounded to the nearest integer, that would give more states an extra senator (adding PA, IL, OH, GA, NC, MI to the list above).
The overall size of the US Senate would increase to ~130-140, not unmanageable, with small states still having disproportionate representation.
The Electoral College would become more representative, though still disproportionately favoring small states.
With PR, the overall balance of parties might not change drastically, since even “blue” states like CA have sizeable GOP minorities. But that might be balanced by Democratic minority representation in the TX delegation.
Anyway, I thought I would put this out there as a topic of discussion, since it would be an interesting application for multiwinner election methods, and would require a technique that could scale up to 6 or 7 winners with around 20 million voters, and who knows how many candidates.