Thanks for this walk through Steven, I found it really helpful. I'm intrigued by the Australian PRCV approach, with its list design. I wonder if you could get much of the same benefit (reducing leakage, organizing choices for voters, etc) just by using the lists and not even including the "above the line" box. Just allowing folks to rank the list if they choose. I don't know if any jurisdiction uses a design exactly like that. For US/state elections, that might be a good way to let voters vote for their party, without putting an explicit party vote on the ballot.
Hi Cosmo, thanks for your thoughts. In Australia's Senate elections, and also in some state legislative elections, voters have the option of either ranking individual candidates or checking a box to select all of one party's candidates. Many Aussie voters do in fact check the box. But as far as I know, it's not a common practice to allow voters who are picking a party in a List system to rank multiple parties. Which would actually be a good idea, when you think about it, because what happens if a voter picks a losing party? Their vote is "wasted," as much as any American voter voting for a Libertarian or Green candidate. Ranked ballots actually have great utility in many different types of elections. In Germany there have been elections (such as in 2013, I recall) in which the number of wasted votes on small parties who did not reach the 5% threshold may well have resulted in a different government formation.
So as US reformers refine their proposals, it would make sense to incorporate ranked ballots where it can add real value. Even if it hasn't been done anywhere else, America's culture, history and traditions are unique enough that it might make sense to design a "never been used before" hybrid that fits the current moment. Thanks, all the best
I'll have to look again, but my understanding was that if you're voting above the line on the Australian Senate ballot, you actually have to rank your preferences among the lists. (Or at least some of the lists.) My suggestion for US (or California!) elections is actually going in the other direction. Keep the lists, get rid of the boxes above the line. If they want to. voters could still rank everyone on a particular list in order. That would effectively make it a vote for the party list, without calling it a party vote.
All good points Tom. But for the US House, even states with three or four seats each won't really see many of the benefits of PR, such as multi-party democracy, minor parties getting elected, etc. It would be more like Illinois' use of cumulative voting for 110 years for state legislature, in which every three-seat district had both Ds and Rs elected, including Chicago Republicans. Certainly that's better than single-seat districts which has led to increased geographic polarization, but only 30 states have 5 or more US House seats, so that's a pretty small playing field for reform.
But at the state legislative level, every state has dozens of seats so that's enough to demonstrate more of the benefits of PR. Given the difficulties of passing a law thru the US Congress to enact PR or even to allow states to experiment with it -- FairVote has been trying to pass a "local option" bill for decades with no success -- and given the bitter partisanship in Congress right now -- I would not recommend putting a lot of effort into passing something at the federal level. For the federal level, I would recommend trying to establish a high profile blue ribbon commission, like New Zealand did and which became the catalyst for the eventual change from WTA to PR.
Agreed. But I think the case for Pro Rep at the federal level will not be sufficiently strong until PR has been passed in several states first. Historically, the states have been the laboratories for new ideas.
states can make their own rules to elect federal members so that is possibility get it in state elections and then get PR in those states at federal level..
maybe even start at city level --
some cities in California use STV, so perhaps that can drift over to state level.
plus STV is used in Cambridge Mass. city elections.
compares STV and FPTP, as alternatives to the old-outdated methods still used in some places -- plurality block voting and the designated seat system (AKA post-seat system).
(interesting to see FPTP as an alternative to bad systems but it is true many jurisdictions still use block voting, which is less good even than FPP, and so much more frustrating as its multi-member districts can so easily do good if voter only had one vote. (even if not transferable --
SNTV works good in Vanuatu elections, if you care to check it out.
haven't quite got into the Benade paper yet but looks like it talks actually about STV, not the IRV that so often is discussed when "ranked voting" is discussed.
with STV , any candidate that picks up Droop worth of votes in the state will be elected.
(quick definition: Droop is even lower than votes divided by seats.)
so say any candidate with a third of votes in a state will take a seat in a state with two seats.
that means likely that two main parties will each be represented, and 90+ percent of vote will be used to secure representation.
this produces multi-party representation (defined strictly - two or more), balance and high percentage of effective votes. those actually are goals of PR.
we must ask ourselves what is goal of elections?
you might seem to say, as many do say, it is to elect an all-powerful government, party with a majority of seats
but it can be said that the goal of elections is to fill seats (so quick and dirty FPTP is good enough ) or to allow the voter a chance to express their sentiments, whether it results in election or not.
PR says voter should have liberty to express their sentiment and also (not coincidentally) that they would have best chance possible to elect someone as well.
but yes some states with just two seats will give both to one party - Hawaii, Idaho - but that is just because two-thirds of voters support one party. it would be unjust not to.
New Hamp. would give one to each, it seems.
in California with 52 Representatives, a candidate with 2 percent of votes would be elected if seats filled in state-wide contest. no third party took that amount of CA votes in 2024 but with lower threshold for representation and fairer system, that might change under PR.
in CA functionally perhaps 17-17-18 seat size is recommended and even then complaints about lack of local representation. (but any 2 percent will get seat whether placed due to local allegiance, or to party, or to religion, or to race, or to gender, etc.)
even with three districts, that 2 percent holds true roughly -- only now that 2 percent of state vote would have to be concentrated in just one third of the state. but still that is much different than trying to get plurality in 1/52nd of the state as now.
list PR operates same as STV in at-large context or in multi-seat districts, except it is based on party label.
two-seat district magnitude is not great but it does offer some advances over present FPTP in two arbitrary or gerrymandered districts.
it may produce mixed, balanced representation that would break down the artificial regionalization and polarization.
with fair voting used (STV or list PR or even SNTV) it gives most voters someone of like mind as a rep., whether that is two thirds of voters for one party that elects two, or one of each major party.
and anyways Greens or Libertarians in states with "flat" representation would be able to look to Greens or Libertarians elected in states with larger district magnitude for a voice, which they do not have now. it is not only the district result we have to look at but also the make-up of the legislature. and PR in large states produces advances for disenfranchised, unrepresented voters in small states
with fair voting, approx. each 347,000 votes (150M/435) would elect someone as long as it was concentrated in one state or district. so based on 2024 results, Libertarian might take two seats, Greens with twice its votes (potentially possible under fair voting) would take one.
hope would be to see PR elect members of several parties (the multi-party culture you mentioned), to break down the polarization and to open up democracy
such as in 1912 when (even under FPTP) Progressives and Socialists received 10 and 8 percent of the vote respectively.
and also if we are talking possibilities, the Senate could be elected in multi-seat contests -- two per states. there is no rule in constitution that they have to be elected separately, just that a third have to be elected each two years. Or change it so there are three per state, with all in a third of the states elected each two years
a party with majority of voters in a state ( if any) get two seats, the largest minority gets one,
or any party with a quarter of votes plus at least 1 vote would get a seat!
150M votes cast in 2024 is about 61 percent turnout based on estimated 245M voters. so that is low for a democracy.
perhaps increase to 500 seats would be acceptable, then Delaware and the Dakotas might have enough population to get two seats.
if turnout increases due to PR, up to say 80 percent, you would have 196M voting and in 500-seat House, 392,000 would be enough to elect a rep. (so both easier and harder than under today's FPTP)
but third parties would get more votes as they would be seen as more likely to get seats so two or three might quickly get seats and no party taking a majority of seats. a multiparty agreement would be needed to pass legislation, breaking down the polarization although there would still be intense battle between Democrats and Republicans. (even in PR NZ, there is still fight between Labour and Nationalists.)
Thanks Tom, your thoughts are not scattered at all, they are very interesting and provocative (in a good way). You make the case for how to design a PR system in the US House among the 50 states without increasing the size of the House. Which is a good plan to have on the table since, as much as many people would like to increase the size of the House, it may not be possible so a Plan B is necessary and you have outlined what that might look like.
I still think that we aren't likely to get PR in the US House until we can pass it first in a few states, and demonstrate how it works, why it's better, etc. If in a state like Pennsylvania with 203 house seats it can be demonstrated why multiparty democracy is better, and then in a few other states too, it becomes much more difficult to say "It will never work at the federal level."
Nevil Shute in his book In the Wet fictionalizes that if one state got fair elections and improved the quality of its representatives its people would eventually arise to leadership of national government and then other states would learn to adopt that fairness too.
In Shute's case this is achieved through giving more votes to (what he thought were) wiser voters.
all adults got one vote
and given one more vote if you have post-secondary educ, family, property, business, military service, public office, children, like that.
elected people from that state proved themselves to be wiser reflecting the changed perspective of most votes cast,
and other states soon copied that.
with PR being case of more diverse leadership, as opposed to wiser (as if there is a right and a wrong), the way to prove the effectiveness of PR would have to be different than that.
We i think rightly believe that voters will be happier (better satisfied anyway) and turnout will rise but will the quality of rep rise under PR?
possibly if we consider that under PR polarization will calm, the more-moderate, not the more-extreme, candidates will have advantage, that local political culture or accidents will have less impact on make up of cabinet - now a local district result can bar a significant player from the chamber while under PR, if that party has substantial support in a district, it will elect, and wil likely elect the most -high status candidate in the district - unless he personally has done something bad in eyes of voter.
while under FPTP even if he personally has done something bad in eyes of voter, he might be elected just because there is no one else from that party running in the district. so the case might be not so much that PR elects better as that FPTP can elect the bad.
and that advantage of PR is bound to have an effect sometime...
That's really interesting Tom. I had not heard of Nevil Shute's "In the Wet." His electoral system would certainly institutionalize various forms of inequality and division. Do you know if any places have actually tried a version of it?
some cities allowed property owners to cast a vote in each ward where they own property, and some allowed them to cast varying number of votes depending value of the property they owned.
a province in Canada gave property owners that right regarding prov. districts. there is an account of one man hurrying around and casting like 24 votes in one day. PEI maybe?
UK used to have university seats where grads would vote to fill the university seat. those people had two votes - one in their local district and one for the Univ seat.
Thanks for this walk through Steven, I found it really helpful. I'm intrigued by the Australian PRCV approach, with its list design. I wonder if you could get much of the same benefit (reducing leakage, organizing choices for voters, etc) just by using the lists and not even including the "above the line" box. Just allowing folks to rank the list if they choose. I don't know if any jurisdiction uses a design exactly like that. For US/state elections, that might be a good way to let voters vote for their party, without putting an explicit party vote on the ballot.
Hi Cosmo, thanks for your thoughts. In Australia's Senate elections, and also in some state legislative elections, voters have the option of either ranking individual candidates or checking a box to select all of one party's candidates. Many Aussie voters do in fact check the box. But as far as I know, it's not a common practice to allow voters who are picking a party in a List system to rank multiple parties. Which would actually be a good idea, when you think about it, because what happens if a voter picks a losing party? Their vote is "wasted," as much as any American voter voting for a Libertarian or Green candidate. Ranked ballots actually have great utility in many different types of elections. In Germany there have been elections (such as in 2013, I recall) in which the number of wasted votes on small parties who did not reach the 5% threshold may well have resulted in a different government formation.
So as US reformers refine their proposals, it would make sense to incorporate ranked ballots where it can add real value. Even if it hasn't been done anywhere else, America's culture, history and traditions are unique enough that it might make sense to design a "never been used before" hybrid that fits the current moment. Thanks, all the best
I'll have to look again, but my understanding was that if you're voting above the line on the Australian Senate ballot, you actually have to rank your preferences among the lists. (Or at least some of the lists.) My suggestion for US (or California!) elections is actually going in the other direction. Keep the lists, get rid of the boxes above the line. If they want to. voters could still rank everyone on a particular list in order. That would effectively make it a vote for the party list, without calling it a party vote.
under STV you can only vote for one candidate but you do mark preferences for multiple candidates if you want to.
your vote can only be used to elect one, or parsed out through fractional transfers altogether is the equivalent of one.
under STV you can mark your preference along party lines if you want.
under closed list list PR you must mark preference for party list.
some above the line marking system do allow backup party lists, i believe.
I disagree with you about the impossibility of having PR at federal level in the U.S.
sure 14 Representatives are elected in states with one or two members each. so scant possibility of PR there
but in 36 states, about 420 Representatives are elected where DM is three or more, sometimes considerably more than 3.
so that gives chance overall for significant PR improvement, even if some states would be left out of PR but they would be no worse than they are now.
and voters of small third parties there could look to members elected elsewhere to carry the ball.
All good points Tom. But for the US House, even states with three or four seats each won't really see many of the benefits of PR, such as multi-party democracy, minor parties getting elected, etc. It would be more like Illinois' use of cumulative voting for 110 years for state legislature, in which every three-seat district had both Ds and Rs elected, including Chicago Republicans. Certainly that's better than single-seat districts which has led to increased geographic polarization, but only 30 states have 5 or more US House seats, so that's a pretty small playing field for reform.
But at the state legislative level, every state has dozens of seats so that's enough to demonstrate more of the benefits of PR. Given the difficulties of passing a law thru the US Congress to enact PR or even to allow states to experiment with it -- FairVote has been trying to pass a "local option" bill for decades with no success -- and given the bitter partisanship in Congress right now -- I would not recommend putting a lot of effort into passing something at the federal level. For the federal level, I would recommend trying to establish a high profile blue ribbon commission, like New Zealand did and which became the catalyst for the eventual change from WTA to PR.
This is why we need to increase the size of the US House!
Agreed. But I think the case for Pro Rep at the federal level will not be sufficiently strong until PR has been passed in several states first. Historically, the states have been the laboratories for new ideas.
states can make their own rules to elect federal members so that is possibility get it in state elections and then get PR in those states at federal level..
maybe even start at city level --
some cities in California use STV, so perhaps that can drift over to state level.
plus STV is used in Cambridge Mass. city elections.
here's an interesting paper:
https://www.newamerica.org/political-reform/reports/the-future-is-proportional/
compares STV and FPTP, as alternatives to the old-outdated methods still used in some places -- plurality block voting and the designated seat system (AKA post-seat system).
(interesting to see FPTP as an alternative to bad systems but it is true many jurisdictions still use block voting, which is less good even than FPP, and so much more frustrating as its multi-member districts can so easily do good if voter only had one vote. (even if not transferable --
SNTV works good in Vanuatu elections, if you care to check it out.
I have a blog on it by the way -- Montopedia https://montopedia.wixsite.com/montopedia/post/what-is-sntv-and-is-it-proportional
haven't quite got into the Benade paper yet but looks like it talks actually about STV, not the IRV that so often is discussed when "ranked voting" is discussed.
with STV , any candidate that picks up Droop worth of votes in the state will be elected.
(quick definition: Droop is even lower than votes divided by seats.)
so say any candidate with a third of votes in a state will take a seat in a state with two seats.
that means likely that two main parties will each be represented, and 90+ percent of vote will be used to secure representation.
this produces multi-party representation (defined strictly - two or more), balance and high percentage of effective votes. those actually are goals of PR.
we must ask ourselves what is goal of elections?
you might seem to say, as many do say, it is to elect an all-powerful government, party with a majority of seats
but it can be said that the goal of elections is to fill seats (so quick and dirty FPTP is good enough ) or to allow the voter a chance to express their sentiments, whether it results in election or not.
PR says voter should have liberty to express their sentiment and also (not coincidentally) that they would have best chance possible to elect someone as well.
but yes some states with just two seats will give both to one party - Hawaii, Idaho - but that is just because two-thirds of voters support one party. it would be unjust not to.
New Hamp. would give one to each, it seems.
in California with 52 Representatives, a candidate with 2 percent of votes would be elected if seats filled in state-wide contest. no third party took that amount of CA votes in 2024 but with lower threshold for representation and fairer system, that might change under PR.
in CA functionally perhaps 17-17-18 seat size is recommended and even then complaints about lack of local representation. (but any 2 percent will get seat whether placed due to local allegiance, or to party, or to religion, or to race, or to gender, etc.)
even with three districts, that 2 percent holds true roughly -- only now that 2 percent of state vote would have to be concentrated in just one third of the state. but still that is much different than trying to get plurality in 1/52nd of the state as now.
list PR operates same as STV in at-large context or in multi-seat districts, except it is based on party label.
this was meant to follow my reply below.
two-seat district magnitude is not great but it does offer some advances over present FPTP in two arbitrary or gerrymandered districts.
it may produce mixed, balanced representation that would break down the artificial regionalization and polarization.
with fair voting used (STV or list PR or even SNTV) it gives most voters someone of like mind as a rep., whether that is two thirds of voters for one party that elects two, or one of each major party.
and anyways Greens or Libertarians in states with "flat" representation would be able to look to Greens or Libertarians elected in states with larger district magnitude for a voice, which they do not have now. it is not only the district result we have to look at but also the make-up of the legislature. and PR in large states produces advances for disenfranchised, unrepresented voters in small states
with fair voting, approx. each 347,000 votes (150M/435) would elect someone as long as it was concentrated in one state or district. so based on 2024 results, Libertarian might take two seats, Greens with twice its votes (potentially possible under fair voting) would take one.
hope would be to see PR elect members of several parties (the multi-party culture you mentioned), to break down the polarization and to open up democracy
such as in 1912 when (even under FPTP) Progressives and Socialists received 10 and 8 percent of the vote respectively.
and also if we are talking possibilities, the Senate could be elected in multi-seat contests -- two per states. there is no rule in constitution that they have to be elected separately, just that a third have to be elected each two years. Or change it so there are three per state, with all in a third of the states elected each two years
a party with majority of voters in a state ( if any) get two seats, the largest minority gets one,
or any party with a quarter of votes plus at least 1 vote would get a seat!
150M votes cast in 2024 is about 61 percent turnout based on estimated 245M voters. so that is low for a democracy.
perhaps increase to 500 seats would be acceptable, then Delaware and the Dakotas might have enough population to get two seats.
if turnout increases due to PR, up to say 80 percent, you would have 196M voting and in 500-seat House, 392,000 would be enough to elect a rep. (so both easier and harder than under today's FPTP)
but third parties would get more votes as they would be seen as more likely to get seats so two or three might quickly get seats and no party taking a majority of seats. a multiparty agreement would be needed to pass legislation, breaking down the polarization although there would still be intense battle between Democrats and Republicans. (even in PR NZ, there is still fight between Labour and Nationalists.)
Thanks for reading these scattered thoughts.
Thanks Tom, your thoughts are not scattered at all, they are very interesting and provocative (in a good way). You make the case for how to design a PR system in the US House among the 50 states without increasing the size of the House. Which is a good plan to have on the table since, as much as many people would like to increase the size of the House, it may not be possible so a Plan B is necessary and you have outlined what that might look like.
I still think that we aren't likely to get PR in the US House until we can pass it first in a few states, and demonstrate how it works, why it's better, etc. If in a state like Pennsylvania with 203 house seats it can be demonstrated why multiparty democracy is better, and then in a few other states too, it becomes much more difficult to say "It will never work at the federal level."
Thanks for your very interesting comments!
Nevil Shute in his book In the Wet fictionalizes that if one state got fair elections and improved the quality of its representatives its people would eventually arise to leadership of national government and then other states would learn to adopt that fairness too.
In Shute's case this is achieved through giving more votes to (what he thought were) wiser voters.
all adults got one vote
and given one more vote if you have post-secondary educ, family, property, business, military service, public office, children, like that.
elected people from that state proved themselves to be wiser reflecting the changed perspective of most votes cast,
and other states soon copied that.
with PR being case of more diverse leadership, as opposed to wiser (as if there is a right and a wrong), the way to prove the effectiveness of PR would have to be different than that.
We i think rightly believe that voters will be happier (better satisfied anyway) and turnout will rise but will the quality of rep rise under PR?
possibly if we consider that under PR polarization will calm, the more-moderate, not the more-extreme, candidates will have advantage, that local political culture or accidents will have less impact on make up of cabinet - now a local district result can bar a significant player from the chamber while under PR, if that party has substantial support in a district, it will elect, and wil likely elect the most -high status candidate in the district - unless he personally has done something bad in eyes of voter.
while under FPTP even if he personally has done something bad in eyes of voter, he might be elected just because there is no one else from that party running in the district. so the case might be not so much that PR elects better as that FPTP can elect the bad.
and that advantage of PR is bound to have an effect sometime...
That's really interesting Tom. I had not heard of Nevil Shute's "In the Wet." His electoral system would certainly institutionalize various forms of inequality and division. Do you know if any places have actually tried a version of it?
some cities allowed property owners to cast a vote in each ward where they own property, and some allowed them to cast varying number of votes depending value of the property they owned.
a province in Canada gave property owners that right regarding prov. districts. there is an account of one man hurrying around and casting like 24 votes in one day. PEI maybe?
UK used to have university seats where grads would vote to fill the university seat. those people had two votes - one in their local district and one for the Univ seat.
forgot to say in my reply below
cities sometimes switched to fewer wards or to at-large district so as to reduce the number of votes property owners could cast.
I've written an MMS vs STV article that approaches things from a more far-right resistance perspective and that heavily favours STV. It's about the UK state of play as well, but I hope it adds some useful perspectives. https://open.substack.com/pub/ewanhoyle/p/the-electoral-reform-debate-cant?utm_source=share&utm_medium=android&r=2u1072