Your argument excluding list or MMP systems for the US House of Representatives makes sense since many states are too small to allow for more than a handful of seats. But I am unpersuaded by your more general misgivings about list or MMP systems. My arguments would apply to the MMP systems in Germany or New Zealand, but here I limit myself to the list system as used in Sweden.
In Riksdag elections, 310 of the members are elected using party-list PR in each of Sweden's 29 electoral constituencies, each with 11 or more seats (except for Island of Gotland with 2). The remaining 39 seats in the Riksdag are distributed amongst the parties in order to constituencies where they were underrepresented in the overall outcome of the 310 in numbers to bring the party distribution in the Riksdag as close as possible to the distribution of the votes nationally.
To win seats in the Riksdag, a party must win at least four percent of the vote nationally, or twelve percent of the vote in one electoral constituency. You argue that this is unfair to those supporting parties unable to meet the requirements, unlike systems that allow the voters second choices to count. This argument really does not apply to Sweden since the polls are very accurate, so the voter knows when his or her part has fallen below 4%, and can choose accordingly.
There is also the matter of voting for the individual as opposed to the party. My sense is that Sweden has found the right balance. Voters can cast personal votes for individual candidates on the party lists. Any candidates who receive a number of personal votes equal to five percent or greater of the party's total number of votes in the constituency will automatically be bumped to the top of the list, regardless of their ranking on the party’s list. But most voters do not do this, which to me means that party supporters tend to have confidence in their party’s leaders, not a bad thing overall these days.
You mentioned that a 4-seat district has a 20% threshold, leading to wasted votes in a party list system. But in a party list system with no legal threshold, parties below this number can still be elected if voters are fragmented. For example, if six parties are all at around 16%, then the three parties with the most votes earn one seat each despite none of them hitting the threshold. Furthermore, STV allows up to 20% of the votes to be wasted, since this many votes can be cast for a loser that makes it to the final round, so the votes never transfer to a winner. Being above 20% guarantees a win in both systems, but being below is not necessarily a loss in either system.
National party list systems being used often waste more votes than necessary by choice, since they enforce a high legal threshold (such as a 5% threshold in Germany, where reaching 5% is good enough for about 40 seats but 4.9% gets you nothing) to prevent too many political parties and fractionalization. I would not enforce a threshold in the US, since small magnitude districts will have the same effect at reducing party count, and if anything we need more party fractionalization instead of less. No legal threshold means fewer wasted votes (though it certainly doesn't bring them to 0).
All this to say, votes will still be wasted under party lists, and probably more than STV, but I think the issue was slightly overstated.
Hello Steven,
Your argument excluding list or MMP systems for the US House of Representatives makes sense since many states are too small to allow for more than a handful of seats. But I am unpersuaded by your more general misgivings about list or MMP systems. My arguments would apply to the MMP systems in Germany or New Zealand, but here I limit myself to the list system as used in Sweden.
In Riksdag elections, 310 of the members are elected using party-list PR in each of Sweden's 29 electoral constituencies, each with 11 or more seats (except for Island of Gotland with 2). The remaining 39 seats in the Riksdag are distributed amongst the parties in order to constituencies where they were underrepresented in the overall outcome of the 310 in numbers to bring the party distribution in the Riksdag as close as possible to the distribution of the votes nationally.
To win seats in the Riksdag, a party must win at least four percent of the vote nationally, or twelve percent of the vote in one electoral constituency. You argue that this is unfair to those supporting parties unable to meet the requirements, unlike systems that allow the voters second choices to count. This argument really does not apply to Sweden since the polls are very accurate, so the voter knows when his or her part has fallen below 4%, and can choose accordingly.
There is also the matter of voting for the individual as opposed to the party. My sense is that Sweden has found the right balance. Voters can cast personal votes for individual candidates on the party lists. Any candidates who receive a number of personal votes equal to five percent or greater of the party's total number of votes in the constituency will automatically be bumped to the top of the list, regardless of their ranking on the party’s list. But most voters do not do this, which to me means that party supporters tend to have confidence in their party’s leaders, not a bad thing overall these days.
Best
Henry
Henry.milner@umontreal.ca
You mentioned that a 4-seat district has a 20% threshold, leading to wasted votes in a party list system. But in a party list system with no legal threshold, parties below this number can still be elected if voters are fragmented. For example, if six parties are all at around 16%, then the three parties with the most votes earn one seat each despite none of them hitting the threshold. Furthermore, STV allows up to 20% of the votes to be wasted, since this many votes can be cast for a loser that makes it to the final round, so the votes never transfer to a winner. Being above 20% guarantees a win in both systems, but being below is not necessarily a loss in either system.
National party list systems being used often waste more votes than necessary by choice, since they enforce a high legal threshold (such as a 5% threshold in Germany, where reaching 5% is good enough for about 40 seats but 4.9% gets you nothing) to prevent too many political parties and fractionalization. I would not enforce a threshold in the US, since small magnitude districts will have the same effect at reducing party count, and if anything we need more party fractionalization instead of less. No legal threshold means fewer wasted votes (though it certainly doesn't bring them to 0).
All this to say, votes will still be wasted under party lists, and probably more than STV, but I think the issue was slightly overstated.