Those who think National Popular Vote is what really matters...

[muon2]

In a parliamentary democracy the party that gets the most votes does not always win control of the government. Does that make a parliamentary system undemocratic? I don't think so. Neither does the fact that a president is elected by states electors make it undemocratic.

Both systems are based on collecting votes in a limited number geographic areas. Then the results of those areas are tabulated independently of each other. Finally the results by area are used to determine the government. They are just different mechanisms for translating votes collected by geographic area.

To put it another way consider if the US used a UK-style parliamentary system. In any rational, non-gerrymandered map of the constituencies, the Dems would find themselves overpacked in big cities compared to the Pubs. If Trump were the party leader of the Pubs last year in a parliamentary election he would be sitting now as PM. Would that be undemocratic?

[muon2]

In a parliamentary democracy the party that gets the most votes does not always win control of the government. Does that make a parliamentary system undemocratic? I don't think so. Neither does the fact that a president is elected by states electors make it undemocratic.

Both systems are based on collecting votes in a limited number geographic areas. Then the results of those areas are tabulated independently of each other. Finally the results by area are used to determine the government. They are just different mechanisms for translating votes collected by geographic area.

To put it another way consider if the US used a UK-style parliamentary system. In any rational, non-gerrymandered map of the constituencies, the Dems would find themselves overpacked in big cities compared to the Pubs. If Trump were the party leader of the Pubs last year in a parliamentary election he would be sitting now as PM. Would that be undemocratic?

That's because of the Rural/City Divide that has developed over the past 20 something years. And one does not have to look at the UK for Parliament Elections.

Sweden for example has an Open Party List System. German has a Mixed Member System. While South Korea has Parallel Voting. 

I wholeheartedly agree on the development of the rural/urban divide as a strongly contributing factor, but it is really that the Dems are more concentrated in the cities than the Pubs are in rural areas that causes the problem. When the the parties are mismatched in the number of precincts that are overwhelmingly in their favor, then no neutral (ie ignoring parties) partition of the country would avoid the NPV-EV mismatch. I used the UK as an example precisely because it has single member constituencies and would suffer the same mismatch between the vote and government if applied to the US. My question is whether that makes the UK system undemocratic? Are all single-member district systems therefore undemocratic?

[muon2]

I don't see why national recounts are so awful.

Painstaking wastes of money that will greatly reduce citizens' trusts in the process, will get very messy quickly, and will take months to complete. The odds of a recount getting done before Inauguration Day, much less in time for the president to pick their cabinet or whatever, are slim to none. The economic impact is huge as well, since over 100 million ballots have to be re-checked.

Seems like a good price to pay for having everyone's vote actually count.

An ideal system would have everyone's vote actually count and avoid that hullaballoo (like the system I detailed in my post would do, for instance).

It would be impossible to enact that on a national level though, and no state really has an incentive to do it with the possible exception of Virginia if Gillespie wins this year.

Another factor to consider in a large recount is the intrinsic error rate of our ballots. A certain number of ballots in any election will be disputable. That is two experts in reading the ballots will draw different conclusions about the intent of the voter on a ballot. Disputable ballots occur with every system whether paper or electronic. At present the best systems get about 1 in 10,000 ballots that are disputable (cf the 2008 MN Senate race).

If a race is so close that it depends on reasonably disputable ballots, it generally comes down to a judge making a choice between competing reasonable experts. In most races an error rate that small has no effect on the perception of fairness. But if it looks like a judge is stepping in and taking sides the sense of fairness may be lost (cf the 2000 FL race for president).

In 2000 FL the margin of 537 out of almost 6 million votes cast was right in that 1 in 10,000 range. For a national recount of 137 M votes a margin of anything less than 13,000 votes would certainly involve judges taking sides with experts on reasonably disputable ballots, and it would likely be higher due to the variety of systems across the states. At that point you might as well ask a judge to flip a coin to determine the winner - it's statistically just as fair as a protracted recount.

[muon2]

In a parliamentary democracy the party that gets the most votes does not always win control of the government. Does that make a parliamentary system undemocratic? I don't think so. Neither does the fact that a president is elected by states electors make it undemocratic.

Both systems are based on collecting votes in a limited number geographic areas. Then the results of those areas are tabulated independently of each other. Finally the results by area are used to determine the government. They are just different mechanisms for translating votes collected by geographic area.

To put it another way consider if the US used a UK-style parliamentary system. In any rational, non-gerrymandered map of the constituencies, the Dems would find themselves overpacked in big cities compared to the Pubs. If Trump were the party leader of the Pubs last year in a parliamentary election he would be sitting now as PM. Would that be undemocratic?

That's because of the Rural/City Divide that has developed over the past 20 something years. And one does not have to look at the UK for Parliament Elections.

Sweden for example has an Open Party List System. German has a Mixed Member System. While South Korea has Parallel Voting. 

I wholeheartedly agree on the development of the rural/urban divide as a strongly contributing factor, but it is really that the Dems are more concentrated in the cities than the Pubs are in rural areas that causes the problem. When the the parties are mismatched in the number of precincts that are overwhelmingly in their favor, then no neutral (ie ignoring parties) partition of the country would avoid the NPV-EV mismatch. I used the UK as an example precisely because it has single member constituencies and would suffer the same mismatch between the vote and government if applied to the US. My question is whether that makes the UK system undemocratic? Are all single-member district systems therefore undemocratic?

Is discussion of single-member districts relevant here, in a discussion of the winner-take-all Electoral College? The districts under consideration aren't single-member, but weighted (roughly, though skewed) by population. You'd run into a similar problem by aggregating single-member results, but there are ways in which it's exacerbated by aggregating state EC allocations.

Larger aggregations potentially have more skew. Many have shown that going to the NE-ME system wouldn't have changed the result, so CD-sized electoral districts in other countries is relevant to the EC.

The problem of aggregating can be viewed from the two extremes. At one extreme is an aggregation of the whole nation, and that's just the NPV. At the other extreme is an aggregation of single voters - every voter is a member of the EC. Both extremes accurately reflect the popular vote. Intermediate aggregations are subject to geographic packing of one party's voters, and that is what happens in the EC or other systems with WTA constituencies.

There's a second factor worth noting for the extreme disaggregation where everyone is a member of the EC. In years like 2016 there would be no winner of the EC vote since no one got to 50%. The Constitution would send the result to the House for a runoff. In a parliamentary system there would be a coalition to form a government. In many presidential system there would be a direct runoff back to the electors. The common feature is a mechanism that doesn't automatically promote the top vote-getter if the total is under 50%. IMO that's the problem with many NPV discussions - there's none of that type of mechanism (though IRV would meet that need).

[muon2]

If Nixon got more popular votes in 1960, he should've won, as Clinton should've won, under an ideal electoral system.

well that not how it works in america we are a republic.

Uh, what about senators?

States are also a republic (republican form of government) and govs are also elected by direct PV.

And what about France? That is no Republic?

And France has a runoff if no one gets 50%. Many US cities do that, too, as do a few other offices. If the US is going to talk about PV for president, then it should talk about runoffs as well.