Sanders and Trump supporters agree on at least one thing ...

[prideandfall]

So, you're saying that differential voter turnout IS the defining partisan issue that decides elections. Republicans drive it better, so they win.

well... sort of.
differential within pre-established voting blocs, not within the general population, and also drive is part of it but not all of it.

basically, trump won because obama was president for 8 years. obama won because bush was president for 8 years. bush won because clinton was president for 8 years. clinton was president because the boner for reagan was so throbbing it lasted through one more republican term. before reagan it was carter. before carter.... are you seeing the pattern here?

R gets a president, the liberals go slowly insane for a term or two, and then get really pissed off about how shitty that R was and come out enough to vote in a D.
D gets a president, the regressives go immediately insane basically forever, and then get really pissed off about the D sort of more or less stabilized the country after the disaster of the R and vote in another R.

back and forth, over and over, it's an endlessly repeating cycle and it always will be.

Gerrymandering is another thing they're good at. I think that if you handed one of those two factors over to Democrats, gerrymandering might have an effect, but turnout would have a bigger one. Just a gut feeling, though.

i think overall that sort of procedural sabotage is a much bigger systemic problem in this country than presidents are, and it's something democrats never pay any attention to.

 

[Elixir]

i think overall that sort of procedural sabotage is a much bigger systemic problem in this country than presidents are, and it's something democrats never pay any attention to.

Agree. Other than failure to motivate their base, the DNC's biggest failure is their myopic focus on the Big Prize. I hope that Cheato teaches them that the Big Prize is fairly meaningless if you can't elect congresscritters who will work with him. And you can't do that unless you get people elected at the local level who can mitigate what you (aptly IMO) call "procedural sabotage".

 

[bilby]

Bold added for emphasis - that number swamps all other factors combined by at least an order of magnitude. Do they even teach Civics in elementary schools these days? Or are elementary schools just qualifiers for vocational schools where an ignorant electorate can be groomed to go to work for the military-industrial complex or auto industry (if there's a difference), then stay home on election day and wash their hands of any dire consequences such as Cheato's gang of thieves?

I think that statistics tell us that it really doesn't matter if 60% of the electorate votes or if 90% of the electorate votes as long as the 60% is a representative slice. Which it has to be since it is such a large part large of the voting population. <snip>.

No, it really doesn't.

if 60% of the electorate votes, and there was a random lottery to randomly select which 60% that is, then you would be correct. But the selection of who votes is NOT random - it is a decision made by the voters, based on their opinions about political subjects - and those opinions are exactly what the election is attempting to measure.

A random sample of sufficient size is likely to be representative of the whole. A selected sample, where the selection criterion might be correlated to the question at hand, is not likely to be representative of the whole even if the sample size is large.

The question is, was the preference distribution of those who voted similar to the preference distribution of those who were eligible to vote? Given that the decision about who to vote for, and the decision about whether or not to vote, are BOTH political decisions, and are BOTH things that political parties put a lot of effort into influencing, it seems highly implausible that the answer to that question would be 'yes'.

In a hypothetical campaign where party A is most effective in persuading voters that their candidate is best, while party B is most effective in persuading their base to actually get out and vote, it is quite possible that the majority of eligible voters will support candidate A, while the majority of votes cast will go to candidate B.

And this phenomenon is well known and frequently remarked upon in US Presidential elections, were pundits talk about the propensity of evangelicals to get out and vote, so that low turnouts tend to skew the vote towards the GOP.

 

[beero1000]

I think that statistics tell us that it really doesn't matter if 60% of the electorate votes or if 90% of the electorate votes as long as the 60% is a representative slice. Which it has to be since it is such a large part large of the voting population. <snip>.

No, it really doesn't.

if 60% of the electorate votes, and there was a random lottery to randomly select which 60% that is, then you would be correct. But the selection of who votes is NOT random - it is a decision made by the voters, based on their opinions about political subjects - and those opinions are exactly what the election is attempting to measure.

A random sample of sufficient size is likely to be representative of the whole. A selected sample, where the selection criterion might be correlated to the question at hand, is not likely to be representative of the whole even if the sample size is large.

The question is, was the preference distribution of those who voted similar to the preference distribution of those who were eligible to vote? Given that the decision about who to vote for, and the decision about whether or not to vote, are BOTH political decisions, and are BOTH things that political parties put a lot of effort into influencing, it seems highly implausible that the answer to that question would be 'yes'.

In a hypothetical campaign where party A is most effective in persuading voters that their candidate is best, while party B is most effective in persuading their base to actually get out and vote, it is quite possible that the majority of eligible voters will support candidate A, while the majority of votes cast will go to candidate B.

And this phenomenon is well known and frequently remarked upon in US Presidential elections, were pundits talk about the propensity of evangelicals to get out and vote, so that low turnouts tend to skew the vote towards the GOP.

The classic example of this is Wald's analysis of battle damage on returning bombers in WWII. A military commission had compiled data on the damage that their bombers had received with the intention of adding armor to the most affected locations. Apocryphally, the Center for Naval Analyses took a look at figures like this one and were about to decide to add the armor on the most commonly damaged areas. Put the armor where the planes are getting shot the most. Makes sense, right?



In comes Wald, who says no, you're missing a major source of selection bias - the areas that aren't showing up as damaged are still getting shot; it's just that the planes that get shot in those areas don't return home for their damage locations to be measured. If a lot of planes are coming home with damage in certain areas, those areas are exactly the ones that are sufficiently well-armored, you really need to put the armor where the damage isn't.