The Democratic Party’s uphill path to 270 electoral votes in 2016

[Mister Mets]

John Sides of the Washington Post argues that some factors favor the Republicans in 2016.

You must be logged in to read this quote.

[eric82oslo]

I just read this article in Washington Post though: The Republican Party's uphill path to 270 electoral votes in 2016.

http://www.washingtonpost.com/politics/the-gops-uphill-path-to-270-in-2016/2014/01/18/9404eb06-7fcf-11e3-93c1-0e888170b723_story.html?tid=pm_pop

There social scientists as well as both Republican and Democratic strategist agreed that most current and future battleground states, as well as demographic trends, are trending Democratic either in a light or heavy fashion.

[IceSpear]

If you run the 2008 numbers in that model, I bet it would show McCain losing by a much bigger margin than he actually did. Due to the weight put on a term limited president's approval rating.

[Foucaulf]

This sentence really irritates me:

You must be logged in to read this quote.

As evidence, he hyperlinked to a study - about the 2008 election. At this point, the lack of data points for presidential elections becomes absurd.

To test this trend, I did a similar experiment to Gelman's, where I calculated the interquartile range of swings in statewide vote proportions to the Democrat. I got back a percentage of 5.1% - higher than his 2008 result but not comparable (because he calculated R vote proportions). This flies with my intuition that the Obama campaign did exploit GOTV campaigns in specific key states.

I can agree that there's a stylized trend of more uniform swings. But I argue that the uniformity is really found at the demographic level - so people of the same sex, income level or racial identity vote similarly. Different demographic proportions in states increase variability in state behaviour. This also fuses the claims made by the other article posted in the thread.

[Skill and Chance]

The curious thing is it should be much harder to do a regional campaign today than at most any other time in American history and yet we are seeing regionalism matter more in elections now than 30-50 years ago.  The Latino vote in the Southwest has something to do with this, but in the era of Google, no one should be able to get away with different campaign messages in different states.  When everyone is watching the same campaign on TV or reading about it on the internet, one would expect uniform national swings. 

The X factor here could be turnout and the difference between emphasizing it in one state and ignoring it in another safe state.  For example the gap between OH and PA was the lowest in a long time in 2012, perhaps because Obama went all out in OH and largely ignored PA while Romney gave it some late attention.

If campaigns are getting really good at fine-tuning turnout, we should expect the electoral college to diverge more and more from the popular vote in future elections.  The next 1876 may not be far off.

[Nichlemn]

The underlying point about demographic changes is that even if it bodes very well for Democratic fortunes, it just doesn't occur fast enough to make a huge difference in only one cycle.

I just read this article in Washington Post though: The Republican Party's uphill path to 270 electoral votes in 2016.

http://www.washingtonpost.com/politics/the-gops-uphill-path-to-270-in-2016/2014/01/18/9404eb06-7fcf-11e3-93c1-0e888170b723_story.html?tid=pm_pop

There social scientists as well as both Republican and Democratic strategist agreed that most current and future battleground states, as well as demographic trends, are trending Democratic either in a light or heavy fashion.

lol, this article is awful. It's not that you couldn't come up with a reasonable case for why you think the Democrats are favoured in 2016 (I personally believe this), it's that it spurts out meaningless statistics about runs of wins (which don't have any correlation with future electoral success), and then misses the forest for the trees by focusing so much on the Electoral College.  It's a very common mistake here and by journalists with limited political science knowledge. That is, the popular vote and Electoral College have a very high correlation, so if certain factors predict one side to win nationally by a few points, they'll almost certainly win the EC.

[I Will Not Be Wrong]

Very interesting article Mets, thanks for sharing!

[I Will Not Be Wrong]

Here are the states that voted at least 55% Democratic, which equals to 179 electoral votes:

And here are the states that voted at least 52% Democratic, equaling 233 electoral votes:

[I Will Not Be Wrong]

As of now, the only state that voted over 52% D that Republicans could win in 2016 is Wisconsin.

[Mister Mets]

As of now, the only state that voted over 52% D that Republicans could win in 2016 is Wisconsin.

In 2004, Florida went to Bush by 52.10%. Virginia went to him by 53.68%. North Carolina was 56.02% for Bush. Indiana was 59.94%.

The '96 election had Perot getting over eight percent of the vote, so there were only five states in which Clinton got more than 55%. But he did end up beating Bob Dole by more than ten points in Louisiana (he won it by 12.07%), West Virginia (he won it by 14.75%) and Arkansas (which he won by 16.94%.)

[I Will Not Be Wrong]

As of now, the only state that voted over 52% D that Republicans could win in 2016 is Wisconsin.

In 2004, Florida went to Bush by 52.10%. Virginia went to him by 53.68%. North Carolina was 56.02% for Bush. Indiana was 59.94%.

The '96 election had Perot getting over eight percent of the vote, so there were only five states in which Clinton got more than 55%. But he did end up beating Bob Dole by more than ten points in Louisiana (he won it by 12.07%), West Virginia (he won it by 14.75%) and Arkansas (which he won by 16.94%.)

I know anything can happen, but I mean as of now. ;0
Nevada and New Mexico could surely flip if Republicans gain ground with Hispanics during the 2016 election.

[bedstuy]

The assumption that the economy will be exactly the same as it is now in 2016 is a poor basis for analysis.  Why not use the economic forecast from the Federal Reserve? 

[old timey villain]

The curious thing is it should be much harder to do a regional campaign today than at most any other time in American history and yet we are seeing regionalism matter more in elections now than 30-50 years ago.  The Latino vote in the Southwest has something to do with this, but in the era of Google, no one should be able to get away with different campaign messages in different states.  When everyone is watching the same campaign on TV or reading about it on the internet, one would expect uniform national swings. 

The X factor here could be turnout and the difference between emphasizing it in one state and ignoring it in another safe state.  For example the gap between OH and PA was the lowest in a long time in 2012, perhaps because Obama went all out in OH and largely ignored PA while Romney gave it some late attention.

If campaigns are getting really good at fine-tuning turnout, we should expect the electoral college to diverge more and more from the popular vote in future elections.  The next 1876 may not be far off.

I think you have it backwards. People used to all watch the same campaign on TV. But now with the growth of cable and internet news/commentary, people can get political news that aligns with their own perspective, be it regional, religious, racial or occupational. Obama and Romney both put out ads on the internet tailored to different types of people in different regions, imbedding them on the unique sites they visit. This is how Obama could run an Ohio campaign and a totally different Florida campaign.