Friday, August 12, 2016

It's The Electoral College, Stupid

The Final Electoral Vote for 2012 - Obama: 332 - Romney: 206
As we head toward the 2016 presidential election, you've probably noticed a lot of attention placed on the national polls.  Well, actually, you've probably noticed a lot of foot-in-mouth disease from Donald Trump who seems determined now more than ever to roll out a new controversial statement about every 24 to 48 hours, but aside from that, you've probably noticed some national polls, right?  Ok then.  Well guess what?  National polls are irrelevant.  Why?  Because we don't elect the President with a national vote.  We have this thing you likely learned about during your high school civics class called the Electoral College.  It was boring then and it is boring now so I won't bog you down with the details.  There's a whole wikipedia page on it HERE if you're feeling particularly academic.  The bottom line is this: the President of the United States is actually elected by each of the 50 States on a State-by-State basis.  In other words, STATE POLLS MATTER!

One other thing you should keep in mind about the Electoral College: not all States are created equal.  What I mean by that is, some States get more votes than others.  Why is that?  Because some States have a higher population than others.  For example, a State like California has the biggest population out of all 50 States, therefore it gets the most votes for President (55 votes).  Conversely, a State like Delaware has one of the smallest populations in the country, so it only gets 3 electoral votes.  There are 538 electoral votes total; a candidate needs 270 of these (50.1% of 538) to become President.

Looking at the polls in all 50 States is hard and takes too much time.  Fortunately for us, there are experts out there who do this for a living.  One such expert, Nate Silver, and the folks at his site (538 dot com) have a pretty solid track record of predicting the Presidential election using regression analysis of all the polls in all 50 States, among other data points such as voting history, demographics, etc. (See comments section in our 2012 post here noting that Nate Silver's model accurately predicted the 2008, 2010, and 2012 elections; and to be sure, he also accurately predicted the 2014 election as well).  Their current electoral map is HERE.

So how do things stand in August 2016?

If we apply all of the current predictions on FiveThirtyEight (key word being "current"), the 2016 election turns out like this:

Click the map to create your own at

with a landslide 374 electoral votes for Hillary Clinton to Trump's 164.  Again, this is August 2016, so it's fair to say that a number of these States will likely switch around, such as Georgia (which hasn't voted for a Democrat since Bill Clinton in 1992), Arizona (which hasn't voted for a Democrat in over 30 years with the exception of Bill Clinton in 1996) and North Carolina (which, although voting for Obama in 2008, hasn't voted for a Democrat since Jimmy Carter in 1976).

But even if we give Trump those 3 States, he still loses by a landslide of 332 to 206, which just so happens to be the exact same result that took place in the 2012 election between Obama and Romney. [Note: Hillary Clinton starts off with a guaranteed 217 electoral votes (needing 53 more votes to reach 270) and Trump starts off with a guaranteed 164 electoral votes (needing 106 more votes to reach 270)]

This would be a good point to discuss the "Battleground States."  That is, the States which can actually change between now and November.  There are certain States like New York (solid Blue State) or Alabama (solid Red State) which, outside of a nuclear level event, are simply not going to change between now and the election.  However, there are currently 12 States which are toss-ups that could actually go either way.  Those States, and the probabilities of each candidate winning, are as follows:

1. Arizona (11 votes) - currently 52% Clinton - 48% Trump
2. Colorado (9 votes) - currently 87% Clinton - 13% Trump
3. Florida (29 votes) - currently 77% Clinton - 23% Trump
4. Georgia (16 votes) - currently 52% Clinton - 48% Trump 
5. Iowa (6 votes) - currently 72% Clinton - 28% Trump
6. Nevada (6 votes) - currently 76% Clinton - 24% Trump
7. New Hampshire (4 votes) - currently 80% Clinton - 20% Trump
8. North Carolina (15 votes) - currently 70% Clinton - 30% Trump
9. Ohio (18 votes) - currently 75% Clinton - 25% Trump 
10. Pennsylvania (20 votes) - currently 87% Clinton - 13% Trump 
11. Virginia (13 votes) - currently 87% Clinton - 13% Trump
12. Wisconsin (10 votes) - currently 92% Clinton - 8% Trump
Total of 157 electoral votes up for grabs.

As of today's post, all 12 of these Battleground States are in Clinton's column.  If we draw an imaginary line in the sand at the 75% probability mark, and give all the States to Hillary Clinton where she currently has a 75% chance or higher of winning, then 8 of these Battleground States (Colorado, Florida, Nevada, New Hampshire, Ohio, Pennsylvania, Virginia and Wisconsin) go to Clinton and 4 States (Arizona, Georgia, Iowa and North Carolina) go to Trump.  That map would look like this on election day:

Click the map to create your own at

with Hillary Clinton winning 326 electoral votes to Trump's 212 electoral votes. [Editor's Note: unless something catastrophic happens between now and November, I think this is the likely outcome.]  Keep in mind, this result assumes that Hillary loses every Battleground State that is not currently giving her at least a 75% chance of winning.  So for instance, as of today, she has a 72% chance of winning the State of Iowa.  To be sure, a 72% chance is a pretty good chance.  Well over 50% right?  If you had a 72% chance of winning the lottery, you'd run out right now to buy a ticket.  However, because 72% falls below our imaginary line of 75%, we're going to assume for the sake of this simulation that Hillary loses Iowa and every other Battleground State where there is not at least a 75% probability for her to win.

But I know what you're thinking: 75% is not good enough this far out from November.  You're willing to accept the fact that States like New York and California, which are each currently showing a 99% chance of a Clinton victory, are not going to change their minds between now and November.  But a 75% chance is too low.  After all, if your doctor said you only had a 75% chance of surviving a medical operation, that means there's a 25% probability that you could die.  Way too risky, right?  So let's up the ante.

How about we only assign Battleground States to Clinton where, as of today, there's an 85% probability of victory or higher.  In that case, only 4 Battleground States (Colorado, Pennsylvania, Virginia and Wisconsin) can be assigned to Hillary right now.   That would look like this:

Click the map to create your own at

In this 85%-or-better simulation, Hillary and Trump actually tie at 269 electoral votes a piece:

Click the map to create your own at

Under such a scenario, since neither candidate reaches 270, the 12th Amendment dictates that the election goes to the House of Representatives to decide.  Since the House is currently controlled by Republicans, arguably this would mean that Trump (a Republican) wins in the event of a tie.  However, this is a highly unlikely scenario as the 12th Amendment has only been used once in the history of the United States (in 1824 to elect Andrew Jackson).   Moreover, look at the Battleground States Trump would need to win in order to make this scenario a reality: Florida (voted for Obama twice and is currently giving Clinton a 77% probability of victory), New Hampshire (voted for Obama twice, voted against Bush in 2004 and is currently giving Clinton an 80% probability of victory), and Nevada (voted for Obama twice and is currently giving Clinton a 76% probability of victory).  All of that to say, this scenario is highly unlikely. 

By now you're probably noticing the same thing that the RNC and Trump's staff have undoubtedly been noticing for some time: the road to 270 for Trump is an unlikely one.  Even prior to the recent meltdown by the Trump campaign, the math was never in Trump's favor.  As we mentioned above, the current make up of the Electoral College gives your generic Democratic candidate somewhere around 217 electoral votes right off the bat (depending on what you consider a "Battleground State"), while the generic Republican candidate only gets 164.  Such is the makeup of Red States and Blue States.  Although there are roughly about the same number of Red States and Blue States in America, the Blue States have more people and thus have more electoral votes.

That's math, folks.
blog comments powered by Disqus