Analyzing The 2016 General Election
Some in the political world shy away from insanely-early election analysis. It is somehow unseemly (or even downright shameful), these people tell us, to indulge in speculation about an election too far in advance. These are the same type of people who point out that there's an election this year to get through first, for Pete's sake.
We don't listen to these people, however, so we'd like to welcome you to our first article to take a hard look at the 2016 election. We're going to engage in some rampant speculation not just (as some are doing) over the primary races, but indeed over how Democrats and Republicans are positioned for the general election itself. In specific, predicting how the Electoral College will vote by delving into a state-by-state statistical analysis of voting trends.
To be clear: when we say "statistical analysis" what we really mean is "taking a hard look at the numbers from the past six elections," rather than applying any sort of formulae or accepted statistical methodology. Just so nobody gets confused. We're going to divide the states up into a few broad categories based on past voting trends, and then at the end add up what it all means in the Electoral College.
A few caveats are in order, before we begin. We only went back six election cycles, which cuts off both George H. W. Bush and Ronald Reagan. The three elections which preceded the period we examined were overwhelming Republican landslides. But you have to draw the line somewhere, and we felt that the 1980s were just too far in the past to have much relation to where the country stands today. Even Clinton's two elections didn't influence us all that much, as we're mostly looking at the last four elections for signs of which way the state is moving, politically.
Barack Obama's 2008 election was a special coming-together of all sorts of demographic groups, which may never happen in the same fashion again. He boosted turnout in ways that may not be repeated in 2016, to put this another way. Then again, for all the talk of how different the 2012 election was from 2008, Obama did manage to hang on to all of the states he initially won, with the only exceptions being Indiana and North Carolina.
Of course, the biggest caveat is that each election is unique, and until we know who the nominees are going to be, we're just engaging in the sheerest of speculation. Entire demographic groups could get excited in unforeseen ways, and shift the whole electoral picture as a result. For instance, the Republicans could nominate a woman with Hispanic roots, while the Democrats nominate an old white guy. Hey, stranger things have happened, right?
But we've got to start somewhere. So today, we're starting with just a pure analysis of how each state has voted in the recent past, to show some sort of baseline before the contest gets underway. Because, at the end, you'll see that the picture which emerges heavily favors one side, which is indeed worth noting even at this insanely-early date. So let's get on with it, shall we?
Solid states
Here are the states which have voted for one party or the other in all six of the past six elections (listed in order of how strongly they voted for their party, on average -- highest vote percentage down to lowest):
Democrats
Washington D.C., Hawai'i, Vermont, New York, Rhode Island, Maryland, Massachusetts, California, Delaware, Connecticut, Illinois, New Jersey, Maine, Washington, Michigan, Oregon, Minnesota, Wisconsin, Pennsylvania.
Republicans
Utah, Wyoming, Oklahoma, Idaho, Nebraska, Kansas, North Dakota, Alabama, South Dakota, Texas, Alaska, South Carolina, Mississippi.
There's not a lot to quibble about in either of those lists. Some are predicting that Texas (due to demographic changes) might eventually become a battleground state, but even the most optimistic of these predictions usually admits "not until 2020, in all likelihood." Some of the Democratic states may come into play depending on the Republican nominee (New Jersey and Wisconsin, for example), but we're taking a broader look today, ignoring any changes due to a home-state candidate.
Conservatively (in more than one sense), there are a few states which Republicans have been lusting after for the last few elections, which could come into play. Michigan, Wisconsin, and Pennsylvania are the most-often cited. Just to be scrupulously fair, we're going to downgrade Wisconsin and Pennsylvania, but we think Michigan's a bridge too far for Republicans in 2016 (Obama carried the state by 16 points in 2008 and nine points in 2012).
Leaning states
This still leaves 19 states which didn't reliably vote for either party. First, we have the states which voted either four or five times (out of six) for the Democrats. George W. Bush won once in New Mexico, Iowa, and New Hampshire, and twice in Nevada and Ohio.
Out of these five states, two seem like a lock for Democrats next time around: New Mexico and Nevada. Changing demographics seem to have shifted these states to a reliable blue, at least for the near future. Of the other three, we're going to put two into the "Lean Democratic" column: Iowa and New Hampshire. This leaves one as a pure toss-up state: Ohio.
Over on the Republican side, the picture is a bit more complicated. Call it the difference between Bill Clinton and Barack Obama. Clinton was from the South, and had a blue-collar appeal that Obama just doesn't. But Obama had a youthful appeal and got excited crowds of minority voters out to the polls. Clinton's elections were the oldest under examination here, but then again if his wife runs they may become more germane.
Five states voted Republican in five out of the six previous elections. Three of these broke for Clinton once: Arizona, Georgia, and Montana. Two broke for Obama once: North Carolina and Indiana.
While Arizona did go for Clinton in 1996, since then it has never voted above 45 percent Democratic, so we're going to call it a solid Republican state for now. Georgia is in much the same category as Texas for Democrats: they like to dream about flipping the state, but it likely won't happen in 2016. But since we downgraded Pennsylvania and Wisconsin, we're also going to put Georgia in only the "Lean Republican" column for now. Montana and Indiana are both pretty solidly Republican (Obama winning Indiana was a fluke that likely won't be repeated soon). North Carolina, however, has to be seen as a pure toss-up state, at this point.
Seven states voted four times for Republicans, and twice for Democrats: Virginia, Missouri, Louisiana, Tennessee, Kentucky, Arkansas, and West Virginia. All but one of these voted both times for Clinton (Virginia was the exception, which voted for Obama both times). The six who voted Republican in the past four elections are all pretty solidly red, at this point. The only exception to this might be West Virginia, but that's just a hunch, not based on any actual numbers. But Virginia is moving more and more towards the Democratic column, again due to shifting demographics. I can't put the state into a Democratic category quite yet, but even though rated a toss-up, it will be one of the easier toss-up states for a Democrat to pick up.
The Toss-Ups
Only two states split their vote down the middle, voting three times for each party. Both Colorado and Florida voted both times for Obama, both times for Bush, and once for Clinton. Out of the two, Colorado shows clear signs of becoming bluer and bluer. Colorado is the heart of the Democratic strategy to take as many of the Mountain West states as they can, and it has to be seen as at least in the "Leaning Democratic" category for the near future. Florida, however, could go either way even without a favorite son on the ballot, so must be seen as the biggest prize among the true toss-up states. It's not going too far out on a limb to predict that the heaviest states for presidential television commercials will once again be Florida and Ohio, to put this another way.
Adding them up
Limiting ourselves to five categories ("Safe" and "Lean" for both parties, as well as "Toss-up"), let's take a crack at adding up the Electoral College votes for a generic Democrat running against a generic Republican. States are listed with their electoral votes in parenthesis.
Safe Democratic
Washington D.C. (3), Hawai'i (4), Vermont (3), New York (29), Rhode Island (4), Maryland (10), Massachusetts (11), California (55), Delaware (3), Connecticut (7), Illinois (20), New Jersey (14), Maine (4), Washington (12), Michigan (16), Oregon (7), Minnesota (10), New Mexico (5), Nevada (6).
Lean Democratic
Wisconsin (10), Pennsylvania (20), Iowa (6), New Hampshire (4), Colorado (9).
Safe Republican
Utah (6), Wyoming (3), Oklahoma (7), Idaho (4), Nebraska (5), Kansas (6), North Dakota (3), Alabama (9), South Dakota (3), Texas (38), Alaska (3), South Carolina (9), Mississippi (6), Montana (3), Indiana (11), Arizona (11), Arkansas (6), Kentucky (8), Tennessee (11), Louisiana (8), Missouri (10).
Lean Republican
Georgia (16), West Virginia (5).
Toss-up
Ohio (18), Florida (29), Virginia (13), North Carolina (15).
When we add up the numbers, Democrats have 19 states locked up, for a total of 223 electoral votes. Five more states lean Democratic, for a total of 49 electoral votes.
Republicans have 21 states locked up, but these states only add up to 170 electoral votes. Two other states lean Republican, adding in 21 electoral votes.
There are four toss-up states, with a whopping 75 votes between them. But, astonishingly, the toss-ups may not even matter.
When you add the safe and leaning Republican electoral votes together, you get 191. But when you add the safe and leaning Democratic states, you wind up with 272 electoral votes -- which is two more than is necessary to win the election. A Democrat could lose Ohio, Florida, Virginia, and North Carolina and still squeak out a 272-266 victory in the Electoral College. Winning any of the four toss-ups would just be icing on the cake.
Of course, it is insanely early. And past elections are not all that reliable as a predictor of future elections. So it would be absurd to suggest to the Republican Party that they have no chance whatsoever of winning, and should really just spare America from all that campaigning next time around by just leaving the field wide open for the Democratic nominee to waltz into the White House.
Kidding aside, though, what cannot be denied is the overwhelming advantage Democrats have in the Electoral College at this point. When the race begins, Democrats will enjoy an automatic headstart, as they can rely on 223 electoral votes to the Republicans' 170. Democrats will only need to pick up 47 more electoral votes to win, while Republicans will need 100. If you classify the states a bit less conservatively, moving Wisconsin and Pennsylvania into the safe category for Democrats, then this advantage gets even steeper, at 253 to 170. Adding in all the leaners, Democrats hold 272 to Republicans' 191. Those are pretty stunning totals to consider, even if it is way too early to be thinking about such things.
-- Chris Weigant
Follow Chris on Twitter: @ChrisWeigant
Chris,
Nice work.
Next week, can we talk about who the candidates will be? :)
To move states from the other party's column to theirs, not just pull one out of the toss-up category, a party has to bring large numbers of people into the electorate who formerly were (or whose parents were) non-voters. That can be a cohort change, but normally it's a racial/religious/ethnic group, or early in our country's history, the lifting of the property requirements.
Any sign that either party is doing so? Who are the non-voters?
The exception, of course, was the flip of the Solid South from solidly Democratic to solidly Republican over the course of 1948 (Strom Thurmond's run as a Dixiecrat) to 1980 (the last time the so-called "Reagan Democrats" were identified as Democrats). Even that was aided by the shift of evangelicals from a stance of withdrawing from the evil of worldly politics to a pattern of wallowing in it.
I can't quibble much with your numbers, and I suspect most objective Republican number crunchers wouldn't either, at least in private.
The next Republican Presidential nominee has an even bigger hill to climb than Romney did. Said nominee can run the swing states and still lose the electoral college.
If a GOP candidate has no reasonable chance of winning the presidency, only an ideologue has much incentive to run.
LizM -
I think I've already dipped a toe into the candidate waters a bit (Christie, mostly), but I don't think I've looked at the Dem side of things at all. Maybe not next week (State Of The Union week...), but soon. Promise!
Think Biden's going to run?
dsws -
A couple of loose observations...
Obama brought in young voters, single women, and he increased minority voting. If Hillary runs, the single women (at least) will likely stick around to vote for her. The youth vote? Maybe, maybe not. Latinos, on the other hand, are becoming a bigger and bigger political presence, and I think they are well aware of their growing clout, so I see this as more of a long-term trend that's not going away any time soon.
Several states either have changed (NV, CO, NM), are in the process of changing (VA, NC) or may change in the near future (TX, GA) based on the demographic changes of their population. People moving around, more suburbanites, Latino presence growing, etc. I gave only Dem examples (states moving towards Dems) here, but I could easily give a few GOP examples in the other direction, too. I think the suburbs of Arlington and Atlanta are doing more to flip VA and possibly GA than gaining new votes from non-voting groups, to put this another way.
TheStig -
That last sentence of yours is kind of scary. The more I think about it, the scarier it gets. Just had to say that...
Of course, if the GOP ever does anoint the second coming of Saint Ronald of Reagan, then they could flip a lot of states back. Unfortunately for them, I don't see Ronnie II anywhere out on the horizon.
-CW
I would like to see an analysis of what happens in 2016 using the 2014 results..
IE..
What happens in 2016 if the GOP shellac the Dems in 2014 and extend their majority in the House and take the Senate..
What happens in 2016 if the Dems trounce the GOP, take the House and extend their majority in the Senate..
Also, what happens in 2016 if the status quo remains the same...
Hmmmmmmmmm??? :D
Michale
Noonan: The Sleepiness of a Hollow Legend
The State of the Union is a grand tradition—but only if people are listening.
http://online.wsj.com/news/articles/SB10001424052702303947904579339142225570548
This, more than anything, will determine the 2016 election..
It won't matter what states lean Left and what states go Right.. If Democrats cannot prove to the American people that they are capable of LEADING, then the GOP will win the day.
To date, the Dem's leadership capabilities in leadership have been... found wanting...
Michale
Chris,
Think Biden's going to run?
I would be very surprised if he did.
Think Biden's going to run?
I would be very surprised if he did.
I am very surprised you said that.. :D
Michale
How so, Michale?
Biden got less than one percent of the vote in the 2008 Iowa caucuses.
Nothing much has changed since then. That is what surprises me.
Ahhh I see...
Good reasoning.. :D
Michale
Indeed.
CW-
Dusting off some of my own spreadsheets from '12...
I would rate most of your Safe Dem picks as at least 0.95 probability of a Democrat win, with some of the lower hanging fruit (Nevada) closer to 0.9.
Your Safe Republican picks seem even safer to me, rate them all 0.99 Republican wins, shoot, make that 1.0 for simplicity.
All of your Democratic Leaning picks strike me as strong leans, 0.75 or better odds of a D win.
Here's another minor quibble, I would place Georgia and West VA in the safe Republican Column, with near certainty of Republican pick up, give them a 1.0 to be on the safe side.
My own analysis indicates state voting shifts towards Red or Blue in any given cycle are highly correlated with each other (national trends strongly dominate local trends) so a good predictor of a Democratic electoral college win is to sum electoral votes in descending order of state probability until you reach or exceed 270. The probability of the last state or states putting Blue over the line is the overall predicted probability of Blue winning the Presidency. I don't claim the model is perfect, but it seems pretty reliable.
So, using my above odds for Blue safe and leaning states (which sum to 272 electoral votes) I'm making A Ridiculously Early Prediction that the Democratic candidate has at least a 75% chance of winning in 2016, based on electoral fundamentals.
That's the hill the Republican candidate must climb.
The remaining swing states don't impact the prediction, including traditionally critical Ohio.
Given these long odds, the most important feature of any Republican nominee ought to be how well he/she tilts a relatively few critical state level races. That's probably much more discipline than the primary/caucus process will allow.
Michale [5] -
That's an interesting question. How do second-term midterms relate to the following presidential election?
My inclination is to say "not much, probably," but this time around it could be more important.
To take your options (out of order), if the status quo wins and neither house flips, then not much will change politically (that's my guess). Senate will pass stuff, House will pass their own stuff, very few things will make it through conference committee, and we all have exactly the same issues to fight about in 2016.
If the GOP wins big and takes the Senate, then Congress will pass a whole bunch of stuff (after the GOP totally throws out the filibuster for legislation in the Senate), and Obama will veto most of it. The lines will clearly be drawn for 2016, since the vetoes will be big headlines each time.
If, however, the Dems win big and take the House, then we will see the dam burst. There will be a pent-up flood of legislation pushed by Speaker Pelosi and Harry Reid, and all sorts of things Dems have been wishing for will pass (immigration reform, voting rights, equality legislation, etc.).
Now, this is the point where partisanship has to enter into your projection. If you believe the Dem agenda is a good one, then you will believe that this will all bode very well indeed for Dems in 2016. But if you believe the Dem agenda is far outside the mainstream, then Dems will pay the price of a big backlash in 2016, possibly losing them the White House.
It's an interesting question, especially (for me) that last one. What does anyone else think Dem control of House and Senate after 2014 would mean (a) for Obama's agenda, and (b) for Dems' chances in 2016? Let's hear your take on it...
LizM -
I asked about Biden because I personally believe that he's the only one who would even be capable of giving Hillary a run for her money. Well, maybe Elizabeth Warren would make for an interesting race, but I don't think she'll run this time around, personally.
But Biden's got name recognition going for him, and he's got a lot of charisma on the stump. Plus, he'll have served 8 years with Obama instead of 4 (for whatever that's worth, good or bad). I think he'd have a shot at the nomination, actually.
TheStig -
Aha! Analysis!
I got my main data chart from Electoral-Vote.com (they have a spreadsheet of every election in the 20th cent., broken down by percent for each state -- very handy -- I can look up the link if you can't find it on their site:
http://www.electoral-vote.com
They haven't updated the spreadsheet for the 2012 election, though, so I added those numbers in by hand. Should have mentioned my sources in the article, sorry.
GA - OK, this was mostly wishful thinking on my part. I wanted to be a little fair, since I had put PA and WI into Dem Lean. But the demographics of northern VA and NC are working the same way (but slower) in the suburbs of Atlanta. Sooner or later, GA will go purple. But you're right, I'm probably being premature.
WV - This one's interesting, because there is a lot of generational love for unions in the state. Also, a lot of racial influence (to be polite). I'll put it this way: if Hillary Clinton had been the Dem nominee, then she just may have taken WV in 2008. Of course, this could be sheer wishful thinking on my part, as by the numbers alone, WV has moved to solid red. Obama was briefly up in the polls in WV a few weeks before the election, and McCain's support was weak, but then there aren't a whole lot of polls out of the state, so it's tough to tell.
Overall, I agree with your call. 75% sounds about right, absent actual candidates' names. You're right -- it's a steep, steep hill.
One thing I should have mentioned in the article is that one shift which counters the blue-state shift to some degree was reapportionment. If Obama had won exactly the same states in 2012, his electoral count would still have been down from 2008. More people have been moving to the South (TX picked up FOUR House seats, while CA picked up none and states in the northeast lost seats). This won't be an issue again until 2024, but it is another trend worth pointing out.
-CW
dsws -
Oh, one thing I forgot to mention. There is a group called (I think) "Blue Texas" or "Turn Texas Blue" (or something similar) who has been putting millions of dollars into a registration drive for Texas Latinos. This is a long-term project, obviously, but the fact is that Texas Latinos vote at a much lower percentage than Latinos in other nearby states. So some smart Dems put some money towards reversing this trend and getting Latinos registered and to the polls. We'll see, in 2014 and in 2016, whether this has any effect at all, but you've got to admire the effort.
-CW
The usual approach is '<group> votes for <party>; there are getting to be more of <group>; therefore <party> will gain.' It involves implicit assumptions about the homogeneity of groups and the stability of their party affinities, and it can be done with different groups to yield different conclusions.
Each cohort of "the youth vote" is a new cohort. One cycle's youth vote doesn't tell much about the next.
The near-term impact of likely-Democratic voter registration in Texas would probably be not to actually win the state, but to make a Republican presidential campaign have to pay attention to it.
CW-
It's hard to imagine political personalities changing the state probability matrix much...it's mostly ones and zeroes for all practical purposes, regardless of who runs on the tickets. It's been that way for decades. The matrix isn't fixed, but change is glacial.
Moreover, if my hypothesis is correct, and regional effects are small compared national effects, picking a candidate based on some critical "leaning states" seems a very marginal strategy. Better to select a candidate who plays well broadly and cross your fingers that the national news breaks your way. Christie struck me (and so many others) as a relatively good Republican pick, but not now given his troubles, and his odds would still have been very long even in the best of post-Sandy times. Hail Mary is not much of a strategy.
Any Republican candidate who isn't deluded or delusional must know this. What sort of candidate runs to lose? When election results are preordained, that's terrible for representative forms of government. Scary indeed.