Five Factors Needed for a GOP Win in 2016

A review of the turnout model first presented prior to last week’s election still shows the road map for victory in 2016 based on five factors.

First let’s look at the model, where 121 boxes represented 242 million American adults. As disappointing as the loss was, 114 boxes were correct and 7 were wrong – so the basic model is still in place even though Nate Silver was right on this campaign and I was wrong on the final winner for 2012:

Turnout	100% Obama	90%	80%	70%	60%	50% Swing	40%	30%	20%	10%	0% (Romney)
100%	Ob	Ob	Ob	Ob	Ob	Ob	Rom	Rom	Rom	Rom	Rom
90%	Ob	Ob	Ob	Ob	Ob	X Ob	Rom	Rom	Rom	Rom	Rom
80%	Ob	Ob	Ob	Ob	Ob	X Ob>Rom	Rom	Rom	Rom	Rom	Rom
70%	Ob	Ob	Ob	Ob	Ob (Lat)	Ob>NV	Rom (Lat)	Rom	Rom	Rom	Rom
60%	Ob	Ob	Ob	Ob (Lat)	X NV (Lat 1/2 Ob)	NV	NV	Rom	Rom	Rom	Rom
50%	Ob	Ob	Ob (Lat)	NV	NV	NV	NV	X NV>Rom	Rom	Rom	Rom
40%	Ob	Ob	X Ob	NV	NV	NV	NV	NV	X NV>Rom	Rom	Rom
30%	Ob>NV	NR>Ob (Lat)	NV	NV	NV	NV	NV	NV	NV	X NV>Rom	Rom
20%	Other/NV	NR	NR	NR	NR (Lat)	NR (Lat)	NR	NR	NR	NR	Other/NV
10%	NR	NR	NR	NR	NR (Lat)	NR (Lat)	NR	NR	NR	NR	NR
0%	NR	NR	NR	NR	NR	NR	NR	NR	NR	NR	NR

The seven boxes I was incorrect on are signified by an “X”, while I have also added “(Lat)” to note where Latino voters are represented.

Factor 1 for victory in 2016 – drop the doom and gloom – the Democratic vote fell off even more than I projected in the model.

I correctly said Obama was losing voters from 2008 because Romney was effectively making the case he had not delivered and could not work with Republicans, and in fact Obama did lose the votes I thought he would (see 30% vertical/100% horizontal and 90%/50%). In fact, he lost even more than those 4 million votes. I did believe box 80%/50% would go all the way from Obama to Romney, whereas they just stayed home after Sandy eased their fears of White House incompetence and partisanship, and Obama additionally surprised me by losing box 90%/50%. So with him losing all those voters, it is clear that the huge registration of African-American voters was fairly maxed out in 2008 and boxes 30%/90% and half of 60%/60% of largely Latino voters were the only additions to the Obama coalition, so they do NOT have a never-ending increase in voters that cannot be defeated.

Factor 2 for victory in 2016 – The opportunity for the Latino Republican vote

Obviously, falling from George Bush’s 44% of the Latino vote to Mitt Romney’s 27% was devastating due to the Latino vote growing to 10% of all voters this year. The true missed opportunity here is that while more non-Latino Catholics were moving toward Republicans this election due to President Obama’s attack on Catholic hospitals and pro-choice appeal, Latino Catholics were weighing their pro-life stance and hard work ethic against the harsh anti-immigrant rhetoric coming from many Republicans and too often picking Democrats as the lesser of two evils.

As you can see from the four boxes together starting at 20%/60%, we are going to end up with another 8 million Latino voters in the not-too-distant future, and it may be that we simply need a Latino candidate at the top of the ticket this year after opting not to shore up Colorado, Florida, and perhaps even Nevada with a Latino candidate for VP this year. I believe half of the new Latino voters will end up as Democrats, but the other half are up for grabs. Winning over those 4 million new voters to the Republican side could be the key in 2016 and beyond.

To understand why we must at least closely split the Latino vote to ever win the White House again, please envision the entire electorate is represented by 100 people standing in a room. Trust me, as a community organizer President Obama and his whole team had a picture like this in their head for two straight elections:

If the Democratic/Republican split of African-Americans, Latinos and single women stay constant and the Latino voters represent 17 of every 100 voters in the near future, then Republicans need to win 73% of all the men and married women who are not African-American or Latino to win the White House. The problem is that when the national vote gets to 17% Latino, then the Texas vote will be about 38% Latino, and at that rate a Republican candidate would need to win 35 of the 37 men and married women in the room to get Texas’s 38 electoral votes, which would never happen. If we let the Latino vote go and Texas goes Democratic, then every future GOP candidate will have to sweep all 8 swing states Romney did not get this year – many of them with heavy Latino votes as well (Nevada, Colorado, Florida).

Factor 3 for victory in 2016 – the chance for a real Republican ground game in 2016

While I picked a 58% chance of a narrow Romney victory, every time I received a text saying “do you believe Romney will be over 300 electoral votes,” my response was always, “Stop watching Dick Morris.” While Morris is brilliant on many fronts, he kept referring to the Republicans having a better ground game – which was absolutely ludicrous. In fact, Republicans made some strides in the ground game, but that was just try to get to a level that was HALF as good as the Democrats’ ground game, and the fact that they could not get additional Republicans to the polls (see X’s in boxes 80/20, 50/30, 40/20 and 30/10) was the difference in losing 63 million to 60 million and winning.

Having worked 14 winning convention efforts against only two losses, and facilitating the knocks on doors of 169,000 voters in 2010, when I saw the numbers of knocks the Republicans had completed I assumed it was targeted like each of mine to maximize every dollar and every volunteer’s time to get potential supporters to the polls. What I have learned in the past week after digesting what workers were actually doing left me very concerned but with a greater understanding of why those 8 million voters did not vote.

Simply put, door-to-door is an art, not a science, and the overbearing push to “have the biggest numbers” in a corporate type accounting led to great inefficiencies, even before considering the Election Day crashes to the Orca system. Too many workers were pushed simply to “get more door knocks” than other groups or coworkers. The fatal flaw in that system is that many have said they were told they could get the door knocks anywhere – even hitting the same door up to three times.

I have taken a lot of teams out door knocking in urban, rural and suburban areas in 8 different states. Certainly we knock on far fewer doors per hour when in more rural areas and in upper middle class suburbs. But guess what – most voters in those two areas are very open to a Republican message, and I’ve often hit the surprised pastor or other leader who cannot believe I made it to his home. When the whole accountability is “how many doors,” then the only logical course for a worker is to find row houses or apartments in a more urban setting and have each person on the team knock on hundreds of doors per shift rather than dozens.

It sounds inefficient to ask people to do FEWER door knocks on doors more likely to vote Republican if asked, but when canvassers this year instead knocked on doors in Democratic areas they were doing things like moving people a little bit right on the grid above (maybe they go from 80% likely to vote for Obama to 70% likely after talking to you), but also making them MUCH more likely to move up the model on how likely they are to vote (maybe the move from 30% likely to vote to 80% likely to vote). By going through largely Democratic areas to ring up numbers, many were being counter-productive – turning out more Democratic voters to vote than they were winning over voters to the Republican side.

The good news is that Republicans were able to engage in grassroots activities that created 37,000 volunteers on election day and many more in the months leading up to it – now they need people in place who understand that anecdotals about winning over voters each day are more important than simply ringing up numbers.

I’ll give just two examples of how Democrats were so much more effective on the ground:

First, millions of emails were sent asking Democratic-likely voters if they had a coach or spare room where one of their many volunteers could stay for a few weeks when they arrived in Ohio or wherever. Even though I’m sure the percent hit on hosting was a low percentage, it engaged the recipient at a new level and got them talking (e.g. “Democrats have so many volunteers helping they asked if they could use my house”).

Second, Democrats’ phoning was much more effective than Republicans’.  While Republicans used tight scripts asking people if they would vote Election Day, Democrats gave their callers much more leeway and found that voters were 8% more likely to turn out if the script closed by asking, “Would you mind if I called back after election day to see how you liked your voting experience?”

The good news is Republicans created a much bigger army this year, and if they can people who know grassroots direct that army next year rather than make it a cold, corporate number-crunching effort, it can deliver the win in 2016.

Factor 4 for victory in 2016 – Photo ID/Republicans must stand for the integrity of each person casting one vote.

The next lesson is tougher, because it is very sensitive.

I’ve run bipartisan ground games to turn out 2,500 people to a school choice event in New Orleans, pull an upset by winning 68% of the African-American vote in a large southern city, help deliver city council votes, and on many occasions get hundreds of people to a zoning hearings.

Therefore, I know that 99% of people who work ground games, whether in urban areas or elsewhere are great, hard-working people and that when they hear “Photo ID” they immediately hear “voter suppression.”

However, the fact is that people are given vote totals they have to hit on any Election Day, and there is always a small percentage who will break the rules to make money in any setting. I do not have in-depth knowledge of the precincts that reportedly had more votes cast than registered voters, or the precincts in which not a single person voted for Romney. However, when I heard about all of the Republican poll observers being kicked out of Philadelphia precincts I remembered a liberal friend who had laughed off my assertion that some votes are stolen somewhere every national election.

This liberal friend moved to Philadelphia years ago. On Election Day, he asked an African-American janitor he had come to know if he had voted.  He was surprised when the response was a suspicious, “Yeah, why?” It turned out several other people had asked that question earlier in the day and offered to take him someplace to vote again. Any of us who have worked a turnout effort on the ground understand this happens. I know of a couple of cases in the south where it was Republicans who were convicted on voter-related fraud. The fact is some people will break the rules to make money in any setting. And as long as we have a divided nation where there are thousands of precincts that are almost 100% one party or the other, we must have a basic Photo ID requirement to ensure the integrity of elections, particularly in a state like Wisconsin where there is no voter registration required before showing up to vote.

We must have Photo ID enforced, particularly in Wisconsin where you do not have to register to vote – just show up Election Day.

Stopping early voting is another matter. Many Ohio pastors of predominantly African-American churches went so far as to tell their congregations to vote for Romney after Obama came out in support of gay marriage. However, when they attempt was made to cut off early voting, it was seen as a direct hit on black voters in many churches who take the bus to the polls to vote as a social event of immense importance in light of our nation’s history. As long as we are enforcing Photo ID laws and are willing to provide a ID in the rare case that someone actually cannot make it to a courthouse to get one, we should pick our battles on early voting.

A great irony occurred on Election Day. For months, the Obama team had labeled Voter Photo ID laws as anti-Latino (among other things) and fought hard in court to block the laws from being enforced in Wisconsin and Pennsylvania. At the same time many Republicans had told the International election observers invited by Hillary Clinton that they were not welcome in voting places. Ironically, one of the first reports regarding the International observers was that they could not believe that Americans were not required to produce a Photo ID to vote.

Factor 5 for victory in 2016 – not have another Act of Nature like Sandy the week before the election

And then there are things you cannot control.

The model I posted had already calculated that Obama would not be able to get close to the 70 million votes in 2008 because of the lower enthusiasm shown in polling. I pegged him at 66 million votes and wrote that even that might be several million off – which it was as he came in at fewer than 63 million votes.

What I did not adjust for was the millions of voters who decided they no longer wanted to fire President Obama after watching him and Gov. Christie deal with Sandy. I failed to make this adjustment based the false belief (hope?) that by Election Day voters would remember why they wanted to fire President Obama pre-Sandy. The Gallup poll of all voters ended up right on again. The race was tied just before our attention turned to Sandy’s approach to New York, and during President Obama’s response shifted steadily to 3-point lead for the President.

Some of my more astute readers noted that in my original post written the weekend before Sandy (and posted October 30 with an extra paragraph added) I stated, “Romney’s chances of becoming President actually rest at about 58% right now.” Then in my November 5 post I wrote:

In a linear sense, Krugman and Silver are right. In my mind, there is no question that several million more adults living in America today prefer President Obama to Governor Romney after watching him and Gov. Christie in the wake of Sandy… but all in all I stand by the 58% probability of a Romney win on Tuesday.

Yes, I realize it didn’t make sense to say that I believed Romney had just lost millions of votes during Sandy and his chance of winning had stayed at 58%. The first half of my assertion proved to be correct, but that meant the second half could not be. I’m guilty of wishful thinking that those couple of million people who were ready to fire President Obama for not doing a good jobs and refusing to work with other people (Republicans) but then gave him a 67% approve/16% disapprove on his handling of Sandy would somehow remember why they wanted to fire him in the first place by Election Day. So yes, many who read both posts let me know it was clear by my own observations that Obama became the favorite during Sandy.

However, in crunching the numbers after Sandy I have calculated that Nate would have still picked the right winner (albeit in a much closer race) if Sandy had not occurred.  

Let’s apply the 3-point swing measured by Gallup due to Sandy to the ACTUAL election results by adding 1.5% to Romney’s percentage in each state and subtracting 1.5% from Obama in each state.  Based on this assumption, Obama’s 272-vote firewall would have held even without the impact of Sandy to give him a narrow win:

Florida easily goes to the Romney column to get him to 235. Ohio, which was surprisingly better for Romney than four states that he appeared to have before Sandy (VA, CO, IA, NH), also easily moves into the Romney column to get him to 253. Virginia would be in a state mandated recount for falling within a half-percent margin, but would also appear to give Romney 266.

So even without Sandy, Obama had a 272-266 win, and my prediction of a very close election would have been correct, but my 58% chance of Romney winning would have been off by a bit.

Reasons Nate’s model was closer than mine in 2012

When it was revealed Election Day that Romney’s pollster actually had him 5 points down in Ohio in his internal polls – contrary to other reports – I was reminded of a former great basketball coach who used to erase missed free throws from scorebooks to give his players a better chance to be drafted in the pros. When the report of the internal poll leaked, I realized immediately that Nate Silver was working off correct data and that I had been working from an altered stat book based on an email from the Romney pollster that stated the Obama numbers did not add up.

To make an analogy, I’ve had some great success working with others to build a model that very accurately projects which college players will produce at the NBA level. Now, suppose I was doing this and calculated that a player was so good at drawing fouls that if he went to the NBA he would average 10 foul shots a game, and as an 80% free throw shooter, that would get his team 8 points. Hearing the Obama +5 leak on Election Day was like learning after a team drafted a player I had suggested that the coach had doctored the stat book and, in fact, the player was only a 20% free throw shooter, meaning he would no longer be able to make an NBA team. Luckily in sports, the data is all public and that can’t happen anymore, but in politics the garbage in-garbage out adage held. Nate had good raw data and was right.

Now that we see public polls have been able to overcome issues with cell phones and Internet vs. landlines to stay accurate, that adjustment will be made.

A couple of years after nailing the 2008 results, Nate admitted the Obama campaign had given him access to their internal polling. If I’d had access to Romney’s internal data this year, I would have reached the same 90% conclusion, and the fact that Nate did have access in 2012 to the raw data compiled from the polling operation of a billion-dollar campaign gave him sound data from which to run his model.  

In between, his model had problems, as he was off on the political impact of Obamacare; his underestimation of how many seats the GOP would pick up in 2010 House races; and his calculation that Rick Perry had a 50% chance of being the GOP nominee. But the fact is he won two straight Superbowls with the spot light on him in the 2008 and 2012 elections. No one is going to care about his 9-7 regular season in between them.

Congratulations to Nate on the correct pick, but more importantly, as conservatives let’s work on the 4 Factors we can control between now and 2016.