State (End) of the Primary: How and Why Hillary is Winning

Inflammatory title aside, what I am setting out to do is not put down Sanders or his supporters, but rather attack implausible projections that make Sanders as still decent odds to win the nomination, and also to put out my own model of the election. There is enough vague boasting and words on this site, so I’ll start out with a table of numbers. The second row is Clinton’s increase in vote %, in other words, in Iowa 2008 she won 30.4% of the vote, but in 2016, she won 49.8, an increase of 19.4%.

Quick initial observations; Sanders also has an obvious regional advantage, same as Clinton, and that advantage is in New England, whether because of how white it is, how many rural Democrats there are, or because its his home region (and like the south, one quite parochial and close-knit). Also like Clinton, his region is mostly tapped out, all but 2 small states plus perhaps 2 congressional districts worth of turf in upstate New York. Sanders has only cut Clinton’s margins back in 6 states out 25. I am not including Michigan for the simple reason that the purpose of this model is working from 2008 to come to an understanding how 2016 is shaping out, including what’s changed and where the fundamentals of the race sit. Michigan, because of a fight about scheduling its primary early, initially had all its delegates taken away by the DNC. Obama wasn’t even on the ballot there, it was a low turnout affair with Hillary Clinton versus non-committed and Dennis Kucinich and winning 54% of the vote. Simply put, Michigan is useless as a data point because it doesn’t represent a regular primary where Clinton was actually on the ballot against Obama, so I don't use it for or against her (and I actually think she would have gotten around 44% in 2008 and that her performance improved there this year, but that is just a speculative piece of info that I don't incorporate into this model one way or the other).

Yesterday I basically introduced a set of rules that explain the entire primary. 1) Check the region of a state. 2) Check nonwhite voting % 3) Check Hillary’s 2008 % 4) Are there a lot of very liberal or very conservative white voters there, 5) Is the state dominated by big urban centers

For all states going forward, 1 no longer really matters. The inclusion of very conservative white voters in #4 my seem odd, but its undeniable that Sanders has been benefiting from the votes of conservative white Democrats that can’t vote in the Republican primaries in their states and so simply vote against whoever they perceive to be the establishment/Obama candidate.  That’s a major part of why Sanders won Oklahoma, it’s why Clinton’s margin slipped a bit from 2008 in Arkansas. The only counties Sanders won in Florida were rural white Demosaur counties in the north part of the state, including one county actually called Dixie, and the only two Parishes he won in Louisiana, were its two whitest, most rural, most Republican Parishes, LaSalle and Cameron. It’s why Sanders has a huge polling lead in West Virginia, where voters are furious about the EPA and the decline of the coal industry.

A secondary issue is whether the primary is an open, semi-open, or closed primary. Sanders has done best in Open primaries, while Hillary does best in Closed primaries. Even in Michigan, had the primary been restricted to Democrats, Clinton would have won 57-43. Pennsylvania, New York, and New Jersey are all closed primaries whose underlying demographics strongly favor Clinton in my model. The fictional argument that people who don’t like reality push is two-pronged: no one has won a majority of pledged delegates, and now that the south is done the rest of the map is a lot more favorable to Sanders. Well my model shreds that.

It also explains why Illinois was so close and Ohio wasn’t; it was basically a battle between factors 2 and 3 and to a lesser degree 5. 2 and 5 formed a strongly favorable demographic setting for Hillary, yet her 2:1 drubbing there in 2008 established a large negative counterbalance to these favorable demographics, especially if in this case, we are considering region and party composition as neutral.

Now, most of the rest of March is favorable to Sanders. It’s almost all caucuses, where party activists have dominated and anti-establishment sentiment is far, far stronger than anywhere else. Clinton has also always struggled, and the series of small, white, rural states + Washington are all states that gave Obama massive margins. The one primary is Arizona, where Clinton should win, but will have to spend time and resources holding off Sanders' late play for the state. Now, going off my model I say, rather matter of fact that regardless, Clinton wins 54-58% of the vote in Arizona if the race simply continues to play as it has been playing. But for good measure, here is a table of the rest of the primary states that Hillary won in 2008, Hillary ‘s 2008 %, and percentage nonwhite. And, just for kicks, I’ll include what I project Hillary to get in each state if it were hotly contested down to the wire, and these projections aren’t final, they are purely a matter of saying this is around what my model would project if everything stays relatively stable, as I anticipate.

*Important to know that despite its reputation, California's Democratic primary voters were actually less liberal and more moderate than Michigan in 2016 or New Mexico in 2008 and by a substantial margin, with 50% identifying as liberal and 37% as moderate, and that Clinton won by 8 points even with Obama getting 78% of the black vote.

*Indiana seems like a state ripe in someways for Sanders, but Obama did win 89% of the black vote there, and if Hillary were to get anything near that, it’s hard to see Sanders peeling off enough of the white vote to win. Also important: in 2008 45% of Democrats in the Indiana primary identified as moderate, the largest group of voters, and Hillary has done best this year with moderate voters and “somewhat liberal.”

*In addition to being a closed primary, Pennsylvania also has a very large block of Democratic voters who identified as “moderate, and a very small block that identified as "very liberal.” 71% of 2008 PA primary voters identified as either somewhat liberal or moderate.

*New Jersey is another delegate rich state with nearly twice as many Democrats identifying as “moderate” than “very liberal.” What’s more, Clinton won this state by 8 points even with Obama winning the 23% of the primary electorate that was black 80-20. New Jersey has all the makings of a total blow out this year, especially with Clinton having the official party endorsement in every single county in New Jersey and Sanders campaign having only a light presence there according to NJ party activists I've asked.

*Only 49% of Rhode Island voters in the 2008 primary identified as liberal.

Furthermore, 3 of the remaining states/districts that Obama won in 2008, strongly favor Clinton in their demographics. These states account for 136 delegates alone.

The states where Sanders is obviously favored by the fundamentals of the race are: West Virginia, Wyoming, Wisconsin, Connecticut, Oregon, Washington, Alaska, Utah, Montana, North Dakota, South Dakota, Idaho, and Hawaii. The problem is that half of those states are extremely small states with low delegate counts, and two of the bigger prizes, Wisconsin and Connecticut, don’t favor him by huge margins he’d need to make up delegates anyway at this point.

I would say Connecticut, Indiana, and Kentucky are the only toss up states left on the map. Sanders is heavily favored to pick off West Virginia and South Dakota, Clinton heavily favored in D.C., Maryland, and Delaware.

It’s wrong to think 2016 is totally different from 2008. The core underlying strength of Hillary in a state does correlate to her 2008 totals, and correlates more strongly when the other main factor, %nonwhite, is calculated, and if you look at other fundamentals such as urban-ness, primary voter ideology and region, you can account even more securely for any variations this year. People desperate to stop Hillary say they can still do it, even with Obama’s 2008 coalition truncated. No one can seem answer the question of how you stop Hillary when the backbone of Obama’s campaign, the deep south, backed Clinton by larger margins than it backed Obama (other than some big, crushing wins for Sanders in New York and California). Last Tuesday, Clinton flipped three Obama states from 2008 to her column. For context, that equals as many Hillary states as Sanders has flipped in the entire campaign (including Michigan, which I think indications are Obama would have won in 2008, making it not a flip), and 2 of Hillary’s flipped states where not in the south, blowing up the argument that she only does better than 2008 in the south. This includes one state, Illinois, that was the crown jewel of Obama’s 2008 wins, and where Hillary increased her vote total by over 50% to snatch a narrow win in a state she took only 32% of the vote in 2008.

At this point with these states in 2008, Obama had a large lead in delegates and vote totals, and that lead never changed even as Hillary wasted time and money continuing on the primary long past the point where there was any sense in doing so. This time around, Hillary’s delegate and vote lead are substantially larger than Obama’s were at the same point, with her consistently sitting on 58-59% of both with almost 15 million Democrats having voted. The absurd question that has to be addressed and this model tries to highlight is this: where do you make up all this ground over 2008?

Sanders supporters: “We just lost 8 Obama states in the south by an average margin of 35%? No big sweat. We’ll just make that up elsewhere.”

Obama beat Hillary by a fraction of a percent in the popular vote in 2008 if we discount Michigan, and by less than 100 delegates out of 3500. So far Hillary has flipped 12 of Obama’s states and lost 2 states she carried in 2008. It’s baffling the stubbornness of people who look at that and then try to spin some convoluted answer for why it doesn’t matter, why the primary is still viable and on the path to win or even have an outside chance at winning (it doesn’t, and Nate Silver said that Hillary Clinton firmly established herself as “the presumptive nominee” with strong overperformances in his baseline model again last Tuesday). The fact that, discounting Michigan since we don’t have a real 2008 baseline there, Sanders has only cut into Hillary Clinton’s margins in 2 states outside of New England (and every state he’s cut into her margin has been either heavily white or with a large contingent of conservative, anti-establishment Democrats backing him), and both were also predominantly rural states. There's ample evidence to establish a consistent pattern of Hillary outperforming her 2008 margins everywhere, including outside the south, and even adding in the Michigan numbers doesn't cut her average increase over 2008. On the contrary, the race has barely moved since January, and Sanders hasn’t changed his numbers with the black vote at all since Nevada, as Hillary’s near 90% numbers in NC and FL indicated. Rather, Hillary consistently gets around 65-70% of the black vote in non-southern states, and 80-90% in Southern states. She’s also consistently winning the Hispanic vote where its big enough to measure, showcasing the strong Hispanic support that stuck with her throughout 2008 against the dynamic Obama campaign remains strong for her in 2016, and complicates Sanders efforts to compete in the southwest and in California and New York and New Jersey (the latter two both have big Cuban and Puerto Rican votes, two Hispanic demos that Sanders has done particularly poorly with).

Nate Silver remarked that the Democratic Primary ‘was remarkably stable, it seems momentum proof.’ Namely that each side has walked off big losses and returned to win comfortably where the fundamentals favor them. Tons of ads, campaigning, microscopic criticism and flame wars from online communities and social media, breathless media coverage, and fluctuating polling hasn’t really mattered. But this was the same in 2008; in 2008 each state fell in line with its fundamentals. The trick for 2016 has just been establishing exactly what the fundamentals are, and how the 2016 lines sliced up both Obama’s coalition in 2008 and Hillary’s coalition in 2008 and recombined them. The key take away from Ohio and Missouri was that Hillary has maintained the core of 2008 support from unions and blue collar white voters, lost even more ground with young voters and a lot of ground with white liberals, and picked up the black vote. Going forward, Hillary is actually strongly favored in the remaining states taken as a whole in this model, and its clear her coalition in 2016 is simply much bigger than it was in 2008 and it gives her a clearer majority than Obama had in 2008.

Nothing changed the fundamentals in 2008 once they were set, though Clinton tried desperately to do so. The fundamentals of this race are baked and rather easy to set into causal relationships now that we have enough states and enough data in place to look at. The bar keeps getting pushed back, by both the Sanders’ campaign (which seems to have no long-term strategy other than playing it one week at a time, and simply throwing oodles of donor-fueled ads at each state, consistently outspending Hillary since January, and sometimes by large margins but without much effect) and supporters. If over the next month, Sanders loses Arizona and New York by decent margins, I hope he does the right thing and downsizes his campaign, changes gear to focusing on the general and party unity, and rides off the remaining states without forcing the Clinton campaign to completely exhaust its cash reserves all the way through June even after it becomes clear she has a majority of pledged delegates.

Furthermore, Sanders message now has to move outside the midwest. You can hardly think of a region better than the Rust Belt for a message based on attacking free trade agreements and big banks as your campaign center pieces. Now we’re moving onto states where other issues are also very important if not more important to Democrats, and also where Sanders has to move beyond big pools of Obama voters and actually pick off substantial numbers of people who supported Hillary Clinton in 2008.

It’s not spiking the football to say continuing to pour money and energy into the Sanders campaign instead of legislative, congressional, and gubernatorial campaigns, is fast approaching the point (I put that point at Mid-April at the latest) where it is actually hurting the advancement of progressive politics by sucking up so much of the oxygen that many progressive politicians and candidates don’t have money or support to go forward and help bring more grassroots representation into the party. Thinking here of Donna Edwards, Carol Shea-Porter, John Fetterman, Lucy Flores, Deborah Ross, and other good Democrats that need support too, including Ann Kirkpatrick and Ted Strickland and Russ Feingold and Roy Cooper and John Gregg and Catharine Cortez Masto and Annette Taddeo. In the meantime, this site needs to focus on being nicer to each other and to tone down the negative campaigning. Not completely stop talking about the primary, or stop posting updates about what each campaign is doing, predictions about upcoming states so long as Sanders is actively running, but refrain from continuously upping the level of mudslinging at the person who is now, to repeat Nate Silver again, the presumptive Democratic nominee, especially as we start entering in a time period that overlaps with the general election campaign where we are likely to be up against the true crazy of either Ted Cruz or Donald Trump, and both Democrats, Sanders and Clinton, honestly need to spend more time attacking their Republican opponents than attacking each other, and this site should take that heart too and focus on following the crazy shit that is the dumpster fire surrounded by burning circus monkeys tossing their feces around, err, I mean the Republican primary. There’s plenty of worthy issues ongoing than to post another "Bernie Sanders is an inept backbencher/idealistic daydreamer" or “Hillary Clinton is a liar/sociopath/racist/misogynistic/genocidal/corporate whore” diaries.

We can critique votes, talk about things we like about each candidate, highlight platforms, criticize specific details (or lack thereof) in platforms, and we can do all of this without personally attacking each others candidates (often on the other side of this debate it sounds like a Republican online forum), or attacking each other in personal terms. Being nice and being civil doesn’t mean agreeing, it means disagreeing based on well-articulated, well-sourced points and not half-ended rants. It means treating our conversation partners with full respect, listening carefully and then explaining or responding with even more care. I’m not saying its impossible for Sanders to win the primary, just that its at the point where out of a hundred possible scenarios, there are only 2 or 3 out of a hundred that lead to him winning and they include a non-stop stream of upsets in bigger primary states and absolutely maximized caucus performances that match or exceed the huge margins Obama set in 2008 when he caught Hillary flat-footed in all the caucuses. That doesn’t mean the discussion will stop, just that the time is now that we can refocus the way we advocate and the way we talk to each other, the time is now that we can start coming together as a community to stop the ever-more-likely global risk that is Donald Trump and his campaign of blatant racism, white grievance politics, and demagoguery. And even if Bernie does manage this 2 or 3 out of a hundred path to the primary win, this is still the time for Bernie’s supporters to move to ease tensions with the massive block of Hillary voters. Too many women and people of color I know take the tone of Bernie supporters as a personal dismissal of women and people of color, when they represent the core of the Democratic coalition. So, for Bernie’s sake too, start building bridges and focusing on party unity. The primary will fall where fundamentals will, and Bernie has a multi-million dollar professional campaign, as does Clinton, neither campaign needs vicious online attacks to win voters, mobilize support, or compete in the primary. Try to keep that perspective in mind.

Thanks for Reading.