Sabato's Crystal Ball

Democrats 2016: From South Carolina to the Ides of March, Part Two

Upcoming Democratic calendar clearly favors Clinton, but Sanders has credible targets, too

Kyle Kondik, Managing Editor, Sabato's Crystal Ball February 25th, 2016

facebooktwittergoogle_plusredditpinterestmail

This is the second part of a two-part series analyzing the flood of primaries and in both parties from now through March 15. Last week we looked at the Republicans, and this week we look at the Democrats.

Unlike the Republicans, who give states some leeway to come up with their own rules for allocating convention delegates, the Democrats have uniform guidelines that apply across their caucuses and primaries. Delegates are apportioned proportionally both at the congressional district and statewide level, and in order to receive delegates at either level, a candidate must pass a 15% support threshold. Given that the Democratic race is now just a two-person contest among Hillary Clinton and Bernie Sanders, both candidates will meet the threshold in nearly all states and congressional districts.

By the time the March 15 contests are over, voters will have allocated close to 60% of all the Republican delegates. The prospect of a knockout blow by Donald Trump over the next few weeks is very real. However, Democrats will have only allocated about 40% of their delegates by mid-March. The Democratic calendar is a little bit more back-loaded, but the main reason for the discrepancy is the large number of superdelegates, the party leaders and elected officials who make up about 15% of the Democratic delegates. These delegates are not tied to the caucus and primary results and can vote their conscience at the convention.

Al Tuchfarber, the founder of the University of Cincinnati’s Ohio Poll and a Crystal Ball contributor, points out two key ways the Democratic method of allocating delegates benefits Clinton over Sanders. We’ll quote him at some length:

“The first rule that benefits Clinton is the use of superdelegates. Of the 4,763 total voting delegates, 712 are superdelegates. They are free agents and can vote any way they want. But, a large majority of them declare well in advance who they plan to vote for and who they support.

“To date, 451 superdelegates have declared for Clinton, with just 19 for Sanders while 242 are still undeclared. The ratio of Clinton declarations to those for Sanders is 96% to 4%. If we assume that Sanders does slightly better with the remaining 242 and gets 10% of them, the superdelegates end up splitting 669 to 43. That is, clearly, an overwhelming margin for Clinton.

“The other rule of special note is that essentially all of the delegates elected by the voters are awarded proportionately. Over-simplifying a bit, if the voters split 50-50, the elected delegates from that state are also split 50-50.

“That contrasts with the Republican allocation process where, after mid-March, the states can decide to go to a winner-take-all format.

“The practical consequence of these two facts is that Sanders needs to win about 58% or more of the votes in EACH of the remaining states. That is a Herculean task as (in effect) he needs a landslide in all those states. A landslide is usually defined as a win 55%-45% or greater.

“Putting this in perspective another way, Sanders must achieve New Hampshire-size wins (61%-38%) in almost every state to get to the 2,383 delegates needed to win the nomination.”

So, given the Democratic rules, Prof. Tuchfarber argues that Sanders’ chances of actually defeating Clinton in a head-to-head matchup barring some new, campaign-altering development are remote, and we agree with that assessment.

Sanders supporters, understandably, are upset about superdelegates, because superdelegates exist for the primary purpose of stopping outsider candidates like Sanders. After outsiders George McGovern (1972) and Jimmy Carter (1976) won the Democratic nomination, “High-ranking Democrats were determined to never again have to sit back and look on helplessly as a candidate outside the control of the established political machinery became their party’s duly elected candidate,” according to Santa Clara University’s Nancy Unger. Superdelegates were created in advance of the 1984 Democratic primary season and have been a significant part of the Democratic delegate picture ever since.

Reasonable minds can differ on whether having superdelegates is appropriate or not. But the reality is that they are a part of the process, and they overwhelmingly support Clinton. Sanders’ challenge is trying to convince them otherwise. This is why the investigation into Clinton’s use of private email is perhaps the biggest hurdle to her winning the nomination: If serious problems emerge for her from that investigation, Clinton’s support among the party leadership could soften, although party leaders would still probably not want to back Sanders. Perhaps that could open the door for another candidate to make a late entry, an off-the-wall scenario we sketched out last year.

A more pressing issue for Sanders, though, is the upcoming caucus and primary calendar. While both candidates will split delegates and score statewide victories, the schedule over the next few weeks appears to favor Clinton strongly so long as her big lead among black voters, a crucial Democratic voting bloc, endures.

With that, here’s a preview of the Democratic contests from now through March 15. Look for a special Crystal Ball next Tuesday night reacting to the Super Tuesday results, as well as a regular issue next Thursday.

— The Editors

Feb. 27

South Carolina Democratic primary

Delegates at stake: 53 (35 based on congressional district vote, 18 based on statewide vote)

2008 primary winner: Barack Obama

2008 nonwhite % of electorate: 57% (55% black)

While Clinton’s support among Hispanics may be eroding — entrance polls showed her losing Hispanics in Nevada, although the Clinton campaign and some outside analysts dispute that — Nevada did confirm that Clinton maintains a rock-solid edge amongst black voters. She won about three-quarters of African Americans in the Silver State. Given that South Carolina should have a Democratic electorate that is at least half black, this bodes well for Clinton, and she leads by about 25 points there in polling averages. Realistically, Clinton should do as well in South Carolina as Sanders did in New Hampshire. While Clinton has revealed plenty of weaknesses in the first month of the race, a solid victory in South Carolina — which looks all but certain at this point — would give her wins in three of the first four contests.

March 1

Alabama Democratic primary

Delegates at stake: 53 (35 CD, 18 statewide)

2008 primary winner: Barack Obama

2008 nonwhite % of electorate: 56% (51% black)

Alabama’s racial demographics are almost identical to South Carolina’s, which means that if Clinton wins the Palmetto State handily, there’s no reason why she shouldn’t be able to perform similarly here.

A Sanders strategy could be to concentrate his efforts in the northern part of the state, where there are relatively few African Americans. For instance, Alabama’s fourth and fifth congressional districts, in the northern part of the state, have smaller black populations than the state as a whole, so maybe Sanders can squeeze a few extra delegates out of these districts. Other fairly white parts of some Southern primary states with large African-American Democratic electorates include northern Georgia, eastern Tennessee, and western Virginia. It’ll be interesting to see how Sanders does in these regions even if he’s doing poorly in the statewide voting.

American Samoa Democratic caucus

Delegates at stake: Six (All six based on island-wide vote)

2008 caucus winner: Hillary Clinton

2008 nonwhite % of electorate: Unknown but very high

Clinton only won two caucuses in 2008: Nevada and American Samoa.

Arkansas Democratic primary

Delegates at stake: 32 (21 CD, 11 statewide)

2008 primary winner: Hillary Clinton

2008 nonwhite % of electorate: 20% (17% black)

On the surface, one would expect the state where Clinton served as first lady to strongly support her, as it did in 2008: She won 70% of the vote, her best primary performance anywhere. But Arkansas is also an interesting test as to whether Sanders’ strength among blue-collar whites will translate outside of places like Iowa and New Hampshire. Arkansas is mostly white and not very prosperous. For instance, New Hampshire has one of the nation’s highest median household incomes, while Arkansas has one of the lowest. Sanders did better among voters with lower income levels in Iowa, New Hampshire, and Nevada, although the difference was more pronounced in the first two states, which probably had something to do with Nevada’s greater diversity.

Still, in addition to her home field advantage and the likely backing of a small black voting bloc, Clinton could benefit from something else in Arkansas: About half its primary voters called themselves moderate in 2008, and Clinton did better with moderates than liberals in both Iowa and Nevada (the split between liberals and moderates wasn’t as clear in New Hampshire, although “very liberal” voters strongly backed Sanders in all three states). See the item on Oklahoma below for more on Sanders’ chances in states like Arkansas.

Colorado Democratic caucus

Delegates at stake**: 66 (43 CD, 23 statewide)

2008 caucus winner: Barack Obama

2008 nonwhite % of electorate: 20%*

It’s clear that Sanders is targeting Colorado as a key state on March 1: He’s outspending Clinton more than two to one on television in the lead-up to the caucus. To the extent that Colorado’s caucus turnout will be diverse, it’ll be much more Hispanic than African American. At the very least, Nevada’s results confirmed that Sanders is competing with Hispanics whether he won them or not. So Colorado is a key target for Sanders on March 1. Colorado’s Democratic caucus had no exit poll in 2008, but in the 2012 general election about 15% of voters were Hispanic versus just about 3% who were black. Given Clinton’s small win in Nevada, which has a larger African-American electorate, Sanders could very well win Colorado.

Democrats Abroad Democratic primary

Delegates at stake: 13 (nine regional, four global)

2008 primary winner: Barack Obama

2008 nonwhite % of electorate: Unknown

This primary for Democrats living outside the United States will take place from March 1 through March 8. Participants can vote by mail, fax, email, or in-person at voting centers across the world.

Georgia Democratic primary

Delegates at stake: 102 (67 CD, 35 statewide)

2008 primary winner: Barack Obama

2008 nonwhite % of electorate: 57% (51% black)

If Alabama and South Carolina go heavily for Clinton, then Georgia likely will too. Three recent polls averaged together show Clinton up a whopping 38 points. Again, watch the margin — and the results in the congressional districts, particularly the ones that have smaller African-American populations.

Massachusetts Democratic primary

Delegates at stake: 91 (59 CD, 32 statewide)

2008 primary winner: Hillary Clinton

2008 nonwhite % of electorate: 15% (6% black, 5% Hispanic, 4% other)

Clinton won Massachusetts by 15 points in 2008, but the state’s demographics don’t fit the current version of her candidacy. In fact, the state looks a lot like New Hampshire, which Sanders won easily. A couple of recent polls either show the state tied or Sanders ahead by a little. While Massachusetts is more diverse than New Hampshire, it still seems like an obvious target for the Vermont senator and arguably a must-win. Both Clinton and Sanders are spending hundreds of thousands of dollars on ads in the Bay State.

Minnesota Democratic caucus

Delegates at stake: 77 (50 CD, 27 statewide)

2008 caucus winner: Barack Obama

2008 nonwhite % of electorate: 11%*

Here’s another Super Tuesday caucus state where Sanders could do well because of a relatively small minority population, plus he’s outspending Clinton on television in pursuit of a victory.

Oklahoma Democratic primary

Delegates at stake: 38 (25 CD, 13 statewide)

2008 primary winner: Hillary Clinton

2008 nonwhite % of electorate: 18%

It seems strange that Oklahoma — one of the most Republican states in the nation at the federal level — might be a key state in the Democrats’ Super Tuesday, but that could be the case next week. Oklahoma, like aforementioned Arkansas, was a strong Clinton state in 2008: She defeated Obama 55%-31% in the Sooner State, with John Edwards getting another 10%. Yet the most recent poll, by the Democratic firm Public Policy Polling, showed essentially a tie earlier this month.

Earlier this week, a West Virginia poll showed Sanders crushing Clinton by nearly 30 points in that once-very pro-Clinton state (granted, it’s just one poll of a primary that’s not coming until May). Oklahoma is similar to states like Arkansas and West Virginia: They are culturally conservative and religious and also not very diverse. They are states that were historically very Democratic at the local and state level.

To the extent that there are conservative, anti-Obama Democrats anywhere in the nation, one might find them in states like these. As friend of the Crystal Ball Richard Skinner of the Sunlight Foundation observed in response to the West Virginia poll, “Maybe anti-Obama Dems see Sanders as anti-Obama candidate.” That might make sense, given how closely Clinton is now tied with Obama through her service as secretary of state. So surprisingly, Oklahoma, one of the most conservative states in the nation, could be fertile ground for the very liberal Sanders — as could West Virginia, which votes in May and overwhelmingly backed Clinton over Obama in 2008 but has now possibly turned against her.

One potential plus for Clinton is that, like Arkansas, Oklahoma’s Democrats were about 50% “moderate” in 2008.

Tennessee Democratic primary

Delegates at stake: 67 (44 CD, 23 statewide)

2008 primary winner: Hillary Clinton

2008 nonwhite % of electorate: 34% (29% black)

Tennessee’s white Democrats are probably fairly similar to Oklahoma’s, but the Volunteer State does have a bigger nonwhite vote, which is almost entirely African American and should strongly support Clinton. The Volunteer State isn’t quite as promising for her as some of the other Southern states, but she still looks like a favorite, and like many other Southern states Sanders does not appear to be really competing for Tennessee, at least on television.

Texas Democratic primary

Delegates at stake: 222 (145 apportioned through state Senate districts, 77 statewide)

2008 primary winner: Hillary Clinton

2008 nonwhite % of electorate: 54% (32% Hispanic, 19% black)

If Sanders truly is vying with Clinton among the Hispanic vote — perhaps not winning those voters, but coming close — Texas suddenly becomes a lot more interesting. Like some of its Southern counterparts, Texas will have a majority nonwhite electorate, but it will be more Hispanic than African American. Clinton has to be favored in the Lone Star State, but this is another test — a much bigger test — of her standing amongst Hispanics.

Vermont Democratic primary

Delegates at stake: 16 (All 16 awarded proportionally statewide because Vermont only has one at-large congressional district)

2008 primary winner: Barack Obama

2008 nonwhite % of electorate: 6%

If recent polling is an indication, Clinton is in danger of not even reaching 15% in Sanders’ home state, which could give Sanders a clean sweep of the state’s delegates. Fortunately for Clinton, Vermont has one of the smallest delegate counts of any state.

Virginia Democratic primary

Delegates at stake: 95 (62 CD, 33 statewide)

2008 primary winner: Barack Obama

2008 nonwhite % of electorate: 38% (30% black)

Geoffrey Skelley took an in-depth look at our home state’s primary as a companion piece in this week’s Crystal Ball. While Sanders will have pockets of strength in the Old Dominion, Clinton is favored here, as she is throughout the rest of the South.

March 5

Kansas Democratic caucus

Delegates at stake: 33 (22 CD, 11 statewide)

2008 caucus winner: Barack Obama

2008 nonwhite % of electorate: 13%*

A small, very white state with a caucus? That sounds like a place where Sanders has an opportunity to post a victory. Clinton lost the nomination in 2008 in large part because she neglected caucuses. This will be a good test as to whether her operation is better this time. She is, narrowly, 2-0 in caucuses so far, winning very white Iowa and more diverse Nevada.

Louisiana Democratic primary

Delegates at stake: 51 (33 CD, 18 statewide)

2008 primary winner: Barack Obama

2008 nonwhite % of electorate: 52% (48% black)

Louisiana shares a similar demographic profile with states like Alabama, Georgia, and South Carolina: If Clinton holds up with African Americans, she should be fine here.

Nebraska Democratic caucus

Delegates at stake**: 25 (17 CD, eight statewide)

2008 caucus winner: Barack Obama

2008 nonwhite % of electorate: 12%*

See entry for Kansas.

March 6

Maine Democratic caucus

Delegates at stake**: 25 (17 CD, eight statewide)

2008 caucus winner: Barack Obama

2008 nonwhite % of electorate: 5%*

Maine, a New England caucus state with a tiny nonwhite population, should be among Sanders’ best states. The same caveat noted in Nebraska also applies to Maine regarding the actual allocation of national delegates.

March 8

Michigan Democratic primary

Delegates at stake: 130 (85 CD, 45 statewide)

2008 primary winner: Hillary Clinton

2008 nonwhite % of electorate: 28% (23% black)

This is a potential Democratic battleground a week after Super Tuesday, although polls have shown Clinton generally ahead (54%-36% in the RealClearPolitics average). Setting aside the importance of Sanders maximizing delegates everywhere, there are symbolic reasons why Michigan is important to him. It comes one week after Super Tuesday, which features many contests that Clinton should win (and a smaller number that Sanders should capture). There are only two primaries on March 8, and Sanders likely has no chance to win the other one (Mississippi). So the big story of this night on the Democratic side will be Michigan, and if Sanders can’t win it, he risks having to deal with questions about whether his campaign remains viable.

Mississippi Democratic primary

Delegates at stake: 36 (23 CD, 13 statewide)

2008 primary winner: Barack Obama

2008 nonwhite % of electorate: 52% (50% black)

As we said in our Republican preview, Mississippi’s result should mirror Alabama’s, a similar, neighboring state that votes a week earlier. This should be another strong Clinton state.

March 12

Northern Mariana Islands Democratic caucus

Delegates at stake: Six (All six awarded based on islands-wide vote)

2008 caucus winner: N/A

2008 nonwhite % of electorate: N/A

It appears that this will be the first time the Northern Mariana Islands will hold a Democratic primary or caucus.

March 15

Florida Democratic primary

Delegates at stake: 214 (140 CD, 74 statewide)

2008 primary winner: Hillary Clinton

2008 nonwhite % of electorate: 34% (19% black, 12% Hispanic)

It may be surprising to some that Florida’s Democratic electorate has more African Americans than Hispanics, but that was the case in 2008 and could be again in 2016. That’s one point in favor of Hillary Clinton. Another is that, perhaps less surprisingly, Florida Democrats skew older: More than 70% of primary participants eight years ago were over 45. Clinton does much better among older voters than younger ones.

Illinois Democratic primary

Delegates at stake: 156 (102 CD, 54 statewide)

2008 primary winner: Barack Obama

2008 nonwhite % of electorate: 43% (24% black, 17% Hispanic)

Clinton got no home state advantage in the state where she grew up in 2008, as she lost handily to Obama in his adopted home state. The Land of Lincoln’s Democratic electorate is almost one-fifth Hispanic, so if Sanders is indeed capable of winning Hispanics he needs to do so here to counteract the big African-American vote that should show up for Clinton. After four years as Obama’s secretary of state and given her home-state status, however, Clinton has to be considered the favorite.

Missouri Democratic primary

Delegates at stake: 71 (47 CD, 24 statewide)

2008 primary winner: Barack Obama

2008 nonwhite % of electorate: 23% (17% black)

Missouri’s rural areas are culturally conservative and largely Southern in heritage, so perhaps the voting in states like Arkansas and Oklahoma will tell us something about how the non-St. Louis and non-Kansas City parts of the state might vote. A decent-sized black population gives Clinton a solid base.

North Carolina Democratic primary

Delegates at stake: 107 (70 CD, 37 statewide)

2008 primary winner: Barack Obama

2008 nonwhite % of electorate: 38% (34% black)

The Tar Heel State is one place where there typically is a lot of polling, and Clinton has led even recent surveys, although her lead might be eroding: She was up just 47%-37% in an Elon University poll that came out Monday. Of the 11 states of the old Confederacy, North Carolina is the last one to vote. Because of the strength of African-American voters across the South, it’s possible that Clinton will win all 11. Any Sanders victories among these states would be a blow to the Clinton campaign.

Ohio Democratic primary

Delegates at stake: 143 (93 CD, 50 statewide)

2008 primary winner: Hillary Clinton

2008 nonwhite % of electorate: 24% (18% black)

The Ohio primary will coincide with a Democratic U.S. Senate primary in Ohio, and the comparison of the Senate candidates to the presidential ones helps illustrate the rift in the Democratic Party, and how it doesn’t break on neat lines. Former Gov. Ted Strickland, the favorite to win the Senate nomination, is from Appalachian Southeast Ohio and is a strong supporter of Hillary Clinton. P.G. Sittenfeld, a young Cincinnati city councilman challenging Strickland, has not endorsed either Clinton or Sanders.

Clinton, with Strickland’s help, won Ohio’s primary by about nine points in 2008. She performed well across the state, carrying all but five counties, although she lost the three biggest ones: Cuyahoga (Cleveland), Franklin (Columbus), and Hamilton (Cincinnati). She ran up the score in Strickland’s Appalachia.

It seems reasonable to think about the Senate race the same way one thinks about the presidential one: Strickland, like Clinton, is the establishment pick, and Sanders, like Sittenfeld, is the outsider. But it’s much more complicated than that.

To the extent that the underdog Sittenfeld has scored hits on Strickland, it’s been through attacking the former governor over his past support from the National Rifle Association. Here again, there’s a similarity to the presidential race, but with Sittenfeld, the insurgent, echoing the argument from the establishment presidential candidate, Clinton, who has aggressively attacked Sanders from the left over guns. On the other hand, Sittenfeld has called for debates (Strickland has passed, thus far), a classic outsider line that mimics Sanders’ call for more debates throughout 2015 (he and Clinton eventually agreed to expand the number of debates).

It’s possible that Strickland will romp to victory statewide without any discernible pattern: There is not much hard data to suggest Sittenfeld is truly pushing Strickland. However, if both the presidential and Senate races are close, Strickland and Sanders might do well in some places outside the urban areas, like Strickland’s home base in Appalachian Ohio, while perhaps Sittenfeld can make inroads with minority voters in the big cities, voters who likely will be supporting Clinton. For a Democrat, Strickland has only a mediocre relationship with black voters in Ohio.

It’s clear that the Clinton 2008 coalition is a lot different from her current one, and Sanders and his supporters don’t necessarily represent Obama 2.0. And even in a state where the Democratic presidential contest had the possibility of creating a down-ballot Senate primary proxy war, that hasn’t really developed because of the peculiarities of the candidates.

Notes: *Estimated nonwhite voting percentages; **Colorado, Maine, and Nebraska’s precinct caucuses start the process of selecting national delegates, but they are not officially allocated until later district and/or state conventions. Iowa and Nevada are also like this. However, the caucus results should give us a fairly good idea of how many delegates each candidate will eventually receive, especially in a two-person race.

Sources: FrontloadingHQ; The GreenPapers; 2008, 2012, and 2016 entrance and exit polls