There’s no handy nickname for this Tuesday’s primary elections (we’ve seen it called “Titanic Tuesday” at least once, which is great if you want to visualize the country sailing straight into an iceberg), but whatever you call it, it’s actually the third most momentous night of the primary season, in terms of number of delegates at stake. In first place, of course, was actual Super Tuesday two weeks ago, and in second place will be the night in June that features, among others, California and New Jersey.
While most eyes here at Daily Kos are going to be on the Democratic side, the biggest deal may actually be Ohio on the Republican side. If John Kasich can pull out the “favorite son” win in his home state, and polls currently show him narrowly on track to do so, that may actually change the trajectory of the whole Republican primary campaign. Ohio is a truly winner-take-all state with 66 delegates; if Donald Trump wins it, he’s still on track to squeak out slightly more than 50 percent of the delegates overall, but if he doesn’t, he’s likelier to fall just short, which would set up the amazing chaos of the first contested convention of most of our lifetimes.
Polls close at 7 PM ET in most of Florida, though it’s likely there won't be any call until 7 PM CT (8 ET), because parts of Florida's Panhandle are in a different time zone. North Carolina and Ohio close at 7:30 PM ET, while Illinois and Missouri close at 7 pm CT (8 ET). Please join Daily Kos Elections starting at 7 PM ET for our live-blogging; beforehand, though, let's look at some of the details concerning delegate allocation, polls, and local demographics.
FLORIDA
Democratic delegates: 46 at-large, 28 party leaders and elected officials, between three and eight delegates in each of the state’s 27 CDs (214 total)
Democratic polls: Clinton 60, Sanders 33 (HuffPo Pollster aggregate)
Republican delegates: 99 statewide
Republican polls: Trump 45, Rubio 24, Cruz 19, Kasich 9 (aggregate)
Florida has the most delegates at stake on both sides of the race, but it may offer the least drama of the night on both sides. Hillary Clinton and Donald Trump are both poised to win by wide margins, unsurprising since Florida contains lots of their key constituencies (African Americans and senior citizens for Clinton, and, well, even more senior citizens for Trump). Florida has closed primaries (in which indies and temporary party-switchers can't participate), which makes the electorates easier for pollsters to model. And, unlike the Midwestern states, we aren’t seeing a wide variation in margins that tell us the pollsters are all over the map; instead, the results are all tightly clustered. For instance, there were 11 polls of the Democratic side taken in March; with the exception of one pro-Clinton outlier, all of Clinton’s results were right in the 57 to 64 range.
On the Democratic side, delegates in each category are awarded proportionately. There is a 15 percent viability threshold, which obviously isn’t going to trip anyone up at the state level, but could pose a problem for Sanders in the congressional district that has the most delegates of any: The mostly African-American FL-24, which has eight delegates at stake. (It’s worth noting, though, that many of the black residents in the 24th are Haitian or from other Caribbean islands, who may perform differently than, say, the residents of AL-07 or MS-02.) If there’s somewhere Sanders might come close, keep an eye on FL-02 and FL-03, where the state’s two major universities are located.
On the Republican side, things are quite simple: It’s the nation’s largest winner-take-all jurisdiction. (California, as we’ll see in June, is winner-take-all at the state level, but most of its GOP delegates are awarded at the CD level.) Whoever gets the most votes statewide wins all 99 of the state’s delegates. And seeing how that in all likelihood won’t be Marco Rubio, for whom Florida was his last hope of staying relevant, Tuesday will probably spell the end of his campaign.
ILLINOIS
Democratic delegates: 34 at-large, 20 PLEOs, between four and nine delegates in each of the state’s 18 CDs (156 total)
Democratic polls: 49 Clinton, 41 Sanders (aggregate)
Republican delegates: 12 at-large, 3 RNC, 3 delegates in each of the state’s 18 CDs (69 total)
Republican polls: 36 Trump, 25 Cruz, Kasich 19, Rubio 13 (aggregate)
A few days ago, we were planning to write about how Ohio was the pivotal state on the Democratic side. Ohio was seeing much closer polls of the Clinton/Sanders race than Illinois, and Ohio is also demographically similar to Michigan (Ohio and Michigan are each other’s most similar states, according to Nate Silver’s “state similarity index”), where, as we all know, Bernie Sanders dramatically over-performed against the polls two weeks ago.
Well, things have a way of changing quickly ... especially in an open primary state like Illinois, where independent voters introduce a layer of unpredictability ... and in the last few days we’ve seen a spate of polls showing a very close race in the Land of Lincoln (a two-point lead for Sanders in a CBS/YouGov poll, plus single-digit Clinton leads from several other credible pollsters like PPP and Marist), while things seem to have stabilized more in Ohio (not that Clinton is showing an impermeable lead there, but no poll in March has seen less than a 5-point lead for her). If you thought only about demographics, that turn of events might be surprising; Illinois is a much less-white state than Ohio, with larger African American and Hispanic populations (though those are both confined mostly to the Chicago area), and Alan Abramowitz’s demographics-only projection sees Clinton winning Illinois but Sanders narrowly winning Ohio.
Instead, this may be one of those cases of effective advertising turning the tide in a race. Sanders has been running ads that are sharply critical of Chicago's unpopular mayor Rahm Emanuel, and tying him to Clinton. While it's frustratingly hard to get real-time information on ad spending, there have been anecdotal reports that Sanders is outspending Clinton on the local airwaves, much as he did in the run-up to the vote in Michigan.
It also helps that Sanders' key local ally is Chuy Garcia, who lost the most recent mayoral race to Emanuel. That, in particular, may be helping Sanders make some inroads into Chicago’s Hispanic communities (which are centered in IL-04). The real question that may make the difference whether or not he gets over the top in Illinois is whether he’s also making some inroads into Chicago’s African-American communities (for those, watch IL-01, IL-02, and IL-07). It’s unlikely Sanders will win or even come close there, but if he can get, say, 25 to 30 percent of their vote, like he did in Detroit, that might be enough to eke out the win statewide. (Unfortunately, it’ll be hard to tell from the results whether that’s happening or not, as they report. Illinois will report at the county level, and Cook County contains a diverse assortment of over 5 million people, with no way to parse out individual neighborhoods.)
For the Democrats, everything is awarded proportionately; there is a 15 percent threshold for delegates at both the statewide level and in each CD. Even if Sanders doesn’t get over the top, he can probably keep things fairly close at the delegate level by winning many of the state’s more rural CDs by a 3-2 margin, or at least hitting the 2-2 tie in the ones with 4 delegates, much the way he did in most of rural Michigan. It’s also worth watching the wealthier suburban districts like IL-06 and IL-10; while these aren’t laden with delegates, they’re likely to be solid for Clinton (as she tends to perform well in areas with household incomes over $100k) … but if Sanders can keep DuPage and Lake Counties competitive the way he did in Oakland County, Michigan, that’ll blunt another avenue for Clinton.
On the Republican side, it’s a sort-of winner-take-all primary. The 15 statewide delegates (12 at-large plus three RNC delegates) are bound winner-take-all to the person who wins the most delegates statewide (who, at this point, looks to be Donald Trump). The other 54 delegates are awarded at the CD level, though, and in each CD, they use a “direct election” method; voters vote directly for their delegates (who have their candidate preferences listed next to them on the ballot). Trump seems likely to get the lion’s share that way, too, but bear in mind there are some regional variations within the state. Cruz may do well in the southern parts of the state; the southern half of the state, like IL-12 and IL-15, is actually fairly evangelical. And Kasich, who seems to be carrying the tattered remains of the Never Trump banner, seems poised to do well in those same establishment-friendly affluent suburban districts like the 6th and 10th. The only candidate likely to walk away with empty pockets looks to be Rubio.
MISSOURI
Democratic delegates: 15 at-large, 9 PLEOs, between four and 10 delegates in each of the state’s 8 CDs (71 total)
Democratic polls: 2 March polls: 47 Clinton, 40 Sanders, and 47 Sanders, 46 Clinton
Republican delegates: 12 statewide, five delegates in each of the state’s eight CDs (52 total)
Republican polls: 1 March poll: 36 Trump, 29 Cruz, 9 Rubio, 8 Kasich
Missouri is probably the state we know the least about of the five states up for grabs. That’s partly because it’s kind of a weird state that's hard to pigeonhole culturally (it's too midwestern to be a southern state, but too southern to be a midwestern state). According to the state similarity index, its most similar state is Ohio, followed by North Carolina (which doesn't help), and based on how its neighbors (Iowa and Oklahoma) have voted, well, all we know from them is that it's on track to be very close. Alan Abramowitz’s model gives Missouri a slight Sanders edge, much the same as Ohio; that’s probably because Missouri has a very similar non-white percentage of the population as Ohio (both were around 83 percent white in 2010). You might think of the large African American populations in St. Louis and Kansas City, but the state’s rural areas (except for the Bootheel in the southeast) are very white, almost at Appalachia or upper New England levels of whiteness.
Mostly we're unsure, though, because pollsters have stayed far, far away from Missouri. The only two polls we’ve seen on the Democratic side are from PPP, who give Sanders a 1 point lead, and Fort Hays State University. FHSU give Clinton a decent lead, but bear in mind they also gave Clinton a similar-sized lead in Kansas, which turned out quite wrong (though, in their defense, it’s even harder to poll a caucus than an open primary like Missouri).
At any rate, between the few polls and all the demographic evidence, it seems like Missouri is Sanders’ likeliest place for a win on Tuesday. Like the other states, delegates are awarded proportionately, subject to a 15 percent viability threshold, at both the state and CD levels. Look for Clinton to do well in the urban MO-01 and MO-05, but Sanders to win the rural portions, especially MO-04, where the University of Missouri is. The pivot point of the whole state may be suburban St. Louis County, which is mostly in MO-02. (St. Louis itself is an independent city, which will report separately from the county. Bear in mind, though, that the county has an increasingly large African-American population; Ferguson, for instance, is located here.)
On the Republican side, Missouri can be a winner-take-all state; if any candidate tops 50 percent statewide, he gets all the 52 delegates immediately. If that doesn’t happen (and Trump doesn't look like he’s polling in that zone), then the 12 statewide delegates are awarded winner-take-all to the statewide winner, and each district’s 5 delegates are awarded winner-take-all to the district’s winner. Cruz might be able to hold his own in Missouri, at least at the CD level; look for him to possibly win in MO-07 and MO-08, the most heavily evangelical portions of the state.
NORTH CAROLINA
Democratic delegates: 23 at-large, 14 PLEOs, between four and eight delegates in each of the state’s 14 CDs (107 total)
Democratic polls: 55 Clinton, 35 Sanders (aggregate)
Republican delegates: 72 statewide
Republican polls: 39 Trump, 25 Cruz, 14 Rubio, 11 Kasich (aggregate)
North Carolina might wind up falling through the cracks on Tuesday night; it’ll probably get called quickly for both Clinton and Trump, it isn’t as delegate-rich as Florida, and it isn’t as competitive as the Midwestern states. It might be interesting to speculate why North Carolina, which has a larger percentage of African Americans than Florida, isn’t performing quite as strongly for Clinton as Florida is (is it because Florida has so many senior citizens? or because North Carolina has so many college students?), but either way, North Carolina, along with Florida, will help Clinton along to huge net gains on the night even if Sanders manages to squeak out multiple wins in the Midwestern states.
On the Democratic side, the delegates are awarded proportionately at both the statewide and district levels, subject to a 15 percent viability threshold. The state’s two majority African-American districts, NC-01 and NC-12, will go strongly for Clinton, but there are probably enough college students and urban hipsters in those districts to help Sanders win a few delegates there. The state’s other Democratic vote sink, NC-04, encompasses both Chapel Hill and Raleigh, and may actually go fairly well for Sanders (probably not to the extent of a win, but possibly a 4-4 split).
On the Republican side, North Carolina is the one state where everybody (even Marco Rubio) gets a participation ribbon. All 72 delegates are awarded statewide, and they’re all awarded proportionately, without any viability threshold. Someone could do as poorly as 1.39 percent of the vote and still walk away with a delegate!
OHIO
Democratic delegates: 31 at-large, 19 PLEOs, between four and 17 delegates in each of the state’s 16 CDs (143 total)
Democratic polls: 53 Clinton, 42 Sanders (aggregate)
Republican delegates: 66 statewide
Republican polls: 40 Kasich, 37 Trump, 15 Cruz, 3 Rubio (aggregate)
As we said earlier, Ohio was originally going to be the linchpin of the night for the Democrats, but with the rise of Illinois—now, not so much. Unlike the rapid convergence in the Land of Lincoln, the trendlines in Illinois have actually been rather flat, and there have been about an equal number of polls in March on the Dem side with double-digit Clinton leads as there were with single-digit leads. (The only poll with a Sanders lead, at any point, was from early February, from Baldwin-Wallace University, giving him a one-point lead. That will either be an outlier, or be prescient.)
The main things that Sanders supporters in Ohio can hang their hats on are, one, Alan Abramowitz’s demographics model points to a narrow Sanders win, largely as a result of its mostly-white population. And two, they can hope that Ohio polls follow the same path as Michigan’s polls two weeks ago. (Again, Ohio and Michigan are very demographically similar; the state similarity index says they’re each other’s most similar states.) Also, Ohio is a “semi-closed” primary, meaning that independents will be able to vote in either party’s primary, so that adds an element of uncertainty and makes it less likely the pollsters are modeling the electorate correctly. (Though, bear in mind, many indies may be taking GOP ballots to try to power local favorite John Kasich past Donald Trump.)
Is there much evidence that Ohio is on track to be as wrong as Michigan, though, other than its sort-of-open status and just the way that everyone is gun-shy about the polls now? That doesn’t seem too likely; in Michigan, the exit polls didn’t show a lot of late deciders or any other element of chaos. Instead, the pollsters were, for the most part, just misjudging who was and wasn’t in the Dem electorate.
Michigan is notorious for having a number of lousy local pollsters (in 2012, for instance, Michigan’s polling average undersold Barack Obama’s performance by over 5 percent; that’s a bigger gap than most other swing states). Polls were off on the Republican side too; it just wasn’t as newsworthy because it didn't affect the ultimate result. Local pollster Mitchell, in particular, is especially bad. Their last poll, which put Clinton up by 28 and skewed the overall aggregate, was based on a sample where 86 percent of the likely voters were over age 50! It’s worth noting that Ohio’s polls, instead, are almost entirely by taken by nationwide firms.
Let's turn to the Democratic delegates. They’re allocated proportionately at both the statewide and district level, subject to a 15 percent viability threshold. As we saw in Michigan, look for Sanders to make up a lot of ground in the rural areas, though note that almost all of the rural CDs have four delegates instead of five, so he might wind up settling for 2-2 splits even where he narrowly wins. And, unusually, there are some extremely large delegate hauls in the few Democrat-heavy districts: there are 12 up for grabs in Columbus’s OH-03, and a whopping 17 in Cleveland’s OH-11, which looks like the most we’ve seen in any district anywhere. Sanders may get a good chunk of the vote in OH-03 in Franklin County (which is where Ohio St. University is), but the African American majority 11th in Cuyahoga County is where Clinton will be running up the score.
And, as we said at the outset, things are simple on the Republican side: the winner truly takes all. There are 66 statewide delegates; whoever gets the most votes, even if it’s by a margin of only a few hundred, gets the whole enchilada covered with the horrible substance known as “Cincinnati chili.” With most pundits’ projections showing Donald Trump barely on track to hit 50 percent of the total delegates, if John Kasich can steal those 66 delegates, that could alter Trump’s trajectory just enough to steer him into a contested convention.
But, if you’re reading this, you’re probably more concerned with Democratic delegate math. If everything goes as expected tonight, what are the implications on the Dem side? If we take the polling averages (or, in Missouri, just take the one credible poll at all), naively extrapolate the undecideds, and then allocate delegates based on statewide percentages (which I've found is usually pretty close to the actual result, even if you skip all the fancy district-level math) the math looks like this:
Coming into Tuesday: Clinton 772 pledged delegates, Sanders 551 pledged delegates (221 lead for Clinton)
Florida: Clinton 64, Sanders 36; Clinton 137 delegates, Sanders 77 delegates
Illinois: Clinton 54, Sanders 46; Clinton 84 delegates, Sanders 72 delegates
Missouri: Sanders 51, Clinton 49; Clinton 35 delegates, Sanders 36 delegates
North Carolina: Clinton 60, Sanders 40; Clinton 64 delegates, Sanders 43 delegates
Ohio: Clinton 55, Sanders 45; Clinton 79 delegates, 64 delegates
Coming out of Tuesday: Clinton 1,171 pledged delegates, Sanders 843 pledged delegates (328 lead for Clinton)