How the Republican Race Plays Out From Here
Obviously, Tuesday night was a good one for Donald Trump. He won over 60 percent of the vote in the New York primary, and carried all but four of the state’s 95 delegates. His main rival, Ted Cruz, finished third.
But lost in the hoopla is that a big Trump win in New York was already baked into most projections. While few prognosticators had him at 60 percent of the vote, most had him winning 70 of the state’s delegates or more. So from a delegate perspective, not much has changed.
Where do we go from here? It’s hard to say, but I do think there are some broad takeaways. I’m still skeptical that Trump will get the 1,237 delegates he needs to be the nominee (as I said, last night’s win was baked in). But this skepticism comes with some important caveats.
First, we’re entering a spate of states where competitive Republican presidential primaries are rare, where pollsters don’t have a lot of experience, and where in one instance (Indiana), much polling is banned by state telemarketing laws. We’re flying blind to a large degree.
Second, a lot still depends on what John Kasich decides to do. The Ohio governor, who still trails Marco Rubio in the delegate count, almost certainly splits the anti-Trump vote. While it’s true that both he and Trump draw from moderate voters, it is a good reminder that the “center” is actually a collection of points off the poles, and not all “centers” are created equal.
Suffice it to say that Kasich’s largely upper-middle-class, largely college-educated moderates are of a different sort than Trump’s working-class, high school-educated moderates. We can see this in the New York exit polling, where only 33 percent of Kasich supporters said that they would definitely not vote for Cruz if he were the nominee, while 72 percent of his voters said that they would not support Trump if he were the nominee. Many would ultimately stay home in a Cruz/Trump primary, but of those who voted, more would probably go to Cruz than to Trump.
In any event, let’s start with where things stand. Delegate counts are difficult to keep track of, but Trump probably has around 850 delegates in his camp right now. The next round of states -- Connecticut, Delaware, Maryland, Pennsylvania and Rhode Island -- are likewise going to be good states for him. While he’ll probably bleed some delegates in Maryland, where Kasich has strength, and Rhode Island, which allocates delegates proportionally, he’ll still pick up another 100 delegates or so on April 26.
At that point, Trump will only be at around 950 delegates, about 285 shy of 1,237. He’ll need to win 58 percent of the remaining pledged delegates to be guaranteed the nomination.
The next race is Indiana, which is a bit of a question mark. While I don’t think it is “Wisconsin-bad” for Trump, it is “Illinois-bad” for him. But didn’t Trump win Illinois? Well, it depends which Illinois you’re talking about.
There are a lot of ways to look at that state, but the following chart utilizes an approach I find useful. For now, just focus on the two left columns, which show Trump’s margin over his closest competitor, by congressional district:
What you’ll notice is that Trump did quite well in the districts with low numbers, but not nearly as well in districts with high numbers. The pattern becomes clearer if you’re familiar with basic Chicago geography, and the structure of Trump’s coalition.
It shouldn’t be a surprise that he ran particularly well in the 3rd Congressional District. This is the non-gentrified remnant of the white ethnic base of Cook County, an area represented over the years by politicians with names like Lipinski, Gorski, Russo, Kluczynski, and Derwinski. Trump’s strength among white working-class voters is well established, and should not surprise us here.
But what about his strength in the 1st (a majority black district) and the 4th (a majority Hispanic district)? Remember, this is a Republican primary we’re examining. These districts both border the 3rd, and a Republican primary electorate probably consists of working-class whites who were brought in to help the district meet the equal population requirement. Trump also did well in the 11th and 14th, which are gerrymandered suburban districts that actually have a lot of descendants of these white ethnic voters. Finally, he did well in the 12th, which is part of the Appalachian diaspora.
On the other hand, Trump performed poorly in the 7th district, which in a Republican primary is probably made up of relatively liberal whites in the downtown area, and had mediocre performances in the 5th, 9th and 10th, which are combinations of silk-stocking voters and working-class whites further inland.
But what’s of most interest to us are Districts 13 and 15 through 18. These represent the non-Chicagoland portions of the state. And with the exception of a one-point win in the 15th, Trump lost all of these districts to Ted Cruz. In fact, the 18th was his worst district in the state.
If we allocate Rubio’s vote 20 percent to Trump, 40 percent to Cruz and 40 percent to Kasich, you see the same tendencies amplified. These are shown in the two right-hand columns:
What does this have to do with Indiana? Indiana is a unique Great Lakes state. Immigration in the United States ran almost exclusively east to west until the latter half of the 20th century, mostly following well-established trails, and then rail lines. Immigrants moved from Europe to New York City before moving westward along these pathways, settling in the industrial centers that sprung up near the Great Lakes. During the New Deal, cities with large white ethnic populations like Erie, Cleveland, Toledo and Chicago took on a distinct class cleavage that layered a new type of politics over these states’ historic Civil War-era cleavages. In the postwar era, the influx of African-Americans to these cities added a racial dimension to their politics.
Indiana was different. It had only a slim slice of lakefront, and never developed a major industrial metropolis outside of the Hammond/Gary areas on that lakefront. It didn’t attract large numbers of European immigrants, and so it never developed a deep New Deal politics; it was one of a handful of states to vote against Franklin Roosevelt twice. It also, outside of the Gary/Hammond area, never developed a heavily racialized post-New Deal politics, at least to the extent that we saw in other major cities during the 1960s, ’70s and ’80s.
So when we think about Indiana, there just aren’t many analogues to the places where Trump ran particularly well. I suspect Trump will run well in the 1st district, which contains Gary, and potentially in the 8th, which contains Evansville and Terre Haute. The 9th is ancestrally Southern and could be friendly to Trump, but it contains the Louisville suburbs and Bloomington; it could likewise be Kasich country. That leaves a lot of landscape that resembles the areas of Illinois where Trump didn’t run particularly well, or that includes Indianapolis and its suburbs, which is still something of a bastion of old-line, Main Street Republicanism.
Reasonable minds can differ on how this state plays out. The real wild card is that Kasich probably remains in the race this far, splitting the anti-Trump vote badly. This didn’t happen in Wisconsin, but Kasich is a reasonably popular governor of a neighboring state. On the low end, I could see Trump winning just six delegates, from the 1st and 8th. On the high end, we might call it 40 delegates. That would leave Trump needing 64 percent of the delegates in the former scenario, and 56 percent in the latter.
A week later, we have West Virginia and Nebraska, which are likely to be Trump country and Cruz country, respectively. Screwy delegate rules may hurt Trump in the former, but let’s give him all of the delegates from West Virginia, and none from winner-take-all Nebraska.
We then come to Washington and Oregon. These states have tended to nominate relatively establishment-friendly candidates for various offices, but it is hard to say how this particular race plays out. The specifics don’t matter so much, since they have relatively few delegates and they are awarded proportionately. Let’s give Trump 35 percent on the low end and 50 percent on the high end.
This takes us to the very last day of voting, with Trump needing 60 percent of the delegates in his best-case scenario, and 75 percent in his worst-case scenario. Fortunately for him, he is far-and-away the favorite in winner-take-all New Jersey. Let’s give him 40 percent in his worst-case scenario in proportional New Mexico, and 50 percent in his best case. Winner-take-all South Dakota is an unlikely Trump state, given how he has fared in neighboring states. I think Montana is likewise Trump-unfriendly, but let’s give him a surprise upset (it is winner-take-all) in our best-case scenario.
This leaves us with California. And at this point, our assumptions really matter. He’s at 1,072 delegates in our worst-case scenario, which means that he would need to win 96 percent of the delegates to reach 1,237. In our best-case scenario, he’s at 1,146 delegates, meaning he needs around half of the state’s delegates.
What will happen in California? That is a very good question. California allocates almost all of its delegates by congressional district, and it is a closed Republican primary. Half of California’s districts gave President Obama almost two-thirds of the vote or more, meaning that there are relatively few districts with large numbers of Republicans; the statewide vote will be heavily clustered in a few districts, meaning statewide polling is of little use. Moreover, there are relatively few white ethnics (there’s no real Staten Island of California).
So who is voting in Nancy Pelosi’s district in a closed Republican primary? I have no earthly idea. One would suspect these are Kasich’s voters, given how he performed in the Chicago and New York silk-stocking districts, but it’s impossible to know. Who votes in a Republican primary in the minority-majority districts in Los Angeles, or the Central Valley?
One way to get at this conundrum is to look at the current RCP Averages, which have Trump at 40 percent of the vote in the state. If Trump had earned 40 percent of the vote in New York, with his remaining 21 percent divvied up evenly between Cruz and Kasich, he’d have won 59 percent of the congressional districts; enough under his best-case scenario, but short under his worst case scenario.
Of course, even this is imperfect. As we noted, he probably doesn’t have the strong concentrations of voters that we saw on Staten Island and Long Island, which immunized him from strong swings. The recent Field Poll, for example, has Trump trailing Cruz by 11 points in Los Angeles County (14 districts), by nine in the Sierra/Central Valley districts (12 districts), while leading by seven in the San Francisco area (11 districts) and by 22 in the remainder of Southern California (16 districts). That’s closer to a 50/50 split. We also have to consider the possibility that Kasich could drop out, and also have to factor in that the most recent poll showed Trump close to 50 percent, suggesting he could have some momentum.
In other words, we really don’t know how this plays out. Trump’s worst-case scenario probably leaves him with around 1,158 pledged delegates (assuming Kasich stays in), while his best-case scenario probably leaves him with 1,283. We also don’t know what the 130 or so unbound delegates will do; at least some of them would vote for Trump on the first ballot.
Regardless, this race seems destined to be decided by Republicans who vote in a closed GOP primary in places like Berkeley, Compton, and Watts, Calif. Which is a fitting end for what can only be described as a screwball process.