Dem’s are rearranging the primary schedule. Possibly adding new states at the beginning. New Jersey was mentioned as a possibility. How would 2008, 2012, 2020 primaries have gone with a different order?
The New Rules
There aren’t any decided yet, but leaked documents and other random bits and pieces (according to this Washington Post article) the proposal we’ve heard about would have states apply to be one of the first five primaries, weighting diversity (“Ethnic, geographic, Union representation” according to link. It doesn’t say what geographical diversity means.), competitive in general elections, and having a primary over having a caucus. The article does not saw what the criteria are, whether they aim to match the country or get as much union and nonwhite as possible; one person quoted says that three of the four states hit at least 2 criteria (Iowa seems the odd one out), suggesting thresholds, probably as much nonwhite and union as possible since South Carolina doesn’t match the country and Nevada is heavily union. It does not say what the boundaries or expectations of these things are.
Why does a Primary Order Matter?
If people picked a candidate before primaries started and stuck with that choice, the order wouldn’t matter. Order matters because primary voters change votes based on how people do in previous states: they see a candidate do well and decide to switch to that candidate, a candidate gets news and more people here about them, a candidate drops out and supporters have to switch to someone else. To see how primaries would play out differently, you’d need to figure out how voters tended to respond to people doing well in particular states.
I can look at polling (in theory) to get some idea of this, but unfortunately it depends on how the story gets told, which I have no way of coming close to measuring. So I’m going to pull some plausible guesses out of my ass.
Another problem is people heavily investing in early states: Iowa voting in the big block of states after Super Tuesday votes differently than Iowa going first and getting lots of attention from candidates. Unfortunately, this means I have to guess how different types of people respond to this sort of attention, so its another pulling things out of my ass possibly.
What order of states? (Skip ahead if you just want to order)
Sequence 1
Since I like self promotion, I’ll look at this sequence I came up with last year:
-Delaware goes first
-Iowa, Mississippi, New Mexico, New Hampshire in a block
-Washington, Arizona, Wisconsin, Virginia, Maryland in a block
-Georgia, Illinois, New Jersey in a block
This obviously doesn’t match the actual proposal, but since the actual changes aren’t fixed yet, I might as well throw it in. This is based on starting with a small state that matches the country’s demographics apart from unions as closely as possibly, than doing blocks of states covering all regions that also match demographics of the country when put together.
Sequence 2 and 3
For these two sequences, I’ll make a list of states that seem to match the criteria, randomly pick a couple collections of 5.
-Smooth score or pass/fail for criteria: The Des Moines register article says that “three states match two of the criteria”, which suggest states either meet it or don’t. So I’ll pick cutoffs.
-Racial Diversity: South Carolina and Nevada don’t match the country’s composition that great, though they may cross some measurement if this is used. Gut tells my they mean more nonwhite, so I marked states as meeting this if they are less than 61.5% non-hispanic white, based on this data. The actual estimate for the total country was 57.8%, but this would cut out a lot of states and leave a smaller pool, so I went with 61% to include Illinois (by almost all measures the most representative state.)
-Unions: The more unions, the better, I’m assuming, so I picked a cutoff of 10% union membership from this data. This is slightly less than the country as a whole, but the Des Moines link says new Hampshire hits 2 out of 2, and it is slightly less than the whole country, so I lowered the cutoff to include it.
Competitive: There are lots of elections and time periods you could chose to estimate this. For 2020, Iowa at about 8% and New Hampshire at 7.4% are quite close, so probably other elections are included. just to keep things simple, I went with “Margin < 8% in either direction in the 2020 election”.
The excel file with all this is here. States that meet at least 2 of the criteria are AZ, AL, CA, FL, GA, HI, MD, MI, MN, NJ, NH, NV, NY, PA, TX. I than randomized the list a bunch of times, and picked a couple orders with states from all 4 census regions.
Order 1:
Nevada, Hawaii, Georgia, New York, Michigan
Order 2:
Texas, Arizona, Hawaii, Minnesota, Alaska
Note 1: Nevada is the only state that meets all three criteria. My criteria might be off, or Harry Reid’s ghost is talking to people in this primary process.
Note 2: I sometimes joke about Hawaii being the first state, so that political people can get some vacation time, but as one of the most unionized and least white states in the country it is a legitimate pick in this system. Might want to use some renewable aviation fuel, though. (Or build a mobile floating campaign ship.)
The Three Sequences
Pickly Primary:
-Delware
-Iowa, New Hampshire, Mississippi, New Mexico
-Arizona, Washington, Wisconsin, Maryland, Virginia
-Illinois, Georgia, New Jersey
Apples, Peaches, and Pineapples Pimary
-Nevada
-Hawaii
-Georgia
-New York
-Michigan
Wheeling Weather Primary
-Texas
-Arizona
-Maryland
-Minnesota
-Alaska
2004: First Woman, first Black man, and some guy who cheated on his wife
The 2008 primary was a very close thing, both in the popular vote and the number of delegates, with Obama coming out very slightly ahead in the popular vote, and ahead about 51/49 in pledged delegates. Any small change could conceivably have effected the result, assuming superdelegates shift to follow the winner. (Superdelegates aren’t really a factor for future primaries, don’t apply to states, and I’d have little way to analyze them anyway, so I’m leaving them out.)
Most polling suggests that more people supported Obama as the first few primaries approached, while Clinton’s support decreased a bit or stayed about the same. A common story is that black voters in particular were seeing if other people liked Obama, though I’ve also seen polls that south Carolina had mostly shifted to its final margin before any other primaries.
Demographically, the older the more Clinton, Latinos went more for clinton, Black went very strongly for Obama.
The big question in this primary is, how does Obama do in early states, and do other people shift to support him as a result?
Pickly Primary
-Delaware:
Delaware’s obscure senator Joseph Biden was running in this primary (anyone remember?). We haven’t had an early state voting for their own politician in awhile: its possible as a result the state would strongly go for its home senator, or that elections are nationalized and he gets more votes than otherwise, but still below the big 2 or 3 candidates. In recent Republican primaries, home states did vote for Kasich, Huckabee, and Gingrich, plus McCain who won anyway, but not for Rubio, Santorum, or a number of candidates who didn’t even get votes. likely, even if Delaware goes Biden, almost all other states vote for the main candidates, as fits the pattern with many home state candidates without much other support.
Delaware is about a fifth black, (compared to about 18-1/10 or so for the whole country), white voters went Clinton and Black voters went strongly Obama, with the state as a whole going Obama. If black voters are seeing if everyone else likes Obama, Delaware’s 2008 white vote isn’t the state to change this, however, if Obama puts a lot more resources into the state, it possibly votes for him anyway and gets things off to a good start.
-Iowa, New Hampshire, New Mexico, Mississippi
This is a mix of two Clinton states and two Obama states in 2008, Edwards would still quite possibly be around during this race unlike New Mexico and Mississippi 2008, though I don’t have a strong sense whether he’d change any results. The most likely changes are Obama getting no individual state to himself like Iowa 2008, and possibly New Mexico shifting if given more early attention. if it does, Obama likely looks stronger and gets some buzz, if not, Obama possibly looks weaker to a lot of people and they vote for him less.
-Washington, Arizona, Wisconsin, Maryland, Virginia
Four strong Obama states, only Arizona went narrowly Clinton. If black voters were seeing if Obama gets nonblack votes, Wisconsin and Washing probably do it. Even if not, Obama probably gets a big boost in attention/buzz from this primary.
-Illinois, New Jersey, Georgia
One pretty good Clinton state, one strong Obama state, one strong Obama home state. These states probably stay the same in this alternate primary. Illinois might be an obvious home state, but these others still tell a story of an even race.
-Overall:
Two main changes here: Obama gets a slightly later boost if one is needed, and Michigan and Florida have actual races since in my theoretical primary the state schedules get much more top down organized. The slightly later boost likely only effects the first few blocks since the Washington-Virginia group is so strongly Obama, all states after it play out similar to 2008 if it has an effect.
For Florida and Michigan, the results were halved, and candidates didn’t campaign that hard. The states theoretically have more than enough delegates to changed who gets more pledged delegates, but not enough at the margins actually received. It’s possible actual campaigning might get bigger margins, though I have no reason to know for sure.
Best Guess: Using this order, election is a bit closer, but still goes Obama. Very slight chance it goes narrow Clinton.
Fruit Primary
-Nevada
Narrow Clinton state. As another early state, probably stays the same, maybe is a bit more Clinton in this alternate primary. If Obama thinks it is really important, he may put more attention on it.
-Hawaii
Strong Obama state in 2008. The big question here is whether it’s treated as “Yea, yea, birth state likes the guy, what about the real elections” or “wow, big Obama win!”.
-Georgia
Strongly Obama, probably still strong Obama in this version. How strong depends on how Hawaii gets reported/told and whether some black voters were really waiting to see how everyone else likes Obama.
-New York
Biggest Clinton margin early on
-Michigan
Gets to vote for real in this version. If like thelower turnout actual 2008 election, is a pretty solid Clinton state.
-Overall
This early lineup is Clinton leaning, and the strong Obama states are his home state and a heavily black state. Good chance of a weaker Obama storyline, if so a small percent shift in the Obama vote seems plausible. Anyone want to take up a bet and test this?*
(Yes, sarcasm, obviously.)
Weather Primary
-Texas
Narrow Clinton state, close enough that Obama plausibly gets more votes with intense campaigning.
-Arizona
Similar deal to Texas, as a smaller state might(?) be better to start Obama off.
-Maryland
Strong Obama state.
-Minnesota
Another strong Obama state, not heavily black so might be better for story purposes.
-Alaska
Strong Obama state like Minnesota.
Summary of all Primaries
Even though its a close election and a different storyline leading to slightly different votes can easily shift the very close race, there are enough strongly Obama states from 2008 to get things going in most orders. This is true even if heavily black states are assumed to go Obama anyway, or if a big enough chunk of black voters is waiting for votes from nonblack states to show strength.
2016? 2020?
See Part 2.