Every four years, the same question comes up during primary season: why do Iowa and New Hampshire get so much say over who becomes the next president? Iowa and New Hampshire are relatively small states, but they go first during the presidential primary season, which lets them bestow momentum on their favorite candidates and weed out many of the others before anyone else can cast a vote.
Having a few states vote before the rest — Democrats will likely have Iowa, New Hampshire, South Carolina and Nevada vote before Super Tuesday — is actually a good way to pare down the field and let strong candidates emerge. But this schedule is flawed. Having Iowa and New Hampshire vote early is, demographically speaking, redundant. Democrats would be wise to keep Iowa (which is similar to other Midwestern states) and trade New Hampshire for a state that’s both more representative of the Northeast Corridor, as well as their increasingly urbanized party.
Putting Iowa and New Hampshire first feels — and, because it gives some voters more power than others, is — undemocratic. But there’s actually a good case to be made for some smaller states going first.
Putting small states first, rather than having one national primary day, initially levels the playing field by rewarding old-fashioned retail politics. Candidates who aren’t as well-known nationally can get off the ground by building an organization and talking to local voters and activists (remember that Ted Cruz, not Donald Trump, won Iowa in 2016). Candidates who want to run a national campaign right off the bat are forced to go local. These states also, maybe counterintuitively, end up separating the wheat from the chaff. If candidates spend months campaigning in these small states and don’t catch fire, they often realize they won’t be able to make it on a larger scale and drop out before the rest of the country votes. And if you put the right states first in line, you can force the candidates to face electorates that are representative of the party as a whole.
And that’s where things start to get dicey for Iowa and New Hampshire. Ideally, the first four states would be a microcosm of the Democratic coalition: that would help candidates who might be popular with the broader party catch fire early. But Iowa and New Hampshire are both white, liberal states. And while white liberals are an important part of the blue team, putting them both at the front of the calendar overemphasizes that group. And I’ve done some math that suggests that it might be good to trade New Hampshire, not Iowa, for another state.
This map shows the basic results of what’s known as “fuzzy c-means clustering” — basically I asked a computer to divide states into into similar groups based on their political, demographic and geographic features. I used race, income, past presidential primary vote, education level among white voters (since that was such a huge divide in the 2016 general election), population density, urbanization, and ratio of liberals to Democrats in the state to determine which states are most like each other. The map shows four “hard clusters” — that is, which group each state belongs to when it’s only allowed to fall into one of four groups.
This map also presents an obvious problem for Iowa and New Hampshire: They’re too similar. If we have four early contests, we might expect them to represent each of the four different clusters, and by extension, a distinct part of the party well. But Iowa and New Hampshire both fall into Cluster 2, a group of mostly white states in the northern half of the country where the (often smaller) cities and suburbs aren’t big enough to dominate the state’s overall politics. New Hampshire isn’t a perfect fit for this group. But there are real similarities between the Granite State and Iowa — both of which have very white, fairly liberal electorates and neither of which contain a huge metro area.
South Carolina and Nevada are more distinctive. South Carolina is part of a broad group of Southern and Appalachian states where many black voters and ancestrally Democratic blue-collar whites live. And Nevada is part of the Southwest-Virginia-Pennsylvania-Minnesota group, states where there is often a large nonwhite (Hispanic, in the case of Nevada and other Southwest states) population and where there are major cities. And Cluster 3 — the Northeast Corridor, Illinois, California and Washington state — doesn’t have an early primary state. That seems odd given how much Democrats rely on major metro areas for their electoral strategy.
These categories aren’t perfect, but they do suggest that having Iowa and New Hampshire both go early is, in some ways, redundant. And more detailed results from this method suggest that New Hampshire, not Iowa, would be the right state to swap out.
The fuzzy c-means clustering also assigns each state “fuzzy” clusters: that is, it describes each state as a combination of the categories. A state might be 50 percent Southern/Appalachian, 20 percent Central Northern, etc.
In these categories, Iowa is strongly part of that Central Northern tier of states that includes Wisconsin, Michigan, Nebraska and a number of other proximate states. New Hampshire isn’t as firmly in that group, and it doesn’t have very much in common with the Northeast Corridor/West Coast group either. New Hampshire does allow independents to have a say, but on balance, it’s mostly an oddball state that fails to represent the Central Northern group, or really any one group, effectively.
If Democrats really wanted to pick the best configuration of early primary states, they might want to trade New Hampshire for a different state in the same region. Massachusetts, Rhode Island or Connecticut would work. It also might be worth swapping Iowa for Wisconsin or Minnesota, and trying out Arizona or Texas instead of Nevada. Having four very different states vote early is a good idea, but it might be interesting to swap out some states to see what the results look like. Swapping Nevada for a more Latino-heavy state might help Democrats nominate someone who has a better ability mobilize Hispanics, and putting more major cities in the early states could make issues such as transportation more salient. It’s hard to predict what might happen, but tweaking the system a bit might produce somewhat better candidates.
But those changes are unlikely to happen. New Hampshire and Iowa have successfully defended their position against other states that have tried to jump in line, and they’ll likely continue to do so. Taking New Hampshire off the calendar would also have ripple effects across the parties. The Republican primary electorates in Iowa and New Hampshire are very different: Iowa is more evangelical/conservative, while New Hampshire is comparatively moderate, so knocking off that delicate balance may not be worth it for the GOP.
But Democrats shouldn’t just keep old rules because they’re difficult to change. The rules of the game shape the outcome — and if Iowa and New Hampshire were even a little bit more flexible and open to giving up their spots, Democrats could have a conversation about how to create the best process — and the best candidates for the country as a whole.
Read more from David Byler:
