For most of my career as a political scientist, representation was the central theme of my teaching and research. Early on, I studied how legislators understood their responsibilities to constituents — whether they saw themselves as delegates carrying out constituent mandates, trustees exercising independent judgment, or politicos combining both. That work asked a question that still matters: What does it mean to represent the public in a changing democracy?
Today I find myself asking a related question: What happens when the electorate changes but the institutions built to represent it do not?
Across the country, voters registered with neither major party are becoming the largest single bloc in state after state. In Massachusetts, roughly two-thirds of registered voters are now "Unenrolled," up from just over half in 2020. In Alaska, more than 60% of voters are registered outside the two major parties. In Colorado, unaffiliated registration has passed the halfway mark. In Oregon, where I live, non-affiliated voters became the state's largest registration category years ago.
This is not simply a story about voters souring on parties. It's also, in part, a story about how states register people to vote.
Since Oregon passed the nation's first automatic voter registration law in 2015, roughly half of U.S. states and the District of Columbia have followed with some version of it. Eligible citizens who interact with the DMV are registered automatically unless they take a separate action to opt out or select a party. In "back-end" opt-out states like Oregon, Colorado, and Alaska, that separate action means filling out and mailing back a postcard — a step a great many people simply never take. The default in these states isn't "unregistered" or "pick a party." It's non-affiliated.
That's not the whole story. North Carolina has one of the largest non-affiliated blocs in the country without any automatic registration system at all, a reminder that plenty of voters are choosing independence on their own, not just defaulting into it. But where automatic registration and back-end opt-out do overlap, the mechanism is direct and measurable, and it's not a criticism of automatic registration itself, which has reliably increased registration rates, particularly among younger and lower-income voters who often fell through the cracks of the older opt-in system. It's an observation about consequences: when a state converts registration from something citizens must actively pursue into something that happens automatically, it also converts non-affiliation from an active choice into a passive default in a meaningful share of cases.
Those consequences show up first in primaries. In states with closed primaries, a rapidly growing share of the electorate is registered in a way that excludes it from choosing party nominees at all, which is a significant reason open-primary campaigns have gained traction in Oregon, Nevada, and elsewhere in recent years.
They also show up in redistricting, where the assumptions behind decades of reform were built for a different electorate. The modern redistricting toolkit — from the Supreme Court's "one person, one vote" line of cases to independent commissions and sophisticated fairness metrics — was developed almost entirely for a political world in which Democrats and Republicans defined the electorate. Districts are not created to represent political parties, though; they're created to represent people. When the largest single group of people in a jurisdiction isn't affiliated with either party, and when many local offices are formally nonpartisan, the two-party lens that redistricting reform was built around becomes a less complete description of who's actually being represented.
A few places are already experimenting with something different. Portland, Oregon, created an independent citizen commission — barred from including party officials or sitting politicians — to draw the city's new council districts, requiring broad supermajority approval and extensive public input before a map could be adopted. Michigan voters did something similar statewide in 2018, creating an independent citizens commission through ballot initiative that likewise bars sitting politicians and party officials from serving. California and Arizona have their own independent redistricting commissions, each built, in part, on the premise that districts drawn by elected officials answering to party organizations aren't the only, or necessarily the best, way to do this.
Not every state has kept pace. Rhode Island's non-affiliated voters are close to half its electorate, yet the state legislature still draws its district maps — a case of a state that modernized its front door to the ballot without modernizing who draws the lines behind it.
None of these experiments proves that independent commissions are the answer everywhere, or that traditional districting approaches have failed. What they demonstrate is more basic: institutions built for a two-party electorate can be redesigned when that electorate changes, and several states are already testing how.
That's worth taking seriously well beyond the states running the experiments. Nearly half the country has now adopted some version of automatic registration, and in the states where that design pairs with a back-end opt-out, it is structurally producing non-affiliated majorities or pluralities. This isn't a regional curiosity. It's a preview of where other states may be headed as more of them adopt similar systems.
None of this means political parties are disappearing, that primaries should be abolished, or that every state needs an independent commission. Those are separate, harder questions, each deserving its own debate. The more basic question comes first: if the electorate a state's institutions were built to represent has changed, should the state's understanding of representation change with it?
After a career spent studying how elected officials represent their constituents, I've come to think that's a question more states should be willing to ask — not because the old institutions have failed, but because every generation inherits institutions designed for an earlier electorate, and periodically it's worth checking whether they still fit the people they're meant to serve.
Kent C. Price is a political scientist and applied public opinion researcher living in Central Oregon. A retired professor from California State University, Bakersfield, he spent more than 30 years teaching voting behavior, elections, and democratic institutions. Over his career, he conducted extensive survey research and advised public agencies on governance, elections, and public policy. His early scholarship focused on theories of legislative representation, including “Instability in Representational Role Orientation in a State Legislature” (1985).
