Legislative Districts

Most state legislators are elected from single-member districts: one geographic area, one seat, one winner. But across ten states, at least one chamber uses multi-member districts, where a single district sends two or more representatives to the capitol. How voters actually pick those representatives varies more than you might expect. The practical consequences are real: bloc-voting districts often produce same-party sweeps and make primary challenges harder to target, while position-based districts behave more like single-member districts for incumbent challenges and strategic recruitment.

Single-Member vs Multi-Member Districts

The single-member district (SMD) is the default in American legislative politics today. The state is carved into districts equal to the number of seats in the chamber, each district elects one legislator, and the candidate with the most votes wins. All 50 state senates use SMDs. So do the lower houses of 41 states.

A multi-member district (MMD) covers a larger area and sends multiple representatives to the chamber. If a state has 60 representatives and 30 districts, each district elects two members. Voters in an MMD are represented by a small delegation rather than a single legislator.

Ten states use MMDs in at least one legislative chamber today: Arizona, Idaho, Maryland, New Hampshire, New Jersey, North Dakota, South Dakota, Vermont, Washington, and West Virginia. Several other states used MMDs historically and converted to single-member districts after the Voting Rights Act and the one-person-one-vote line of Supreme Court cases in the 1960s and 70s.

The Three Variants of MMD Voting

Calling a chamber "multi-member" only tells you the seat-to-district ratio. It does not tell you how voters actually pick their legislators. That is determined by the voting method, and across the ten MMD states there are three distinct approaches.

Method	How voters cast ballots	Where used
Bloc voting	Voters cast as many votes as the district has seats, choosing from a single at-large pool of candidates. The candidates with the most votes fill the seats.	AZ House, MD HoD, ND House, NH House, NJ Assembly, SD House (most), VT House & Senate
Position-based	Each seat is its own race on the same ballot. Candidates file for a specific position; voters cast one vote per seat.	ID House, WA House
Position-based, staggered	Each seat is its own race, but the seats appear on the ballot in different years.	WV Senate

States can also subdistrict a multi-member district into smaller pieces. Maryland subdistricts some of its three-member House of Delegates districts into one-member or two-member sub-pieces. South Dakota does the same for its Districts 26 and 28. Subdistricting sits one level above the voting method (covered below).

Bloc Voting

Bloc voting is the most common MMD method. It is sometimes called plurality-at-large or multiple non-transferable vote. In a two-seat bloc district, every voter gets two votes that they cast for any two of the candidates running. The two candidates with the most votes win. There is no requirement that voters split their two votes between parties; a voter can cast both votes for the same party's candidates, or one for each party, or write someone in. New Jersey's General Assembly describes this on the official statehouse page as "plurality block voting."

The candidate names appear in one combined list. There is no "Seat A" and "Seat B" on the ballot in a bloc-voting district. When you see seat designators in our database or in other trackers for these districts, they are an external tracking convention, not a designation the state uses on the ballot. They exist so that vacancies can be assigned to a specific seat for succession purposes and so that records can track who has held each seat over time. (Position-based MMDs are different: as we'll see below, those seat or position labels are real, ballot-level designations the state uses to keep the two races distinct.)

Bloc voting tends to amplify party slate effects. In a heavily partisan district, the two top vote-getters are typically the two candidates from the dominant party, which is why uncontested same-party generals appear in places like rural Arizona, rural North Dakota, and parts of New Hampshire: only one party seriously contests the district, and both seats go to that party's two nominees.

Many municipal elections in New England (selectboards, school committees) use bloc voting, so for legislators from those states it is the familiar election method even outside the statehouse.

Position-Based

Position-based MMDs split the multi-member seats into separate concurrent races. Idaho House districts elect two representatives, but each candidate files for either Seat A or Seat B (Idaho calls them "Position 1" and "Position 2"), and the two seats run as independent races on the same ballot. A voter casts one vote in the Seat A race and one vote in the Seat B race.

From the voter's perspective, position-based MMDs look like two normal SMD races stacked on the same ballot. The seat designators are real and consequential: a candidate must choose which seat to run for, and they cannot win the other seat even if they outpoll its winner.

Position-based MMDs are sometimes adopted explicitly to surface head-to-head challenges: a challenger can target Seat A's incumbent directly rather than running in a pool. Washington's House uses this method for its 49 two-seat districts.

Staggered position-based is a variation. West Virginia's State Senate elects two senators from each of its 17 districts, but the two seats are on a four-year staggered cycle: Class I senators are on the ballot in midterm years (2022, 2026, 2030), and Class II senators are on the ballot in presidential years (2024, 2028, 2032). In any given election year, a West Virginia voter only sees one of their two senate seats on the ballot. The two seats are genuinely distinct races; they simply never share a ballot. Most state senates with four-year terms use some version of staggered classes, but West Virginia is the only state that does this with two senators sharing a district.

Subdistricted MMDs

Some states carve a multi-member district into smaller sub-pieces, each electing one or more members. The parent district still exists as a unit on the map, but the actual election happens at the sub-piece level.

Maryland's House of Delegates is the most elaborate example. Each of the 47 legislative districts is a three-member territory under Article III, Section 2 of the state constitution, but the General Assembly chose to split many of them into lettered subdistricts. District 1 in western Maryland is divided into 1A, 1B, and 1C, each of which is a single-member district. District 11 is split into 11A (one seat) and 11B (two seats, bloc-voted within the sub). District 12 has 12A (two seats) and 12B (one seat). The subdistrict choice is made district by district based on geographic and demographic factors, and the result is a state where some delegates run in three-way bloc races, some in two-way bloc races, and some in clean one-on-one contests.

South Dakota uses subdistricting more sparingly. Districts 26 and 28 are each split into A and B sub-pieces, each electing one of the district's two representatives. The other 33 South Dakota House districts use straight bloc voting for both seats. District 28's split predates 2004; District 26's split was ordered by a federal court in Bone Shirt v. Hazeltine, 336 F. Supp. 2d 976 (D.S.D. 2004), which held that South Dakota's at-large districting on the Pine Ridge and Rosebud Reservations diluted Native American voting strength in violation of Section 2 of the Voting Rights Act. The court-ordered remedy created subdistrict 26A as a Native-majority opportunity district.

State Profiles

The ten states that currently use multi-member districts for at least one chamber.

New Hampshire's Floterial Wrinkle

New Hampshire is unusual enough to deserve its own section. The state draws district lines by combining and splitting towns until population balance is achieved, and the result is a tangle of overlapping districts. A voter might cast ballots in a small town-level district AND in a larger "floterial" district that covers several towns and elects one or more additional representatives. Floterials exist to equalize population without breaking town boundaries: a 2006 New Hampshire constitutional amendment authorized them explicitly, and the current map has 39 floterial districts holding 57 seats.

A New Hampshire voter typically sees two separate House races on their ballot when their town is covered by a floterial: one for the base town district and one for the floterial overlay. Each is bloc-voted within itself. A voter in Brentwood, for example, would see two races on the ballot: the single-seat Rockingham 6 base district covering Brentwood alone, and the single-seat Rockingham 32 floterial district covering Brentwood, Danville, and Fremont together.

New Hampshire is the only state that uses floterials in this systematic way.

Where MMDs Come From

Multi-member districts were once the norm rather than the exception. In the early and mid 20th century, dozens of states used at-large or county-based MMDs for their lower chambers, often with seat counts that varied by county population. This was efficient for legislatures organized around county delegations but it created persistent population inequalities and, in many southern states, was used to dilute Black voting strength.

The U.S. Supreme Court's 1964 decision in Reynolds v. Sims, which required legislative districts to be roughly equal in population, did not directly outlaw MMDs but it forced states to redraw maps and prompted many to switch to single-member districts in the process. The 1965 Voting Rights Act, and especially subsequent litigation like Thornburg v. Gingles (1986), went further by treating at-large bloc voting as suspect where it had the effect of submerging minority voting strength inside a white majority. By the 1990s most southern states had abandoned MMDs in their state legislative chambers under VRA pressure.

The states that retained MMDs into the 21st century did so for reasons that mostly predate or sit alongside the VRA debate: New England town-based traditions, federalist preferences for delegations over individual representation, or geographic constraints that made small single-member districts impractical. Vermont and New Hampshire retained MMDs largely on town-government grounds; Maryland subdivided rather than abandoning the three-member model; the Dakotas, Arizona, and Washington kept two-member House districts for reasons more administrative than ideological. What these states share is less ideological coherence than path dependence: each had a working multi-member arrangement at the time of the post-Reynolds redraws and chose not to disturb it.

The Voting Rights Act and Subdistricting Today

Section 2 of the Voting Rights Act remains the primary legal protection that subdistricting and minority-opportunity districts depend on. Its scope has been actively contested at the Supreme Court over the past decade. Shelby County v. Holder (2013) struck down the Section 4(b) preclearance coverage formula that had required certain jurisdictions to get federal approval before changing voting laws. Brnovich v. DNC (2021) narrowed how Section 2 applies to vote-denial claims (rules that allegedly burden the ability to cast a ballot).

The structural districting framework was reaffirmed in Allen v. Milligan (2023), where the Court left the Gingles three-part test for vote-dilution claims intact and required Alabama to draw a second majority-Black congressional district.

That consensus did not hold for long. In Louisiana v. Callais (2026), the Court ruled 6-3 that Louisiana's second majority-Black congressional district (drawn to satisfy the Milligan framework) was an unconstitutional racial gerrymander, holding that Section 2 had not actually compelled the second district and therefore race-predominant line-drawing lacked a sufficient justification. Section 2 itself was not formally struck down, but the Gingles framework was substantially narrowed: plaintiffs now face heavier burdens, including the requirement that illustrative remedial maps satisfy all legitimate state interests (including incumbency protection). The Court also denied a stay of mandate in May 2026, giving the decision immediate effect for pre-2026-election redistricting.

The practical effect on subdistricting depends on why each subdistrict exists. Arrangements adopted explicitly as Section 2 remedies, like South Dakota's District 26A from Bone Shirt, are now exposed to fresh racial-gerrymander challenges and harder to defend if race predominated in the line-drawing. Subdistricts adopted for non-race reasons (population balancing inside a county delegation, traditional districting criteria, or simple geographic convenience) are not directly implicated. Most of Maryland's lettered subdistricts fall in that second category, but the lines between race-conscious and race-neutral districting are precisely what Callais has put back into active litigation. Expect new challenges through the 2030 cycle.

Why the Voting Method Matters

The choice between bloc and position-based voting changes campaign dynamics in subtle but important ways.

Slate effects. Bloc voting rewards parties that can run two strong candidates together: voters who select both of the party's candidates deliver both seats. Position-based voting separates the races and lets challengers concentrate fire on a specific incumbent.

Recruitment. In bloc districts, the second slot is often easier to recruit a candidate for because they ride the slate. In position-based districts, every challenger has to mount a full campaign against a named incumbent.

Minority representation. Bloc voting in racially polarized environments can submerge minority voters in a white majority and produce all-white delegations even where minority candidates could win SMDs. This is the structural reason VRA litigation targeted at-large bloc voting in the 1970s and 80s. Subdistricting (Maryland's solution) and position-based MMDs (where minority candidates can target one position) are both ways to keep multi-member districts while preserving minority representation.

Incumbency. Bloc voting protects incumbents who run as a slate. Position-based voting exposes them individually.

Data and reporting. For anyone tracking these races (this site included), bloc voting creates significant complications: a single election produces multiple winners, candidates must be assigned to administrative seat designators after the fact, and historical data about who held which seat depends on conventions that did not appear on the ballot. Position-based MMDs are easier to track because each seat is its own race throughout.