House Expansion¶

A Primer Series Article of the Congressional Elections Modernization Act (CEMA)¶

Prepared by Albert Ramos for The American Policy Architecture Institute

A republic governs through representatives, and the raw material of representation is the seat. Each seat is a portion of the public given a voice in the national legislature, and how many there are sets how faithfully the whole body can mirror the country. Think of it the way a screen renders an image: too few points and the picture is coarse; add points and the same subject comes through in sharper, truer detail. The number of seats is the resolution of the People's House.

That number has been frozen since 1929. The House has held exactly 435 seats for nearly a century -- a figure set by a federal law passed when the country had fewer than half as many people as it has today. The population has more than tripled since; the House has not grown by a single seat. A body built to render one country is now asked to render a far larger one at the same low resolution.

The Congressional Elections Modernization Act unfreezes that number. Under the Act's House Expansion framework, the House grows from 435 to roughly 720 seats -- not a figure picked by negotiation, but a size set by a fixed rule tied to the nation's population, so the chamber keeps pace with the country it represents. It is one of the changes that make up CEMA, supported by the national elections body the Act creates, The Electoral Science Office.

A 1929 Number Governing a 2026 Country¶

Most people assume the number 435 means something. It does not. It is not in the Constitution, and it reflects no principle about how many people one representative should serve. It is simply where the count stopped. A 1929 law capped the House, and Congress has never revisited the cap. Every census since has redistributed those same 435 seats among the states; none has added to them.

That freeze was the exception, not the rule. For most of American history, Congress enlarged the House after nearly every census, keeping pace -- imperfectly, but in the right direction -- with a growing country. The chamber was meant to grow. The 1929 decision to stop was a product of its moment, and it was simply never undone.

The cost shows up in a single ratio. When the cap was set, each member represented roughly 210,000 people. Today, each represents about 760,000. That is the resolution of the People's House falling in real time: the same fixed number of seats stretched across a country half again as large, and then larger still.

The freeze did something else worth naming. Holding the number of seats fixed did not only stretch each member's share of the public -- it fixed the number of hands the chamber's work is divided among. A House that never grows slowly becomes a more exclusive body, its membership a scarcer share of the country. The People's House was not meant to drift toward the character of an exclusive club; keeping it sized to the nation keeps it the broad, populous chamber it was designed to be.

That ratio is arithmetic, not opinion. Grow the chamber to roughly 720 seats and the same arithmetic runs the other way: the population each representative answers to falls from about 760,000 to about 465,000 -- a reduction of roughly 39 percent. House Expansion begins from this fact and asks what it would take to make the chamber grow with the country automatically, on a rule rather than a whim.

A Rule, Not a Number¶

CEMA's answer is to stop setting the size of the House by hand. Instead of a fixed figure that requires political will to change -- and that, predictably, never gets changed -- the Act ties the size of the House to the population through a fixed formula. The chamber grows because the country grows.

The starting point is a principle, not a number. A republic needs enough seats to render the public faithfully, but not so many that the body stops being a working legislature. Both edges are real. Push the seat count very low and representation degrades toward rule by a small circle -- a handful of people deciding for everyone is not a republic reflecting its public. Push it to the limit the other way, one seat per person, and there is no representation left at all; that is direct democracy, not the government the framers built. A republic lives in the range between: enough seats for high-resolution representation, few enough to remain a chamber that can conduct the nation's business.

CEMA finds that range with the cube root rule: the House is sized at roughly the cube root of the national population. For today's United States that comes to about 692 seats, which the Act rounds to a convenient target of roughly 720. The rule is not arbitrary -- the world's national legislatures tend to cluster near the cube root of the populations they serve, a pattern observed in the data, not a convention anyone set out to follow. CEMA adopts that pattern deliberately, as a fixed engineering rule rather than a one-time bargain. The point is not that the cube root is sacred; it is that a principled, fixed rule beats a number frozen by accident, and the cube root puts the chamber squarely in the right range.

Expansion does not happen overnight. The Act phases it in over roughly fourteen years -- about three presidential terms -- with an initial jump from 435 to 510 seats, then steady increments each Congress until the target is reached, after which it adjusts once a decade to track the population. And no sitting member loses a seat. Expansion works entirely by adding seats, distributed among the states by population through the same process -- apportionment -- that already divides the existing 435 every ten years. Today, because the total is fixed, one state's gain is another's loss; expansion simply gives that process more seats to divide. The House gets bigger; it does not get rearranged at anyone's expense.

Why the Size of the House Is the Lever¶

Raising the resolution of the House is worth doing on its own. But within CEMA's design the size of the chamber does something more specific: it sets how open the whole system can be -- how small a group of like-minded voters can be and still earn a seat.

CEMA pairs expansion with STAR Voting, under which most representatives are elected from larger, multi-seat districts rather than one-winner districts. In a multi-seat district, the share of support needed to win one seat depends on how many seats the district has: a three-seat district takes roughly a third of the vote, a five-seat district roughly a fifth, a seven-seat district roughly a seventh. More total seats in the House means more seats per district, and more seats per district means a smaller slice of support is enough to win one. The size of the House is the dial that sets how low the bar to win representation can fall.

This is what resolution looks like in practice. A coarse House renders the public in a few broad blocks; a higher-resolution House can render the finer grain -- the cohesive groups of voters who, under a small chamber and winner-take-all districts, get filtered out in a primary or absorbed into one of two big coalitions and never appear at all. Grow the chamber, and a bloc that could never assemble a local majority can still claim one seat among several. The people who were rounded off come through.

That makes expansion the lever behind the whole structure. If the threshold to earn a seat is still too high for smaller groups to clear, the remedy is not to tinker with the voting method or redraw any lines -- it is a bigger House. Grow the chamber, and every threshold in the system comes down automatically, because the arithmetic of seats-per-district does the work. The mechanics belong to STAR Voting; what House Expansion contributes is that the size of the chamber governs all of them. Expand the House and you open the system; hold it small and you keep it closed. That is what lifts House Expansion from housekeeping to something central to the design.

Addressing Concerns¶

"Isn't the real problem that Congress is too remote? Wouldn't smaller districts let representatives actually know the people they serve?"¶

This concern comes from a reasonable place. Congress does feel distant, and the instinct that representation should mean a genuine connection between a community and the person who speaks for it is sound. The feeling is correct. What is worth examining is the fix it reaches for -- the idea that the answer is a district small enough that a representative can personally know it.

That fix does not survive the arithmetic. At the founding, each member represented about 30,000 people, and it is tempting to read that as the lost ideal to recover. But no achievable House comes close. Even at roughly 720 seats, each member represents about 465,000 people -- not a community anyone can personally know, just a smaller very-large number. Following the logic all the way would mean a House of many thousands of members, which stops being a legislature and becomes an unworkable crowd. The closeness instinct, taken literally, has no stopping point short of a chamber that cannot function.

The deeper reason it misfires is that it asks the federal House to be something the federal level was never meant to be. Direct, know-your-neighbor representation is real and it matters -- but its home is state, county, and city government, the levels built for the close texture of local life. The national legislature operates at a different altitude: it deliberates on national questions -- defense, the borders, the country's largest collective choices. What makes a person well represented there is not whether their representative could pick them out of a crowd. It is whether a coalition they belong to actually helped put someone in a seat -- whether the body, taken as a whole, renders the country they are part of. That is what expansion improves: not intimacy, which no workable chamber can deliver, but fidelity.

"If this is about fair representation, why do some states still get left out?"¶

This is the most important question this change has to answer, and the honest answer begins with where the limit comes from. It does not come from CEMA. It comes from the map.

State borders are accidents of history -- drawn over two centuries by charters, surveyors' lines, treaties, and bargains, almost never by any theory of what makes a sensible unit of representation. CEMA takes those borders as it finds them. But it has to live with a consequence of them: the Constitution guarantees every state at least one seat, and seats are apportioned by population, so a state with very few people gets very few seats no matter how large the House grows. Proportional representation works by dividing a district's seats among its groups of voters according to their support -- and a state holding only one or two seats has nothing to divide. One or two seats is one or two winners; there is no remainder to allocate.

The numbers are specific. At roughly 720 seats, 42 of the 50 states -- 84 percent -- reach the three-seat minimum that proportional representation requires, up from 38 states today. Of the eight that fall short, six hold exactly two seats, and two -- Vermont and Wyoming -- remain single-seat states. This is not a transition problem that more growth would solve. Vermont, the closest of all, would not gain a second seat until the House passed 728; lifting Wyoming to three seats would take a chamber of roughly 1,400 -- nearly twice the size CEMA proposes. These are structural, and they would persist at any House size the country could plausibly adopt. CEMA does not pretend it can reach these states. It names them as a defined exception and is honest that the design has a floor it cannot cross.

"What would it take to get every state to fair representation -- 100 percent across the country?"¶

Reaching every state would require setting state borders aside for this one purpose -- keeping them for everything else -- and dividing the whole country fresh into rational units of roughly equal population. Only then would every part of the country sit inside a district large enough to render its voters proportionally. That is beyond what an act of Congress can do; it would take a constitutional amendment, not a statute. And even then it would be genuinely hard, because people are not spread evenly across the land -- some regions, the dense and irregular Northeast in particular, resist being carved into equal pieces no matter how the lines are drawn. So 100 percent is a ceiling set first by the Constitution and second by geography. CEMA goes as far toward it as a statute can reach, and is candid that the ceiling is there.

What House Expansion Does and Does Not Reach¶

House Expansion changes the size of one chamber and nothing else. It grows the House; it does not touch the Senate, where every state has two seats regardless of population, and it does not alter the Electoral College. It does not change any state's borders, does not displace any sitting member, and federalizes nothing beyond the size of the federal House, which the Constitution already commits to national resolution. The change is deliberately contained: one number, set by a rule instead of by accident.

Conclusion¶

A republic is only as good as the reflection it gives the people of themselves, and the seats of the House are what that reflection is made of. For nearly a hundred years the country has tripled in size and grown in variety while the number of seats held still -- the resolution of the People's House frozen at a 1929 setting. House Expansion fixes that at its root, replacing a figure frozen by accident with a chamber that grows on a fixed rule as the country grows.

And the size of the House turns out to be more than a measure of detail. It is the lever that sets how open American representation can be -- the dial that, paired with STAR Voting, determines how small a group can be and still earn a seat. That is why this change is central rather than cosmetic. House Expansion works alongside the Act's other changes -- STAR Voting, Hands-Off Mapping, and Full-Field Elections -- each supported by The Electoral Science Office, to deliver a Congress sized to the nation it serves.

Other articles in the CEMA Primer Series:

Works Cited¶

Taagepera, Rein. "The Size of National Assemblies." Social Science Research 1, no. 4 (1972): 385-401.

U.S. Census Bureau. "Historical Apportionment Data (1910-2020)." Accessed June 2026. https://www.census.gov/data/tables/2020/dec/2020-apportionment-data.html

U.S. Census Bureau. "2020 Census Apportionment Results." April 26, 2021. https://www.census.gov/newsroom/press-releases/2021/2020-census-apportionment-results.html

U.S. Constitution, Article I, Section 2.

U.S. Permanent Apportionment Act of 1929, Pub. L. No. 71-13, 46 Stat. 21.

The Congressional Elections Modernization Act (Rev 6.2) is an internal APAI document available on the same website and is referenced in prose by name. The 84-percent-at-720, Vermont-728, and Wyoming-~1,400 figures are APAI Huntington-Hill apportionment analysis computed on the 2020 Census apportionment data cited above. The ~210,000-per-member 1929 figure is computed from the 1930 Census population over 435 seats; the ~760,000 and ~465,000 figures from the 2020 Census population over 435 and 720 seats respectively.

Revision history available in the raw file.

📥 Download this document (opens on GitHub -- click the ⬇ download button)