Note 18. Why 435?
Reapportionment and the Size of the House
The House has been at 435 members since 1913, when one seat was added for the admission of Arizona and one for New Mexico. That number was retained with the enactment of the Permanent Apportionment Act of 1929. The obvious result of capping the size of the House over the last century was to steepen the increase the size of House districts (see Figure 18-1).
How did we get here? This is one of the most common questions asked of me over the years. Most people who are attentive to politics know that the decennial census counts the number of Americans, is followed by “reapportionment” to determine how the 435 seats are distributed among the 50 states, and the states then engage in “redistricting” to draw district lines to adjust for the number of seats and population shifts within each state. I address the size of the House and apportionment in this Note.
Apportionment may seem like a mechanical issue of little political interest, but that is hardly the case. State and party interests pervaded discussions of apportionment throughout the history of the Republic. It is true that some basic issues appear to be settled. The size of the House has not varied except for the temporary increase to 437 for the admission of Alaska and Hawaii (1959-1962). Apportionment of House seats to the states has been done under a formula first adopted 80 years ago. Recently, however, calls for changes in the size of the House have become more common.
The easiest way for me to give a sense of the way the apportionment decisions have been infused with politics is to highlight a few events in chronological order.
The Constitution and the First Apportionment
The Constitution provided for 65 House seats and specified the number of seats for each of the original 13 states. It also provided that “the number of representatives shall not exceed one for every thirty thousand, but each state shall have at least one representative.” The Constitution also provided a sinful provision that counted slaves as three-fifths of a person, allowing slave-holding states to increase their population for the purpose of representation in the House and electoral college but without disadvantaging non-slave holding states too much. Subsequent apportionment acts increased the number of districts and the admission of states added a minimum of one seat each time a state was admitted.
Controversy was present from the start. The first apportionment legislation was the subject of President George Washington’s first veto in 1792. Congress passed a bill that assigned seats by dividing 30,000 into the U.S. population, but, because that yields fractions as a remainder, assigned the leftover seats to states on the basis of the size of their remainder. This meant that some states received more than one seat for every 30,000 people. In the view of Thomas Jefferson, Washington’s secretary of state and others, that approach violated the constitutional provision limiting representatives to no more than one for every 30,000. Alexander Hamilton favored the congressional approach, but Jefferson persuaded Washington to exercise his first veto. Congress responded with a new approach suggested by Jefferson, which was signed into law.
The episode had political implications. Jefferson’s approach produced one additional House seat for Virginia, his home state, over the initial approach adopted by Congress and reflected a tension between Hamilton and Jefferson that would intensify. Although Washington was a Virginian, he probably was persuaded by the constitutional argument.
The episode generated labels for the two apportionment strategies. The approaches deal with the remainder problem in different ways. The remainder problem is due to the fact that 30,000 does not divide evenly into the population number and fractions of seats cannot be allocated.
“Hamilton’s method” divides the population size by the number of seats and allocated seats to the states based on their sizes, which left some seats to be allocated. They were allocated in the order the size of the remainder or fraction of a seat not assigned in the first step, starting with the state with the largest remainder and working down until all seats were allocated.
“Jefferson’s method” also starts by dividing the population size by the number of seats to get a “divisor” and then, for each state, divides the population by the divisor and rounds down to get the number of seats for that state. If the seats total less than the correct number, this method adjusts the divisor until the seat total is correct.
I list the methods in Table 18-1 and illustrate the current method at the end of this Note.[1]
The 1840s and 1850s
The Jefferson method was used as long as his party (the Democratic-Republicans) and the Jacksonian Democrats retained a House majority. That changed in the 1840 elections. Figure 18-1 shows a dip in the number of House seats in the 1840s. It was the direct result partisan conflict.
In 1840, a new census was conducted and Whigs won a new House majority, placing Whigs in a position to pass an apportionment bill to their liking. The subsequent debate over reflected the view of majority party Whigs that Democrats had stacked the decks in a few states by using a general-ticket, at-large system (most notably, Alabama, Georgia, and Mississippi).[2] In a general-ticket system, there were no single-members districts; rather, voters could cast a many votes as the state had House seats and the top vote-getters would win the seats. In practice, voters use paper ballots provided by the parties so that the more popular party statewide would win all the seats.
Fearing that the Democrats would win a disproportionate number of any new districts, the Whigs proposed cutting the size of the House. They rationalized this on the grounds that the House would be more deliberative legislative body if its size was curtailed. They also proposed to ban the general-ticket system. Remarkably, the bill passed the House with ban on general-ticket system but not before Democrats and some Whigs increased the number of House seats. The Senate, also controlled by Whigs, reduced the number of House seats in the bill, returned it to the House, and enough Whigs supported the Senate version to approve it. Whig President John Tyler, who already had proven to be a turncoat, in the view of most Whigs, signed the bill nevertheless. The House was shrunk to 242 from 223 and single-member districts were mandated.
Thanks to the erratic Tyler presidency, the Whigs lost 70 House seats and their House majority in the 1842 elections. After the next census, a Democratic majority increased the size of the House to 234 in 1852 and a Republican majority increased it again to 241 in 1862. The House continued on its upward trajectory spurred by the adoption of the 14th Amendment (counting the black population fully), the admission of new states, and population growth until 1913. A minor irony is that the Democrats adopted the Hamilton method in 1852, which Congress left in place until 1901. During most of that period the Hamilton and Webster methods would have produced the same state allocations.
Into the 1900s
In 1901, Republicans increased the size of the House to 386 so that no state would lose a seat. They also substituted the Webster method for the Hamilton method. In 1911, they again increased the size of the House to protect all states—this time to 433 plus two pending the admission of Arizona and New Mexico.
Remarkably, no apportionment act was adopted following the 1920 census. Despite the constitutional mandate, the Republican Congress did not pass an apportionment bill, leaving the 435 seats in place in 1913 until the 1930s. The Republicans and rural legislators figured correctly that the huge growth in northern urban areas, the product of foreign immigration and the migration from the South, would make the Democrats big winners from increasing the size of the House and reapportionment. Indeed, the House of the late 1920s operated with a strong bias in apportionment and districting that favored the Republicans. Had reapportionment occurred, even without increasing the size of the House, more than ten seats would have shifted from rural to urban states.
By 1929, the need to reapportion the House following the 1930 census was obvious and required by the Constitution. Some northern cities had House districts with more than twice the population of the average district. With Republicans still maintaining a majority in the House and Senate, Congress left the House at 435 members. That number, of course, was the inherited number and somewhat arbitrary. It did not, and still does not, not reflect a coherent view of some proper size for the House.
By the time the 1929 act was considered, statisticians worked on new methods for apportionment. Among them were Joseph Hill, the chief statistician of the Census Bureau, who proposed a method that later came to be known as the method of equal proportions. A friend of Hill, Edward Huntington, a Harvard mathematician, revised Hill’s proposal so the method is now often called the Huntington-Hill method (Table 18-2). The 1929 act required the Secretary of the Interior to send to Congress the state allocations under both the Webster and Huntington-Hill after each census.
Oddly, Congress punted on the question of the method of apportionment. The 1929 act required the president to send to the Congress apportionments based on the Webster method, the Huntington-Hill method, and on the method used in the previous apportionment. If Congress remained silent on the matter (as it did), the method last used would be used. As it turned out, Webster the last used, but, following the 1930 census, Webster and Huntington-Hill produced the same apportionment of the 435 seats. And, of course, the number of seats going to Democratic areas increased.
In 1941, with Democrats in the majority, Congress made the method of equal proportions the required apportionment method. It has been used since then without any increase in the number of seats. Congress also made apportionment self-executing so that it would not have to pass legislation every decade.
The Small-State Bias of the Current Method
At times, the Huntington-Hill method (H-H), or method of equal proportions, produces a small state advantage due to the fact that it rounds up more frequently for small states than large states. This may seem technical, but it is worth understanding. The H-H method provides for rounding up when the remainder is greater than the geometric mean. In 2021, New York’s quota was 26.49526 and its geometric mean is the square root of 26 x 27, or 26.49528, so its remainder is less than the geometric mean and H-H rounds down to 26. Minnesota’s quota was 7.48334 and its geometric mean is the square root of 7 x 8, or 7.48331, so the quote is greater than the geometric mean and H-H rounds up to 8 for Minnesota. The difference between the states is due to the size of the states—26 and 27 are much larger than 7 and 8 so the geometric mean is higher for the larger state—even when New York’s quota remainder is larger than than Minnesota’s remainder. The result: New York lost a seat and Minnesota retained its 8 seats.
Another method, the Webster method, does not have a small state or a larger state bias. In 2021, it would have given 27 seats to New York and 7 to Minnesota. The result under H-H is that New York ended up with districts with an average population of 777,528 and Minnesota’s districts averaged 713, 719.
The big picture after the 2020 apportionment is that two large states (New York and Ohio) had one fewer seat allocated to each of them that went to smaller states (Montana and Rhode Island) than would have happened with the Webster method.[3] In fact, since the H-H method was adopted after the 1940 apportionment, eight of nine reapportionments through 2021 reapportionment produced a small-state bias—a seat loss for at least one large state that would not have been lost with the Webster method.[4] Reverting to the Webster method is likely to be advocated by large-state legislators again, particularly Democrats, when apportionment legislation is considered again in the late 2020s.
The Size of the House
Arguments about the size of the House are resurfacing. Proponents of enlarging the House always start by noting that House districts are well over 760,000 people on average, by far the largest districts of any national legislature in the world. To over generalize just a little, arguments in favor of enlarging the House center on improving representation. Several specific arguments are made, although not all arguments are made by all advocates of expansion:
Large House districts, which may reached a million in a couple decades, violate the spirit of the Constitution, which provided for no more than one representative for every 30,000 people.
House districts are far too large for a single legislator to understand and fairly represent.
Citizens will have greater access to their representatives if districts are smaller.
Smaller districts will allow some states with just one representative to have more and move the House closer to proportional representation of states.
More districts would allow House districts to better match local units of government, such as counties and cities.
Enlargement would make it easier to construct more districts that are majority black or Latinx.
Enlargement would make it easier for minor parties to gain representation in Congress.
Enlargement would increase the number of representatives in most states and create a better basis for instituting the proportional representation of parties in each state’s delegation to the House.
Counterarguments emphasize the effect of size on the House as a lawmaking body. “Too many cooks in the kitchen” is the central argument, but there are others:
A central purpose of a legislature, deliberation on issues important to the public interest, will be even more difficult in a larger House and its committees.
A larger House is likely to be even more dependent on central party leaders, like the British House of Commons (650 members) and be more partisan.
A larger House will not be as attractive to highly qualified individuals who want to have an impact on public policy.
A nation as large as the modern U.S. cannot abide by even the spirit of the Framer’s plan to initially limit the size of House districts to 30,000 people (which would create a House of over 11,000 members).
Final Thoughts
I favor both reinstitution the Webster method and expanding the size of the House. The Webster method outperforms the current H-H method on a criterion for good formula—being unbiased with respect to state size. What tipped the balance in 1940 for majority party Democrats in favor of H-H over Webster was that H-H would take a seat from the larger and Republican Michigan and give it to the smaller and Democratic Arkansas (state partisanship has changed dramatically since then). It is time to adopt a method that does not systematically favor large or small states.
Reform of the apportionment formula is not easy, as the decades since 1940 have proven. While large states might squeeze legislation to adopt the Webster method through the House, the Senate has a strong small-state bias built into the constitutional allocation of two senators to each state without regard to size. That small-state bias, along with the ease of filibustering legislation, give small states, and Republicans who currently benefit from the small-state bias, the ability to block action on apportionment reform.
The same political logic affects the odds of enlarging the House. While the advantage gained by small states by virtue of having one House seat whatever the size of the state may seem small, the number of House seats currently translates into electoral votes in presidential elections, too, where small states and Republicans may not want to give up some of their arithmetic advantage. Greatly increasing the size of the House—say, to 650, the size of the British House of Commons—would dilute the power of incumbent legislators and have somewhat unpredictable consequences for how the House does it work, all of which generates uncertainty for incumbent members who would have to endorse the legislation. We are likely to have the somewhat arbitrary number of 435 representatives for decades to come.
[1] For more details, there are several accurate online sites. Census Bureau: https://www.census.gov/history/www/reference/apportionment/methods_of_apportionment.html. American Mathematical Society: http://www.ams.org/publicoutreach/feature-column/fcarc-apportionii1.
[2] Kenneth C. Martis, The Historical Atlas of United States Congressional Districts (Macmillan, 1982).
[3] https://www.ianrmcdonald.com/posts/2021-05-10-new-york-and-house-apportionment-in-2020_update/
[5] Ibid.
Professor Smith, please visit https://thirty-thousand.org/ — perhaps you could lend a hand with this. - Jeff Quidam