Commentaries on Representation

The Blog for Thirty-Thousand.org

Commentaries on Representation

The Blog for Thirty-Thousand.org

The NY Times’ proposals for enlarging the House would be ineffectual

The New York Times identifies many of the problems resulting from the House of Representatives being “far too small”. Unfortunately, their preferred solutions not only fail to solve those problems, they could even worsen them to some extent.

In 2018, the New York Times Editorial Board published a two-part series on “how to make the House of Representatives more representative”. The first part explained why having a House of Representatives that is “far too small … poses a big danger to American democracy”. In order to ameliorate the resulting problems, it then recommended increasing the number of Representatives based on a formulation known as the “cube-root law”. However, though the cube-root formulation is commonly believed to be a viable solution, our analysis reveals that the resulting increase in the size of the House would be completely ineffectual. The second part of their editorial promotes the theoretical benefits of multimember congressional districts while glossing over its apparent drawbacks.

Part One: America Needs a Bigger House

Part one of their editorial series, America Needs a Bigger House,¹ explains why our House of Representatives is far too small to properly represent the American people, a position which was recently affirmed again by their Editorial Board.² In those editorials, the New York Times provides compelling arguments for why our representation in the House should be greatly enlarged. This commentary focuses only on the efficacy of the solutions they proposed for overcoming the problems created by our grossly undersized House.

After providing some background regarding Congress’s failure to expand the House along with the population, as intended by the framers, the editorial observes that “the House today is far too small, and that poses a big danger to American democracy”. The editorial focuses on three of the problems that result from massively oversized and unequal congressional districts:

For starters, how does a single lawmaker stay in touch with the concerns of three-quarters of a million people? The answer is she doesn’t. Research shows that representatives of larger districts are more likely to take political positions at odds with what a majority of their constituents want. These representatives are also ripe targets for lobbyists and special interests, whose money enables them to campaign at scale, often with misleading messages. Special interests are more likely than regular voters to influence policy positions and votes.

Second, the cap on the number of House members leads to districts with wildly varying populations. These discrepancies violate the basic constitutional principle of one-person-one-vote, causing voters to be unequally represented in the chamber that was designed to offset the Senate, where every state gets two seats regardless of population.

Third, the size of the House determines the shape of the Electoral College, because a state’s electoral votes are equal to its congressional delegation. This is one of the many reasons the college is an unfair and antiquated mechanism: States that are already underrepresented in Congress have a weaker voice in choosing the president, again violating the principle that each citizen should have an equal vote.

As the editorial states, “[t]here’s a simple fix: Make the House bigger”. Specifically, the editorial recommends increasing the number of Representatives from 435 to 593 which, they assert, would make the House “proportionally similar to most modern democracies”. The basis for choosing that size is a questionable proposition known alternatively as the “cube root law” or the “cube root rule”,³ which postulates that the ideal size of the House is the third root⁴ of the total population. Though the the “cube root rule” sounds both precise and authoritative, it is actually neither, and is therefore identified herein as the cube root model.

Rather than getting into the unsoundness of the cube root model itself, this commentary will focus on how a House size derived therefrom would impact the three problems identified in the editorial. And, to be fair, let’s update the cube root used in the 2018 Times’ editorial: Their recommended House size of 593 would assume a total apportionment population of 208.5 million;⁵ however, the U.S population hasn’t been that small since the 1970s! Presumably they actually calculated a cube root of 693 (based on an interim population estimate), and then subtracted the 100 Senators from that total.⁶ Even taking that into account, the Times should have derived the cube root from the 2010 census,⁷ which would have indicated a total of 676 legislators.⁸ Subtracting the 100 Senators would then indicate a total House size 576.

However, in order to maximize the potential benefit to be derived from the cube root formulation, let’s update that number to the cube root of the 2020 census without subtracting the 100 Senators, which would indicate a total House size of 692.⁹ That size will be used in this commentary to evaluate the efficacy of the cube root model relative to solving the three problems identified in the editorial.

Returning to those problems, the first one mentioned is the obvious inability of any single Representative to “stay in touch with the concerns of three-quarters of a million people”. Based on the 2020 census, the average constituency size for a 435 Representative House is now approximately 760 thousand. However, if the size of the House were increased to 692, the average constituency size would be reduced to approximately 478 thousand.¹⁰ Though that would be an improvement, it would still be impossible for a single lawmaker to “stay in touch with” nearly a half a million constituents. In addition, in such large districts, these representatives would still be “ripe targets for lobbyists and special interests” due to the tremendous cost of political campaigns in massive districts.

The second problem of a small House is that it “leads to districts with wildly varying populations” (thereby violating the constitutional principle of one-person-one-vote). To illustrate that point, the recent apportionment of 435 Representatives, based on 2020 census, results in a disparity between the largest and smallest districts of 82%, and the relative standard deviation of all 435 districts is a massive 2,986%. Based on the same census, increasing the House size to 692 would reduce that disparity only slightly (to 75%), and the relative standard deviation of all the districts would be a whopping 1,623%. So obviously, increasing the number of Representatives to 692 would accomplish very little relative to reducing the problem of “wildly varying populations”.

The third problem is that because “the size of the House determines the shape of the Electoral College … States that are already underrepresented in Congress have a weaker voice in choosing the president, again violating the principle that each citizen should have an equal vote”. Presumably this ambiguous statement is referring to the fact that the unequal representation suffered by citizens from one state to another (referenced in their second objection) is effectively paralleled relative to their presidential elections. It could also be referring to what is known as the “small state advantage”, which is a concern expressed by many critics of the Electoral College. However, the increase indicated by the cube-root formula would have little impact on either of these inequities, as it would require a much larger House of Representatives for the Electoral College to achieve true interstate parity.

The point is, though the cube-root formulation is a favored solution, it would accomplish very little relative to ameliorating the “big danger to American democracy” that results from having a House that is far too small, and there is no reason to expect the average citizen would be any better off if the size of the House were increased by only a few hundred.

Rather than the “cube root rule”, we should adhere to the Founders’ Rule: The average constituency size should be between 30,000 and 50,000. That would eliminate all of the “danger to democracy” identified in the NY Times editorial. In fact, it would completely unshackle democracy in the United States.

Part two: A Congress for Every American

The second part of the Times editorial, A Congress for Every American,¹¹ advocates making the congressional districts even larger in order to accommodate multimember districts wherein the elections would be based on “ranked-choice” voting (“RCV”). RCV is an electoral system in which ballots list all the candidates who are running for multiple seats in a particular district. The voters “vote” by ranking the candidates according to their preference. The voters’ rankings are then aggregated and run through an algorithmic process to determine which of those candidates are elected to office.¹²

The main argument for RCV is that a greater diversity¹³ of representation can be elected to office than otherwise would happen in single-member districts. To make this point, the editorial uses two states to graphically illustrate their hypothetical examples. In Massachusetts they argue that, because of RCV, Republican representatives could be elected in a state that hasn’t elected one since 1994. And their Texas hypothetical contends that a greater proportion of Democrats would be elected as a result of implementing RCV. Both of these examples rely on a comparison of the expected election outcomes of single-member vs multimember district elections. And though their analyses appear to confirm that multimember districts produce greater diversity than single-member districts, their analyses were flawed: Their multimember-district scenario was predicated upon a House of 593 Representatives, while their single-member scenario was predicated upon the current House size (435), whereas it should have also been predicated upon 593-member House for an apples-to-apples comparison.

The table below summarizes the different scenarios relevant to this discussion. Referencing that table, the NY Times editorial compared scenarios A and C for each state. Obviously, a proper analysis would have first compared scenarios A and B, to determine how much additional diversity results from increasing the number of single-member districts. Next, scenarios B and C should have been compared to isolate any additional diversity resulting from switching from single-member to multimember districts.

Table 1: 435 vs. 593 Representatives

		A		B		C
State	Population 2010 Census	435 Representatives		593 Representatives
		Single-Member Districts		Single-Member Districts		Multimember Districts
		Districts	Average Pop.	Districts	Average Pop.	Districts (Total Reps)	Average Pop.
MA	6,559,644	9	728,849	13	504,588	3 (13)	2,186,548
TX	25,268,418	36	701,901	51	495,459	11 (51)	2,297,129

So, in their Massachusetts example, they compared 13 Representatives seeking office in three multimember districts to nine single-member district elections – why not use 13 single-member districts for their hypothetical comparison? And in their Texas example, they compared 51 Representatives seeking office in 11 multimember districts to 36 single-member district elections – why not use 51 single-member districts as a basis for their comparison?

Note: To be consistent with the examples illustrated in the editorial, the table above reverts to its 593-Representative scenario which, coincidentally, would be nearly equal to the cube root of the 2020 census (692) minus 100. (As a point of reference, the same table illustrating the 692-Representative scenario is available here here.)

However, even if this analysis could be done in a completely impartial manner, an inherent problem with such analyses is that, for a variety of reasons, historical data cannot be extrapolated to reliably predict future voter behavior under a completely different electoral district configuration. First, citizens with minority political views may not have previously bothered to vote in their districts thinking it to be pointless, or others of the majority view may also not have bothered to vote if their district was certain to prevail in their favor. Second, having so many additional candidates/parties in such massive districts would dramatically change the way the political ground game is conducted, especially to the extent that any coalitions are formed. Related to that, the emergence of third parties would likely attract new voters and alter a historical pattern based on duopolistic voter dynamics. And finally, given that these multimember districts would be even more massive than single-member districts, the ability of the candidates/parties to raise huge sums of money becomes an even more significant factor. That is probably why elections in multimember districts tend to be party-centric rather than candidate-centric.

Putting aside the problematic nature of the analysis contemplated by the NY Times editorial, how would the creation of multimember districts impact the three problems identified therein? Revisiting those, their first objection was that there was no way for a single lawmaker to “stay in touch with the concerns of three-quarters of a million people”, referring to the national average. (Column A provides the averages for each state.) Given that their proposed multimember districts in Massachusetts and Texas are triple that size, how could a Representative in such a district do any better? As shown in the table above (column C), the multimember districts in would contain approximately 2.2 million residents. If those districts were cities, they would be the fourth largest in the country!

However, as stated above, a more sensible comparison would be between columns B and C: What the editorial is really proposing, under the larger House scenario, is to replace single-member congressional districts containing half a million people (B) with multimember districts quadruple that size! Even with three to five Representatives “representing” each of those massive constituencies, are each of them expected to represent all 2.2 million constituents on some overlapping basis? Clearly, a Representative in a single-member district of half a million would generally be more accountable and could do a better job of providing constituent services.

The NY Times glosses over the considerable disadvantages of having massive multimember congressional districts. The most conspicuous one is the amount of funds that candidates would have to raise to mount effective campaigns. Given the huge amounts currently raised to fund two-party races in single-member districts, imagine how much more would have to be raised by dozens of candidates in multiple parties competing for three to five House seats in districts containing over two million people. Moreover, in such massive districts, incumbents would still enjoy a nearly insurmountable advantage with respect to be reelected. That being said, it should be up to each state to decide how it wants to implement its federal congressional districts.

The next two objections expressed in the first part of their editorial concerned the egregious violation of the Constitution’s one-person-one-vote equality requirement, and the related problem of Electoral College inequality. The recommended increase in representation proposed by the NY Times editorial board would diminish these problems only slightly (as explained earlier in this commentary). However, instituting multimember congressional districts would not further reduce these problems in any way.

The New York Times Editorial Board should be lauded for calling attention to why having a House “far too small … poses a big danger to American democracy”. However, not only would the solutions they propose have a negligible impact on the problems they identified, they would exacerbate them to some extent. A House size determined by the cube root formulation might improve representational diversity somewhat, but it would nonetheless preserve massive congressional districts, thereby perpetuating America’s political ruling class. And those huge districts would be quadrupled in size if multimember districts were imposed—which would be a cure that is worse than the disease! Perhaps it was the Times’ enthusiasm for the latter concept that led them to state that “districts that send multiple members to Congress … was the norm at the nation’s founding”, an assertion which is, at best, an extraordinary exaggeration.

As it turns out, the only reliable way to solve the problems identified in the editorial is the solution proposed by our nation’s founders, especially in combination with single-member districts.

Table 2: 435 vs. 692 Representatives

		A		B		C
State	Population 2020 Census	435 Representatives		692 Representatives
		Single-Member Districts		Single-Member Districts		Multimember Districts
		Districts	Average Pop.	Districts	Average Pop.	Districts (Total Reps)	Average Pop.
MA	70,33,469	9	781,497	15	468,898	3 (15)	2,344,490
TX	29,183,290	38	767,981	61	478,415	13 (61)	2,244,868

Note: In order to maintain approximately the same proportions as illustrated in the Times’ 593 scenario, the total number of multimember districts in MA was kept the same, but the total number of districts in TX was raised from 11 to 13.