Post 21: 20 May 2025
We should admit that the American democracy is broken and do what we can to fix it.
Congress is abdicating its constitutional authority. This isn’t a Trump thing, it has been going on for a century, most notably when Congress ceded its power to declare war to the President after World War Two. But Trump is trying to finish the job by eliminating the power of Congress to control appropriations and taxation (tariffs), or to exercise any meaningful oversight of the executive branch. After he has done that, there won’t be much left. In his disastrous wake, Congress will have to rebuild almost from scratch.
To that end, today I offer a set of recommendations that would significantly strengthen the US House as a representative institution capable of checking executive overreach. This will be different from my usual social policy and philosophy posts but will still draw on the same store of civic republican principles.1
By acts of Congress, that is without Constitutional amendment, Congress can change the number of House members, establish standards for federal elections, and establish its own internal ethical standards. My proposed reforms fall into these three areas.
The U.S. House of Representatives should be significantly larger.2 The House has been locked into its present size of 435 members since 1930.3 At that time, the average district included 282,000 people; in 2020, it was 762,000. Compared to our peer nations, this is way out of line. Across the 36 OECD nations with populations greater than one million, the median size of a lower-house legislative district was 77,000. US districts are nearly ten times larger. This is bad for simple representation—Members of Congress (MCs) are too distant from their constituents—but it has other consequences for governing. I propose that the US House of Representatives expand to one thousand members. This would make the size of the average district fall to 331,000 (2020 populations), comparable to what it was in 1930. Even after this change, U.S. districts would still be the largest in the OECD (see chart, note the log scale) but managing a very large assembly carries its own issues, so there are probably limits. The UK has the largest legislature in the OECD with 650 members.4
Smaller districts mean a more dynamic and more diverse set of representatives, and campaigns with a more local, grass-roots style. Member knowledge of their own districts would improve, along with their ability to advocate for their constituents collectively and individually. The shift is not intended to skew partisanship in Congress, and I do not expect that this shift alone would favor one party over the other.5 It is true that expanding the House dilutes the overall Congressional and electoral college mal-representation embedded in the design of the Senate. This is a good thing in itself, especially for the larger states. The faster election cycle of the House also means that Congress as a whole will be more responsive to shifts in the preferences of the electorate.
The most important benefit of a bigger House would be that its subcommittees would be more capable of monitoring the work of executive branch agencies. One reason for the excessive amount of delegation by Congress—I’ve already called it abdication—is that the scale and, especially, the complexity of the US government overwhelms 435 monitors. Expanding the number of committees at work will significantly improve Congress’s ability to check the Executive.
Perhaps unintentionally, the Supreme Court has recently added another important justification for this shift. In 2024, the Court overturned the so-called “Chevron doctrine” by which courts routinely deferred to executive agencies on technical matters about which the law is silent or ambiguous. This could eventually negate a great swath of executive branch rule-making, especially at agencies like the EPA. Congress must respond. The best thing it can do is to provide much more detailed legislative direction than has been its practice in the past, including updating legislation related to regulatory actions much more frequently in order to deal with changing conditions and technologies. More, and more active, Congressional subcommittees, working directly with agency staff—this will always be where the needed expertise is—can meet this challenge. Since more democratic oversight is always better, this would actually be a good outcome of the bad decision to overturn Chevron.6
Only about a third of all Americans are satisfied with the way that democracy is working (Gallup).7 They are not wrong: We should admit that the American democracy is broken and do what we can to fix it. I doubt whether the typical American, or American politician, could articulate what’s wrong. In political science parlance, there is a failure of preference aggregation. Somehow, what people want does not get reflected in the policies of the governments they elect. People who want higher minimum wages, access to abortion, gun control, healthcare, environmental protections, and so on, routinely vote for Members of Congress who work tirelessly to block all these things. Incumbency advantage, partly an artifact of such large districts, and the protection of heavily gerrymandered districts help to insulate Representatives, party branding and skillful redirection of focus to seemingly irrelevant issues does the rest. The complexity of the American state—a good thing overall—does not help the situation: cause and effect are disconnected by the separation of powers, bicameralism, federalism, and even agency independence. Elections really don’t seem to matter—except of course that they do. This may be a lesson we are doomed to relearn again and again.
For most would-be reformers, the biggest problem is money, and the solution is public financed federal elections. Sometimes these folks say there is just too much money in politics. This is probably wrong. The issue is where the money comes from. It actually takes a lot of resources to ensure access for lesser known candidates, though much of that could be through “free” public access to media rather than paid campaign ads. The barrier here is the Supreme Court. In a series of decisions, culminating in Citizens United, the Court has concluded that money is speech, so the first amendment allows anyone to spend any amount to elect/buy their preferred candidate. Cue Elon. While patently ridiculous, Citizens is also entirely consistent with the traditional liberal approach that the Court has almost always taken on other issues. While Congress can create a more transparent financing system, it is probably powerless to exclude from or limit the role of private money in the process. My other suggestions assume that this is impossible in the near term. So what else can we do?
Congress should require that all members of the House and Senate are elected by ranked-choice voting. Congress has the power to regulate federal elections but rarely uses it. For example, in 1842 it mandated that members be elected from single-member districts. Today, most elections are decided by plurality (known as “first-past-the-post” (FPTP)), meaning that the candidate with the most votes wins, even if that is not even close to a majority of votes cast. But federal law does not require FPTP and some states use majoritarian methods today. Georgia and Louisiana use two-round systems in which the second election has only two candidates. Maine uses the modern version of this, known as “instant runoff” or ranked-choice voting. Counting sequentially eliminates the candidate with the fewest highest-rank votes and allocates their votes based on the next ranking until one candidate has a majority of votes cast.
The benefit of ranked-choice voting is that voters can express their sincere preferences without fear of “wasting” their votes on (obviously) losing or, generally, on minor party candidates. Now, we require the supporters of so-called “fringe” candidates vote strategically for their second-choice candidate in order to prevent a much worse outcome. We call this “sophisticated” voting. Elections are supposed to record the actual preferences of voters, however, so any pressure on voters to be “strategic” (to lie) is a pathology that we should avoid if we can.
Misrepresentation in the party-primary phase of elections is at least as bad as in general elections. A few states have adopted systems that bring the “primary” into the general election cycle. California and Washington have two-round systems in which the first round (called a “jungle primary”) serves as a kind of primary. In my opinion, Congress cannot dictate how parties choose their candidates, so this is not an area where federal action is appropriate. Beyond that, I fail to see how the jungle primary solves the preference aggregation problems we face. Voters are still pressured to vote strategically in the first round in order to preclude an unacceptable set of choices in the second. [The system in Alaska is better. It is also two rounds, but the second round includes four candidates in an instant runoff. This presumably allows for more “sincerity” in the first round.]
We can and should have an American version of a multi-party system. The mathematics of single-member districts virtually require a two-party system but it does not follow from this that it has to be the same two parties nationwide.8 Minor parties, or regional variations of the major parties, may be able to earn majorities regionally and join coalitions in Congress. Even when they fall short of local majorities, the use of ranked-choice voting will allow voters to express their support for minor parties, which will pressure their major party representative. Congress should also require that federal elections allow for cross-party nominations. Known as “fusion”, this is the process by which a third-party endorsement of a major-party candidate is recorded on the official ballot. For example, the Green Party might endorse a Democrat, or the Libertarian Party might endorse a Republican. This ability further helps clarify voters’ preferences and provides a bridge toward developing more robust minor party slates.
Finally, Congress needs to get its own house in order. For want of space, I won’t write much about this now, but there are some “easy” reforms that Congress can adopt right away. High on my list: Congress should prohibit members from buying and selling individual stocks and bonds. This would eliminate the appearance of one of the most egregious sources of corruption now afflicting the institution. There is an active proposal to effect this reform now. Congress should pass it today.
A functional and empowered Congress would also be better able to enact laws to check the President as a co-equal branch. I’ll leave my detailed reflections on this topic to another time but, in summary, we need to codify into law the guardrails that had previously held mainly because of custom and precedent, especially rules ensuring the independence of certain agencies, especially the Department of Justice. In the wake of the ridiculous Presidential immunity case, some of these needed repairs will require Constitutional amendments. Our best strategy is to limit our use of that approach, but sometimes it will be unavoidable.
Next week: TBD (27 May 2025)
Notes
Image: Library of Congress https://www.loc.gov/item/2009631449/ (9 May 1861., Wood, John, 1806-1866, photographer)
This is connected in a basic way to the social citizenship agenda, which depends on having a much more functional legislative capacity than we do now. Social minima can only be set through democratic conversation in Congress and, more often, in state legislatures. The courts do not have the competence to perform this role properly. To make matters worse, the courts are utterly trapped by the traditional liberal vocabulary through which they approach every issue and this language simply does not support social citizenship.
After drafting this essay, I saw a report from the Commission on the Practice of Democratic Citizenship, “The Case for Enlarging the House of Representatives”. Their recommendation is modest compared to mine (+150 seats to start), but their argument is similar. Many well-regarded individuals and institutions seem to be behind the Commission’s work and therefore endorse, at least weakly, this recommendation. I have not seen any discourse in the actual public about it, however; that needs to change. (Indeed, I follow the news more or less, but this was the first I have heard of this Commission.)
The oldest unratified amendment to the Constitution, proposed by Madison as part of the Bill of Rights (it would have been the first amendment!) read:
After the first enumeration required by the first article of the Constitution, there shall be one Representative for every thirty thousand, until the number shall amount to one hundred, after which the proportion shall be so regulated by Congress, that there shall be not less than one hundred Representatives, nor less than one Representative for every forty thousand persons, until the number of Representatives shall amount to two hundred, after which the proportion shall be so regulated by Congress, that there shall not be less than two hundred Representatives, nor more than one Representative for every fifty thousand persons.
The second article in this proposed amendment prevented Congressional pay raises from taking effect until an ensuing election. It was enacted in 1992. So Madison’s proposal is still on the table! [The prospects for passage are poor: a 50,000 target would require a House of over 6000 members in 2020.]
I have been unable to find a detailed study of this critical decision—it deserves one. The “Permanent Apportionment Act of 1929” set the total at 435 and the current procedure for apportionment. Population growth after the Civil War, combined with a strong incumbency protection bias, drove the total size of the House, but the influx of new states provided a steady catalyst for action and this process wound down in the early 20th century. The present cap is effectively based on the 1911 Reapportionment, which had expanded the House by 42 members to reach 433, plus two slots for the (last?) soon-to-be states of New Mexico and Arizona (1912). In 1920, the whole process of apportionment failed, a significant constitutional failure. This created an urgency to act around the 1930 census; hence the “permanent” part of the act.
It is not a coincidence that the process broke down in 1920, the first year in which the urban population was greater than the non-urban. Then as now, there was some rural advantage to be gained through statis, mainly because states were guaranteed at least one MC. This is a valid “political” cause of the decision, but I am not sure it is powerful enough to explain the outcome. I am told, but have not researched myself, that some Jim Crow disputes based on Section 2 of the Fourteenth Amendment are relevant. This provision says that states can be penalized seats for disenfranchisement. Some Republicans wanted to invoke it to reduce the number of Southern seats. They would have been right to. This feels like a very strong status quo pressure. (In American politics, if there is a race-based answer, it is usually the strongest one.) We should also accept that there was a “practical” cause as well: expanding the House created a management problem. This complaint was common after the 1911 expansion. Personally, I am not convinced that concerns about unwieldiness weren’t just rationalizations—these things are usually about power. A bigger House would require restructuring how the leadership and committee system works. My reaction: figure it out.
The OECD is the Organisation for Economic Co-operation and Development, a set of developed nations often used as a peer comparison group. If we include all 38 OECD countries, the median falls to 47,000. This is one question where Wikipedia was useful: https://en.wikipedia.org/wiki/List_of_legislatures_by_number_of_members. (I did update the population figures at the OECD site.) Globally, the largest districts are in India, with an average of 2.6 million constituents per representative.
The UK cannot seat all 650 MPs at once as the chamber holds only 427 comfortably. The US House chamber is 2.5 times larger (by area) however, so it ought to be possible to fit 1000 in the room (with benches, not nice chairs; if you want a nice chair, and a desk, run for Senate).
I ran only the most cursory check on this. Presently, the partisan divide in Congress (House and Senate taken together) is tolerably close to that in the American public—that is, evenly split. Increasing the number of districts ought not, or need not, change this. Smaller districts are geographically smaller, of course. My intuition is that smaller districts are less likely to be distorted by gerrymander, but I have not tried to prove this. The Commission on the Practice of Democratic Citizenship (note 2 above) did a pretty deep dive on this question and similarly found very little partisan advantage either way in a bigger House.
The doctrine was established in Chevron v. NRDC, 1984. It was overturned in Loper Bright Enterprises v. Raimondo. A better Court would back off a little on this, but ultimately Loper Bright requires that legislation be much more specific than it is now, and revisited more often—not inherently bad things.
The surest path to multi-party democracy would be to drop the single-member district standard in the United States, but this would be hard to reconcile with the fact of federalism (even at 1000 members, a state like Wyoming probably only has two representatives).