Should we expand the House?

Should we expand the House?

By Matthew Yglesias. SlowBoring.com. 

April 15, 2021. 

11 min ago

When the first U.S. House of Representatives met in 1789 it had 65 members. The size of the House then expanded steadily over time as the population grew and new states were admitted, up until the Apportionment Act of 1911. That law set the size of the House at 433 members but also made a provision to add two new seats upon the admission of New Mexico and Arizona to the union.

Then came the controversial 1920 Census, after which Congress refused to reapportion the House at all in a huge breach of constitutional norms.1

But a new Reapportionment Act did happen in 1929, and after that, we went back to the routine of changing up House apportionment after each Census to account for shifts in the population. But we didn’t go back to the old habit of changing the overall number of seats, which instead stayed fixed at 435. Even the admission of Alaska and Hawaii didn’t lead to the creation of new seats for new states the way it had for New Mexico and Arizona.

People sometimes talk about this as if the population per member of Congress started soaring after 1911. The truth is that it had been rising steadily throughout the 19th century, but certainly it has continued to rise over the past 100+ years with no increase in seats.

(Graphic by Tk62619)

It’s become sort of fashionable to argue that this was a mistake and we should increase the size of the House and go back to routine increases as part of the decennial routine. A lot of Slow Boring readers have asked me what I think about this, and my feelings are ambivalent.

I don’t think it’s a bad idea. I wouldn’t lift a finger to block it from happening or oppose pursuing the idea as part of a coalitional strategy or a logroll. But I also would not donate money to a group pursuing this idea nor spend time and energy on it. The main benefit I can see is a small improvement in proportionality. The biggest thing is that while I don’t see any big downsides to expanding the House, I do think enthusiasm for it is based in part on a false idea that has harms elsewhere in the discussion of political institutions — a misguided sense that smaller-scale electoral entities are in some sense “closer to the people” or more democratic.

Cases for expanding the House

The way a lot of people like to argue about politics is they take an idea and then come up with lots of benefits that they think the idea would have. I always find this model a little bit off-putting and prefer to start with a problem and then think of well-tailored solutions to the problem.

House expansion definitely seems like it falls into the first bucket of ideas — it’s not that there’s some concrete issue that folks think House expansion solves. It’s more like folks think House expansion has an array of benefits. Surveying the arguments, we see a lot going on.

The American Academy of Arts and Sciences argues that large units make elected officials less responsive to constituents, and they note that House expansion would reduce Electoral College skew and that “the current Capitol Building could easily accommodate an additional fifty members.”

David Litt in the Atlantic says an expanded House would be more responsive, and he also notes that we have seen a growing ratio of population to House staff.

Chris Wilson in Time Magazine says an expanded House would make gerrymandering more visible to the courts, echoes the Fordham point about campaigns being cheaper, suggests you’d get more occupational diversity in Congress, and also that it would be harder for leadership to whip votes.

The NYT Editorial Board cites Electoral College skew, disproportionality, and responsiveness, and they also argue that there would be more toss-up districts with a larger House.

Payton Young for Brookings says it would be easier for challengers to fundraise, you’d have less gerrymandering, there’d be less temptation to buy off individual members with picayune pet projects, and you’d get more representation of diverse communities and more emphasis on local concerns and door-to-door canvassing.

My main observation about all of this is that for most of these problems, it doesn’t seem like House expansion is the optimal solution.

Consider the Electoral College. Wyoming has 582,328 residents whereas South Dakota has 892,717, but they both have three electoral votes thanks to having one House seat. But Maine — with 1.3 million people — has four electoral votes. If you made the House bigger, you would draw finer-grained distinctions and South Dakota would fall between Maine and Wyoming and its number of EVs rather than being exactly identical to Wyoming.

That said, this is not at all the main source of Electoral College skew. The Senate is what creates small-state EC bias, and then swing-state bias comes with the fact that winning 50.1% of the vote in Pennsylvania versus winning 49.9% of the vote creates a huge 20 point swing in EVs. What’s more, if you think the winner of the popular vote should become president, the way to accomplish that is to scrap the electoral college, or get enough states to agree to have their electors all go to the winner of the national popular vote, not by tinkering with House apportionment.

Similarly, if you’re worried about gerrymandering (and I think you should be), then the solution is anti-gerrymandering rules.

Litt’s point about there being too few congressional staffers is extremely well-taken, and it suggests the obvious remedy of increasing congressional staff budgets. Indeed, I would actually argue that the shoe is on the other foot here. One thing congressional staff does is constituent work, in which case spreading the work out across more offices would be good. But they also do policy work. If you hold the number of House members steady but increase the staff budget five-fold, then members will be able to get more specialized policy advice from more experienced staffers. But if you pair a five-fold increase in budgets with a five-fold increase in the size of the House, then you’re mostly diluting the impact of more staff spending.

The bona fide issue here seems to be responsiveness.

Helping challengers win

There’s a refrain running through these pieces which suggests that most incumbents are sitting fat and happy and not facing much electoral competition because it’s too taxing for challengers to raise money and give them a tough fight.

This seems to me to reflect a view of politics that’s probably 20 years out of date and getting worse.

It used to be that you had really strong incumbency effects, money for challengers was scarce, and politics was not that nationalized. You could plausibly tell a story about a long-time incumbent who was maybe a bit long in the tooth losing to a vigorous challenger, but also about challengers not being able to wage a vigorous campaign thanks to lack of money. These days politics is just much, much more nationalized.

Maciej Ceglowski made a really wonderful effort in the 2018 cycle to raise money from progressive-minded tech people to back vigorous campaigns by energetic progressives in rural districts. He sent a bunch of money to a bunch of great candidates who ran really admirable campaigns and then just lost in droves, even in a really solid national political environment.2 The exception that proves the rule here — Jared Golden —represents a Maine district that was Dem-leaning in national politics in the very recent past.

A more typical rural Dem story is Collin Peterson, who had a ton of seniority in rural Minnesota, a moderate voting record, and a history of beating the odds in the tough 2010 and 2014 cycles. But even though the national climate for Democrats was much better in 2020 than it had been in 2010 or 2014, politics became much more nationalized. By 2020, Peterson’s constituents didn’t care about Peterson — they cared about national partisan politics.

Now I’m sad Peterson lost. But from a pure standpoint of “do incumbents have it too easy?” this is progress. A challenger to any halfway vulnerable incumbent can easily raise money and mount a viable challenge. As we saw in 2018, given a favorable national environment, it’s really not hard to beat GOP incumbents even in districts that have a modest pro-GOP skew. There’s no daunting incumbent advantage anymore.

Nonetheless, most incumbents are perfectly safe because most seats are gerrymandered to be non-competitive.

Expansion doesn’t help with gerrymandering

A lot of these takes assert that an expanded House would help with gerrymandering, but I think this is plainly not the case.

For starters in the limiting case, you can’t gerrymander the governorship or presidential elections, which is why Democrats can elect governors and win presidential elections in Pennsylvania, Wisconsin, and Michigan, but can’t break GOP control of the state legislatures.

Indeed, with 99 seats, the Wisconsin State Assembly has incredibly small districts. The equivalent would be something like a 5,000 seat U.S. House of Representatives. So how are they doing on gerrymandering? Well, Democrats won 38 seats there even while Joe Biden won state-wide. Back in the 2018 blue wave year, Wisconsin Democrats scored a 7.5 percentage point swing in their favor, which netted them exactly one more seat.

As far as these things go, it’s not even a particularly egregious map in terms of being “funny-looking” or anything.

It’s just very well-done. The Green Bay area is divided into one super-blue district where Republicans don’t even bother mounting a challenge, plus two safely padded suburban seats. Nothing about the small size of the seats prevents the basic gerrymandering tactics of packing and cracking from working.

Non-electoral responsiveness

This leaves us with the idea that smaller units would lead to more responsiveness in a non-electoral sense.

There’s obviously something to this — in D.C., we have a mix of at-large and ward-based councilmembers, and if you email your ward rep you are much more likely to get a prompt response than if you email one of the at-large members. If you made House districts really small, you’d get better constituent service — especially if you kept staff budgets equal on a per-member basis. Against this, if you greatly increase the number of districts, the average quality of a member of Congress is going to go down. Imagine the typical safe seat backbencher. Now imagine a very similar person, but all-around less impressive. That’s your marginal new member of Congress.

But then here’s the other thing. Deeply embedded in the American political tradition is the idea that more local institutions are better than more distant ones. There was the idea that the Electoral College would be a bunch of local notables who are sent to the capital to deliberate on the choice of the president because constituents know a lot about their local notables but not a lot about national political figures.

Is that how contemporary politics works?

I know the name of Joe Biden’s dog, Major. And I know that Kamala Harris went to high school in Montreal. These are very present characters in my life as a 21st-century person who is constantly immersed in media. I remember the kerfuffle about Nancy Pelosi’s ice cream, and I have a nuanced take on the differences between Bernie Sanders and Alexandria Ocasio-Cortez, and I also understand why a lot of Elizabeth Warren fans were hurt and disappointed that AOC backed Bernie in 2020 but I think I also understand why they fundamentally misunderstood AOC. By contrast, I keep forgetting how to spell my D.C. Council member’s name even though I’ve actually met her IRL, emailed with her, etc. Most people probably couldn’t even name their state senator, and there is nobody who has more knowledge about the state senator’s political positions than they do about Joe Biden’s.

The bigger units get more media coverage and are most present in our lives in a real-world way. I don’t think that’s a reason to oppose expanding the House. But I do think House-expanders misunderstand this point in a way that is connected to a broader question about political reform.

Political reform for the real world

I think gerrymandering is a real problem, and the solution is anti-gerrymandering rules. Problem —> solution.

And I think another big problem is that too much governance in the United States happens as a second-order consequence of national trends. Over and above the problems induced by gerrymandering, state legislators simply aren’t judged on the basis of what they do on the issues that are actually under state control. Instead, people assess their feelings about the president and then vote accordingly for tons of down-ballot races they don’t really pay attention to. The only mitigating factor is incumbency and inertia. This is why realignments start at the top and then trickle down to state legislatures.

Governors seem to be able to lean against this. A Laura Kelly or a Larry Hogan can win based on people’s assessments of state issues.

Yet the governorship is generally a weak office, with executive authority in most states split among tons of separately elected officials — attorneys general, insurance commissioners, and all kinds of other weird stuff. That’s a bad idea. In general, more centralization of authority would create a healthier system in which people have a stronger sense of what’s going on. I also think we should resist the multiplication of different levels of local government. It’s good when big cities are their own county (like San Francisco or Philadelphia) rather than being embedded in a larger unit (like Los Angeles or Chicago). I like mayoral control of school systems rather than separately elected boards. And for places that aren’t urban, I think it’s generally better to just have a strong county government rather than lots of little incorporated towns.

For Congress, beyond the gerrymandering point, I think this ought to be a better-paying job. It’s not that members are currently making starvation wages or anything. But it should be a job that a successful professional — a doctor or a lawyer or an executive or a school principal or whatever — considers highly desirable. Something where you’d be strongly motivated to go through the hassle of taking the risk to try to run. And where once you’re in office, the job is so good that you really don’t want to lose rather than just kind of shrugging and saying “well, if I lose I’ll make more money on K Street anyway.”

Andrew Hall wrote a great underrated book called “Who Wants to Run: How the Devaluing of Political Office Drives Polarization” showing that basically the worse you make the job, the more the only people who want it are mediocre ideologues. This is admittedly a bigger issue in state legislatures than in Congress, but it’s present everywhere.

Last but by no means least, my real white whale is proportional representation via single transferrable vote. In a state like Ohio, there should just be a statewide contest for the 12 seats. You might need to split New York and Florida in half and Texas and California into thirds. But proportional representation really does make gerrymandering harder. It also means that currently unrepresented people like “Republicans who live in Los Angeles” and “Democrats in Arkansas” would have representation in Congress. It would also of course genuinely mean that third parties would have a shot at representation. You’d have a less polarized landscape and a healthier politics.

In that specific context, I think there’d be a strong case for significantly expanding the House. That’s because a proportional election in a state like Idaho that only has two House members gets a little silly, and it’s totally meaningless in the six states who only have one rep. But as things currently stand, it feels to me mostly like a solution in search of a problem.