There’s been some interesting debate in the last week about expanding the size of the House. Lee Drutman has been on the case in favor of this for a while, the New York Times editorialized in favor of it this week, and Jonathan Bernstein made the case against it today. I tend to side with Bernstein in this argument, but there are important and good arguments both ways. I would encourage you to read all three pieces if you are interested in the topic.

What I want to take up here is a much more narrow concern: how much would this cost? In essence, what would be the price tag, per new seat, of increasing the size of the House?

Now, I want to make it perfectly clear up front that (1) the total cost is chump change in the big picture if there are good normative reasons to increase the size of the House; and (2) I’m personally in favor of spending billions more on legislative branch capacity, regardless of whether the size of the House is increased or not.

That said, voters hate when Members of Congress spend money on anything that can be construed to be related to themselves, so it is politically difficult to do things like raise members’ salaries, increase legislative branch staff or pay, or provide for increased funding for legislative support staff. Every member of Congress I’ve ever talked to would love to have more resources on the Hill, but many of them are terrified of taking those votes. So the cost of expanding the House is a relevant political issue.

I bring all this up because the New York Times estimate in their editorial was laughably small:

Nor would growing the House cost too much. Salaries for lawmakers and their staffs would total less than one million dollars per representative — which means a couple hundred representatives could be added for the price of, say, five F-14 fighter jets.

I’m not sure where they got this figure, but it’s either poor estimation or intentional lowballing. The overall cost of adding 158 seats—which the Times suggests—would be much, much higher. Even the F-14 comparison is silly; jets are mostly one-time capital purchases, whereas 158 new members would create annual recurring costs—forever.

Let’s add it up.

Start with the annual, recurring costs. The most basic cost is the salary of the Members. This is currently $174,000 annually. The next cost is the Members Representational Allowance (MRA), which is the money given to each member to pay for staff, mail, travel, district office rent, and officer supplies. It varies slightly depending on how far a Members home district is from DC and how many postal addresses are in their district, but in 2017 the average MRA was $1.32 million.

So on the MRA and Members’ salary alone, we’re already at $1.494 million per additional member, per year, which is 50% higher than the Times estimate. Now the Times estimate does only mention the salaries of Members and staff, not the other aspects of the MRA (or other costs we’ll get to in a moment). But that’s a pretty crappy dodge: if the cost of having an office is more than the salaries (it is), then you should count the total cost of having the office, not just the salaries.

Could the MRA be cut if we expanded the House? In theory, perhaps. Smaller districts have fewer postal addresses, and fewer constituents would mean a decrease in casework for each office. But we’re talking about increasing the house by 36%, not 300%. Each district would go from about 750k residents to about 550k residents. Congressional staff are already horribly overworked, and a fair amount of the work is district-size independent. Maybe the MRA takes a small haircut—to account for the loss of postal addresses—but I really can’t see that.

Next up is the agency contribution costs. The MRA pays the salaries of the staff of Members, but it only covers the base salary, not the benefits side. The government’s side of the retirement, TSP matching, and health costs all come out of a separate account. It’s a $249 million account right now in the legislative branch appropriations bill. Now, that covers all House employees (including committee staff, officer staff, and support staff), so it’s not like we’re looking at 249 x .36 more millions of dollars, but it’s a chunk.

Which brings us to support staff. Would expanding the House require an expansion of non-Member staffing? Almost certainly. More Sergeant-At-Arms staff. More Legislative Counsel staff. More Capitol Police officers. More Architect of the Capitol employees to care for the offices. What about committees? In theory, you could get by with the same committee structure, but my colleague Mark Harkins made the smart point that a 36% bigger House probably creates new committees—divide up Energy and Commerce? Split Financial Services in two?— which probably entails new staff. What about CRS and CBO? A bigger house that didn’t increase the capacity of the support agencies would inherently see diminished service capabilities for each member. So you might see increases there.

So that’s a sizeable annual cost per new member: the Members salary + the MRA + the government contribution for the new Members staff + increased support staffing + possibly new committee staff. There’s simply no way this is anywhere south of $2m/year/member for each additional member. And it’s probably higher. But let’s call it $2m, which means $318 million per year going forward. That’s not too bad as a raw dollar amount, but it is a 7% increase in the total budget for the legislative branch. And it’s a conservative estimate.

Rayburn, under construction, early 1960s

But the real headache is the one time fixed cost of doing this: all of these new members and staff need physical space to sit in. There’s already a space crunch on the House side of the Capitol; some committee staff have been annexed out of the Cannon, Longworth, and Rayburn buildings, and now sit down at the Ford and O’Neil buildings. There’s just no way around the fact that 159 new Members would require a new building on the House side. That’s not easy to price out, but we have some guidelines. The Capitol Visitor Center cost $600 million. The Rayburn building cost just under $100 million—in the mid 1960s—which is roughly $800 million in today’s dollars. Given the increased costs of security technology, I don’t see how you are doing a fourth building for less than a billion dollars. In addition, the existing Capitol complex would probably need renovation: the House chamber probably isn’t suitable as is for 600 members, though that’s small beer in the grand scheme of things. A new building is also going to create a sizeable increase in Capitol Police and Architect staff to care and secure it, so there are additional annual costs associated as well.

Anyway, that’s my conservative estimate for increasing the size of the by 159 Members. A billion dollars up front in capital costs, and then ~$318 million in annual recurring costs. Over ten years, we’re looking at something like $4.2 billion, minimum. And those annual costs could be much higher depending on how much support staffing is ramped up out side of the Member offices, but take it as a baseline estimate.

This may be worth it—as I said at the outset, I don’t think these costs are high enough to be, in and of themselves, damning objections to our normative judgements about the proper size of the House—but it’s a big enough price tag that it’s going to weigh heavily on the political calculations about chamber expansion.