Congressional capacity—or the lack thereof—is on the minds of a lot of people in Washington these days, left and right. And almost all of them are very concerned. Flat personal office staffing numbers, combined with an ever-growing elecotrate and radical increase in electronic mail to Congress, has stretched resources thin and forced members to shift staff away from policy and toward constituent work. Increased executive branch power and responsiblity post-9/11 has created increased oversight demands, but committee staff levels remain strikingly lower than they were 30 years ago. Saff pay continues to stagnate, fueling a brain-drain that lands a significant portion of the institutonal conressional knowledge in lobbying shops and think tanks, rather than on the Hill. Congress’s own non-partisan repositories of institional knowledge—CBO, CRS, GAO and other legislative branch agencies—also have suffered from significant staff reduction in the last generatoin. All of this sums to a gloomy picture, in which large numbers of senior Hill staffers do not believe Congress has the resources or knowledge base to compete in the separation of powers system.

It’s no secret that one of the biggest obstacles to increased congressional capacity is politial: Members of Congress are loathe to take votes to increase spending on their own staff. It’s not that they personally oppose it—I’ve never talked to a member in private who didn’t wish they could have more staff and pay them better—but instead a simple iron electoral logic: voters hate it. Rational or not, many voters can’t stand anything that looks like Members spending money on themselves. And so electoral challengers latch onto it like demogogues, and many Members do all they can to burnish their image as frugal spenders.

This downward democratic pressure on congressional spending has predictable results. Memebers themselves have not received a salary increase since January 2009; in fact, the automatic COLA system they set up in 1989 to avoid tough votes on their own pay has been virtually scuttled. Instead, they take a vote each year to deny themselves the COLA. Since the great recession, staff pay hasn’t fared much better. The House appropriations account that pays for personal office staff (the “Members’ Representational Allowance,” or MRA) is down over 14% since it peaked in FY10, and in real dollars is lower than it was 20 years ago. Things are almost as bad in the Senate. Committee funding shows a similar trajectory. It’s just really tough to get Members to vote for these increases; they are more inclined, politically, to want to “lead by example” and “tighten their own belts.”

A decade ago, I did a stint as a staffer on the House Appropriations Legislative Branch Subcommittee. During that time, I noticed something odd. Our bill, the annual Legislative Branch Appropriations bill (FY19 bill here; FY19 report here), funded all the accounts in the legislative branch, including those for congressional staff. Some of these accounts and sub-headings were well-known by the members and the press, like the “Members’ Representational Allowance” heading (which funds staff and resources for Members’ personal staff) and the “Committee employees” heading, which funds staff and resources for, well, committee employees.

But there were other accounts—like the “Government Contributions” account—that literally no one talked about, ever. And it’s not like this was a size issue; the “Government Contributions” account is the second largest account ($233m in FY2018) of all accounts related to the House of Representatives, second only to the MRA account ($562m) and larger than the committee funding account ($150m).

Here’s the kicker: the Government Contributions” account is entirely used to compensate congressional staff. It pays the employer side of social security, retirement contributions, health insurance, and other benefits. In effect, it operates like any employer does in the private sector, where your salary is only part of your total compensation package. You may make $65k as your salary, but it costs your firm a lot more than that in total compensatoin to employ you. The upshot of this is something of a revelation, however, in the context of the Congress: staff compensation comes in a variety of forms, through a variety of different accounts, with a variety of levels of visiblity, both absolute and in relation to the indiviual Members.

Consider the contemporary arrangement: the Legislative Branch Appropriations bill provides a single account for the MRA ($562m in FY19), which the Comitteee on House Administration then authorizies each Representative to draw off of, up to a specific amount ( about $1.3m, depending on travel distances to their district and number of mailing addresses).

All of the expenditures from this money are publicly reported in the Statement of Disbursements of the House, organized by Member. It is very easy to look up how much your Representative spent on staff salaries and travel and office supplies.

But the “Government Contributions” account is almost impossible to link to indiviual Members. The costs are reported, but mostly in aggregate and never linked to individual Members. You can see the House spent close to $3M on employee health insurance in the last quarter, but it’s just an aggregate lump.

To the degree that Members want to minimize being seen as spending money on themselves, this system works pretty well. If the $1.3m you spent last year running your office only inclues actual salaries and not total compensation, that sure beats having to report the total compensation number, which might upset the public and hand your electoral opponent an even shiner issue. But it also raises a tantalizing possibility: why not move more staff expenditures “off-book” from the MRA and into this sort of arrangement, as a way of increasing staff compensation without running head-on into the downward democratic pressure that keeps Members from approving such increases?

In fact, this is exactly what the House has done several times in the last decades. Prior to June 2009, Members could choose to provide staff in their office with a transit benefit to cover the cost of commuting to the Hill; the benefit was paid for out of the MRA. Effective in June 2009, however, the Committee on House Administration made the transit benefit available to all House employees, and paid it out of a central account. Ditto with student loan repayment. Begun in May 2003, repayments for eligible staff are paid out of a central House account, not the MRA. And you guessed it—the transit benefit and student loan repayment programs are funded by, yes, the “Government Contributions” account.

These sorts of benefits help Members in two ways: they make employment in the House more attractive to candidates and current employees, and they free-up the MRA to be used for other purposes (including increasing regular salaries). Numerous obvious extensions exist: the Senate just approved funding to pay interns, which will not be paid out of the SOPOEA (the MRA equivalent in the Senate); new child-care or maternity/paternity leave benefits could be paid outside the MRA; health-related benefits could also be addressed this way.

But even more importantly, the House might fundamentally rethink what the MRA pays for. For example, right now Members use the MRA to pay the costs of leasing their district offices. But Senators do not; leases for Senate home-state offices are paid for through funds appropriated to the Senate Sergeant-at-Arms. If the House were to adopt that policy, they could free up tens of thousands of dollars for staff salaries in each office, without increasing the MRA a dime.

There are, of course, limits and democratic costs to these sorts of transparency moves. While it seems perfectly justifiable to take fixed employee benefits like transit or retirement that the employing members have no discretion over and hide their visibility, it would be another thing entirely to start playing these sorts of visibility games with actual discretionary salary that the individual members have control over. Part of the rationale of the MRA is to make Members accountable to a fixed buget for their operations, and part of that accountability is public disclosure of their choices. But in a lot of areas, it makes sense to “off-book” new benefits for employees. And in the case of things like district office-space, it’s not entirely clear why Members should be individually held to financial account, particularly since they aren’t charged rent on their DC office space.

In sum, the structure of compensating staff has implications for how much the downard democratic pressure of the voters impacts the willingness of Members to provide additional funding for congressional capacity. There are obviously limits here, too. It would not be politically feasilbe to double the “Government Contributions” account overnight and load it up with staff compensation; voters would see right through that. But on the margins, moving compensation out of the MRA when it can reasonably be justified as a general benefit that should be paid for by the House is a strong strategy for improving staff retention without creating electoral firestorms about Members spending money on themselves.