When Did Revenue Stop Mattering?
Decoupling federal revenue from spending policy has taken us exactly where an unplanned build always ends up: buried in debt.
Early in my career, I was hired to stand up New York City’s Independent Budget Office, a municipal agency modeled after the Congressional Budget Office with a mission to enhance official and public understanding of fiscal issues facing New Yorkers. Importantly, the agency was designed as a counterweight to the mayor to bring an analytically rigorous and nonpartisan perspective on city budget matters.
During the agency’s initial months, when we were developing a research agenda, one piece of advice came up time and again: prepare an independent revenue estimate, separate from the mayor’s. The reasoning was straightforward. Budget participants in the City Council, as well as others involved in the budget process, shouldn’t be wholly dependent on a single forecast produced by a person with the most incentive to slant it to accommodate political goals. An independent estimate would provide an alternative against which to consider the mayor’s number, the starting point for crafting a budget.
That suggestion echoes how budgeting actually works not only in NYC and other state and local governments, but for most organizations and households. Make a responsible forecast of how much money you have to spend in a given period, then develop a spend plan consistent with that estimate.
The revenue estimate isn’t a side calculation; it’s the foundation upon which everything else gets built. There’s no procedural escape hatch, no alternate track, no way to spend first and figure out the revenue consequences later. The number sets the ceiling, and every serious actor in the budget process — the mayor’s OMB, the Council, the comptroller, the press — treats it that way. That’s precisely why so many of my IBO advisers thought an independent estimate would matter so much. At a minimum it would arm non-mayoral players in the budget process with information to push back on the mayor’s forecast when a significant difference in estimates existed. That would lead to a more informed debate than simply and blindly accepting whatever the mayor’s team deems to be the correct number.
The Federal Contrast
Though I left New York in the early 2000s (not entirely due to my now-relieved despondency over the Knicks) and moved to the DC area, I think about that experience often when I look at how our national government budgets. The good news here in the nation’s capital is that independent revenue estimates are routinely developed by CBO and the Joint Committee on Taxation and are of excellent quality: professionally rigorous, technically serious, largely insulated from political pressure. That’s not the problem.
The problem is that, unlike at the local level, the federal revenue estimate has become something of an afterthought. It’s a number that’s meticulously calculated and reported but no longer treated as foundational to the budget process. Spending and revenue policy are built without any blueprint for what the finished structure is supposed to look like. Instead, each gets built on its own with no one checking whether the two pieces are architecturally consistent.
Nobody is asking, “given what we expect to collect, what can we afford to do this year, even if we have to borrow some reasonable amount to cover the gap?” Instead, they’re asking “given our apparently inexhaustible capacity to borrow, what do we want to do this year?” The revenue number only informs the bill that arrives afterward, telling you how large the gap turned out to be.
This casual attitude toward borrowing was affordable for a long time because it was, in a literal sense, cheap. For most of the period since the 2008 financial crisis, and especially during the pandemic, the U.S. government could borrow at historically low interest rates. Deficits grew, but debt service didn’t bite the way conventional fiscal logic says it should have, and there was no market signal — no spiking yields, no rollover crisis — punishing our elected officials for letting spending and revenue drift apart. That cushion is thinner now. Borrowing costs have risen substantially, debt service is consuming a huge share of the budget, and the assumption that decoupling carries no real penalty is becoming harder to sustain with each passing year.
This is not a story about bad faith or a particular administration. It’s structural, and it’s bipartisan, because both parties have learned to use the same gap for opposite purposes. One side runs tax policy through the gap, the other runs spending through it. And neither has much incentive to close it, at least right now, because closing it would mean accepting a constraint neither wants.
Why the Decoupling of Revenues and Spending Is Structural
A few features of the current process make this kind of decoupling almost inevitable rather than incidental.
The budget resolution was originally supposed to be the moment when Congress looked at spending and revenue together and set a coherent top-line relationship between them. In practice, the resolution has been hollowed out into little more than a procedural trigger for reconciliation — a formality cleared so that a tax or spending bill can later move with a simple majority, rather than a genuine occasion for Congress to reconcile the two sides of the ledger.
Baseline budgeting compounds the problem. Both spending and revenue are projected forward as a continuation of current law and, at times, current policy, which means the baseline itself often embeds assumptions about extensions and renewals that were never actually voted on as a package. The baseline becomes a kind of fiscal autopilot, and decisions to deviate from it are made one at a time, without much pressure to ask how the deviations on the spending side relate to deviations on the revenue side.
And a large and growing share of spending simply isn’t part of an annual conversation with revenue at all. Entitlement spending grows according to formula and demographics, not appropriations choices, which means it expands or contracts with essentially no annual look-back at what revenue is doing.
Tax expenditures make the same point from the other direction. A tax deduction, credit, or exclusion is functionally spending. It’s the government deciding to direct resources toward a favored purpose, just as a grant or a subsidy would. But because it shows up as foregone revenue rather than an appropriation, it never passes through the spending side’s annual review at all. It simply sits in the tax code, growing or shrinking with economic activity, treated by the process as a revenue matter when it is, in substance, a spending decision nobody is re-examining.
A Consequence Problem, Not an Estimating Problem
None of this means the government needs to import a state or local-style balanced budget mandate. That’s a different argument, with its own tradeoffs, and I’m not making it here. What the local government comparison illustrates isn’t that balanced budget rules are the answer. It’s that revenue can function as a real constraint on spending decisions when institutions are deliberately built to make it one. At the local level, revenue matters because the process requires it. At the federal level, crafty legislators and staff have figured out how to move fiscal legislation without a conversation about how spending relates to revenue. So, increasingly, the relationship doesn’t matter and the conversation doesn’t occur.
Washington still estimates revenue carefully. It just stopped asking the number anything.
This is one of the core problems the Reform for Results budget process working group’s recent manifesto, Radical Reform of the Federal Budget Process: A Call to Action, is organized around addressing: rebuilding a process where spending and revenue are forced back into the same conversation, on the same timeline, with real consequences attached to the gap between them.
Until Washington rediscovers what every city budget office, every corporation, every household already knows — that the amount of money you spend has to have some semblance to the amount you expect to come in — the number on the revenue side will be about as important as a “Don’t Walk” sign in NYC. Everyone sees it, no one really cares.