How the Federal Budget Process Is Supposed to Work

Last updated 2 months ago. Our resources are updated regularly but please keep in mind that links, programs, policies, and contact information do change.

Since the modern budget process took effect in the mid-1970s, Congress has finished all of its regular spending bills on time exactly four times: fiscal 1977, 1989, 1995, and 1997, according to a Pew Research Center review of every appropriations bill since 1976.

Here’s the short answer to why it almost never happens: the process has a clean legal design and a messy political reality bolted on top of it. The gap between the two is not an accident of any single year.

The 1974 Blueprint and Its Moving Parts

The framework everyone still argues over comes from the Congressional Budget and Impoundment Control Act of 1974, Public Law 93-344, enacted July 12, 1974. Its stated purpose, in the compiled text, is “To establish a new congressional budget process; to establish Committees on the Budget in each House; to establish a Congressional Budget Office; to establish a procedure providing congressional control over the impoundment of funds by the executive branch; and for other purposes.”

Impoundment matters here. As the Government Accountability Office puts it, the impoundment title created the procedural means by which the Congress considers and reviews executive branch withholdings of budget authority. Translation: the President can’t simply refuse to spend money Congress appropriated.

The key term is budget authority. That’s the legal permission for an agency to enter obligations that turn into spending, and appropriations bills grant it for most discretionary programs, from defense procurement to education grants.

Discretionary is the operative word. Congress decides that spending each year. The rest, Social Security, Medicare, interest on the debt, flows automatically under permanent law. According to the Bipartisan Policy Center, annual appropriations only steer about 25 percent of federal spending.

The rest of the toolkit rounds out the picture. A Congressional Research Service overview lists the pieces: “the President’s budget submission, the budget resolution, reconciliation, sequestration, authorizations, and appropriations.” For a closer look at how authorizing a program differs from funding it, see our explainer on authorizations versus appropriations.

The Calendar Nobody Keeps

The 1974 Act sets a detailed schedule, and it reads like a well-run office.

The President submits a budget request by the first Monday in February. The CBO delivers its economic and fiscal outlook by February 15. Other committees send their “views and estimates” within six weeks. The Senate Budget Committee reports a budget resolution by April 1, and both chambers agree on it by April 15.

That resolution is the pivot. It isn’t a law. It’s a concurrent resolution that never goes to the President, so there’s no veto, and it doesn’t create binding legal rights.

What it does is set a ceiling on total discretionary spending, the 302(a) allocation, then split that pot among the appropriations subcommittees as 302(b) allocations. In plain terms: the resolution decides how big the pie is, and how it gets sliced, before anyone writes the checks.

Then appropriations begin. The House can start floor action by May 15, its committee reports the last bill by June 10, and the House finishes by June 30. Everything is supposed to be signed by October 1, when the fiscal year starts.

There are twelve regular appropriations bills, each matched to a pair of House and Senate subcommittees (there used to be thirteen). Defense, Agriculture, Homeland Security, Labor-HHS-Education, and the rest. When they move through committee hearings, floor amendments, and a House-Senate reconciliation, that’s what insiders call “regular order.” For the full step-by-step walk-through, we covered it in how Congress funds the federal government.

It’s a genuinely elegant design. Deadlines cascade so agencies know their budgets before the year begins.

And then it doesn’t happen.

How Often Regular Order Actually Runs

The numbers are stark. Since 1996, Congress has never passed more than five of its twelve regular bills on time, and in 13 of the past 15 fiscal years reviewed by Pew, it passed zero by October 1.

The budget resolution, that pivotal step, is skipped or late almost as a rule. Pew found its April 15 deadline missed in 45 of the past 51 fiscal years, including fiscal 2026.

When the final spending bill does arrive, it arrives late. Pew clocks the average gap since fiscal 1998 at 117 days between the start of the fiscal year and the last bill becoming law.

So what keeps the lights on in the meantime? Two workarounds have become the real system.

Sources: Pew Research Center; Congressional Budget Act timetable.

The Continuing Resolution Becomes the Norm

A continuing resolution is the emergency patch. The Bipartisan Policy Center defines it as “a temporary ‘stop gap’ by which Congress funds the federal government for a limited period to avoid a lapse in appropriations (more commonly referred to as a government shutdown).”

A CR mostly freezes each covered program at last year’s dollar level and bars new starts. It’s an appropriations law itself, just a formulaic one, so it still needs both chambers and a presidential signature.

The reason a CR is the difference between open and closed is the Antideficiency Act, codified at 31 U.S.C. §§ 1341, 1342, and 1517. It bars officials from spending or committing money before there’s an appropriation. It even forbids accepting free labor, except in emergencies “involving the safety of human life or the protection of property.” In practice: when the money lapses, agencies can’t just carry on and settle up later. They have to stop.

The BPC identifies six standard parts of any CR: which agencies it covers, how long it runs, the funding rate, the ban on new activities, “anomalies” that let a specific account deviate, and any attached policy provisions. That fifth one, anomalies, is where the fights hide.

Stopgaps were meant to be rare. They aren’t. The Bipartisan Policy Center reports an average of five CRs enacted per fiscal year from 1998 through 2025, with a peak of 21 in fiscal 2001, and the government running on temporary funding for roughly four months of the average year.

Fiscal 2026 followed the familiar script. Congress passed multiple CRs, then finally cleared full-year funding in two consolidated packages: H.R. 6938, signed January 23, 2026, and H.R. 7148, signed February 3, 2026, which absorbed the House-passed H.R. 7006. Even then the funding was uneven: a partial shutdown hit the Department of Homeland Security on February 14, 2026, when its money lapsed while other departments had full-year funding.

When negotiations fully stall, Congress sometimes just extends prior-year funding for the whole year, a full-year CR. The BPC’s objection is blunt: agencies get forced to operate as if the current budget year is largely the same as the last one, when no two budget years are the same.

The other workaround is the omnibus, a single bill that bundles many appropriations measures, sometimes all twelve, into one giant package. It moves fast and lets leadership strike one grand bargain instead of twelve. The CRS notes the cost, though: limited debate and the difficulty of evaluating so many provisions at once.

The Tax Nobody Sees on the Budget

Averting a shutdown is not the same as being funded well, and this is where the abstraction becomes concrete for real people.

The GAO has studied what repeated CRs do inside agencies. Its examination of six of them, the Administration for Children and Families, the FDA, two Veterans Affairs administrations, the Bureau of Prisons, and the FBI, found the same pattern everywhere: delayed hiring and repetitive work, like issuing a string of short-term grants and contracts instead of one clean annual award. Michelle Sager, GAO’s managing director of strategic issues, has testified that this uncertainty delays contracts and reduces the level of service agencies can deliver.

The cost is real even when nothing dramatic happens. Consider the facilities projects at Joint Base San Antonio delayed under fiscal 2024 CRs: GAO’s 2026 defense work found the delays drove an average 66 percent increase in contract costs, with one contract more than doubling from $579,000 to $1.4 million. Multiply that logic across agencies and decades, and the BPC’s phrase for it, an inefficiency tax, starts to look literal.

Grant recipients feel it downstream. The Department of Education notified states in 2025 that it was withholding roughly $6.2 billion in approved K-12 formula funds for migrant education, English learners, and after-school programs — money Congress had already provided in a full-year CR. The administration released the funds in late July after lawsuits and bipartisan pressure, but districts had spent a month planning fall budgets without them. An Urban Institute survey found one-third of nonprofits hit by some funding disruption in early 2025. These are organizations with thin reserves, waiting on money Congress already approved.

Riders and the Art of the Poison Pill

Now add the leverage points. A rider is a policy provision hitched to a funding bill, restricting how money can be used or forcing a policy change that would struggle to pass on its own.

The House and Senate both have rules meant to keep policy out of spending bills. House Rule XXI bars provisions that “change existing law” in general appropriations bills. Senate Rule XVI bars “general legislation” in appropriations amendments. In practice, these rules get waived, worked around, or simply overridden by omnibus negotiation.

Because appropriations bills are must-pass, a rider that a bloc of members won’t swallow can kill an entire bill. The informal term is a “poison pill.”

The White House plays the same lever from the other end. The Biden administration threatened to veto a House GOP defense bill over riders on abortion travel benefits and transgender health care, calling them divisive policy provisions. Once a veto threat is tied to a rider, compromise requires stripping it, which then angers whoever demanded it.

Political scientists Hans Hassell and Samuel Kernell, in their study “Veto Rhetoric and Legislative Riders,” analyzed hundreds of riders attached to appropriations bills from 1985 through 2008 and found them concentrated in periods of divided government, with their numbers rising sharply over time. Riders aren’t anomalies. They’re a routine weapon.

Earmarks are the friendlier cousin, rebranded as Congressionally Directed Spending in the Senate and Community Project Funding in the House. Members steer money to specific local projects, now under tight rules: capped at one-half of 1 percent of discretionary spending, with a per-member limit on requests (15 in recent cycles, raised to 20 for fiscal 2027), and no for-profit recipients. The House Appropriations Committee requires a “Federal Nexus Requirement” tying each project to federal purposes, plus public online disclosure of every request.

The Debt Ceiling: A Separate Fight Over the Same Dollars

Here’s the distinction that trips up almost everyone. The debt ceiling is not the budget.

Treasury defines the debt limit as “the total amount of money that the United States government is authorized to borrow to meet its existing legal obligations,” including Social Security, military salaries, and interest on the debt. And it stresses the part people miss: “The debt limit does not authorize new spending commitments.” It “simply allows the government to finance existing legal obligations that Congresses and presidents of both parties have made in the past.”

So a shutdown and a debt-limit crisis are different failures. A shutdown means agencies lack legal authority to spend. A debt-limit breach means the obligations exist, the spending is authorized, but Treasury can’t borrow the cash to pay for all of it. One is about permission; the other is about financing.

When the limit binds, Treasury buys time with accounting maneuvers it calls extraordinary measures. Its January 2025 description spells out the mechanics: a two-month “debt issuance suspension period” frees up roughly $17.6 billion of headroom through the Civil Service Retirement and Disability Fund. Treasury Secretary Janet Yellen, in her January 2023 letter to Congress, warned these measures offer only temporary relief and are subject to substantial uncertainty.

The two fights increasingly bleed together. The Fiscal Responsibility Act of 2023 suspended the debt limit through January 1, 2025, and threw in discretionary spending caps and rescissions of unobligated COVID funds. When the suspension ended, the limit was reinstated on January 2, 2025, at $36.1 trillion. A debt deal, in other words, quietly set the terms for the next two years of appropriations. For what happens operationally when the funding actually lapses, see our coverage of a missed funding deadline.

Why the Design Keeps Failing

None of this is a story of lazy or corrupt legislators. The breakdown is structural, and thoughtful analysts on both sides largely agree on the mechanics even when they disagree on fixes.

Start with the Senate. Most appropriations bills need 60 votes to overcome a filibuster, so a determined minority can block funding even with a majority against them. That’s not theoretical: in 2026, a DHS appropriations bill failed four cloture votes with 51 to 52 senators in favor, short of 60. Cloture votes, once rare — 49 in total from 1917 through 1970 — hit 168 in the 115th Congress (2017-2018) alone. Governing by supermajority is simply harder than obstructing.

James Capretta of the American Enterprise Institute points to a deeper incentive trap: because appropriations need 60 votes but reconciliation needs only a simple majority, each party holds out for the next election, betting it can pass its agenda alone rather than cut a bipartisan deal now.

Brookings scholars add that the budget resolution has “morphed from honest funding proposals to party position papers,” so the very step meant to set shared limits has become a messaging exercise. And a SAGE analysis of congressional polarization notes that leaders increasingly bundle everything into omnibus bills negotiated at the top, “circumventing the role of the appropriation committees” and sparing rank-and-file members a long series of politically dangerous votes.

Alice Rivlin, the founding CBO director, and former Senate Budget Committee chair Pete Domenici spent years arguing the same point from a bipartisan perch: the current rules practically invite brinkmanship. Even the empirical work outside Washington points the same way. A study of state legislatures found a one-standard-deviation rise in polarization lengthens budget delays by roughly 43 percent.

Put the pieces together and the pattern isn’t a bug. Supermajority thresholds, must-pass leverage, and the shrinking slice of the budget still up for annual grabs all reward waiting over dealing.

How to See Where Your Government Actually Stands

Because the textbook and the reality diverge so sharply, the useful skill isn’t memorizing the calendar. It’s checking, in any given month, which parts of the government are funded and how.

Congress publishes a live scoreboard. The “Appropriations Status Table” sits on Congress.gov under the “Bill Searches and Lists” area, with separate tabs for regular appropriations, “Continuing Resolutions,” and supplemental appropriations. Pick a fiscal year, and each row shows how far a bill has traveled: subcommittee markup, committee report, chamber passage, and a “Presidential Approval” column. A date in that last column means that slice of the government has full-year funding. A blank means it’s on a CR or waiting.

The debt-ceiling question has its own primary sources. The CBO’s recurring “Federal Debt and the Statutory Limit” reports state the current limit, whether Treasury is using extraordinary measures, and when the money might run out. In March 2025, CBO projected exhaustion sometime between August and September absent action.

Here’s the tension worth carrying forward. Every reform that would smooth the process, an automatic CR, a cleaner debt-limit mechanism, weaker filibuster rules on spending, would also strip away leverage that some faction currently prizes. The dysfunction isn’t a glitch anyone forgot to fix. It’s a set of tools people keep choosing to use.

Which is why the next appropriations deadline, and the next X-date after it, will tell you more about who holds leverage this year than about whether Congress has finally learned to keep its own calendar. It hasn’t, and the design gives it little reason to.

Our articles make government information more accessible. Please consult a qualified professional for financial, legal, or health advice specific to your circumstances.