How could a legislature that voted on spending bill by bill, program by program and session by session, using numbers prepared by the president, ever control a budget as a whole? The Congressional Budget Act exists because Congress, by the early 1970s, could not answer that question with the machinery then in place. Members could amend a line item, defeat an appropriation, raise a tax or lower a tax, and each act carried real consequences. Added together, the acts produced a total that no member had voted on, supported by estimates that no congressional office had prepared, under a calendar that expired before the work was finished. Control of the parts coexisted with loss of the whole.

The statute that closed the distance was the Congressional Budget and Impoundment Control Act of 1974, Public Law 93-344, 88 Stat. 297, signed on July 12, 1974, codified principally at 2 U.S.C. sections 601 and following, with the impoundment title at 2 U.S.C. sections 681 and following. Enactment joined two reforms that a single confrontation had fused together. One reform was internal, addressed to the way Congress decided: committees charged with totals, an office charged with estimates, a resolution stating a complete fiscal position, allocations carrying that position down to the committees that drafted bills, and points of order policing the result. The other reform was external, addressed to the way the executive executed: procedures governing presidential withholding of enacted funds, so that an appropriation enacted over a presidential veto could not be set aside by administrative refusal to spend.

The internal problem had a shape that ordinary diligence could not cure. Appropriations bills moved through subcommittees organized around departments and agencies. Revenue bills moved through the tax committees. Authorizing committees created programs, and a growing share of spending arose outside the annual bills altogether, through entitlement formulas, contract authority that obligated the Treasury in advance of an appropriation, and borrowing authority that financed activity without an annual vote. No stage of that sequence required a decision on total spending, total revenue, the deficit or surplus, or the public debt. Each committee answered for a jurisdiction. No committee answered for the aggregate, and the aggregate was where fiscal policy lived. Layered over the fragmentation stood dependence on executive arithmetic. The Budget and Accounting Act of 1921 had given the president a consolidated budget and a budget bureau to prepare the document, so debate about a presidential program began from presidential figures, forecasts and cost estimates. Disputing a number is a weaker act than producing a rival number by a stated method, maintained across bills and years by an office answerable to the legislature.

The namable claim of the statute is “Numbers before rules”: the most consequential thing the statute did was give Congress its own estimating capacity, because every enforcement rule that followed depends on an authoritative number, and a legislature that must accept the executive’s arithmetic cannot meaningfully constrain the executive’s budget. Allocations and points of order mean little if the cost of a bill breaching an allocation is measured by the branch proposing the spending, and a resolution means little if the baseline against which its totals are judged shifts with presidential assumptions. The statute therefore created the Congressional Budget Office in Title II, under a director appointed jointly through the two chambers, charged with estimates, projections and scorekeeping offered without policy recommendations. Enforcement elsewhere in the act drew strength from that construction.

Around the estimating office, the remaining creations formed a connected system. Title I created a Committee on the Budget in each chamber, composed so that the established money committees held a guaranteed presence inside the new committees. Title III supplied the timetable, the concurrent budget resolution, the allocation chain and reconciliation, a late stage device for bringing mandatory spending and revenue law into line with adopted totals. Title IV reached spending authority created outside appropriations. Title V moved the fiscal year from a July start to an October start, and Title VI amended the required content of the presidential budget. Title VII assigned program review and evaluation functions to the Comptroller General. Title X, the Impoundment Control Act, separated permanent cancellation from temporary delay and assigned each category a procedure, a deadline and a congressional default. Evaluation divides along a line the sections ahead keep in view. A restraint reading holds that deciding totals in advance disciplines piecemeal incentives. A responsiveness reading answers that rigid aggregates can force blunt choices and shift influence toward leadership and the Budget Committees. The record through the close of 2019 supports both readings in part: the estimating capacity endured as the working arithmetic of fiscal debate, while ten fiscal years had closed without an agreed budget resolution, and the fiscal 2020 cycle, completed in 2019, added an eleventh.

Why the Congressional Budget Act Had to Exist

The Constitution assigned the power of the purse to Congress in plain terms, providing that money could be drawn from the Treasury only in consequence of appropriations made by law. By the early 1970s, possession of that authority and capacity to exercise that authority as a coherent fiscal policy had become two different things. Congress voted constantly on money, raising revenue, authorizing programs, appropriating funds, setting borrowing limits, and creating obligations that bound the Treasury for years beyond a single session. Each vote could be defended on its own terms by the committee that reported the bill. Added together across a session, however, the votes did not amount to a budget in any sense that a finance ministry, a state legislature, or a household would recognize. No vote decided how much the federal government would spend in total. No vote decided how much revenue the federal government would collect in total. No vote decided, in advance and in one place, what difference between spending and revenue Congress accepted for the year, or what addition to the public debt that difference required. The deficit or surplus for a fiscal year emerged as a residue, calculated after the separate decisions had been made, rather than as a decision made by the legislature charged with making it. That gap between formal authority and practical capacity is the problem to which the Congressional Budget Act responded, and the gap had four parts that reinforced one another: an absence of totals, an absence of a workable timetable, an absence of an independent scorekeeper, and an absence of an effective answer when a president declined to spend money that Congress had appropriated.

Consideration of spending bills one at a time, in sequence and in isolation, structured the entire process before 1974. The House and Senate Appropriations Committees divided annual funding among subcommittees, each subcommittee held hearings on the requests of a set of departments and agencies, each subcommittee reported a bill for a portion of the government, and each bill moved through passage, conference, and presidential signature on a separate track and often on a separate calendar. A decision to increase funding for one department did not, within the formal process, require a decision to decrease funding for another department, to raise revenue to cover the increase, or to record a larger accepted deficit. Revenue legislation moved through the House Committee on Ways and Means and the Senate Committee on Finance, committees distinct from the Appropriations Committees and from the authorizing committees that shaped the programs being funded. The Congressional Research Service characterized congressional action in exactly those terms: Congress considered authorization, revenue, and appropriations legislation separately, with no formal mechanism to develop, consider, and approve a comprehensive congressional budget plan.

Fragmentation of jurisdiction produced fragmentation of responsibility, and participants understood that defect well before legislation corrected it. The Joint Study Committee on Budget Control, created in 1972 to examine procedures for improving congressional control of outlay and receipt totals, identified the central defect in its 1973 work as the absence of any legislative committee with responsibility for deciding whether total outlays were appropriate in light of economic conditions and anticipated revenues. Appropriations subcommittees guarded the programs within assigned jurisdictions. Authorizing committees defended the missions entrusted to them. Tax committees judged revenue measures by standards that had little formal connection to spending totals taking shape elsewhere on Capitol Hill. No committee owned the aggregate. Allen Schick, whose analysis accompanied the Joint Study Committee work and whose later study Congress and Money, published in 1980, became a standard account of congressional budgeting in that era, located part of the weakness in resources as well as in structure, pointing to insufficient information, staff, and time for the scale of the task. Schick also described the longer bargain that had produced the weakness. After passage of the Budget and Accounting Act of 1921, Congress had in effect enlisted the president as an instrument of spending control, reasoning that a fragmented legislature could rely on an executive budget to impose coherence that Congress had difficulty imposing on itself. By the early 1970s, that bargain had inverted the constitutional expectation: the branch with a unified budget was the executive, and the branch with the power of the purse reacted to it piece by piece.

A second defect compounded the first, because a substantial and growing share of federal spending never passed through the annual appropriations process at all, or passed through that process only after the real decision had been taken elsewhere. Reformers in the early 1970s grouped those routes under the label of backdoor spending, a term that captured the institutional complaint rather than an accusation about any single program. Contract authority illustrated the mechanism with particular clarity. Under contract authority, an authorizing statute empowered an agency to enter contracts or other obligations in advance of an appropriation. The obligation was real from the moment the agency incurred it, and the Treasury would eventually have to pay. Congress later enacted an appropriation to liquidate the obligation, meaning to provide the cash to honor a commitment already made, and that later appropriation received, in practice, largely perfunctory review, because refusal to fund a lawful obligation carried consequences quite different from refusal to start a program. The sequence reversed the intended order of control: authorization became a decision that money would be spent, and appropriation became a bookkeeping step that ratified a commitment taken in an earlier bill, by a different committee, under different pressures.

Borrowing authority operated through a related route. A federal entity received statutory permission to borrow, from the Treasury or in some arrangements from the public, and to obligate and expend the borrowed funds. Enterprises and credit programs could in that manner finance activity at a scale set by the borrowing grant rather than by an annual appropriation, without a corresponding annual vote weighing the activity against competing claims on revenue. Entitlement authority completed the pattern on the largest scale. An entitlement statute defined a class of eligible persons and a formula or standard for payment, and obligated the government to pay every person who met the statutory conditions, in whatever total amount eligibility and economic conditions produced. Social Security, established in the 1930s, supplied the early model, and broad based entitlement programs expanded after the Second World War, with the result described in a Brookings Institution study of the Congressional Budget Office creation period: mandatory spending, most of it entitlement spending outside annual appropriations review, accounted for about a third of federal spending in 1962 and about half by the middle of the 1970s. The budget process problem was separate from the merits of any program. A legislature that set the total only for the portion of spending passing through annual bills, while entitlement formulas, contract commitments, and borrowing grants generated the remainder outside that reckoning, had no reliable method for determining the size of the government being financed until spending had in large measure determined itself.

Even for the portion of the budget that did move through annual bills, Congress lacked a comprehensive set of numbers of its own, and that informational dependence shaped every debate about fiscal policy before 1974. The Budget and Accounting Act of 1921 had required the president to submit a consolidated budget to Congress each year and had created the Bureau of the Budget to assist the president in assembling, correlating, revising, and reducing or increasing the estimates of the departments and establishments. Before that act, agencies had carried requests more directly to Congress, and a consolidated executive budget promised discipline within the executive branch and a single document against which Congress could measure proposals. The bureau moved to the Executive Office of the President in 1939, and Reorganization Plan No. 2 of 1970 restructured and renamed the bureau as the Office of Management and Budget. Across those changes, one fact held constant for half a century: the only complete, internally consistent presentation of federal spending, revenue, and economic assumptions arrived as the president’s document, prepared by the president’s staff, organized around the president’s program, and supported by economic forecasts that the president’s advisers had selected.

Debate conducted on that footing carried a structural handicap unrelated to the intelligence or diligence of individual members. To contest a presidential proposal for a department, a committee could hear testimony, examine agency officials, and adjust line items, and appropriations subcommittees performed that work with considerable expertise within assigned areas. To contest the president’s fiscal policy as a whole, Congress needed an alternative aggregate: an independent estimate of revenue under existing law, an independent projection of spending under existing rules, an independent reading of economic conditions, and a consistent method for estimating what a reported bill would cost in later years. Congress possessed audit capacity through the General Accounting Office, headed by the Comptroller General and charged with examining how enacted appropriations had been spent, a function oriented toward accountability for past spending rather than prospective scoring of pending legislation against an aggregate plan. Committee staffs reproduced the fragmentation of the committee system itself, each serving a jurisdiction rather than a whole budget, and no office held a mandate for continuous scorekeeping against a set of totals, because no congressional totals existed. The practical consequence was the one Schick identified in connection with the Joint Study Committee report: reliance on executive agencies, the Office of Management and Budget above all, for financial information weakened Congress, because the legislature argued about the president’s program using the president’s arithmetic, and any disagreement about assumptions or program costs began from numbers supplied by the branch whose proposals were under examination.

Timing turned those structural defects into an annual operational failure. Before the 1974 act changed the calendar, the federal fiscal year ran from July 1 to June 30. A session that convened in January therefore had roughly six months to complete authorizations, pass the full set of regular appropriations bills through both chambers, resolve differences in conference, and present the bills for signature before the new fiscal year began. Authorization legislation, which in principle preceded appropriations, frequently lagged. According to the Congressional Research Service, during the twenty five fiscal years from fiscal 1952 to fiscal 1976, Congress did not once enact all of the regular appropriations acts by the beginning of the fiscal year. Continuing resolutions, interim measures that kept agencies operating at restricted rates and under prior year conditions, filled the gap in every year of that period except fiscal 1953. Agencies planned hiring, procurement, grants, and construction under temporary authority that might last weeks or months. Late appropriations were not a mere embarrassment about punctuality. A process chronically unable to finish on a compressed calendar had no slack for the prior step that a genuine budget process required, namely a deliberate decision on totals taken early enough in the session to guide the bills that followed.

Experience had already shown that declaring an aggregate intention, without building machinery to support the declaration, accomplished little. The Legislative Reorganization Act of 1946 had attempted a legislative budget through a joint committee drawn from the tax and appropriations committees of both chambers, charged with comparing estimated receipts and expenditures and recommending a ceiling on total expenditures. The attempt failed in practice after efforts in 1947, 1948, and 1949, and Congress abandoned the plan, leaving the provision dormant until the Legislative Reorganization Act of 1970 repealed it. The episode mattered for the reform debates of the early 1970s because the failure was diagnostic. A ceiling voted by a body that had no independent estimates to test the ceiling, no timetable sequenced around the ceiling, no allocations carrying the ceiling down to committees, and no procedure for policing bills against the ceiling was a statement of aspiration rather than an instrument of control.

The fourth part of the capability gap concerned execution, because congressional control that ended at enactment left the final size and content of federal programs open to revision by the branch charged with carrying laws into effect. Impoundment, the withholding by a president of funds that Congress had appropriated, had precedents reaching across the history of the republic, invoked by presidents of both parties and in circumstances that ranged from changed military need to economy in administration to outright disagreement with the policy Congress had enacted. Examples ranged from President Thomas Jefferson, who declined to spend funds appropriated for gunboats on the Mississippi River after notifying Congress that changed circumstances had removed the immediate need, to President Harry Truman, who withheld funds added to expand the Air Force beyond the size the administration had requested, and President Lyndon Johnson, who withheld appropriations to restrain inflation during the Vietnam War. The scale, breadth, and stated constitutional basis of withholding under President Richard Nixon in the early 1970s raised the practice to a different order of institutional conflict. That history receives full treatment in the account of the impoundment confrontation that follows in the next section. For the problem statement, the structural point stands apart from the merits of any single withholding. Before 1974, no statute supplied a regular procedure by which a president proposed the permanent cancellation of budget authority and Congress approved or rejected the proposal within a defined period, no statute governed temporary deferral through a defined congressional check, and no officer of the legislative branch held a statutory mandate to report withholdings that the executive had failed to disclose and to pursue release of funds withheld without authority. A legislature that enacted an appropriation therefore lacked assurance that the enacted figure described the program that would operate.

Each defect made the others harder to repair in isolation. Totals without independent estimates would rest on executive numbers and would bend whenever an assumption changed in the executive branch. Independent estimates without totals would inform debate without binding any decision. A timetable without totals would organize the sequence of piecemeal bills without changing the piecemeal character of the result. Control of backdoor spending without an aggregate process would close one route around appropriations while entitlement formulas and borrowing grants pressed against whatever boundary a single rule drew. Reformers who examined the system in 1972 and 1973, through the Joint Study Committee and the committees that handled the resulting legislation, confronted a system in which the missing elements formed a single missing capability rather than a list of unrelated annoyances. Congress lacked an institution charged with the whole, numbers that belonged to the legislature, a calendar arranged around a decision on the whole, coverage that reached spending committed outside annual appropriations, and a statutory procedure that made execution answerable to enactment.

Stated in negative form, the condition of congressional budgeting on the eve of reform can be summarized without exaggeration. No totals existed, because no resolution, vote, or committee action set aggregate spending, revenue, and the resulting deficit or surplus before the component bills moved. No timetable existed, in the sense of a statutory sequence keyed to an aggregate decision and to a fiscal year with enough room for the sequence to run, and the July 1 start had been paired for twenty five fiscal years with appropriations that finished late. No scorekeeper existed, because no legislative office maintained independent estimates, baselines, and running tallies for the legislature as a whole, and the Bureau of the Budget and the Office of Management and Budget served the president whose budget Congress was supposed to judge. No answer to impoundment existed, because withholding rested on asserted executive discretion and on ad hoc congressional responses, without defined categories, defined time limits, or a defined officer charged with exposure and enforcement. The problem was never that Congress lacked opinions about spending or revenue, nor that members lacked votes to pass bills. The problem was that a legislature deciding a national budget one bill at a time, on numbers supplied by a rival branch, against a calendar that expired before the work finished, for only part of the spending the government was committed to make, and without recourse when enacted amounts went unspent, could control many things with great precision while the budget as a whole escaped control entirely.

The Impoundment Confrontation

Budget reform had been studied in Congress for years before 1974, in hearings, committee reports, and proposals for better information and better timing. Study alone produced no statute. The force that converted study into law was a concrete confrontation over money already enacted. Between 1970 and 1974, the Nixon administration withheld funds across a wide range of domestic programs, asserted a constitutional entitlement to do so, and defended that assertion after Congress had overridden a presidential veto to enact one of the affected programs.

What is impoundment and why did it provoke Congress?

Impoundment is executive withholding of funds that Congress has provided, by refusing allotment, apportionment, or expenditure. It provoked Congress under President Nixon because withholding ceased to be routine saving or delay and became a claimed constitutional right to nullify enacted programs, including programs enacted over a presidential veto, at government-wide scale.

In ordinary administration, not every dollar provided by Congress is spent at once. The Office of Management and Budget apportioned funds to agencies, agencies allotted funds to programs, states, or projects, and obligations followed when the government committed itself to grants, contracts, or loans. A withholding could intervene at any step. Money could be placed in reserve, an allotment to states could be capped below the enacted figure, or a program could be announced as terminated even though the statute creating the program remained in force. The practical effect did not depend on the label. A state that received a smaller allotment planned smaller projects, and a locality that received no grant built no treatment works.

The pedigree of the practice reached back to the early republic, and defenders of presidential withholding invoked that pedigree often during the Nixon years. The example most frequently cited dated to 1803, when President Thomas Jefferson reported to Congress that a sum of 50,000 dollars appropriated for gunboats remained unexpended because a favorable and peaceable turn of affairs on the Mississippi River had rendered immediate execution of the appropriation unnecessary. The amount was small, the reason was a change in circumstances that removed the immediate need, and Jefferson reported the decision to Congress in an annual message. In the account of impoundment history prepared by the Congressional Research Service, impoundments of that early kind were typically routine and relatively small in scale, and presidents and agency heads had long treated appropriated amounts as permissive ceilings rather than as commands to spend every dollar regardless of need. Twentieth century practice added a statutory vocabulary for that discretion. Presidents and agencies pointed to the Antideficiency Act, enacted to prevent agencies from spending at a rate that exhausted resources early and then presenting Congress with a demand for more money. That statute allowed reserves for contingencies and for savings achieved through efficiency. The distinction embedded in that justification was purpose. A reserve tied to a contingency or a genuine saving administered a program. A reserve imposed to defeat a program pursued a different object.

Larger precedents existed, and each carried features that limited the lesson drawn from the precedent. During the Second World War, President Franklin D. Roosevelt curtailed public works spending on the ground that resources were needed for the war effort. The Congressional Research Service account of that episode stressed a feature absent from the Nixon confrontations, namely that actions of that scale were generally taken in consultation with congressional leaders and therefore produced no confrontation between the branches. President Harry Truman supplied a sharper precedent in 1949. Congress appropriated funds for an Air Force of 58 groups rather than the 48 groups the administration had requested, and Truman directed that the additional 735 million dollars for the extra groups be placed in reserve, defending the action in terms of a balance between national security and a sound economy and presidential authority as Commander in Chief. Even that confrontation ended as an episode, with no general statute following and a single defense appropriation at issue. President Lyndon B. Johnson also withheld amounts appropriated above the levels in the administration budget request, and in 1966 the Johnson administration limited spending under the federal highway program, beginning a line of highway withholding that continued under President Nixon. The precedents established that presidents had withheld funds before, for economy, during war, and in direct disagreement over the size of a program. The precedents did not establish a settled power to terminate domestic programs, to refuse allotments commanded by a statute, or to substitute an executive ceiling for the totals enacted in law.

The change under President Richard Nixon was one of scale, of stated purpose, and of legal theory. Withholding reached across the domestic government rather than resting in one department. A list supplied to Congress in February 1973 under new reporting legislation described reserves concentrated in transportation, defense, and agriculture, with the largest single item in federal aid for highway construction, and administration and congressional estimates of the government-wide total differed sharply in public reporting during that month. A Congressional Research Service history of rescission proposals recorded a moratorium on subsidized housing programs, a suspension of community development activities, and a reduction in disaster assistance, with several farm programs targeted for elimination. Environmental grants under the water pollution statute formed the largest and most visible single dispute. Administration officials described many withholdings as devices for holding total spending under a ceiling, restraining inflation, and avoiding a tax increase. Those were government-wide fiscal purposes, not findings that a particular grant was unnecessary or a particular project complete under budget. In some instances the announced purpose went further. Agriculture Secretary Earl Butz announced that funds for the Rural Environmental Assistance Program had been impounded and that the program had been terminated, an announcement reported in February 1973. A termination asserted that executive action alone could end a program that Congress had created and funded and had not repealed.

Legal theory completed the change. Earlier withholdings had usually rested on a statute or on claimed administrative discretion within a program. The Nixon administration advanced those arguments as well, and the February 1973 reporting to Congress cited authority item by item. Layered over the itemized claims stood a broader claim of constitutional right. At a press conference on January 31, 1973, President Nixon described the claimed constitutional right to impound appropriated funds as, in his word for the clarity of the right, absolutely clear, in circumstances in which expenditure would raise prices or taxes. A president who claimed a clear constitutional entitlement placed the dispute beyond ordinary statutory correction, because a constitutional right, if accepted on the terms claimed, would survive a clearer appropriation and would limit what any later appropriation could command.

The executive argument, stated at full strength as administration officials and defenders stated the argument during 1973 and 1974, began with national fiscal responsibility. The president, in that argument, was the only officer charged with administering the whole budget rather than a part of the budget. Individual committees saw individual programs, and no single committee bore responsibility for the aggregate of spending, taxes, borrowing, and prices. President Nixon made a version of that point at the January 31, 1973 press conference, contrasting committees and the interests pressing on committees with a presidency obliged to speak for the general interest. Administration defenders argued that a president confronted with inflation could not be reduced to a cashier for every appropriation enacted in isolation. Spending authorized in separate bills arrived in the economy as a single total. Prices responded to the total, the Treasury borrowed against the total, and any statutory spending ceiling or debt limit operated on the total. A president who possessed no power to slow or refuse expenditure, defenders argued, possessed no effective means of keeping faith with ceilings that Congress itself had enacted. The same argument drew on the permissive character of appropriations language, which in that reading fixed maximum amounts rather than mandatory floors. Administration defenders also invoked the Take Care Clause from the opposite direction to the direction taken by congressional opponents. Faithful execution, in the executive reading, meant execution of the laws as a whole, including fiscal limits, rather than mechanical disbursement under one statute in disregard of economic conditions that made full disbursement harmful.

The congressional argument, stated at full strength as Senator Sam J. Ervin, Representative Jamie L. Whitten, Senator Edmund Muskie, and other members stated the argument in hearings and debate, began with the text of Article I. The Appropriations Clause provided that no money could be drawn from the Treasury except in consequence of appropriations made by law, and the legislative power vested in Congress included the power to decide, by law, what programs existed, what purposes public money served, and what amounts supported those purposes. An appropriation enacted over a veto by the required two thirds of each House was not a suggestion offered to the executive branch for acceptance or rejection program by program. The veto was the constitutional instrument assigned to a president who opposed a bill, returnable with objections for a decision by the heightened majority the Constitution specified. Once that process had run, the congressional argument held, a second and unwritten veto exercised through the Treasury would give the president a power the Constitution had deliberately conditioned, exercisable after enactment and without a two thirds safeguard. The Take Care Clause anchored the second half of the argument. Article II charged the president to take care that the laws be faithfully executed, and members argued that a statute directing an allotment or a grant program was one of the laws covered by that charge. Faithful execution did not, in the congressional reading, confer a power to refrain from executing a law because the president preferred a different policy or a lower total, whatever room the clause left for administration and for discretion that a statute itself conferred. Senator Ervin pressed a further point. Impoundment, Ervin argued, did not save public money in the manner of a repeal debated and enacted in public, because withheld funds reallocated national resources toward executive priorities while programs disfavored by the administration starved, and the choice among programs was precisely the choice the appropriations power assigned to Congress.

The water pollution confrontation turned those competing arguments into a record on which courts, members, states, and cities could act. President Nixon vetoed the Federal Water Pollution Control Act Amendments of 1972 on October 17, 1972, objecting in the veto message to a staggering and budget wrecking price and warning of pressure toward higher taxes, and the message promised that provisions conferring a measure of spending discretion would be used to put brakes on expenditures if the bill became law over that objection. Congress overrode the veto, and the amendments became law on October 18, 1972. The sequence removed ambiguity about congressional intent, because the constitutionally specified supermajorities had enacted the program after hearing the objections. Title II of the 1972 amendments provided federal assistance for municipal sewers and sewage treatment works. Section 207 authorized appropriations of 5 billion dollars for fiscal 1973, 6 billion dollars for fiscal 1974, and 7 billion dollars for fiscal 1975. Section 205(a) provided that sums authorized under section 207 would be allotted by the Administrator of the Environmental Protection Agency among the states. The administration read the qualifying phrase not to exceed in section 207 as preserving discretion to allot less than the full amounts, treating each figure as a ceiling without a floor. Acting on presidential direction, the Administrator capped allotments at 2 billion dollars for fiscal 1973 and 3 billion dollars for fiscal 1974, and the cap for fiscal 1975 stood at 4 billion dollars. Across the three fiscal years, the caps withheld 9 billion dollars, the difference between 18 billion dollars authorized and 9 billion dollars allotted under the caps.

Litigation followed promptly. The City of New York and other prospective recipients sued the Administrator, Russell E. Train, seeking a declaratory judgment that the Administrator was obligated to allot the full authorized amounts and an order directing the allotments to be made. The District Court ruled for the city, and the Court of Appeals for the District of Columbia Circuit affirmed. The Supreme Court decided Train v. City of New York, 420 U.S. 35 (1975), on February 18, 1975, and affirmed. The Court resolved the case as a matter of statutory interpretation. Reading section 205(a) together with section 207 and consulting the legislative history of the amendments, the Court held that the 1972 amendments did not permit the Administrator to allot less than the entire amounts authorized. The qualifying phrase not to exceed, in the setting of the allotment command, did not create the open ended discretion the administration claimed. Equally significant was the ground the Court declined to occupy. The Court made no categorical constitutional ruling on impoundment and announced no general test for presidential withholding under other statutes. The Congressional Budget Act was already law by the date of the decision in Train, having been signed on July 12, 1974, and the decision therefore confirmed a statutory reading without supplying the general rule that Congress had concluded a general statute had to supply.

Other flashpoints filled out that demonstration, each on a separate record. Highway funds produced State Highway Commission of Missouri v. Volpe, 479 F.2d 1099 (8th Cir. 1973), in which the Eighth Circuit held that the Secretary of Transportation could not withhold approval of state highway projects for reasons beyond those Congress had supplied in the highway statute. Rural programs produced Sioux Valley Empire Electric Association v. Butz, 504 F.2d 168 (8th Cir. 1974), arising from withheld rural electrification assistance, in which the Eighth Circuit compelled disbursement in accordance with the governing statute. Rural water and waste disposal grants, a comparatively small line of 120 million dollars in the February 1973 agriculture reporting, mattered out of proportion to the amount, because the beneficiaries were small communities with no alternative source of capital for systems that public health required. Housing withholding operated partly through a moratorium, suspending subsidized housing activity that local housing authorities and low income families had treated as an operating program. Agriculture withholding reached loan and grant programs and the terminated Rural Environmental Assistance Program. Together, the disputes presented a pattern that no single judgment could settle.

The pattern explains why impoundment, rather than a general theory of budget process, carried reform from study to statute during 1973 and 1974. A theory of congressional budgeting asked members to imagine a better sequence of decisions in a future fiscal year. Impoundment confronted members with enacted decisions already nullified in the present fiscal cycle. A member who had voted to override the water veto in October 1972 and then watched the administration allot half or less of the authorized schedule had direct evidence that a supermajority vote had failed to guarantee that a program would operate. Governors, mayors, and local authorities who had committed matching resources pressed members for a remedy that reached beyond a single program. Delay compounded the injury. The water case required more than two years from the October 1972 override to the February 1975 judgment. Fiscal years closed while cases remained pending, construction seasons passed, and states could not prudently commit funds that a court might later order released. Enactment itself had ceased to guarantee operation.

Congress therefore treated impoundment as the urgent half of a larger settlement, and the Congressional Budget Act joined budget process reform and impoundment control in a single statute, Public Law 93-344, signed on July 12, 1974. Title X of the statute, the Impoundment Control Act, addressed withholding through reporting and through procedures under which Congress rather than the president alone would decide whether enacted budget authority ended, and the design of those procedures belongs to the treatment of Title X in a later section. By the middle of 1974, Congress had experienced executive action under which enactment, appropriation, and even a veto override did not assure execution, had heard that action defended as a clear constitutional right, and had watched courts answer the claim one statute at a time on statutory grounds that left the constitutional claim formally undecided. A legislature in that position could continue to litigate program by program, could tighten the verbs of each new appropriation, or could enact a general law governing withholding itself. The choice of a general law recorded the judgment of the 93rd Congress that the confrontation had come to concern whether legislation remained legislation after Congress enacted a bill over a veto.

Drafting and Passage

Who wrote the Congressional Budget Act?

No single author wrote the Congressional Budget Act. The Joint Study Committee on Budget Control, created in 1972, supplied the 1973 blueprint. Richard Bolling guided the House bill, H.R. 7130, through the Rules Committee, and Senator Sam Ervin sponsored the Senate companion, S. 1541.

The reform had a long gestation because the defect that reformers described was structural. Under the arrangements in place since the Budget and Accounting Act of 1921, the president submitted a single, comprehensive budget, while Congress acted on the parts in sequence and in isolation. Appropriations bills moved through subcommittees organized around agencies and programs. Revenue measures originated in the Committee on Ways and Means in the House and the Committee on Finance in the Senate. Authorizing committees created programs and, in a growing number of cases, provided spending authority outside the annual appropriations process. Yet no stage of the process required either chamber to vote on total spending, total revenue, or the size of the resulting deficit. Congress, in the formulation repeated throughout the drafting period, lacked a budget of its own. The aggregates emerged from the accumulation of separate choices rather than from a deliberate choice among competing totals. Conflict with President Richard Nixon then converted an institutional complaint into an urgent one. Presidential impoundment, the refusal to spend funds that Congress had appropriated, had a history extending beyond the Nixon administration, but the scale and the asserted basis of the practice between 1969 and 1973 gave the dispute a different character, and at a press conference on January 31, 1973, Nixon asserted a constitutional right to impound appropriated funds. For members of both parties, a president who could decline to spend after enactment possessed, in practical effect, a second vote on the substance of appropriations law. Budget procedure and impoundment control therefore developed together during 1972 and 1973, and the eventual statute carried both in its formal title, the Congressional Budget and Impoundment Control Act of 1974.

The vehicle for the first stage of drafting was the Joint Study Committee on Budget Control, created in 1972 and active during 1972 and 1973. A joint committee was a deliberate choice, because budget reform touched the jurisdiction of committees in both chambers, and a design produced by one chamber alone would have arrived at the remaining centers of power as an external threat. Composition shaped the result. Members drawn from the appropriations and tax committees predominated on the study committee, and analysts of the process have traced the character of the 1973 recommendations to that predominance. A proposal to consolidate budget authority in a new and dominant committee, displacing the established money committees, had little prospect among members whose standing rested on the established structure. The recommendations of 1973 therefore layered a coordinating process on top of the existing committee system rather than replacing that system. Congress would acquire Budget Committees, a budget resolution, and a professional estimating staff, while the Appropriations Committees retained the appropriations bills and the tax committees retained revenue legislation. The study committee also settled a question of constitutional form. The instrument for stating congressional totals was a concurrent resolution, adopted by both chambers and not presented to the president for signature. A resolution presented for signature would have given the president a formal role in a document whose purpose was to establish the position of Congress, and a veto over that document would have reproduced the dependence that reformers opposed. The resolution would instead bind through the rules and points of order of each chamber. Self binding was the point of the exercise. External restraint on presidential withholding formed one part of the statute, in the impoundment title, while internal restraint on congressional fragmentation formed the larger part, in the budget titles.

The recommendations of 1973 were introduced as H.R. 7130 in the House and as S. 1541 in the Senate. In the House, jurisdiction went to the Committee on Rules, the panel that ordinarily governs the terms of floor debate. The assignment reflected the character of the bill, which amended the rules of the House as well as enacting statutory procedure, and the assignment placed the measure in the hands of Representative Richard Bolling of Missouri, a senior member of the Rules Committee with a long record of advocacy for congressional reorganization. The legislative history identifies Bolling as a central architect of the House bill, and the committee record supports that description. The Rules Committee reported H.R. 7130 with the report numbered H. Rept. 93-658, and Bolling sponsored the special rule, H. Res. 715, introduced on November 20, 1973, that provided for consideration of the bill on the House floor. Sponsorship of the rule was more than a scheduling function. Bolling managed the bill as a settlement among committees as much as a program of reform, and Bolling operated from a clear view of what the House would accept, formed through earlier campaigns for institutional change in the chamber. Proposals that read as a transfer of power from the standing committees to a new fiscal authority drew resistance from precisely the members whose support passage required, including senior members of Appropriations and Ways and Means, whose committees stood to lose the most from consolidation.

The House bill therefore presented the Budget Committee as a coordinator rather than as a successor to the money committees. The new committee would report a resolution setting totals for revenue, budget authority, outlays, the deficit or surplus, and the public debt, and the resolution would divide those totals among the committees of the House through allocations. Legislation that breached an allocation or an aggregate would face points of order. The Appropriations Committee, however, would continue to write the appropriations bills, subdividing its allocation among its subcommittees, and Ways and Means would continue to write tax legislation. The Budget Committee received authority to state the size of the whole, while the committees with established expertise retained the drafting of the parts. That division was the central design bargain of the statute, preserved in the allocation structure of Title III rather than in any grant of line item authority to the new committees. The composition of the House Budget Committee expressed the same bargain in personnel. As enacted, the committee consisted of five members drawn from the Committee on Appropriations, five members drawn from the Committee on Ways and Means, eleven members drawn from other standing committees, one member from the leadership of the majority party, and one member from the leadership of the minority party. The money committees held a large and guaranteed bloc, the remainder of the committee system held the largest single share, and the party leaderships held a direct presence without control. Closely connected to composition was rotation. The statute provided that no member could serve on the Committee on the Budget during more than two Congresses in any period of five successive Congresses beginning after 1974, and that selections were to be made without regard to seniority. Rotation prevented the emergence of a permanent cadre of budget members who might develop into a rival seniority hierarchy, reassured the established committees that Budget Committee service would remain a temporary assignment, and circulated experience with aggregate budgeting through the wider membership. Defenders of rotation emphasized that a committee meant to coordinate the House had to remain representative of the House. Skeptics observed that representativeness purchased at the price of continuity would limit the authority the committee could exercise against committees whose members served without comparable limits. The enacted text chose representativeness, and the choice was a condition of enactment.

The third major bargain concerned impoundment, and the legislative history credits Bolling with insistence that control of impoundment belong in the bill. A budget process statute standing alone would have given Congress a resolution, committees, and an estimating office, while leaving untouched the practice that had made executive dominance of spending decisions a live constitutional grievance. Congress could hardly assert control over budgetary totals against a president who retained an unrestricted capacity to nullify enacted appropriations by withholding. Attachment of impoundment control also changed the politics of the bill. Budget procedure, considered alone, was an internal and technical subject. Impoundment control spoke to every member with a program or an appropriation affected by executive withholding, which by 1973 described a large share of both chambers. The combined bill could therefore assemble a coalition far broader than the coalition for procedural reform alone. Title X, the Impoundment Control Act, supplied the machinery, distinguishing a rescission, under which the president proposes permanent cancellation and Congress must enact approval within a defined period or the funds must be released, from a deferral, a temporary withholding subject under the 1974 text to disapproval by either House.

The House completed action first. Following the report of the Rules Committee in H. Rept. 93-658 and adoption of the special rule for consideration, the House debated H.R. 7130 at the beginning of December 1973 and passed the bill on December 5, 1973, by a vote of 386 to 23. The margin deserves attention because of the character of the measure. Members were voting to subject future appropriations and revenue decisions, including decisions of the committees on which many of them served, to aggregate limits and to points of order that colleagues could raise against bills on the floor. Opposition of only twenty three votes indicated that the committee bargains had done the work assigned to them during drafting. The money committees had been protected in jurisdiction, rotation had answered fears of a new permanent power center, and impoundment control had given the bill a constituency beyond the reformers.

The Senate route ran through two committees in sequence, giving the bill a second full examination. Senator Sam Ervin of North Carolina sponsored S. 1541, introduced on April 11, 1973, and the bill was referred to the Committee on Government Operations, the committee that Ervin chaired. Government Operations reported the bill on November 28, 1973, with the report numbered S. Rept. 93-579. Consideration in that committee placed the reform within a wider inquiry into executive organization and congressional control, consistent with the impoundment strand of the legislation and with the standing of Ervin, whose chairmanship of the Senate select committee investigating the Watergate affair during the same period had made Ervin a prominent voice on the limits of executive power. Sequential referral then took S. 1541 to the Committee on Rules and Administration, which received the bill on January 31, 1974, and reported the bill on February 21, 1974, with the report numbered S. Rept. 93-688. The second referral mattered because large parts of the bill operated as rules of Senate procedure. Senator Robert C. Byrd of West Virginia played a leading part at that stage. In a later statement printed in the Congressional Record, Byrd described assembling a staff working group to make extensive revisions to the bill that Government Operations had reported, drawing representatives from the chairmen of the standing committees of the Senate, from joint committees, from the House Appropriations Committee, from the Congressional Research Service, from the Office of Senate Legislative Counsel, and from the Parliamentarian of the Senate. The composition of that working group reproduced, at staff level, the strategy that had governed the House bill and the study committee. Every center of authority affected by the revisions had a representative in the room where the revisions were drafted, so the text that reached the Senate floor in March 1974 arrived as a negotiated instrument. The Senate debated the measure during March 1974 and completed passage on March 22, 1974, by a vote of 80 to 0. In form, the Senate passed H.R. 7130 as amended, in lieu of S. 1541, so the House bill became the vehicle for enactment and the Senate text traveled as an amendment to that vehicle.

Differences between the House and Senate texts required a conference committee, and the route of H.R. 7130 through passage in one chamber, amendment and passage in the other, and reconciliation of the two versions in conference followed the standard course by which major legislation of the period became law, the ordinary course by which a bill becomes law in practice. Conference on a bill of this kind involved more than the settlement of verbal differences. The conferees had to produce a single structure that both chambers could administer, because a budget process exists only in the repeated practice of the chambers. The conference report was filed as H. Rept. 93-1101 in the House and as S. Rept. 93-924 in the Senate. The House agreed to the conference report on June 18, 1974, by a vote of 401 to 6, a margin larger in absolute terms and smaller in opposition than the margin for original passage in December 1973. The Senate agreed to the conference report on June 21, 1974, by a vote of 75 to 0, recording a second unanimous result. Across the four recorded votes, opposition totaled twenty three votes in the House in December, no votes in the Senate in March, six votes in the House in June, and no votes in the Senate in June. The near unanimity signified more than agreement on a text. The bill regulated Congress at least as much as the president. Members voted to bind themselves, accepting aggregate limits, allocations, and points of order against the piecemeal habits of the institution to which the voters belonged, and members accepted that discipline in part because the discipline remained self administered, through rules that a majority of either chamber could waive. That combination, real restraint under congressional control, explained margins that procedural legislation rarely achieved, and the figures formed part of the authority of the new process during the first budget cycles conducted under the statute.

President Nixon signed the bill on July 12, 1974, and the measure became Public Law 93-344, printed at 88 Stat. 297. The signature completed a sequence that contains an evident tension. A president whose withholding of appropriated funds had supplied much of the impetus for the legislation signed into law a statute that restricted presidential impoundment, required the president to seek congressional approval for permanent cancellations of budget authority, and equipped Congress with committees and an estimating office designed to end reliance on executive figures. A veto of a bill passed by margins of 401 to 6 and 75 to 0 would have invited an override vote with every prospect of success. The calendar gives the enactment a closing fact that no drafter could have incorporated into the design. On August 8, 1974, four weeks after the signature of July 12, 1974, President Nixon announced his resignation, and the resignation took effect on August 9, 1974. The president who had signed the Congressional Budget and Impoundment Control Act left office less than a month after the signature, as the Watergate affair reached the conclusion that had been approaching during the whole period of drafting and passage. The coincidence of dates can mislead if pressed into a causal account. The bill had passed the House in December 1973 and the Senate in March 1974, before the final crisis of the Nixon presidency, and the votes rested on institutional grounds that extended beyond the fate of one administration. The departure of the president four weeks after signature nevertheless fixed the historical setting of the statute, an enactment through which Congress reasserted authority that members believed had migrated to the executive branch over several decades. Measured by the drafting record alone, passage stands as an exercise in institutional negotiation. The Joint Study Committee negotiated between a congressional budget and the survival of the committee system. Bolling and the Rules Committee negotiated between that design and the distribution of power in the House. Ervin, Government Operations, Rules and Administration, and the staff working group described by Byrd negotiated between the Senate bill and the procedural traditions of the Senate. Conferees negotiated between the two chambers. At each level the price of agreement was a limitation on the new machinery, in protected jurisdiction, in rotating membership, or in enforcement through points of order that each chamber retained the power to waive. The drafters accepted those limitations as the difference between a statute and a proposal, and the bargains of 1973 and 1974 remained legible in the structure of the process as the process stood at the end of 2019.

The Scorekeeper: A Budget Office of Its Own

What does the Congressional Budget Office do?

The Congressional Budget Office provides Congress with independent estimates of the cost of legislation, projections of federal spending and revenue under existing law, and scorekeeping reports that measure pending measures against the levels set in the budget resolution, without offering policy recommendations.

That short description understates the change Title II worked in the balance of the federal government. Before the Congressional Budget Act, a committee that wanted to know what a bill would cost, what the deficit would be under a set of proposals, or whether a tax change would raise the revenue claimed for it depended in large part on figures supplied by the executive branch. The administration prepared the budget, the administration housed the estimators, and the administration defended the arithmetic on which its program rested. Title II closed that gap by design. Sections 201 to 209 of the statute, codified principally at 2 U.S.C. sections 601 and following, established an office of Congress itself, placed its director under a joint appointment by the two chambers, and defined its duties in law.

The claim that runs through the statute, numbers before rules, receives its fullest meaning in this title. A point of order has no force in the abstract. A point of order under section 302 against a bill that breaches a committee allocation, or under section 311 against a measure that breaches the aggregates in a budget resolution, depends entirely on a measurement: the budget authority or outlays a bill carries, the revenue it raises or loses, and the comparison of those amounts against a level fixed in advance. If the measurement belongs to the president, a Congress attempting to restrain presidential spending inherits the president’s figures for the size of the restraint required, the cost of the alternative offered, and the consequence of doing nothing. Title II answered that problem at its root. Enforcement points of order would be only as strong as the number they referenced, and the number, the statute provided, had to be Congress’s.

The design of the office followed from that purpose. The Congressional Budget Office was located in the legislative branch, not attached to a single committee and not housed in either chamber alone. A joint office, serving the House and the Senate under a director appointed by officers of both, was built to produce one set of figures that floor managers in either chamber could cite without first translating a rival institution’s assumptions. The office was also made professional rather than advisory in the political sense. Its product was analysis: estimates, projections, reports, and studies. The statute did not charge the director with recommending a level of spending, a level of taxation, or the passage or defeat of any measure. That absence was the design’s central protection, because an office that advises Congress on what to enact spends credibility each time its advice is rejected, while an office that states what a proposal would cost under stated assumptions preserves credibility for the next estimate regardless of the vote that follows.

Appointment carried the same logic. Section 201 provided that the director would be appointed by the Speaker of the House and the President pro tempore of the Senate, after considering recommendations received from the Committees on the Budget of the House and the Senate, without regard to political affiliation and solely on the basis of fitness to perform the duties of the office. Joint appointment prevented either chamber from treating the director as its appointee alone and gave the selection a bicameral character matching the office’s bicameral service. Consultation of the Budget Committees tied the choice to the committees that would rely most heavily on the office’s work and could judge whether a candidate commanded the technical subjects the office covered. The instruction to appoint without regard to political affiliation, unusual in a statute creating a position of consequence, stated the standard against which every later appointment would be judged in public argument.

The term reinforced the appointment. The statute set a term of four years, expiring on January 3 of the year preceding each presidential election, with a director appointed to fill a vacancy serving the unexpired portion and a director permitted to remain in place after expiration until a successor was appointed. Removal lay with either House acting by resolution, a further statement that accountability ran to the legislature that had created the office and to no executive officer. No president could dismiss the director for an estimate that complicated an administration budget, and no secretary could direct the office to revise a projection. Independence in that setting did not mean absence of accountability. It meant accountability of a particular kind: to both chambers, through a public appointment, for the quality and impartiality of analysis rather than for agreement with any program.

The statutory duties, read together, describe a complete circuit between a plan and the legislation tested against it. The first duty was the cost estimate. Under the statute, committees reporting a bill or resolution were to receive an estimate from the office of the cost of carrying out the measure, projected over the years the statute specified, together with a comparison against spending under existing law and a description of the method used. The requirement attached to reported bills as a class rather than to bills a committee chose to submit, and that universality changed the status of the estimate in debate. An estimate prepared as a routine incident of reporting a bill functioned instead as part of the legislative record, available to supporters and opponents on equal terms and available to the Budget Committees when a point of order turned on the figure.

The mechanics by which an individual estimate is assembled, the treatment of behavioral responses, the handling of administrative costs, and the conventions that govern the estimating window belong to a specialist treatment, and the detailed account appears in how the Congressional Budget Office scores legislation. At the level of the pillar statute, the essential point is narrower. Title II assigned the estimating function to an office with no program to defend, required estimates as a regular product of the legislative process, and left a published method that later estimates could be checked against. Consistency across bills carried a value separate from accuracy in any single case, because a member comparing two alternatives needed figures built on the same assumptions more than figures built on shifting assumptions that happened to favor each alternative in turn.

The second duty was the baseline itself. The office was charged with projecting federal spending, revenue, deficits or surpluses, and the debt under existing law, and with reporting on the economic and budgetary outlook in which those projections sat. A baseline is easy to mistake for a prediction, and the distinction governed much of the office’s early explanation of its work. A projection under existing law asks what would follow if statutes on the books remained in place and economic conditions developed as projected, and it thereby provides a measuring rod rather than a forecast of political decisions. Without a baseline maintained by Congress’s own office, every debate over an increment would have reopened the prior debate over the whole, and an administration baseline built to showcase an administration program would have set the terms on which alternatives were measured.

The third duty closed the circuit. Scorekeeping reports, prepared for the Budget Committees, tracked the state of congressional action against the levels in the budget resolution: spending and revenue legislation enacted, legislation pending, and the amounts by which action to date stood above or below the resolution’s aggregates and allocations. Scorekeeping turned the resolution from a statement adopted in the spring into an account maintained through the session. A committee considering a bill late in the year could be told, from a running record rather than from memory or assertion, how much room remained under its allocation and what earlier bills had consumed. The Budget Committees, in turn, gained an evidentiary basis for raising a point of order or for warning a committee before a breach occurred. One office performing the estimate, maintaining the baseline, and keeping the score gave the three figures a consistency that three separate sources, each with a distinct method, could not have matched.

Beyond those core duties, the statute contemplated analysis on request. Committees other than the Budget Committees could seek studies of program alternatives, of the budgetary effects of proposals under consideration, and of subjects within the office’s competence, and the statute set a priority that served the Budget Committees, the Appropriations Committees, the House Committee on Ways and Means, and the Senate Committee on Finance before other requesters when capacity ran short. It did not confine the office’s influence to those rooms. A study prepared for an authorizing committee on the cost of alternative program designs could shape a bill long before the bill reached the estimating stage, and testimony by the director placed analysis in a public hearing record that any member could cite.

An institution of that kind required a founding director who understood that method would be judged before results were known, and the office received one. Alice M. Rivlin took office in February 1975 as the first director and served eight years, to 1983, spanning the administrations of Presidents Gerald Ford, Jimmy Carter, and Ronald Reagan. Rivlin arrived with an economist’s training and experience in federal budget analysis, and accounts of the office’s history generally credit her direction with the choices of the founding period: recruitment of analysts on professional competence, organization that protected longer range studies from displacement by the daily demand for estimates, and a strict refusal to attach policy recommendations to analytical products. A refusal to recommend, held even when members pressed for a judgment on the merits, taught committees that a request for analysis would return analysis and that disappointment with a figure could not fairly be answered by accusing the office of campaigning for a rival policy.

The norm of identical service to both parties belonged to the same founding settlement. The office offered the same numbers to the majority and the minority, to the House and the Senate, and to a committee chair and a ranking member who intended to use a figure for opposite purposes. Criticism of unwelcome estimates tested the norm in a useful way, because the office’s defense rested on showing the method, the data, and the treatment of comparable proposals rather than on claiming a political balance in outcomes. An estimate favorable to one side in one debate and unfavorable in the next, each traceable to the same baseline and conventions, made the case for impartiality more persuasively than any statement of intent could have done. Rivlin’s tenure established that pattern early enough for successors to inherit it as an expectation rather than invent it as a preference, and later directors maintained the procedural norms of no recommendations and equal access as settled features of the office.

Contrast with the Office of Management and Budget explains why a second estimator altered politics rather than merely adding expertise. The Office of Management and Budget serves the president. Its estimates support the president’s budget, its analysts work within an administration that has proposed the policies under examination, and its director is a presidential appointee charged with advancing the president’s program through the budget process. The structural difference lies in alignment. When the only authoritative estimate of a presidential proposal came from the president’s budget office, congressional skepticism had no institutional footing, and congressional assent had no independent verification. After Title II, a committee could set two estimates side by side, identify the assumptions on which they diverged, and argue about assumptions rather than about access.

Two estimators also changed the burden in public debate. An administration defending a costly proposal after 1975 had to reckon with a legislative estimate that journalists, outside analysts, and members of the president’s own party could quote without quoting the opposition. A Congress defending a costly alternative faced the mirror image of that discipline, since the Congressional Budget Office scored congressional proposals by the same conventions applied to presidential ones. The discipline cut in both directions, and that symmetry supplied the office’s political protection. Former directors themselves made that argument in common when the office came under attack. In July 2017, eight former directors, drawn from service under presidents and congressional majorities of both parties, wrote jointly to congressional leaders urging respect for the reliance Congress had placed on the office’s estimates in developing and scoring bills, and stating that such reliance had served Congress and the American people well during the preceding four decades.

By the close of 2019, the succession of directors illustrated the appointment system’s operation across changes of party control. Rivlin was followed by Rudolph G. Penner, who served from 1983 to 1987, by Robert D. Reischauer, who served from 1989 to 1995, by June E. O’Neill, who served from 1995 to 1999, by Dan L. Crippen, who served from 1999 to 2003, by Douglas Holtz-Eakin, who served from 2003 to 2005, by Peter R. Orszag, who served from 2007 to 2008 before directing the Office of Management and Budget under President Barack Obama, by Douglas W. Elmendorf, who served from 2009 to 2015, and by Keith Hall, whose service began in 2015. Phillip L. Swagel began service as director on June 3, 2019, and held the office at the end of that year. An economist, seated at the Office of Management and Budget, estimated in support of a president’s budget, and seated at the Congressional Budget Office, estimated in support of a legislature’s decision. Title II made the post, not the person, carry the allegiance.

Standing, in that dated sense, can be described without superlatives. By the end of 2019 the office’s estimates had become the working arithmetic of fiscal debate in the precise sense that budget resolutions were drafted against its baseline, points of order were argued from its scorekeeping, and legislation was revised in committee to change its measured cost under its conventions. Measures were redesigned to fall within an allocation, effective dates were moved to alter a score, and offsets were selected because the office would credit them under its method, and each of those behaviors testified to authority more concretely than praise did. Authority of that kind has a limit worth naming. An estimate is a projection under assumptions, not a measurement of a settled fact, and directors of the office stated that limitation repeatedly in testimony and in the explanatory material that accompanied major products. Error in a projection did not by itself impeach the system, because the alternative to a stated projection with stated assumptions was not certainty.

The pillar significance of the Congressional Budget Office therefore lies in sequence. The Budget Committees came first in the statute’s organization, the budget resolution supplied the levels, and the points of order supplied the sanctions, but none of those elements could operate on trust alone. Each required a number that members who disagreed about policy could accept as the figure to argue from, produced by an office that neither chamber owned alone, that no president directed, that offered no recommendations of its own to defend, and that served every requester by the same standard. Numbers before rules was not a slogan attached to the statute after enactment. In Title II, the statute built the number first, housed it, staffed it, and protected it, and every enforcement device elsewhere in the Congressional Budget Act drew strength from that prior act of construction for the remainder of the period ending December 31, 2019.

The Resolution and the Allocation Chain

How does a budget resolution control individual spending bills?

A budget resolution controls a spending bill indirectly, through allocations and points of order inside each chamber. Once adopted, its totals are divided among committees and appropriations subcommittees, and a bill or amendment that would breach the applicable allocation faces a point of order that blocks consideration unless the chamber waives the objection.

Title III of the Congressional Budget Act built that indirect control into the center of the congressional budget process, by a deliberate choice about the kind of instrument a budget plan ought to be. The plan took the form of a concurrent resolution, a measure adopted by the House of Representatives and the Senate in identical form. A bill, upon passage by both chambers, went to the president for signature or veto, and a signed bill became law. A concurrent resolution stopped at adoption. No presentment to the president occurred, no signature followed, and no statute resulted. The Congressional Research Service, in reports on the budget resolution issued through 2019, drew the consequence in procedural terms. Because the resolution never became law, the resolution could not amend existing law, could not create new law, and could not compel any action outside the two chambers that had adopted the resolution. Legal force governing agencies, programs, taxes, and borrowing remained in separate legislation, each enacted through the ordinary lawmaking process. That formal weakness was the design, rather than an oversight in the design.

Section 301 of the 1974 act, codified in Title 2 of the United States Code, directed Congress to complete action on a concurrent resolution on the budget for the fiscal year beginning on October 1 of the year in question, stating a complete fiscal position for that year and for at least each of the four ensuing fiscal years. For each year covered, the resolution set forth totals of new budget authority and outlays, the total of federal revenues and the amount, if any, by which the aggregate level of federal revenues should be increased or decreased by bills and resolutions to be reported by the appropriate committees, the surplus or deficit in the budget, and the public debt. For purposes of Senate enforcement under Title III, the resolution also set forth outlays and revenues of the old age, survivors, and disability insurance program established under Title II of the Social Security Act, amounts that stood apart from the budget totals because of the off budget status accorded to Social Security. Aggregate figures alone would have described an aspiration without assigning responsibility for any part of the aspiration, so section 301 required a second level of statement. The resolution set forth new budget authority and outlays for each major functional category, based on allocations of the total levels, so that the aggregate figure appeared as the sum of amounts assigned to broad purposes such as national defense, international affairs, energy, agriculture, transportation, education and training, health, income security, and administration of justice. The functional amounts had to add to the totals.

The resolution nevertheless stopped well short of legislating program by program. The Congressional Research Service explained in its introductions to the federal budget process, in versions issued before the end of 2019, that the resolution did not allocate funds among specific programs or accounts. Major program assumptions underlying the functional amounts often appeared in the reports accompanying a resolution, but those assumptions carried no binding force against the committees that later wrote the actual bills. The committee with jurisdiction retained authority to distribute the committee allocation differently, provided the committee respected the enforceable ceiling that the allocation chain imposed. The distinction between an assumption and an allocation defined the boundary of Budget Committee power. The Budget Committees proposed the shape of the whole. The legislative and appropriations committees retained the drafting of the parts, subject to a ceiling. Reformers who had complained before 1974 that Congress decided spending in fragments, without ever voting on a total, received a vote on a total.

Drafting responsibility rested with the two committees that Title I had created for the purpose, the Committee on the Budget of the House and the Committee on the Budget of the Senate. Each Budget Committee marked up and reported a resolution for its own chamber, and the two chambers then had to reach agreement on a single text, ordinarily through a conference report that each chamber adopted. The Budget Committees did not draft in isolation. Other committees of each chamber submitted views and estimates reports on matters within their jurisdiction, addressing the budget submitted by the president and stating the preferences and legislative plans of the submitting committee for the year ahead. The Congressional Budget Office supplied estimates and analysis that belonged to Congress rather than to the executive branch.

The timetable in section 300 gave that drafting task a place in the calendar, and the timetable changed in a way that revealed a shift in the theory of the resolution itself. As enacted in 1974, the process called for two resolutions in each cycle. Under the original design described by the Congressional Research Service in its historical reports on budget resolutions and in its reports on reconciliation timing, the House and Senate were scheduled to complete action on a first resolution by May 15, setting advisory targets, and to adopt a second resolution by September 15, shortly before the fiscal year began on October 1, setting binding limits. Reconciliation belonged naturally to the second stage, as a device for making last minute adjustments in pending legislation or in existing law so that enacted policy conformed to the binding figures. The Balanced Budget and Emergency Deficit Control Act of 1985, enacted as Public Law 99-177 on December 12, 1985, eliminated the requirement for a second budget resolution and required Congress to complete action on a single budget resolution by April 15 of each year. Before the fiscal 1987 cycle, the deadline for reporting the resolution had been April 15 and the deadline for final adoption had been May 15, and the amended timetable moved final adoption to April 15. A second resolution remained possible as an additional resolution, although the Congressional Research Service noted that no second resolution had been adopted after the fiscal 1982 resolution. The single resolution, adopted in spring, was expected to govern legislation written during the months that followed.

Adoption by April 15 remained a target rather than a reliable description of practice. In the historical information report on congressional budget resolutions, the Congressional Research Service recorded that Congress met the target date only six times for fiscal years 1976 through 2015, twice under the original May 15 deadline, in 1975 and 1976, and four times after the 1985 act moved the deadline to April 15, in 1993, 1999, 2000, and 2003, corresponding to the resolutions for fiscal 1994, fiscal 2000, fiscal 2001, and fiscal 2004. Late adoption weakened the resolution mechanically, because appropriations subcommittees drafted bills during the spring and early summer. The House could consider annual appropriations bills after May 15 even without an adopted resolution, and interim subdivisions adopted without a resolution lacked the enforceable character of subdivisions made under an adopted allocation.

The allocation chain converted the adopted totals into enforceable discipline, and section 302 supplied the links. Under section 302(a), the totals in the resolution were allocated among the committees of each chamber with jurisdiction over spending or revenue, so that each committee received an allocation of new budget authority, entitlement authority, and outlays consistent with the resolution. In practice the allocations appeared in the joint explanatory statement accompanying the conference report on the resolution, assigning to each committee a share of the whole for the first fiscal year covered and for the total of the years covered. The Appropriations Committee of each chamber received the allocation covering discretionary spending to be provided in appropriations acts, stated as a single committee total. Authorizing committees received allocations covering mandatory spending within their jurisdiction, and revenue legislation stood under the aggregate revenue level and the allocation applicable to the revenue committees. A total that no single committee could observe in drafting thus became committee figures that each committee could test a reported bill against, using estimates supplied by the Congressional Budget Office. Scorekeeping reports tracked legislation against the allocations, so that a breach appeared as an arithmetic finding.

The second link belonged to the Appropriations Committees alone. Section 302(b) required each Appropriations Committee, after receiving the allocation under section 302(a), to subdivide the allocation among the subcommittees of that committee. Each subcommittee held responsibility for one of the regular appropriations bills, and the subdivision, commonly called the suballocation, fixed the amount of new budget authority available to that subcommittee for the fiscal year. The sum of the subdivisions could not exceed the committee allocation. Section 302(c) barred consideration of a measure within the jurisdiction of the Appropriations Committees that provided new budget authority for a fiscal year until the committee had made the subdivisions for that year. The sequence therefore ran in a strict order. The resolution stated the total, the allocation assigned the total to the committee, the subdivision assigned shares to the subcommittees, and only then did bills emerge for floor consideration with a ceiling attached to each bill in advance. A subcommittee that wished to provide more for the activities in the subcommittee bill had to find room by providing less for other activities in the same bill, persuade the full committee to revise the subdivisions within the committee allocation, or seek a larger allocation.

Enforcement depended on points of order, and the character of that enforcement deserves exact statement, because a point of order differs from a prohibition in law. A statute that forbids an agency action binds the agency whether or not any official raises an objection. An allocation bound a bill only through the procedures of the chamber considering the bill. The Congressional Research Service stated the point directly in its reports on allocations and subdivisions. The allocation limits were not self enforcing, and a member had to raise a point of order for an allocation to be enforced. If no member objected, a bill that exceeded an allocation could proceed. If a member did object, the presiding officer ruled on the objection under the precedents of the chamber, and a sustained objection stopped the offending measure, amendment, or provision from proceeding in that form.

Section 302(f) furnished the principal objection tied to the chain. Under section 302(f), consideration was barred for a measure or amendment that would cause the allocation under section 302(a) or the subdivision under section 302(b) to be exceeded, for the first fiscal year covered by the resolution or for the total of all fiscal years covered. Section 311(a) furnished the companion objection tied to the aggregates, barring consideration of a measure that would cause the total levels of new budget authority or outlays in the resolution to be exceeded, or that would reduce revenues below the revenue level in the resolution, for the applicable year. In the House, section 311(c) narrowed the aggregate objection through the exception that the Congressional Research Service associated with Representative Vic Fazio. A House measure fell under the section 311 point of order only if the measure also breached the relevant committee allocation, so House enforcement concentrated on the allocation figures. The architecture therefore enforced the whole mainly by enforcing the parts, on the reasoning that totals composed of allocations could not be breached if every allocation held. Section 306 limited consideration of matters within the jurisdiction of the Budget Committees that had not been reported by a Budget Committee, protecting the committees that maintained the allocations against circumvention.

The practical consequence of a sustained point of order was severe. A bill barred by a point of order received no vote on the merits of the programs the bill funded or the policy the bill enacted. Debate on substance became irrelevant once the arithmetic objection succeeded, because the procedural ruling removed the vehicle for the substance. An amendment that breached a subdivision met the same fate even when every other provision of the underlying bill complied, and the sponsor of such an amendment had to redraft within the subdivision, pair an increase with an offsetting decrease inside the same bill, or seek a waiver. The offsetting route preserved a genuine choice within discipline. An amendment that raised one activity and lowered another by a commensurate amount in the same fiscal year did not exceed the subdivision, an example structure that the Congressional Research Service used in explaining section 302(f). A member proposing additional spending had to name the reduction that paid for the addition, or confront the objection that the proposal spent the same budget authority twice.

Waiver supplied the safety valve, and the valve differed by chamber. In the House, points of order under the Budget Act could be waived by a special rule reported by the Committee on Rules and adopted by the House for the consideration of a specified measure, or by unanimous consent. A special rule commonly waived all points of order against consideration of a bill and against specified provisions, and adoption of the rule required a simple majority of the House. The majority that controlled the Rules Committee therefore held a reliable means of releasing a favored bill from the allocation chain, at the price of a recorded vote on a rule that disclosed the release. In the Senate, waiver of a point of order under section 302(f) or section 311 generally required a motion under section 904 of the act and an affirmative vote of three fifths of the senators duly chosen and sworn, sixty votes when no more than one seat stood vacant, or waiver by unanimous consent. The supermajority threshold gave a minority that could hold forty one votes power to sustain an allocation against a majority that wished to breach the allocation.

The honest limit of the system appeared in the years when the chain had no first link, because the House and Senate failed to agree on a resolution at all. The Congressional Research Service, in the deeming resolutions report in the version updated on April 29, 2019, and in the frequently asked questions report on the budget resolution, recorded the fiscal years without an adopted resolution as fiscal 1999, fiscal 2003, fiscal 2005, fiscal 2007, the five consecutive years fiscal 2011 through fiscal 2015, fiscal 2019, and fiscal 2020, for which no resolution had been adopted by the close of 2019. Eleven fiscal years without agreement established that failure was a recurring feature of the process rather than a rare accident. Without identical adoption by both chambers, the reconciliation process could not be triggered for both chambers, because reconciliation instructions required a resolution agreed to in the same form. Each chamber also lost the common frame that a joint resolution supplied, and later negotiation between the chambers had to settle totals that a resolution existed to settle in advance.

Congress legislated in those years under deemed levels. A deeming resolution, as the Congressional Research Service defined the term, was a measure or provision deemed to serve in place of an annual budget resolution for the purpose of establishing enforceable budget levels for the upcoming fiscal year. The device had no prescribed content in the 1974 act and no formal definition, and chambers improvised the vehicle to fit the year. In some years a chamber adopted a simple resolution solely for deeming purposes. In other years deeming language rode inside a special rule that also provided for consideration of an appropriations bill. In other years a statute supplied the levels. For fiscal 2012, section 106 of the Budget Control Act of 2011, Public Law 112-25, enacted on August 2, 2011, required the chair of the Senate Budget Committee to file enforceable budgetary levels for that year. For fiscal 2019, sections 30103 and 30104 of the Bipartisan Budget Act of 2018, Public Law 115-123, enacted on February 9, 2018, required the chairs of the Senate and House Budget Committees respectively to file enforceable levels for fiscal 2019 if Congress had not agreed to a budget resolution by May 15, 2018, and no agreement followed by that date. For fiscal 2020, the House adopted a deeming resolution, H.Res. 293, on April 9, 2019, and enforceable levels also rested on the Bipartisan Budget Act of 2019, Public Law 116-37, signed on August 2, 2019. Deemed levels allowed appropriations to proceed under a ceiling, and a ceiling of statutory origin could bind more firmly than a resolution, because a statute bound as law until amended by law. Deeming nevertheless surrendered the central act of the Title III design. No joint vote on a complete fiscal position occurred, and each chamber often enforced levels that the other chamber had never adopted. The process retained arithmetic without retaining agreement.

Reflection on that record explains why the designers of 1974 chose internal discipline over law. A budget plan cast as law would have required presentment, and presentment would have given the president a veto over the fiscal position of Congress, in a statute written largely to restore congressional control of the budget against a president who had claimed authority to refuse to spend appropriated funds. Members who had fought the impoundment battles of 1973 and 1974 had no intention of handing the executive a veto over the instrument designed to state the legislative alternative to the executive budget. A concurrent resolution avoided presentment by its nature and allowed Congress to state a position, amend the position in a later resolution, and enforce the position through rules that each chamber controlled under the constitutional authority of each House to determine the rules of proceedings. The price followed from the same nature. A concurrent resolution could not bind a future Congress, because a later Congress could adopt different levels, waive the points of order, or decline to adopt any resolution, and no court stood behind the allocations to compel compliance. Even within a single Congress, enforcement required a member willing to object and a chamber unwilling to waive. The framers of the 1974 act accepted that exposure deliberately, on a judgment about where budget discipline could realistically live. The record through the end of 2019 supported parts of that judgment and rebuked other parts. In years with an adopted resolution, the allocation chain gave every spending bill a ceiling known in advance. In the fiscal years without agreement, ten closed through fiscal 2019 and the fiscal 2020 cycle completed in 2019 made an eleventh, the same record showed that a system built on self discipline failed exactly when agreement on totals proved hardest to reach. Title III therefore stood at the end of 2019 as a structure of genuine procedural force and acknowledged fragility, a chain strong at every link after adoption and dependent, for the first link, on the willingness of two chambers to vote together on a single set of numbers.

Reconciliation as Conceived and the End of Impoundment

Two parts of the Congressional Budget Act addressed the same problem from opposite directions. Section 310 gave Congress a controlled method for revisiting mandatory spending and revenue law already enacted, so that permanent law could be brought into line with the totals adopted for a fiscal year. Title X addressed executive action against law already enacted, and replaced an asserted power to withhold appropriated funds indefinitely with proposal procedures bounded by deadlines and by congressional decision.

Reconciliation entered the statute in section 310, codified at 2 U.S.C. 641, as an optional step. Nothing in the Congressional Budget Act required a budget resolution to contain reconciliation instructions. A resolution could set aggregate levels for revenue, budget authority, outlays, the deficit or surplus, and the public debt, divide those levels among committees through allocations, and leave permanent law untouched. Reconciliation became necessary only when the gap between permanent law and the resolution became the point, when mandatory programs on a given statutory course would produce outlays above the resolution level or revenue law unchanged would produce receipts below the resolution floor. Section 310 supplied the mechanism for closing that gap by legislation through the committees with jurisdiction over the law concerned.

The design placed reconciliation late in the sequence, and the lateness mattered. Congress first adopted targets, made allocations, and acted through appropriations and ordinary authorizing legislation. Reconciliation stood at the end as a conforming device. Committees named in the instructions were directed to report changes in the laws within the jurisdiction of each committee, in amounts and by dates specified in the resolution, so that enacted law would conform to the resolution. The operative idea was conformity, not origination. The resolution supplied the fiscal judgment in aggregate numbers, and the authorizing and revenue committees supplied the drafting needed to make permanent law match those numbers.

The deficit orientation of the original conception followed from the circumstances of the early and middle 1970s. The planning problem that produced the Congressional Budget Act was the tendency of Congress to enact spending and revenue decisions in fragments, each defensible on program grounds, without a vote on the combined fiscal result. A conforming device attached to that planning system served principally to reduce the deficit relative to the path implied by existing law. Section 310 did not, by its own terms, restrict instructions to deficit reduction. The text authorized instructions directed at changes in revenues, in mandatory spending, and in the statutory limit on the public debt, and the fiscal direction of any given set of instructions depended on the resolution that contained the instructions. Deficit reduction was the principal purpose in the conception and in the early practice, rather than a permanent textual limit, and that distinction between purpose and text explains much of the later history.

The assembly rule reinforced the subordinate character of the device. When more than one committee received instructions, each committee reported recommended changes to the Budget Committee of its chamber, and the Budget Committee assembled the submissions into a single omnibus bill without substantive revision. Substantive choices about which benefits, eligibility rules, payment formulas, tax rates, credits, or deductions would change remained with the committees of jurisdiction, subject to the dollar instructions in the resolution and to action on the floor. A Budget Committee could package the responses, but the package could not become a second drafting of health, welfare, agriculture, or tax policy by a committee created in 1974 to manage aggregates.

The Senate procedure attached to reconciliation transformed a modest conforming device into a potent legislative instrument. Debate on a reconciliation bill in the Senate was limited to 20 hours, with debate on a conference report limited to 10 hours. Because debate ended by operation of the statute, extended debate could not block a reconciliation measure indefinitely, and passage required a simple majority of senators voting, with the Vice President available to break a tie, rather than the three fifths vote ordinarily needed to close debate under Senate rules. A measure that qualified as a reconciliation bill and held the support of a Senate majority could therefore reach enactment on majority support alone, provided that the House agreed and the president signed the measure, or that Congress overrode a veto.

Congress first used reconciliation in 1980, for fiscal 1981. The Omnibus Reconciliation Act of 1980, Public Law 96-499, enacted on December 5, 1980, gathered committee responses that reduced spending and adjusted revenues in conformity with the fiscal 1981 resolution, at a scale close to the conception. Reconciliation worked as an omnibus, forcing committees with separate jurisdictions into one measure governed by one set of aggregate instructions, and the Senate time limit converted committee decisions that might have stalled as separate bills into a single measure with a guaranteed path to a final vote.

The center of gravity shifted quickly during the 1980s. The Omnibus Budget Reconciliation Act of 1981, Public Law 97-35, enacted on August 13, 1981, applied instructions on a far larger scale, changing mandatory spending programs across committee jurisdictions in a single measure. The Tax Equity and Fiscal Responsibility Act of 1982, Public Law 97-248, enacted on September 3, 1982, and the Omnibus Budget Reconciliation Act of 1982, Public Law 97-253, enacted five days later, used the process for deficit reduction through revenue increases and spending restraint, and near annual omnibus measures followed through the decade. The Byrd rule, section 313 of the statute, added in 1985 and made permanent in 1990, entered that development as a Senate content constraint on material extraneous to budgetary purposes, noted here only in passing as the boundary that later practice placed around the vehicle.

Practice during the 1990s confirmed the enlarged role. The Omnibus Budget Reconciliation Act of 1990, Public Law 101-508, enacted on November 5, 1990, carried the legislative substance of a deficit agreement reached after extended negotiation. The Omnibus Budget Reconciliation Act of 1993, Public Law 103-66, enacted on August 10, 1993, again placed the central fiscal program of a budget resolution into a reconciliation bill, enacted after narrow votes in both chambers, including a tie breaking vote in the Senate. In 1996 reconciliation carried a restructuring of cash welfare and related programs in the Personal Responsibility and Work Opportunity Reconciliation Act of 1996, Public Law 104-193, enacted on August 22, 1996. In 1997 the process divided into two measures under one budget plan, the Balanced Budget Act of 1997, Public Law 105-33, centered on mandatory spending programs including Medicare and Medicaid, and the Taxpayer Relief Act of 1997, Public Law 105-34, centered on revenue, both enacted on August 5, 1997. Reconciliation measures also drew vetoes in 1995, 1999, and 2000, and a further measure reached a veto in January 2016.

During the 2000s the direction of the instructions broadened in a way that completed the departure from the original deficit orientation. The Economic Growth and Tax Relief Reconciliation Act of 2001, Public Law 107-16, enacted on June 7, 2001, was a revenue measure enacted under instructions written for a budget resolution that contemplated a reduction in revenues within a projected surplus, so reconciliation served a deliberate increase in the deficit relative to the baseline rather than deficit reduction. The Jobs and Growth Tax Relief Reconciliation Act of 2003, Public Law 108-27, enacted on May 28, 2003, followed the same revenue pattern during a period of enacted deficits. In the middle of the decade the two track pattern reappeared, with the Deficit Reduction Act of 2005, Public Law 109-171, enacted on February 8, 2006, addressing mandatory spending, and the Tax Increase Prevention and Reconciliation Act of 2005, Public Law 109-222, enacted on May 17, 2006, addressing revenue. The College Cost Reduction and Access Act of 2007, Public Law 110-84, enacted on September 27, 2007, applied the vehicle to student aid.

Practice during the 2010s placed reconciliation at the center of health and revenue legislation. The Health Care and Education Reconciliation Act of 2010, Public Law 111-152, enacted on March 30, 2010, amended health and student lending legislation through a reconciliation measure adopted after the principal health statute had been enacted through ordinary procedure, a sequencing that used reconciliation for a defined set of budgetary amendments within the Byrd rule boundary rather than for the entire program. The reconciliation measure enacted on December 22, 2017, Public Law 115-97, commonly known by the short title Tax Cuts and Jobs Act, rewrote large portions of the Internal Revenue Code under instructions in the fiscal 2018 budget resolution. By the close of 2019, the Congressional Research Service could describe reconciliation as used in most years since the first use in 1980. A reader seeking the full sequence of adoption, instruction, committee response, floor consideration, and conference action can find a full account of reconciliation procedure treated as a process in its own right. Section 310 was drafted to make permanent law conform to a resolution, and four decades of practice converted the conformity device into the principal Senate vehicle for major revenue and entitlement legislation.

Title X of the same statute, the Impoundment Control Act, sections 1011 to 1017, codified at 2 U.S.C. sections 681 to 688, governed the reverse situation, executive action against enacted spending law. An impoundment, in the vocabulary of Title X, covers executive withholding of budget authority from obligation, whether permanent or temporary, and the act divided every impoundment into one of two categories with different congressional defaults. A rescission proposed the permanent cancellation of budget authority and required affirmative congressional enactment before the cancellation took effect. A deferral proposed a temporary delay in availability and took effect unless Congress acted, under the original design, to disapprove the delay. The asymmetry was the settlement in miniature. Permanent executive economy required a new law, while temporary executive delay received a presumption of validity, subject to congressional reversal and to strict reporting.

A rescission proposal under section 1012, 2 U.S.C. 683, had to identify the amount of budget authority proposed for cancellation, the account, department or establishment, and any specific project or governmental function involved, state the reasons for the proposed cancellation, and supply the estimated fiscal, economic, and budgetary effects of the proposal. The special message went to both chambers, was referred to the appropriate committees, was printed as a document of each chamber and in the Federal Register, and was transmitted to the Comptroller General for review. While Congress considered the proposal, the president could withhold the identified authority from obligation, but only within a fixed window. If Congress enacted a rescission bill approving the proposal, in whole or in part, the approved portion was cancelled and the remainder released. If Congress enacted nothing, the withholding authority expired by operation of the statute.

What happens if Congress ignores a rescission proposal?

If Congress takes no action, the proposal fails by operation of law. The President may withhold the identified budget authority only during the statutory review period. Once forty-five days of continuous session pass without enactment of a rescission bill, the authority must be released and made available for obligation for the purposes Congress enacted.

The phrase continuous session carried technical meaning. Days on which either chamber stood in recess for more than three days did not count toward the period, so the 45 days of continuous session ordinarily spanned a longer stretch of calendar time, and expedited procedures for a rescission bill applied while the withholding window remained open. The logic inverted the older practice. Before Title X, congressional silence left a withholding in place and placed the burden on Congress to compel release through new legislation over a likely veto. After Title X, silence defeated the proposal, and the consequence favored the enacted appropriation. Congress reinforced the deadline in 1987 against repetition, when section 207 of Public Law 100-119, enacted on September 29, 1987, barred a president, after one rescission proposal had lapsed, from submitting a new proposal covering identical or substantially similar authority and thereby restarting the withholding period for the same funds.

The deferral half of Title X began from a weaker congressional position, and later litigation and amendment stripped the weakness down to routine administration. Under section 1013 as enacted in 1974, the president transmitted a special message proposing to defer budget authority, stating the amount, the account, the period, and the reasons, and the deferral took effect for the period proposed, subject to a limit preventing a deferral from extending beyond the end of the fiscal year in which the proposal was reported. Either chamber could end the deferral at any time by adopting an impoundment resolution disapproving the proposal, a one house veto that required no action by the other chamber and no presentment to the president. A deferral, unlike a rescission, therefore placed the burden of action on Congress.

The Supreme Court destroyed the foundation of that design in Immigration and Naturalization Service v. Chadha, 462 U.S. 919 (1983). The case arose under an immigration statute rather than under Title X, and the Court held that action by one chamber altering legal rights and duties outside that chamber violated the bicameral passage and presentment requirements of the Constitution. The reasoning reached every one house veto in the United States Code, including the impoundment resolution in section 1013. In City of New Haven v. United States, 809 F.2d 900 (D.C. Cir. 1987), the Court of Appeals for the District of Columbia Circuit applied Chadha to the deferral provisions and treated presidential deferral authority as inseverable from the invalidated one house check, with policy deferrals falling together with the mechanism Congress had chosen for controlling deferrals. Congress resolved the resulting doubt by legislation. Section 206 of Public Law 100-119, enacted on September 29, 1987, at 101 Stat. 785, rewrote section 1013, eliminated the one house disapproval mechanism, and provided that deferrals shall be permissible only to provide for contingencies, to achieve savings made possible by or through changes in requirements or greater efficiency of operations, or as specifically provided by law. Deferrals based on policy disagreement with an enacted program fell outside the permissible list, and a deferral could not extend beyond the end of the fiscal year of the proposal. After 1987 a deferral resembled an accounting and management tool, suited to changed requirements, efficient operations, or a genuine contingency, and unsuited by text to executive revision of enacted priorities.

Oversight of both procedures fell to the Comptroller General, the head of the Government Accountability Office. Section 1014, 2 U.S.C. 685, required transmission of each special message to the Comptroller General, who was directed to review the message and report to both chambers on the facts surrounding the proposed action and the probable effects, including a determination, for a deferral, of conformity with existing statutory authority. Section 1015, 2 U.S.C. 686, addressed gaps and errors in executive reporting. If the Comptroller General found that action, or failure to act, amounted to a rescission or a deferral without the required special message, the Comptroller General reported the impoundment to Congress, and the report carried the same effect as a presidential special message, starting the statutory clock. If the Comptroller General concluded that a message transmitted as one category belonged in the other, the Comptroller General reported the reasons to both chambers, allowing Congress to treat the action under the correct procedure. Classification carried consequences because the defaults differed, and an executive preference for the deferral label over the rescission label could otherwise defeat the requirement of affirmative enactment for permanent cancellation.

Section 1016, 2 U.S.C. 687, supplied an enforcement remedy unusual in budget law. If budget authority was required to be made available for obligation under Title X and an agency failed to release the authority, the Comptroller General was expressly empowered, through attorneys selected by the Comptroller General, to bring a civil action in the United States District Court for the District of Columbia to require release, and the court was empowered to enter any decree, judgment, or order necessary to make the authority available. No action could be brought until 25 calendar days had expired after the Comptroller General filed an explanatory statement with the Speaker of the House and the President of the Senate. Congressional Research Service accounts record a single early use, a 1975 action seeking release of deferred authority for a low income home ownership program, in which the district court denied a motion to dismiss and granted preliminary relief before the parties stipulated to dismissal. Through the end of 2019 the authority otherwise remained unused.

The working record of the rescission procedure through 2019 matched the design. Presidents used special messages as a routine instrument for proposing cancellations, and Congress accepted a minority of the proposals by number and by dollar amount. In testimony delivered in 1999, GAO reported that presidents had proposed 1,172 rescissions totaling 75.8 billion dollars since enactment of the Impoundment Control Act in 1974, and that Congress had accepted 459 of the proposals, 39 percent by number, totaling 24.9 billion dollars, about 33 percent of the amount proposed. A special message forced disclosure, started a clock, and gave a cancellation proposal expedited consideration, but enactment remained an ordinary act of legislation. Presidential proposals succeeded when congressional majorities already agreed with the cancellations proposed, or could be persuaded within the short window, and failed without drama when agreement was absent, because failure required nothing more than the passage of time.

The rescission package of May 2018 illustrated the default in operation late in the period covered here. On May 8, 2018, the president transmitted a special message proposing rescissions of roughly 15 billion dollars in budget authority, drawn largely from unobligated balances. The House passed a rescission bill, the Spending Cuts to Expired and Unnecessary Programs Act, H.R. 3, by a vote of 210 to 206 in June 2018. In the Senate, a vote on proceeding with the bill failed by 48 to 50 in June 2018, no rescission bill became law, and the proposed cancellations did not take effect. The Congressional Budget Office had estimated that the package would reduce outlays by only about 1 billion dollars over the following decade, because much of the budget authority identified would have gone unspent in any event. Each branch had performed the role assigned by Title X, and the enacted appropriations, as modified by other law, governed.

The closing point of the impoundment settlement is conceptual. Title X converted an asserted unilateral power into a proposal procedure. The assertion, pressed during the Nixon administration, had treated appropriations as ceilings that the executive could undershoot at discretion for policy reasons, indefinitely and without a vote. The statute answered with categories, messages, deadlines, review by the Comptroller General, and defaults that ran in favor of enacted law. Judicial resolution of the largest Nixon era dispute ran in parallel rather than settling the constitutional question. Train v. City of New York, 420 U.S. 35 (1975), covered in full in an earlier section, held as a matter of statutory interpretation that the water pollution statute at issue required allotment of the full sums Congress had authorized, without a categorical ruling on the constitutional limits of impoundment. The constitutional question therefore remained, at the end of 2019, formally open in the abstract and practically governed by the statute.

Creation	Problem it addressed	Standing at the close of 2019
The budget committees in both chambers	No committee in either chamber owned total spending, total revenue or the deficit, so aggregate fiscal policy had no institutional home	Committees in both chambers had reported budget resolutions and maintained allocations and scorekeeping across the decades through 2019, with membership rotation and money committee representation preserved from the 1974 design
The Congressional Budget Office	Congress judged presidential proposals using presidential estimates, forecasts and cost figures, with no estimating office of the legislature	By the close of 2019 the office supplied the estimates, baselines and scorekeeping against which resolutions were drafted and points of order were argued, under a director whose service began on June 3, 2019
The concurrent budget resolution	No vote stated a complete congressional fiscal position covering revenue, budget authority, outlays, the deficit or surplus and the public debt	The resolution had stated that position in most cycles, but ten fiscal years through fiscal 2019 had closed without an agreed resolution, none had been adopted for fiscal 2020 by the close of 2019, and deemed levels had substituted for agreement in those years
The allocation chain under sections 302(a) and 302(b)	Adopted totals had no path down to the committees and subcommittees that drafted individual bills, and no objection policed a bill against a total	When a resolution was adopted, allocations and subdivisions gave each committee and appropriations subcommittee a ceiling enforced by points of order, subject to waiver by each chamber under chamber rules
Reconciliation	Mandatory spending and revenue law enacted in permanent statutes could drift away from the totals a resolution adopted for a fiscal year	First used in 1980 and used in most years after that date through 2019, reconciliation had grown from a conforming device into a principal Senate vehicle for major revenue and entitlement legislation, with Senate debate limited by statute
Impoundment control through the rescission and deferral procedures	A president could withhold enacted funds indefinitely on an asserted constitutional claim, with no reporting duty, no deadline and no congressional default	Permanent cancellation required an enacted rescission bill and deferrals were confined to routine administrative grounds after 1987; in the May 2018 test the House passed its bill 210 to 206, a Senate vote on proceeding failed 48 to 50 in June 2018, and the Congressional Budget Office estimated about 1 billion dollars of outlay reduction over a decade

A New Fiscal Year and the Record of the Timetable

Title V of the Congressional Budget Act addressed a problem of the calendar before the statute addressed any question of totals. A process built around an annual resolution, committee allocations, and a sequence of authorizing and appropriations bills required time in a fixed order, and the calendar in force at enactment gave that sequence less time than the sequence consumed. The drafters of Public Law 93-344 treated the starting date of the fiscal year as a working part of the machinery. Moving the date changed no spending level and created no sanction, but the change lengthened the runway on which every other procedure had to operate.

Why did Congress move the fiscal year to October?

Congress moved the fiscal year to October to gain three additional months between the president submitting a budget and the year beginning, because appropriations and authorizing work under the July start routinely finished late, forcing stopgap funding and compressing orderly consideration of the totals, allocations, and individual bills.

The July calendar had produced a documented pattern of lateness. The Congressional Research Service, in reports on continuing resolutions and federal funding gaps, recorded that during the 25 fiscal years from fiscal 1952 to fiscal 1976 Congress did not once enact all of the regular appropriations acts by the start of the fiscal year, and that one or more continuing resolutions were enacted in nearly every year of that period, fiscal 1953 excepted on the resolution count. Appropriations bills drafted against a deadline already passing absorbed floor time in the opening weeks of the new year, crowding authorizing and revenue work scheduled for the same months, and agencies began each year uncertain of final funding for programs already operating.

Section 501 of the statute, at 88 Stat. 321, rewrote the definition directly. The provision amended the Revised Statutes so that the fiscal year ran from July 1 through June 30, holding that arrangement in place through June 30, 1976. Beginning October 1, 1976, the fiscal year would run from October 1 through September 30. The definition enacted by that amendment appears at 31 U.S.C. 1102. The drafting was deliberately prospective. Congress legislated in the summer of 1974 for a change taking effect more than two years later, giving the executive branch a full cycle in which to rebuild budget formulation schedules, accounting periods, and agency planning calendars before the new dates governed any actual appropriation.

Between the two calendars stood a single bridging device, the transition quarter, running from July 1, 1976 to September 30, 1976. The quarter covered the three months belonging to neither the old fiscal year ending June 30, 1976 nor the new fiscal year beginning the next day, and the quarter received its own budget estimates and its own appropriations treatment. Federal activity in those months therefore rested on enacted authority rather than on an assumption that funding from the expired year carried forward. The device was used once and then exhausted.

The three added months were allocated, in effect, to specific stages of the congressional sequence. Under the July start, a presidential budget arriving early in a calendar year left roughly half a year for committee review, adoption of a budget resolution, subdivision of allocations among subcommittees, drafting and passage of a dozen or more appropriations bills in two chambers, conference agreement on each disputed bill, and presentment, all before money had to be available on July 1. The October start placed a full additional quarter between those pressure points and the moment funding lapsed. House completion of appropriations action, set for June 30 in the timetable at section 300, changed character under the new calendar. A missed target at the end of June left a summer of working time rather than an immediate funding gap. The added quarter also gave the intended order of decisions a plausible schedule, with aggregate choices in a resolution preceding allocations, allocations preceding subcommittee drafting, and drafting preceding floor consideration, each stage resting on Congressional Budget Office estimates that required time to prepare and time to contest.

Nothing in Title V compelled Congress to use the added time as the timetable directed. The fiscal year provision changed a definition, and a definition carries no point of order, no sanction, and no consequence for delay beyond the funding gap that delay eventually produces. The new calendar reduced accidental lateness, produced by an overloaded schedule in a chamber acting against a tight deadline. The new calendar could not reduce deliberate lateness, produced when postponement itself served a negotiating purpose for participants who preferred a later decision, a different baseline, or a larger package in which disputed items could be traded against one another.

The procedures of the Budget Act were first implemented fully for fiscal 1977, the opening fiscal year under the October calendar, and fiscal 1977 supplied the strongest single year in the subsequent record. The Congressional Research Service recorded that all of the regular appropriations acts for fiscal 1977 were enacted on time. One qualification from the same accounts belongs beside that result. Two continuing resolutions were enacted for fiscal 1977, not as interim funding for unfinished regular bills, but to fund certain unauthorized programs that had been dropped from the regular appropriations acts. The fiscal 1977 record therefore showed a completed set of regular acts by October 1, 1976 together with a narrower use of the continuing resolution device for programs lacking authorization. Later reliance on continuing resolutions took the opposite form, covering core programs whose regular bills remained unfinished after the year had begun.

Measured from fiscal 1977 through fiscal 2019, the fiscal 1977 result was an exception rather than a pattern. The Congressional Research Service identified only four fiscal years in which all regular appropriations acts were enacted by the beginning of the fiscal year: fiscal 1977, fiscal 1989, fiscal 1995, and fiscal 1997. In every other fiscal year of that span, at least one regular act remained unenacted on October 1, and funding for the affected accounts depended on a continuing resolution, on enactment inside a later package, or, in years when agreement failed entirely for a period, on a lapse in appropriations. Three of the four successes fell within a single decade, and no fiscal year after fiscal 1997 joined the list through the end of 2019, a stretch of 22 fiscal years without one complete on time outcome.

A continuing resolution operates mechanically as a temporary appropriations act with a restricted content. In the form described by the Congressional Research Service, a continuing resolution generally funds activities at a rate for operations derived from the prior year level, retains conditions and limitations carried in the prior year acts, and generally bars initiation of new activities, because authority extends to projects and activities conducted in the preceding year rather than to a redesigned program. Congress may write anomalies into the text, adjusting a particular rate or condition where the prior year formula would produce an unworkable result, and each anomaly is an express legislative exception rather than an administrative discretion. Duration is set by the text as well. A resolution may run for days, for weeks, or, in the full year form, for the remainder of the fiscal year, converting a stopgap into the operative appropriations law for accounts that never received a regular act.

The omnibus package developed as the companion device for completing overdue work. Mechanically, an omnibus appropriations measure combines the unfinished regular bills, or the conference products derived from them, into a single legislative vehicle moving through passage and presentment as one act. Combination changes the decision structure facing each chamber. Separate bills permit separate votes and separate veto calculations for each area of spending. A single package presents one vote on the whole, in which an objectionable provision must be weighed against funding for every other account in the vehicle and against the consequences of a funding lapse if the vehicle fails. Negotiation concentrates in the offices able to assemble and clear the full text, and floor amendment opportunities narrow because reopening one component threatens the balance of the whole. By the end of 2019, a missed October deadline, one or more continuing resolutions, and eventual completion through consolidated packages had repeated often enough to function as the operating routine for annual appropriations, alongside the separate routine described in the statute.

The costs of that routine were described in concrete administrative terms by participants in the process. Alice Rivlin, the first director of the Congressional Budget Office, observed in discussion of continuing resolution reliance reported in 2004 that departments and agencies unable to learn a final budget for a fiscal year already underway faced difficulty planning a budget for the following fiscal year, and that the core problem was uncertainty. An agency drafting estimates while operating under a continuing resolution was building the next request on a base that Congress had not finally settled, and the unsettled base propagated forward into the following cycle. Lateness, on that account, reproduced the conditions for further lateness.

Congress answered the timetable record by layering enforcement regimes on top of the 1974 framework, and each layer altered a different variable. The amendments of 1985 attacked the deficit total directly. The amendments of 1990 abandoned fixed deficit totals in favor of limits on the discretionary decisions Congress voted each year, paired with a rule for mandatory spending and revenue changes. The amendments of 2011 returned to a cap structure and added a further automatic reduction mechanism. A separate statute of 1996 attempted to give the president a cancellation power over enacted items, and the Supreme Court decision invalidating that statute in 1998 marked the constitutional edge of enforcement design.

The Balanced Budget and Emergency Deficit Control Act of 1985, known by the names of Senators Phil Gramm, Warren Rudman, and Ernest Hollings, consolidated the timetable and attached an automatic enforcement device to fixed deficit targets. The original 1974 schedule had contemplated two budget resolutions, a first resolution to be completed by May 15 and a second, binding resolution by September 15. The 1985 act replaced that two stage structure with a single budget resolution, with completion of congressional action set for April 15. Against the targets, the act provided sequestration, an across the board cancellation of budgetary resources in covered accounts, triggered when estimated deficits exceeded the statutory path. Enforcement no longer depended solely on points of order raised and sustained on the floor. A calculated excess produced a calculated reduction, scored from budget estimates, and the Gramm-Rudman-Hollings profile carries the design, the litigation over the Comptroller General role in the original mechanism, and the revisions that followed in 1987.

The Budget Enforcement Act of 1990 changed the object of enforcement. Fixed deficit targets had exposed the enforcement system to economic and technical estimating changes beyond congressional control, since a recession or a revised forecast could breach a target without any new legislation passing. The 1990 act replaced the deficit target structure with statutory caps on discretionary spending, divided by category in the opening years of the regime and unified later, together with pay-as-you-go discipline for legislation affecting mandatory spending and revenues. Under that discipline, enacted changes increasing the deficit on the pay-as-you-go scorecard required offsetting changes, and a net deficit increase remaining on the scorecard at the end of a session produced a sequester of covered mandatory programs. Alice Rivlin, director of the Office of Management and Budget in the middle 1990s after directing the Congressional Budget Office in the decade after 1975, later credited the 1990 framework, in published retrospectives dated through 2010, as the starting point of the bipartisan effort that moved the budget from deficit to surplus by the end of the 1990s. The caps expired after fiscal 2002, and no successor statutory cap regime governed the middle years of the following decade.

The Budget Control Act of 2011 constructed a further cap regime on the foundation left by the expired 1990 structure. The act set statutory limits on discretionary budget authority extending across the decade following enactment, divided between defense and nondefense categories in the design, and paired the caps with a second sequestration mechanism calibrated to produce additional automatic reductions. Administration ran through the scorekeeping channels the earlier statutes had used, with the Office of Management and Budget issuing sequestration reports and the Congressional Budget Office supplying estimates against the cap levels. Congress adjusted the cap levels by later statute in several years between enactment and the end of 2019, raising the limits for specified fiscal years through negotiated agreements while leaving the statutory structure in place.

The Line Item Veto Act of 1996 attempted enforcement from the executive side of the process. The act authorized the president, after signing a bill into law, to cancel specified dollar amounts of discretionary budget authority, new direct spending items, and limited tax benefits, subject to a congressional disapproval procedure. In Clinton v. City of New York, 524 U.S. 417 (1998), the Supreme Court held the cancellation authority unconstitutional, reasoning that cancellation altered the legal effect of enacted text through a procedure outside the passage and presentment requirements of Article I. The decision left intact, by contrast, the rescission procedure of the Impoundment Control Act of 1974, under which a president proposes a cancellation and the proposal takes effect only if Congress enacts a rescission bill through the ordinary legislative process. The boundary drawn in 1998 confirmed the allocation of roles the 1974 act had assumed. Enforcement could be made automatic by formula inside a statute Congress enacted, but alteration of enacted law required a further enacted law.

Two readings of the record assembled by the end of 2019 stood in direct opposition, and each rested on evidence the other did not dispute. The disagreement concerned which set of facts described the process accurately: the facts of the calendar, or the facts of the institutional structure inside which every annual fight had been conducted. An assessment of budget process reform and its consequences weighs the same record from the standpoint of reform outcomes, and the two readings frame the question that any such assessment must answer. The failure reading held that a process missed at nearly every deadline, in nearly every year, over four decades, had ceased to describe congressional behavior in any meaningful sense. The framework reading held that the calendar was one component of a larger construction, and that the larger construction had organized federal budgeting continuously from fiscal 1977 through fiscal 2019 even in years when the calendar failed completely.

Allen Schick gave the failure reading an influential scholarly form. The third edition of The Federal Budget: Politics, Policy, Process, published in 2007, treated the deterioration of the congressional budget process as a collapse to be explained rather than as a run of isolated mishaps, examining how the termination of discretionary caps and the weakening of resolution discipline had removed the constraints that once connected aggregate decisions to subsequent bills. In Schick analysis, Congress retained the procedural advantages conferred by the 1974 act, including privileged consideration for budget measures and protection against filibuster delay for reconciliation, while ceasing to exercise the aggregate control that formed the purpose of the act.

The dated record supplied the failure reading with specific exhibits beyond the appropriations count. The Congressional Research Service recorded that no budget resolution was adopted for fiscal 1999, the first such failure after adoption of the Budget Act, and that the failure recurred for fiscal 2003, fiscal 2005, and fiscal 2007. Congressional Research Service listings further recorded no adopted budget resolution for any of fiscal 2011 through fiscal 2015, five consecutive fiscal years, and none for fiscal 2019. In those years the opening move of the statutory sequence, agreement between the two chambers on a single set of aggregates and allocations, did not occur at all, and enforcement proceeded through substitute devices rather than through the resolution the timetable placed first. Completion dates in years when a resolution was adopted told a parallel story against the April 15 deadline, with adoption sometimes falling deep into the fiscal year the resolution governed.

The framework reading answered from the same exhibits and drew the opposite conclusion about institutional survival. In years without an agreed budget resolution, Congress legislated enforceable budget levels through deeming devices, chamber rules, statutory filings by the chair of a Budget Committee, and cap statutes, each device performing the allocation and enforcement functions that a resolution would have performed, and each drafted by reference to the categories, aggregates, and points of order of the 1974 act. The Budget Control Act of 2011 supplied enforceable levels by statute for years in which no resolution was adopted, including a provision directing the chair of the Senate Budget Committee to file allocations for fiscal 2012. Appropriations bills in every year through 2019 were drafted against section 302 allocations or against a statutory substitute for an allocation, scored by the Congressional Budget Office, and exposed to Budget Act points of order on the floor of each chamber.

Alice Rivlin published assessments pointed in the framework direction on estimating capacity, the element of the 1974 design Rivlin had built at the Congressional Budget Office from February 1975. In a published defense of the office dated March 14, 2017, Rivlin described the Congressional Budget Office as the resource that allowed Congress to legislate on complex proposals against a single set of estimates rather than against competing estimates selected by each participant, and warned that undermining the credibility of the office damaged the capacity of Congress to perform legislative work. A reform proposal Rivlin prepared with former Senator Pete Domenici for the Bipartisan Policy Center, discussed in published Brookings Institution proceedings in February and March 2016, proposed that the budget process include all federal spending and revenues, operate transparently, and conclude on time. The proposal treated the existing process as a structure to be completed and repaired, locating the defects in coverage and timeliness rather than in the existence of congressional budgeting itself.

By the close of 2019, the factual core common to both readings could be stated compactly. Title V moved the fiscal year, and the October calendar remained in force at 31 U.S.C. 1102 without interruption from October 1, 1976 through the end of 2019. The timetable built on that calendar was met in full, on the appropriations measure that mattered most, in four fiscal years out of the 43 fiscal years from fiscal 1977 through fiscal 2019. Continuing resolutions and consolidated packages carried annual funding in the remaining years. Every enforcement regime enacted between 1985 and 2011 operated through estimates, scorekeeping, and allocation machinery descended from the 1974 act, and the Supreme Court decision of 1998 left the constitutional allocation of the 1974 design undisturbed. Whether that record amounted to a failed process operating inside a durable framework, or to a durable process with a failed timetable, remained the contested question in the literature dated through 2019. The calendar reform of Title V succeeded permanently. The schedule the calendar was built to serve did not.

What the Statute Did Not Do and Where to Read Next

A pillar account earns trust by stating the limits of the statute as plainly as the creations. The Congressional Budget Act appropriated nothing. No dollar of spending authority issued from the act for a department, a program or a grant, and the totals in a budget resolution authorized no payment by the Treasury. Money moved only when separate appropriations acts and other spending legislation supplied budget authority for a stated purpose. The resolution framed those later decisions, allocated room for those decisions and armed objections against decisions that exceeded the room. Framing, allocation and objection are powers over legislation rather than legislation itself, and the distinction explains both the reach and the boundary of the design.

The act also set no binding limit on deficits. A budget resolution stated a surplus or deficit figure for the years covered, and points of order policed legislation against the aggregates the resolution contained. No provision of the 1974 text forbade a resolution that accepted a large deficit, and no provision penalized Congress for adopting totals that produced borrowing on any particular scale. Enforcement regimes that attached automatic consequences to deficit figures arrived in later statutes, in 1985 and after, resting on choices made by a later Congress rather than on a limit embedded in the 1974 act. Within the framework, discipline depended on the content of the resolution a given Congress adopted and on willingness to sustain objections when legislation pressed against the adopted figures.

Binding a future Congress lay beyond the design as well. The budget resolution was a concurrent resolution, an instrument of the two chambers that adopted the resolution, enforceable through chamber rules and points of order. A later Congress could adopt different totals, waive the points of order in a particular case, replace the timetable by statute, or decline to adopt a resolution and legislate under deemed levels instead. Even within a single Congress, each chamber retained authority over the rules that gave the allocations force. The drafters accepted that exposure as the price of keeping the congressional budget inside Congress, free of presentment and of a presidential veto over the fiscal position of the legislature.

The timetable record supplies the clearest measure of the limit. Title V moved the fiscal year to an October start, and that calendar change held without interruption from October 1, 1976 through the close of 2019. The schedule built on the calendar fared far worse. All regular appropriations acts were enacted by the beginning of the fiscal year in only four fiscal years of the span from fiscal 1977 through fiscal 2019, namely fiscal 1977, fiscal 1989, fiscal 1995 and fiscal 1997. Continuing resolutions carried unfinished accounts in the remaining years, consolidated packages completed overdue bills, and ten fiscal years closed without an agreed budget resolution, with the fiscal 2020 cycle, completed in 2019, adding an eleventh. Deadlines in the act functioned as targets that organized work and exposed delay, rather than as commands backed by a sanction.

Those limits qualify the achievement without cancelling the achievement. “Numbers before rules” survives the qualifications because the estimating capacity changed the terms on which every later argument proceeded. After 1974, and through the close of 2019, a member disputing a presidential figure could answer with a figure produced by an office of Congress, built on a stated baseline and maintained through scorekeeping that tracked action against adopted levels. Committees redesigned bills to change a measured cost, and opponents raised points of order that rested on an arithmetic finding rather than on assertion alone. The framework organized budgeting even in years when the calendar failed, because caps, deemed levels and enforcement statutes enacted between 1985 and 2011 all operated through estimates, allocations and scorekeeping descended from the 1974 act.

Reading outward from the pillar follows the same sequence. The budget process sits inside a wider lawmaking process of introduction, committee action, floor passage, conference and presentment, and how a bill becomes law in practice traces that wider process through which every appropriation, reconciliation bill and rescission bill governed by the Congressional Budget Act had to travel. A reader building a study file around the statute, with the identity anchors of Public Law 93-344, the codification at Title 2 of the United States Code, the Train litigation and the later enforcement statutes kept in one ordered place, can keep your statute notes, citations, and case chronologies together free on VaultBook. The statute rewards that kind of ordered reading, because the meaning of each creation appears in the operation of the whole: committees without estimates would deliberate in the dark, a resolution without allocations would state an aspiration, allocations without scorekeeping would fade during a session, and impoundment control without a clock would return cancellation to the silence that the 1974 settlement replaced with a deadline and a default in favor of enacted law.

Frequently Asked Questions

Q: What did the Congressional Budget Act create?

The Congressional Budget Act created the institutional machinery of congressional budgeting. Title I established a Committee on the Budget in the House of Representatives and in the Senate. Title II created the Congressional Budget Office, an estimating and scorekeeping office serving Congress. Title III created the congressional budget process itself, including a statutory timetable, the concurrent budget resolution stating totals for revenue, budget authority, outlays, the deficit or surplus and the public debt, allocations of those totals to committees, and reconciliation for bringing mandatory spending and revenue law into line with adopted totals. Title V changed the federal fiscal year to an October start, and Title X, the Impoundment Control Act, created procedures governing presidential withholding of enacted funds.

Q: Which president signed the Congressional Budget Act?

President Richard Nixon signed the Congressional Budget Act on July 12, 1974, and the measure became Public Law 93-344, printed at 88 Stat. 297. The signature carried a tension that the legislative history records plainly. Presidential withholding of appropriated funds under the Nixon administration had supplied much of the impetus for the legislation, and the statute restricted that practice, required congressional approval for permanent cancellations of budget authority, and equipped Congress with committees and an estimating office designed to end reliance on executive figures. A veto would have faced a House that agreed to the conference report by 401 to 6 and a Senate that agreed by 75 to 0. Nixon announced resignation on August 8, 1974, four weeks after signature.

Q: Why did impoundment lead to the Congressional Budget Act?

Impoundment led to the Congressional Budget Act because presidential withholding under President Richard Nixon converted a study question about budget procedure into a live dispute over enacted law. The administration withheld funds across domestic programs, capped allotments under the Federal Water Pollution Control Act Amendments of 1972 at half or less of the authorized schedule after Congress enacted that statute over a veto, and asserted at a press conference on January 31, 1973 that a constitutional right to impound was absolutely clear. Members of both parties concluded that a president able to decline spending after enactment held a practical second veto, exercisable without the two thirds safeguard that governed a returned bill. Reformers therefore joined impoundment control, in Title X, to budget process reform in a single statute.

Q: Is a budget resolution a law under the Congressional Budget Act?

A budget resolution is not a law under the Congressional Budget Act. The resolution takes the form of a concurrent resolution, adopted by the House of Representatives and the Senate in identical form and never presented to the president for signature or veto. Because no statute results, the resolution cannot amend existing law, create new law, appropriate money or compel any action outside the two chambers. Legal force over agencies, programs, taxes and borrowing remains in separate legislation enacted through the ordinary lawmaking process. Inside Congress, the resolution carries real force of a different kind. Adopted totals are allocated to committees, and points of order inside each chamber block consideration of bills or amendments that would breach an allocation or an aggregate, unless the chamber waives the objection.

Q: Did the Congressional Budget Act change the fiscal year?

The Congressional Budget Act changed the fiscal year directly. Section 501, at 88 Stat. 321, amended the Revised Statutes so that the fiscal year ran from July 1 through June 30 through June 30, 1976, and beginning October 1, 1976 ran from October 1 through September 30. The definition appears at 31 U.S.C. 1102. The change gave Congress three additional months between submission of a presidential budget and the start of the fiscal year, after a record in which Congress, during the twenty five fiscal years from fiscal 1952 to fiscal 1976, had not once enacted all regular appropriations acts by the beginning of the fiscal year under the July calendar. A single transition quarter bridged the two calendars in 1976.

Q: How often does Congress finish appropriations on time under the Congressional Budget Act?

Congress finished appropriations on time rarely under the Congressional Budget Act. The Congressional Research Service identified only four fiscal years in the span from fiscal 1977 through fiscal 2019 in which all regular appropriations acts were enacted by the beginning of the fiscal year: fiscal 1977, fiscal 1989, fiscal 1995 and fiscal 1997. Fiscal 1977 was the first fiscal year under the October calendar and the first cycle in which the budget procedures of the act were implemented fully, and even that year required continuing resolutions for certain unauthorized programs dropped from the regular acts. In every other fiscal year of the span, at least one regular act remained unenacted on October 1, and funding depended on continuing resolutions or later consolidated packages. No fiscal year after fiscal 1997 joined the on time list through the close of 2019.

Q: What is a rescission under the Congressional Budget Act?

A rescission under the Congressional Budget Act is a presidential proposal to cancel budget authority permanently, governed by section 1012 of Title X and codified at 2 U.S.C. 683. The president transmits a special message identifying the amount proposed for cancellation, the account and establishment involved, the reasons for the proposal and the estimated fiscal, economic and budgetary effects. The Comptroller General reviews the message and reports to both chambers. While Congress considers the proposal, the president may withhold the identified authority from obligation, but the withholding window is fixed. Unless Congress enacts a rescission bill approving the proposal within forty five days of continuous session, the authority must be released and made available for obligation for the purposes Congress enacted.

Q: What is the public law number of the Congressional Budget Act?

The public law number of the Congressional Budget Act is Public Law 93-344. The formal title of the statute is the Congressional Budget and Impoundment Control Act of 1974, and the text is printed at 88 Stat. 297. Congress enacted the measure as H.R. 7130 in the 93rd Congress, after the House passed the bill on December 5, 1973 by 386 to 23, the Senate passed a version on March 22, 1974 by 80 to 0, the House agreed to the conference report on June 18, 1974 by 401 to 6, and the Senate agreed on June 21, 1974 by 75 to 0. President Richard Nixon signed the bill on July 12, 1974. The statute is codified principally at 2 U.S.C. sections 601 and following.

Q: What do the budget committees do under the Congressional Budget Act?

The budget committees draft and report the concurrent budget resolution for each chamber, stating totals for revenue, budget authority, outlays, the deficit or surplus and the public debt, divided by major functional category. After adoption of a resolution, the committees prepare the allocation of the totals among the committees of the chamber under section 302(a), ordinarily presented in the joint explanatory statement accompanying the conference report. The committees receive scorekeeping reports from the Congressional Budget Office tracking legislation against the resolution, assemble committee responses into an omnibus reconciliation bill without substantive revision when a resolution contains reconciliation instructions, and defend the allocations through points of order on the floor. Title I created one committee for the House and one for the Senate, with the first chairs held by Brock Adams in the House and Edmund Muskie in the Senate.

Q: How is the director of the Congressional Budget Office chosen?

The director of the Congressional Budget Office is appointed by the Speaker of the House and the President pro tempore of the Senate, after considering recommendations received from the Committee on the Budget of the House and the Committee on the Budget of the Senate. Section 201 directs that the appointment be made without regard to political affiliation and solely on the basis of fitness to perform the duties of the office. The term is four years, expiring on January 3 of the year preceding each presidential election, and a director may remain in place after expiration until a successor is appointed. Removal lies with either House acting by resolution, so accountability runs to the legislature that created the office and to no executive officer. Alice M. Rivlin, the first director, took office in February 1975 and served to 1983.

Q: What is an allocation under the Congressional Budget Act?

An allocation under the Congressional Budget Act is an assignment of a share of the totals in an adopted budget resolution to a committee of a chamber, made under section 302(a). The allocation states the new budget authority, entitlement authority and outlays available to that committee for the first fiscal year covered by the resolution and for the total of the years covered, so that each committee can test a reported bill against a figure rather than against an aggregate that no single committee controls. The Appropriations Committee receives a single committee total for discretionary spending and must then subdivide that total among appropriations subcommittees under section 302(b). Section 302(f) bars consideration of a measure or amendment that would cause an allocation or a subdivision to be exceeded, and Congressional Budget Office scorekeeping tracks action against each figure.

Q: What is a deferral under the Congressional Budget Act?

A deferral under the Congressional Budget Act is a temporary withholding of budget authority from obligation, governed by section 1013 of Title X and codified at 2 U.S.C. 684. The president transmits a special message stating the amount, the account, the period and the reasons, and transmits the message to the Comptroller General for review. As enacted in 1974, a deferral took effect unless either House adopted an impoundment resolution disapproving the proposal. The Supreme Court decision in Immigration and Naturalization Service v. Chadha in 1983 destroyed the one house veto mechanism, and Congress rewrote section 1013 in 1987. After that amendment, deferrals were permissible only to provide for contingencies, to achieve savings made possible by changes in requirements or greater efficiency of operations, or as specifically provided by law, and a deferral could not extend beyond the end of the fiscal year of the proposal.

Q: What does the Comptroller General do under the Congressional Budget Act?

The Comptroller General reviews every presidential special message proposing a rescission or a deferral under Title X and reports to both chambers on the facts surrounding the proposed action and the probable effects, including whether a deferral conforms to existing statutory authority. Under section 1015, if the Comptroller General finds that executive action, or failure to act, amounts to a rescission or a deferral that the president failed to report, the Comptroller General reports the impoundment to Congress, and the report carries the same effect as a presidential special message by starting the statutory clock. The Comptroller General also classifies disputed withholdings that the executive placed in the wrong category. Under section 1016, after filing an explanatory statement and waiting twenty five calendar days, the Comptroller General may bring a civil action to compel release of budget authority withheld in violation of Title X.

Q: Why did the original act require two budget resolutions each year?

The original act required two budget resolutions each year because the design of 1974 separated an early statement of targets from a late statement of binding limits. Under the timetable described by the Congressional Research Service, Congress was scheduled to complete action on a first resolution by May 15, setting advisory targets for revenue, budget authority, outlays, the deficit or surplus and the public debt, and to adopt a second resolution by September 15, shortly before the fiscal year began on October 1, setting binding figures. Reconciliation belonged naturally to the second stage, as a device for adjusting pending legislation and existing law so that enacted policy conformed to the binding totals after appropriations action was largely complete. The Balanced Budget and Emergency Deficit Control Act of 1985 eliminated the second resolution requirement and consolidated the process into a single resolution due April 15.

Q: When was the first budget resolution adopted under the act?

The first budget resolution under the act covered fiscal 1976 and was adopted during 1975, the first cycle in which the congressional budget process was implemented. The Congressional Research Service recorded that the process was first implemented in 1975 for fiscal 1976, with full implementation of the procedures following for fiscal 1977, the first fiscal year under the October calendar created by Title V. The fiscal 1976 cycle operated under the original two resolution design, with a first resolution setting targets in the spring of 1975 and a second, revising resolution later in 1975 carrying reconciliation directives for that fiscal year. Adoption in 1975 placed the first congressional statement of aggregate totals a full cycle before the new fiscal year calendar took effect on October 1, 1976.

Q: What was the transition quarter in the Congressional Budget Act?

The transition quarter was the single three month period that bridged the old and new fiscal calendars under Title V of the act. The old fiscal year ended on June 30, 1976, and the first fiscal year under the October calendar began on October 1, 1976, leaving July, August and September of 1976 belonging to neither fiscal year. The act treated that interval, running from July 1, 1976 to September 30, 1976, as a distinct period with separate budget estimates and separate appropriations treatment, so that federal activity during those months rested on enacted authority rather than on an assumption that funding from the expired fiscal year carried forward. The device was used once and then exhausted, and the fiscal 1976 budget resolution and the fiscal 1977 resolution each addressed the quarter as part of the calendar change.

Q: What happens when Congress misses the budget resolution deadline?

Missing the budget resolution deadline carries no automatic sanction under the Congressional Budget Act, and the consequences are procedural rather than penal. The deadline, April 15 after the 1985 consolidation of the timetable, is a target in section 300 rather than a condition that voids later action. A late resolution leaves appropriations subcommittees drafting bills during the spring without an adopted allocation, and interim subdivisions adopted without a resolution lack enforceable character. When the House and Senate fail to agree on a resolution at all, as occurred in ten fiscal years through fiscal 2019, with no resolution adopted for fiscal 2020 by the close of 2019, reconciliation cannot be triggered for both chambers, and each chamber legislates under deemed levels supplied by a simple resolution, a special rule or a statute. The Congressional Research Service recorded that Congress met the adoption target only six times for fiscal years 1976 through 2015.

Q: Where is the Congressional Budget Act codified in the United States Code?

The Congressional Budget Act is codified principally at 2 U.S.C. sections 601 and following, in Title 2 of the United States Code, the title covering Congress. The Congressional Budget Office provisions in Title II of the statute occupy sections 601 and following, the congressional budget process provisions in Title III include the timetable at section 300 of the act and allocations and reconciliation in the same codified range, and the Impoundment Control Act in Title X is codified at 2 U.S.C. sections 681 to 688. The fiscal year definition created by section 501 of the act sits apart from that range, at 31 U.S.C. 1102, in the title covering money and finance. Codification collects the enacted text into the Code arrangement, while the session law text remains Public Law 93-344 at 88 Stat. 297.

Q: Does the Congressional Budget Act itself spend money?

The Congressional Budget Act does not itself spend money. The statute created committees, an estimating office, a timetable, a resolution, allocations, points of order and impoundment procedures, and none of those creations supplied budget authority for a department, a program or a grant. A budget resolution adopted under the act stated totals and divided the totals among committees, but the resolution was a concurrent resolution rather than a law, and no payment by the Treasury could rest on the resolution alone. Spending occurred when separate appropriations acts and other spending legislation, enacted through the ordinary lawmaking process, provided budget authority for stated purposes and agencies obligated that authority under the enacted terms. The act governed the conditions under which spending decisions were made, measured and policed, and governed executive withholding of amounts after enactment, without functioning as an appropriation.

Q: How was the Congressional Budget Act changed by later laws?

Later laws changed the Congressional Budget Act framework in three principal waves, stated here at orientation depth. The Balanced Budget and Emergency Deficit Control Act of 1985 consolidated the two resolution timetable into a single budget resolution due April 15 and attached sequestration, an automatic cancellation of budgetary resources, to fixed deficit targets. The Budget Enforcement Act of 1990 replaced fixed deficit targets with statutory caps on discretionary spending and pay-as-you-go discipline for legislation affecting mandatory spending and revenues, under which a net deficit increase on the scorecard at the end of a session produced a sequester of covered mandatory programs. The Budget Control Act of 2011 built a further regime of discretionary caps extending across the following decade and paired the caps with an additional sequestration mechanism. Each wave operated through the estimates, allocations and scorekeeping machinery of the 1974 act.