Unaccountable: Exploring the Lack of Budgetary Transparency for U.S. Nuclear Security Spending

Unaccountable: Exploring the Lack of Budgetary Transparency for U.S. Nuclear Security Spending

Want to dive deeper?

Visit the Education Center

Stephen I. Schwartz

The James Martin Center for Nonproliferation Studies

With U.S. federal government spending, including defense spending, now expected to decline sharply over the next decade—even as the Obama administration has pledged to invest more than $210 billion to upgrade the U.S. nuclear arsenal and its supporting infrastructure—it becomes increasingly important to know where nuclear security dollars—money spent on nuclear weapons and weapons-related programs (such as cooperative threat reduction)—are going on both an annual and a cumulative basis. [1]

Yet today, as has been the case for more than six decades, no one in the federal government has any clear idea how much is being spent on nuclear weapons and weapons-related programs. This is especially true for programs managed by the Department of Defense (DOD), where most nuclear security dollars are spent. This situation exists because there is no comprehensive budget or budgetary line item that tracks all these programs and costs within and across government organizations and over time. Ensuring accountability and transparency for nuclear security spending has never been a priority, and some program and congressional officials may have actually preferred to keep their colleagues in the dark about especially large or controversial expenditures.

But as cost cutting becomes the modus operandi of the federal budgeting process, it becomes increasingly important to know where the money is going, and why. Failure to understand the full measure of costs associated with nuclear weapons and weapons-related program will contribute, as it has in the past, to poor budgetary and programmatic decisions and wasteful or misguided spending, affecting not just the nuclear arsenal but a variety of nonproliferation and arms control programs that contribute to U.S. and global security.

Why Don’t We Know What We Spend? [2]
For the more than 69 years that the U.S. nuclear weapons program has existed, there has never been an official comprehensive accounting of its expenditures, either on an annual or a cumulative basis. Nor has Congress demanded such an accounting as a prerequisite for future funding.

When the Atomic Energy Commission (AEC) sought to present its first budget to Congress in 1947 (after assuming control of the nuclear weapons program from the Manhattan Engineer District run by the U.S. Army Corps of Engineers), chairman David Lilienthal lamented his inability to justify the request because “we did not have a set of books showing costs, since the Army’s Manhattan District didn’t have or keep any.” When the situation had not improved three years later, Representative Francis Case (Republican of South Dakota) admonished the AEC: “No other agency of the Government has been able to get by the Appropriations Committee with the lax presentation of detailed estimates that the [AEC] has. No other agency of the Government, so far as I know, has been able to come up and cloak itself with the aura of a scientific subject and get by with such general justifications.”[3]

In 1952, as a massive expansion program was underway, AEC Commissioner T. Keith Glennan warned his colleagues that the Commission was then spending more than $100 million a month (about $940 million today) yet, “I am continuously at sea as to the status of our budgets, of our expenditures and particularly of any real basis for judging the quality of the financial performance of our various offices.” [4]

In 1950, Senator Brien McMahon (Democrat of Connecticut), chairman of the powerful Joint Committee on Atomic Energy (JCAE) and a powerful proponent of an accelerated nuclear weapons program, grew concerned that insufficient funding was being allocated to nuclear armaments, despite his assertion that the “cost of military fire power based upon atomic bombs is hundreds of times cheaper, dollar for dollar, than conventional explosives.” A simple analysis by the JCAE staff reported that just three cents out of each military spending dollar was being expended on atomic bombs (a figure derived by examining the AEC’s budgets alone and excluding any of the significantly larger expenditures by the military itself, which were then consuming one-quarter to one-third of the entire DOD budget). [5]

In September 1951, McMahon introduced a resolution calling on the United States to “go all-out in atomic development and production.” [6] (Interestingly, one day after this speech, on September 19, 1951, McMahon finally permitted the AEC to brief him for the first time on the size and disposition of the U.S. nuclear stockpile.) McMahon’s efforts alarmed some military officials. As an official Air Force history of the period records, the resolution “focused attention on the fact that there existed no itemized record of the military expenditures, either direct or indirect, for the atomic energy program” (emphasis added). The Air Force, at the request of the DOD controller, began to compile data to disprove McMahon but: “At once it became apparent that it was difficult to draw sharp lines between those activities which pertain to the atomic energy program and those which were outside its limits. This lack of clarity was one of several factors which made it impossible to arrive at exact sums either spent or to be spent for atomic energy.” [7] Although the secretary of defense subsequently directed the services to show nuclear costs itemized and broken out in its fiscal 1954 budget, Air Force officials continued to find the exercise “inherently impossible” given the overlap between programs, the scope of the effort, and the fact that nuclear weapons were not a discrete budgetary category. In the end, after providing sufficient evidence to the JCAE that the military was hardly shortchanging its atomic endeavors, the effort to provide an accurate accounting was discontinued.

Although most in Congress were relatively complacent about the high costs of nuclear weapons (not least because those costs were never fully presented to or understood by them), some members occasionally voiced concern and frustration. On May 24, 1957, Representative Errett P. Scrivner (Republican of Kansas), a leading member of the House Appropriations Committee, called for some explanation—either by the Defense Department or “perhaps some outsiders”—of the current and long-term costs of nuclear weapons:

"One plane today can carry more potential death and destruction by 1 drop in 1 bomb than all of the planes carried in all of the sorties during all of World War II in both the Pacific and the Atlantic. . . . Somebody ought to tell us how much it is going to cost to deliver a megaton bomb, whether it is an atom bomb or a hydrogen bomb, by missile, by bomber, by carrier, by submarine. Somewhere, some place, there ought to be an answer as to which is the best and the most economical method. You know and I know that all of our services cannot be kept perpetually at peak effort to fight an atomic war all the time, one which may never come. . . . When you see some of these figures for a 10-year program I would expect your hair to sizzle." [8]

In February 1960, recently retired army chief of staff, General Maxwell D. Taylor, testified before a Senate subcommittee that the DOD made little effort to assess its overall needs and that both its planning and budgeting were haphazard: “We never look at our forces; we never build our forces in a budget sense in terms of military functions such as atomic retaliation, limited war capability, antisubmarine warfare, continental air defense. We don’t case our books in that form. So as a result, I never know, and I doubt personally that anyone knows, exactly what we are buying with our budget.” [9]

This began to change in 1961, when Secretary of Defense Robert McNamara introduced the Five-Year (now Future Years) Defense Program and assigned unique program element numbers to every major program in the DOD budget. [10] Building on a system that previously grouped programs by service, McNamara created a set of eleven Major Force Programs (MFPs). For the first time, officials were able to make accurate, long-range forecasts of budgetary expenses and assess historical trends. Still, no provision was made to account for nuclear weapons as a separate category. The closest approximation was MFP 1, Strategic Forces (see Figure 1 [24]). But MFP 1 does not include a number of relevant line items: most overhead and support costs; most research and development costs for delivery systems and supporting equipment; airlift and sealift costs for strategic and tactical nuclear weapons programs; most centralized command, control, and communications programs associated with nuclear weapons; all intelligence programs supporting nuclear weapons; tactical nuclear weapons; some training; nuclear-related civil defense efforts; and the costs of monitoring, verifying, and complying with various nuclear arms control and reduction agreements. When MFP 1 was created in the early 1960s, strategic bombers, intercontinental ballistic missiles, and ballistic missile submarines were exclusively nuclear delivery systems. So creating a special budgetary category just for them made sense. But over the years, things have changed. Today, there are 113 B-2 and B-52H bombers, of which 18 and 76, respectively, are nuclear capable. But only 60 total bombers are believed to be nuclear tasked.[11] Nevertheless, the costs for all 113 aircraft are counted under MFP 1. Similarly, there are 18 Ohio-class Trident submarines, but between 2002-08, four were converted to carry non-nuclear Tomahawk land attack cruise missiles and support Navy SEAL missions.[12] Yet the costs of those four converted submarines are almost certainly included in MFP 1 (though the weapons they carry may not be). In short, MFP 1 both overstates the costs of the US nuclear arsenal (by including systems that no longer perform a nuclear weapons mission) and understates those costs (by failing to include a host of essential support programs, as well as the costs of tactical nuclear weapons).

On May 3, 1962, President John F. Kennedy, during a meeting on nuclear weapons requirements, “stressed the extreme importance of holding down expenditures and requests for additional appropriations” even as a large nuclear buildup was well underway, and called on the Bureau of the Budget (now the Office of Management and Budget), in cooperation with the Secretary of Defense, to develop a procedure for compiling “a statement of the costs of nuclear weapons provided for the national defense, including both the cost of delivery systems provided in the Defense budget and costs of weapons funded in the budget of the Atomic Energy Commission.” [13] Despite a direct presidential request, no such procedure was ever developed. Although it is possible that successive presidents and senior government officials made similar requests, this is the last documented instance of such an effort.

Why Does Knowing What We Spend Matter?
There are many reasons why the U.S. nuclear weapons program has consumed at least $8.7 trillion (in adjusted FY 2010 dollars) since its inception, and why it grew as large as it did as quickly as it did, from a handful of hand-built weapons in 1945 to 31,255 advanced warheads and bombs 22 years later (or, to put it another way, from a total explosive yield of 40 kilotons in 1945 to a peak of nearly 20,500 megatons in 1960, equivalent to 1.7 million Hiroshima-sized bombs). [14] These include: perceptions of the Soviet and Chinese threat, the indeterminate nature of deterrence, redundant targeting, technological advances and obsolescence, interservice and inter-laboratory rivalry, nuclear weapons as “free goods” for the military services (because the cost of warheads and bombs was borne by the AEC, not the military), pork barrel politics, corporate lobbying, secrecy, political considerations, and arbitrary decision making, to name a few. The profound lack of financial accountability and transparency from the very beginning was a critical factor.

The entire rationale behind Sen. McMahon’s call for an “all out” nuclear weapons program in 1951—as well as President Dwight D. Eisenhower’s subsequent “New Look” approach—was that nuclear weapons were believed to be significantly less expensive than conventional ones. McMahon, Eisenhower, and others strongly believed that the United States could not compete with the Soviet Union on the conventional level without risking a national economic collapse. In a 1951 floor speech, McMahon said: “Money spent upon the atomic bomb could pulverize a dozen enemy war plants at no more expense than destroying a single plant with TNT, to say nothing of the fact that one plane can deliver one A-bomb as against the huge armadas needed to deliver an equivalent cargo of blockbusters. . . . If we mass-produce this weapon, as we can, I solemnly say to the Senate that the cost of a single atomic bomb will become less than the cost of a single tank.” According to McMahon, relying on nuclear weapons would be economically advantageous because “in all logic and common sense, an atomic army and an atomic navy and an atomic air force ought to mean fewer men under arms. They ought to mean a major reduction in the tens of billions of dollars we would otherwise spend upon stacks and stacks of conventional armaments. They ought to mean a sloughing off of outmoded operations and outdated expenses.” [15]

The Army conducted two-sided war games during the winter of 1952-53 and discovered, in the words of the commander of the Seventh Army Corps General James M. Gavin, that “more rather than less manpower would be required to fight a nuclear war successfully” (findings that were confirmed in subsequent war games in 1955 and 1958). [16] Yet even after this, the nuclear buildup based on the assumption of “a bigger bang for a buck” continued unabated. The lack of readily available data to validate that assumption, along with a surprising lack of inquisitiveness on the part of government and military leaders across the board, enabled this approach to continue far longer than might otherwise have been the case. Far from being inexpensive, nuclear weapons were responsible for 29 percent of all military spending and almost 11 percent of all government expenditures from 1940 to 1996. Nuclear weapons were the third largest government program, after all non-nuclear national defense and social security spending, and just ahead of income security, or welfare spending. [17]

Official confusion about the full costs of nuclear weapons is not just ancient history. For several years in the mid-to-late 1990s, the DOD included two charts in the annual report of the secretary of defense purporting to show total expenditures on strategic nuclear weapons, over time and as a percentage of the entire DOD budget. But these charts were highly misleading, in that they only counted the readily available but incomplete costs found in MFP 1. The charts in the 1999 report (see Figure 2 [25]) showed total strategic costs falling from $18 billion in 1990 to less than $6 billion in 1999, while their percentage decreased from a little over six percent to a little more than two percent over the same period of time (in reality, the costs and percentage were at least three times higher).

This matters today because the number of nuclear weapons the United States still has, and assumptions about them, are influenced by what has come before. Although the total size of the nuclear stockpile has declined significantly over the last 20 years, and while nuclear weapons now account for some seven percent of the DOD budget (versus 67 percent of the DOE budget) [18], until recently most observers expected spending on both delivery systems and warheads to increase over the next decade by more than $210 billion as a consequence of the ratification of the New START agreement. But the new financial outlook brought on by the state of the economy and the Tea Party’s insistence on cutting deeply into government spending to reduce the federal budget deficit has upended budgetary assumptions across the board. Among the more radical approaches is Sen. Tom Coburn’s (Republican of Oklahoma) proposal to slice $79 billion out of nuclear weapons delivery systems over 10 years. [19] In June 2011, the House Appropriations Committee passed a funding bill cutting $100 million from the budget for a new plutonium processing plant at Los Alamos, and zeroing out funding for an associated waste facility, after raising concerns about their necessity and safety (the bill was subsequently passed by the House of Representatives in July). [20] And in September 2011, the Senate Energy and Water Development appropriations subcommittee cut the Obama administration’s requests for nuclear weapons by more than $400 million and nonproliferation and threat reduction programs by $167 million (the amounts approved, however, still mark an increase over fiscal 2011 funding). [21]

Budgetary transparency and accountability is a particular problem for the DOD. At least 65 percent of all U.S. nuclear weapons and weapons-related spending ($34 billion) flows through DOD. [22] Not only are there few good numbers available from DOD about what it spends on nuclear weapons and related programs, but what numbers there are can be inconsistent and contradictory. An inquiry about what it costs, for example, to operate the ballistic missile submarine fleet, will likely produce different answers from the fleet’s operating bases, the commander of submarine forces, the Chief of Naval Operations, U.S. Strategic Command, and the DOD.

Financial ignorance impedes accountability and facilitates waste. The long-term viability of nuclear weapons and weapons-related programs may be jeopardized if this state of affairs is allowed to continue. In a fiscally constrained world, Congress, and the taxpayers it represents, should know what it is paying for.

This applies to threat reduction and incident preparedness programs as well, which have lower costs but provide significant benefits. In its drive to cut federal spending, Congress may inadvertently scale back or eliminate highly effective programs in these areas. Program managers and supporters need to be able to demonstrate the value of these programs to skeptical lawmakers, but that is difficult to do if one-half of the equation, the costs, is not readily available or well understood.

Significantly, the 1999 Commission to Assess the Organization of the Federal Government to Combat the Proliferation of Weapons of Mass Destruction, headed by former director of central intelligence, John Deutch, noted, “There is no system for tracking resource expenditures for combating proliferation. Doing so is essential to an effective interagency effort.” Consequently:

"No one in the Federal Government knows how much money we are spending to combat proliferation. The success of any campaign depends on the resources available to wage it, and on the ways in which those resources are brought to bear. Currently, however, no one decides what level of resources should be devoted to proliferation-related efforts, there is no overall plan for how those resources should be allocated and no consistent evaluation of the effectiveness of these expenditures. [23]"

If adopted, said the commission, such a system would result in a “more transparent process for tracking the application of resources to their intended purposes.” Twelve years later, no such system has been implemented.

A Solution
If Congress (and the interested public) had a clear understanding of what it costs to sustain the nuclear arsenal, or of, for example, the annual expenditures required to secure vulnerable nuclear materials in the United States and overseas, it would be possible to have an informed discussion about the costs and benefits of these programs. Because such an understanding does not exist, rhetoric and assumptions tend to substitute for facts when it comes to making important decisions about the future of U.S. nuclear security spending.

To rectify this decades-long problem, the following proposal offers a way to create financial transparency and accountability in the nuclear weapons program. Proponents of increased nuclear weapons spending could support such a measure, not just because it is good government but also because it will help them monitor spending and support their desire to invest heavily in sustaining the nuclear arsenal and its associated support infrastructure. Similarly, critics of nuclear weapons, and fiscal conservatives, would be better able to identify waste and demonstrate how reductions in the size of the arsenal, or changes in how weapons are deployed, might help reduce costs. Under such a system, both advocates and critics of various nonproliferation and arms control measures would have new, consistent metrics to gauge the costs of such programs relative to their actual benefits.

Congress could pass a law requiring the executive branch to prepare and submit each year, with the annual budget request, an unclassified and classified accounting of all nuclear weapons and weapons–related spending for the previous fiscal year, the current fiscal year, and the next fiscal year (this could also be done by the president via an executive order, although that would not require or encourage congressional support and would not have the permanence of a law). As it already does, the DOD would project its nuclear weapons–related spending five or six years into the future, only in much greater detail. A senior White House official would oversee this annual exercise, in coordination with relevant officials of the Office of Management and Budget and senior budget officials of key departments and agencies, ensuring that the bureaucracy fully complied with this new requirement.

After an initial draft accounting was prepared, Congress would direct the Government Accountability Office and the Congressional Budget Office, both of whom have significant expertise in this area, to audit the nuclear budget for accuracy and completeness, and then ensure that the executive branch incorporates its recommendations for future years. To the maximum extent possible, this accounting should be unclassified, like the 2010 Nuclear Posture Review. There is little if any need to keep secret the specific amounts of money being expended for nuclear weapons and weapons-related programs, particularly after the Obama administration revealed for the first time, in May 2010, the total number of nuclear weapons in the arsenal. In light of such changes, the ongoing cost of these programs should not remain classified, or at least largely inaccessible, even to those inside the government charged with overseeing them.

Once a framework for identifying and allocating costs was agreed upon (and this would take some time, especially for programs that cross over areas of responsibility), new numbers could be added each year. These new data should lead to more insightful debates and informed decisions. At the very least, government officials, and the people they represent and work for, would have a significantly better appreciation of the actual costs of nuclear security programs. True cost-benefit analyses would become possible, and the risk of making important programmatic decisions based on misleading (or non-existent) budgetary data would decline. Officials would also be able to accurately forecast savings from projected changes in the deployment of nuclear weapons, something not possible at present.

[1] Stephen I. Schwartz, “New START and the Maintenance and Modernization of U.S. Nuclear Weapons,” CNS Feature Story, 22 December 2010,
[2] This section draws upon Stephen I. Schwartz, “Introduction,” in Stephen I. Schwartz, ed., Atomic Audit: The Costs and Consequences of U.S. Nuclear Weapons Since 1940, (Washington, DC: Brookings Institution Press, 1998), pp. 4-12.
[3] Lilienthal and Case are quoted in Harold P. Green and Alan Rosenthal, Government of the Atom: The Integration of Powers (New York: Atherton Press, 1963), p. 225.
[4] Memorandum, T. Keith Glennan to Chairman Gordon Dean and others, 9 May 1952, Record Group 324, Records of the Atomic Energy Commission, Office of Secretary, General Correspondence, box 73, File: Organization and Management 8, Progress Reports by Division, vol. 1, National Archives.
[5] Stephen I. Schwartz, “Introduction,” in Schwartz, ed., Atomic Audit: The Costs and Consequences of U.S. Nuclear Weapons Since 1940, (Washington, DC: Brookings Institution Press, 1998), p. 8.
[6] S. Conc. Res. 46, 82 Cong. 1 sess., 18 September 1951.
[7] Lee Bowen, Robert D. Little, and others, A History of the Air Force Atomic Energy Program, 1943–1953, vol. 3: Building an Atomic Air Force, 1949–1953 (Washington, DC: U.S. Air Force Historical Division, 1959), pp. 471–472 (formerly top secret; declassified in 1980).
[8] Congressional Record, 85 Cong. 1 sess., vol. 103, pt. 6 (24 May 1957), pp. H7619–20.
[9] Missiles, Space, and Other Major Defense Matters, Hearings before the Preparedness Investigating Subcommittee of the Senate Committee on Armed Services in Conjunction with the Committee on Aeronautical and Space Sciences, 86 Cong. 2 sess., (Government Printing Office, 1960), p. 199.
[10] For an overview of the FYDP see, Department of Defense, “Future Years Defense Program (FYDP) Structure,” DOD 7045.7-H, April 2004,
[11] Hans M. Kristensen and Robert S. Norris, "US Nuclear Forces, 2011," Bulletin of the Atomic Scientists, 67 (March/April 2011), pp. 66-76,
[12] Ronald O'Rourke, "Navy Trident Submarine Conversion (SSGN) Program: Background and Issues for Congress," Congressional Research Service, RS21007, updated May 22, 2008,
[13] Memorandum, Charles E. Johnson for the record, “President’s Decisions at the Meeting on Nuclear Weapons Requirements on May 3, 1962” (draft), 4 May 1962 (formerly top secret/restricted data), National Security Archive, p. 1. Officials present at this meeting included national security adviser McGeorge Bundy, military representative General Maxwell D. Taylor, Deputy Secretary of Defense Roswell L. Gilpatric, and AEC chairman Glenn T. Seaborg.
[14] See Department of Defense, “Fact Sheet: Increasing Transparency in the U.S. Nuclear Weapons Stockpile,” May 3, 2010,; Department of Energy, "Declassification of Certain Characteristics of the United States Nuclear Weapon Stockpile,” June 27, 1994,
[15] Congressional Record, vol. 97, pt. 9 (September 18, 1951), p. S11496-509.
[16] James M. Gavin, War and Peace in the Space Age (New York: Harper & Brothers, 1958), p. 139. Robert Endicott Osgood, NATO: The Entangling Alliance (University of Chicago Press, 1962), pp. 118-19.
[17] See Stephen I. Schwartz, “The Costs of U.S. Nuclear Weapons,” Nuclear Threat Initiative Issue Brief, October 2008,
[18] Stephen I. Schwartz with Deepti Choubey, Nuclear Security Spending: Assessing Costs, Examining Priorities, Carnegie Endowment for International Peace, 2009,
[19] Senator Tom Coburn, "Back in Black: A Deficit Reduction Plan," July 18. 2011, pp. 121-22,
[20] Kingston Reif, Center for Arms Control and Non-Proliferation, “Summary of the House Appropriations Committee Version of the FY 2012 Energy and Water Appropriations Bill,” June 16, 2011,
[21] Kingston Reif, Nukes of Hazard, “Summary of the Senate Energy and Water Appropriations Subcommittee Markup: Good News for Nonpro,” September 6, 2011,
[22] Schwartz with Choubey, Nuclear Security Spending.
[23] Commission to Assess the Organization of the Federal Government to Combat the Proliferation of Weapons of Mass Destruction, “Combating the Proliferation of Weapons of Mass Destruction,” July 14, 1999, pp. vi, 19,
[24]  Department of Defense, Office of the Under Secretary of Defense (Comptroller), National Defense Budget Estimates for FY 2012, March 2011, p. 94,
[25] William S. Cohen, Annual Report to the President and the Congress, 1999,

Stay Informed

Sign up for our newsletter to get the latest on nuclear and biological threats.

Sign Up

The 2023 NTI Nuclear Security Index


The 2023 NTI Nuclear Security Index

“The bottom line is that the countries and areas with the greatest responsibility for protecting the world from a catastrophic act of nuclear terrorism are derelict in their duty,” the 2023 NTI Index reports.


My Resources