The Cost of Intelligence


Viscerally, in the wake of the Cold War, many Americans believe the costs of intelligence should go down. Indeed, since 1989, the resources allocated to intelligence have gone down-by about 21 percent in real terms-but, nonetheless, they remain substantial. Budget projections show spending for intelligence holding relatively constant in real terms through the rest of the decade. Reflecting a sense of unease, Congress asked this Commission to determine "- whether the existing levels of resources allocated for intelligence collection and intelligence analysis are seriously at variance with United States needs. -"

To answer this question, the Commission undertook an extensive review of the intelligence budget and analyzed the changes to that budget since 1980. In doing so, the Commission attempted to ascertain what basis, if any, had been used over time to arrive at the resource level for intelligence. Was there a discernable standard or criteria that might help those responsible for resource allocations in the future to determine how much intelligence is enough?

Recognizing that pressure to reduce spending is apt to continue, the Commission attempted to assess whether and how the costs of the existing intelligence capability could be reduced without damaging the nation's security.

The Recent History of Intelligence Funding, in Brief

In recent decades, intelligence funding has been treated preferentially when compared to other parts of the Defense budget. As Figure 1 illustrates, non-intelligence defense funding in real (constant dollar) terms grew by 40 percent from 1980 to 1986, leveled off, and then declined to its current level, four percent below its 1980 level. In contrast, total intelligence funding grew by 125 percent in real (constant dollar) terms from 1980 to 1989 but declined thereafter to its current level of 80 percent above 1980. Reductions taken in the intelligence budget since 1989 have been at a rate to allow the intelligence agencies to continue most of their basic activities. Each agency has taken its share of the reduction from 1989, but no major structural change was required.

Graphic: Figure 13-1 (80k)

Graphic: Figure 13-2 (87k)

Personnel strength at NSA, CIA, and DIA has also remained significantly above the 1980 level despite across- the-board reductions over the last four years. NSA is 22 percent above its 1980 level; CIA, 8 percent; and DIA, 80 percent, primarily because DIA assumed major new functions which involved the transfer of additional personnel. In 1991, Congress in concert with the Bush Administration imposed a 17.5 percent across-the-board reduction in intelligence personnel to be accomplished between 1991 and 1997. This is less than 3 percent per year and has already largely been accomplished through attrition. This agreement had the Intelligence Community already on track when President Clinton directed that overall government personnel be reduced by 12 percent from 1993 to 1997, or about 3 percent per year. The Community has, in fact, extended these reductions through 2001, resulting in an anticipated total reduction from 1991 to 2001 of about 24 percent. This pace of reduction is consistent with the level of reductions that the President has directed for non-intelligence agencies.

Although intelligence funding remains classified, Figure 2 depicts the relative funding levels of the major intelligence agencies and their personnel levels for FY 1996. NSA, CIA, and DIA (and the Service intelligence units) have the largest number of personnel; the NRO, on the other hand, has the highest level of funding of any program in the Community, but virtually no federal workforce. Its work is accomplished primarily by contractors in the private sector.

The Community Management Staff (CMS) is the small staff of the DCI used to assist in the execution of his Community functions.

What Conclusions Can Be Drawn from Recent Experience?

In general, from 1980 until the present, intelligence grew at a faster rate than defense when defense spending was going up and decreased at a slower rate when defense spending was going down. As a result, intelligence funding is now at a level 80 percent above where it was in 1980, while defense overall (other than intelligence) is now 4 percent below its 1980 level.

Because the Secretary of Defense in consultation with the Director of Central Intelligence has largely determined1 the size of the annual budget for intelligence vis-à-vis the remainder of the defense budget, one conclusion that might be drawn is that successive Secretaries since 1980 have believed that intelligence should be funded at a somewhat higher rate than defense (regardless of the rate of increase or decrease for defense as a whole). Based upon the Commission's interviews, it appears that, in practice, most Secretaries have begun with the amount appropriated for intelligence the previous year, taken into account whether the overall defense number is increasing or decreasing, examined the additional initiatives needed for intelligence, and arrived at a somewhat preferential number for intelligence spending. Compounded over a period of years, this practice has led to the relatively large disparity between where intelligence is now funded, relative to its 1980 level, as compared with where defense spending other than intelligence is now funded relative to its 1980 level.

The Commission does not conclude the practice followed by successive Secretaries of Defense is necessarily right or wrong, but only that it has been the case. Nor does the Commission conclude that the wide disparity between intelligence spending and defense spending, relative to their 1980 levels, necessarily means that intelligence spending should be cut. But what, then, should be used to gauge the level of spending for intelligence?

The Commission struggled to find a substantive standard or criteria that might serve as a basis for answering this question. Some suggested that the spending level for intelligence should be what is needed to support the military operational requirements of a particular Administration, e.g. the "two major regional conflicts" strategy of the current Administration, and all other intelligence needs should be met by the same intelligence capabilities. The capabilities needed to support military operations, however, will not satisfy all intelligence needs. For example, the President and other users (including defense officials) rely on information produced by intelligence capabilities that are not principally used to support military operations, e.g., HUMINT, some imagery, and some signals intelligence activities. Moreover, the operational needs of the military provide no real limits to intelligence spending. As military commanders seek to win battles while minimizing casualties by knowing where the enemy is at all times-what the Pentagon has been referring to as "dominant battlespace awareness"-the need for intelligence grows exponentially. The costs of collecting signals and imagery intelligence from satellites, processing it at ground stations, and transmitting it to foxholes, tanks and aircraft in the "battlespace," all in a matter of seconds, are potentially unlimited and, regardless of how much capability is available, it may never provide total coverage.2

The Commission found it is equally difficult to assess the overall resource level for intelligence based upon the perception of the "threat" at any given point in time. If one target is no longer considered a "threat," intelligence collectors usually can be moved to others that are. Indeed, since intelligence collection capabilities, both human and technical, take years to deploy once resources for them have been provided, they are designed to be flexible in order to adapt to new needs.

Ultimately, the Commission concluded that developing a precise criterion for measuring the right level of intelligence resources would inevitably be too simplistic and perhaps unwise. The reality, as for many functions of government, is that intelligence capabilities are determined by whatever the nation chooses to spend on them, not by some rigorous calculation which attempts to precisely balance threats against capabilities. Like the conduct of diplomacy, controlling commercial air traffic, monitoring weather, or defending our borders, there is always more that could be done. Unlike the precision that the government can attach to the cost of delivering a letter, or printing and delivering a Social Security check, there is no precise means to determine how much the nation should spend on intelligence. Just as with other aspects of our national security, determining the appropriate level for intelligence funding requires an assessment of various criteria such as foreign threats and the advantages a particular capability can provide against such threats. These must then be weighed against what the nation can afford, given other government spending requirements and priorities.

In any event, how much the nation can afford to spend on intelligence has been and will continue to be constrained. While the need for such capabilities is compelling, so too is the need to reduce Government spending. Over the next decade, there is likely to be strong and persistent pressure to reduce the costs of Government across the board. Given the fact that the President and the Congress have now agreed in principle to balance the federal budget by the year 2002, deficit reduction casts a particularly long shadow over future intelligence investments. Though the calculations thus far agreed upon for balancing the budget assume that defense spending is capped and intelligence funding remains flat within this cap, changes in the economy or other spending priorities could re-open the debate on the level of spending for defense and hence impact intelligence. Therefore, reducing intelligence funding due to external pressures may be unavoidable in the long run. Furthermore, internal Defense Department pressures to reduce intelligence are also growing as the military services vie for funds to modernize their forces, maintain current levels of readiness, and pay for an increasing number of peacekeeping missions. Even within the existing cap for intelligence in the projected budget, there will be pressures to hold costs down in some areas and increase them in others.

What Needs to be Done

In view of these pressures, it behooves those with responsibility for intelligence resources to begin planning how such resources might be further reduced and/or reallocated to meet future intelligence requirements. Unfortunately, while the Commission found evidence of such planning in a few agencies, most intelligence agencies seemed to lack a resource strategy apart from what is reflected in the President's current six-year budget projection. Indeed, until the Intelligence Community reforms its budget process, it is poorly positioned to implement such strategies. The Commission's recommended actions to improve the budget process, set forth in Chapter 7, are thus a necessary first step towards more rational resource planning. The Commission also believes that certain of its other recommendations, if implemented, would produce costs savings:

The Commission also explored the feasibility of reducing costs by allowing intelligence agencies to charge their consumer departments and agencies for the support they receive. Many witnesses pointed out that intelligence is a "free good" to most departments and agencies. Whether the support provided by intelligence is a map for a State Department analyst, an intelligence report on the economic conditions in a particular country for a Treasury analyst, or imagery necessary for precisely targeting Air Force cruise missiles, the cost of producing these intelligence products is free to the user. Because it is free, the appetite of consumers is essentially insatiable and undisciplined. Consumers who appeared before the Commission conceded that if they had to pay for intelligence support out of their agency budgets, they would, in fact, be more judicious in the number and type of requests they levied.

While charging for intelligence support may, indeed, bring greater discipline to the system, the Commission found many practical difficulties in implementing such a system. One is assigning cost to intelligence support. How does one assign a monetary value to an intelligence analysis, or a map, or a photograph? Another is providing intelligence support to departments and agencies who need it but have not asked for it. What do intelligence agencies do when they have crucial information for a particular department or agency, but the customer agency has used up its annual allocation for intelligence support? How would such a system be administered? Would each intelligence producer keep its own accounts? Would departments and agencies who were delinquent in their payments have their intelligence spigots turned off?

The Commission concluded that the benefits of bringing greater discipline to the current system by permitting intelligence agencies to charge for their services were far outweighed by the difficulties apparent in implementing such a proposal.

The Commission did, however, make one additional effort to assess intelligence costs by undertaking a review of the projected budget for intelligence from FY 1996-2001 in order to ascertain whether there appeared to be intelligence programs that were duplicative or providing marginal value. The conclusions of this review are set forth in the next section.

The Commission's Budget Review

The purpose of the Commission's review was to determine whether there were existing intelligence capabilities, developed during the Cold War, which appeared to be duplicative of each other or of other government capabilities, or which otherwise provided capability excess to the country's needs in the post-Cold War era. The intent of this review was not to arrive at recommendations for cuts to specific programs, but rather to judge whether such problems were apparent, and, if so, to what extent.

To perform this review, the Commission's staff organized the President's FY 1996 to FY 2001 budget for national and tactical intelligence programs into seven "business areas:" 1) signals collection and processing, 2) imagery collection and processing, 3) human source collection and processing, 4) measurement and signature collection and processing, 5) production and analysis of intelligence products, 6) multidisciplinary intelligence (those programs that contribute to multiple intelligence disciplines), and 7) intelligence infrastructure (buildings, support staffs, telecommunications, etc.). All intelligence programs and activities funded by the National Foreign Intelligence Program, the Joint Military Intelligence Program, and the Tactical Intelligence and Related Activities aggregation were included.

Within each of the business areas identified above, further divisions were made to identify the end use, target, or ultimate purpose of the expenditure. For example, the business area of "imagery collection and processing" was divided into three further categories: 1) imagery of the battlefield; 2) imagery for indications and warning intelligence, science and technological developments, and other needs; and 3) imagery management and dissemination. The purpose of this approach was to align expenditures for intelligence programs with their ultimate purpose or target in order to make informed decisions concerning possible duplication or excess capability.

Ultimately, the Commission staff assessed the contributions made by each intelligence program or activity with respect to each of the selected end uses or targets, and judged whether the programs were appropriate given the end of the Cold War, whether they duplicated other programs, and whether they provided excessive capability. These were necessarily subjective judgments based upon the staff's appreciation of the Government's need and what the respective programs and activities contributed in particular areas.

The Commission nonetheless believes the staff's review demonstrated that reductions to the existing and planned intelligence resources may be possible without damaging the nation's security. Indeed, finding such reductions is critical if funds are to be found for the investments in intelligence capabilities that the nation will need in the future, capabilities that are not now funded in the proposed program and budget. Precisely where such reductions should be made and at what level are judgments which the Commission is not in a position to make. Nonetheless, it is clear a more rigorous analysis of the resources budgeted for intelligence is required. In the Commission's view, this analysis should be performed jointly by the DCI, the Secretary of Defense, and the Director, OMB. It should span all three sources of intelligence funding (NFIP, JMIP, and TIARA) and assess the total U.S. intelligence capability against particular targets or types of targets.

In sum, the Commission believes cost savings can be achieved if the Intelligence Community adopts the management practices and implements the cooperative arrangements summarized earlier in this chapter. Those actions, together with pruning unnecessary requirements and unproductive systems and activities, could free significant resources. At the same time, the Commission recognizes that its proposed reforms to the budget review process could result in the identification of shortfalls between programmed resources and needed capabilities, or identify areas where new developments and investments are needed but are not now programmed. This might require the expenditure of most, perhaps all, of the funds freed up by cost saving measures. The Commission itself is not in a position to make this assessment. 1


1 Congress annually makes adjustments to the President's budget request.

2 There is, in fact, a process within DoD for trading off intelligence expenditures needed for the support of military operations against non-intelligence expenditures needed to support military operations within the overall limits of the DoD budget. Support to military operations does provide, in any case, the principal justification for the overall level of expenditure for intelligence within DoD.



Page INT017 February 23, 1996