1.1 The big picture: comprehensive measures of military resources and activity

Proposals to increase defense spending by between $110 billion and $160 billion over six-years only make sense as remedies for long-term, wide-spread, and systemic readiness problems. Yet, neither the Administration nor the Joint Chiefs describe today's problems in this way. And it would be difficult to reconcile such an alarming assessment of the military's condition with general measures of resource change since the Cold War's end:

• Comparing average defense outlays during the five fiscal years preceding the Gulf War with the average for the years 1994-1998 shows a 27% decrease in real terms. However, during the same period, the size of the military measured in terms of full-time uniformed personnel decreased by 29%. In other words: real, per person spending has increased. (Tables 1 and 2; Figure 1.)



• Comparing 1994-1998 with the "hollow force" years 1976-1980 shows total real spending to be higher in the recent period despite a significantly smaller force. Recent per person expenditure has been almost 40% higher than in 1976-1980. Indeed, in real terms, the recent five years show a bigger margin of improvement over the "hollow force" period than do the Reagan-Bush years 1986-1990.



• Looking at change in the Operations & Maintenance (O&M) account: During 1994-1998 O&M spending per full-time person in uniform was 20% higher than in 1986-1990 and 63% higher than in 1976-1980 (1999 USD).

Of course, resources are only half of the story. The notion that America's military suffers deep or systemic readiness problems also depends on the fact of increased operational activity (Optempo). An increase in resources, as demonstrated above, is not inconsistent with the emergence of a readiness crisis if operational activity increases by a greater degree. Most of the recent concern about increased Optempo has focused on contingency operations.

• In recent years, contingency operations have grown to an average level of about 40,000 military personnel continuously deployed. Although several times greater than the level common in the late-1980s, this nonetheless represents only 2.8% of today's active military.

Personnel serving on temporary deployment constitute only a portion of the U.S. military overseas. A more comprehensive view of change would take into account not only those personnel temporarily deployed, but also those who are based or stationed overseas.

• During the 1980s, more than 24% of the U.S. military was typically stationed or deployed on foreign soil or waters. After 1992 that proportion declined to approximately 17% of the force. In other words, the transition to a post-Cold War posture has involved a decrease in the proportion of the U.S. military overseas.

These comprehensive measures of change in the levels of resources and activity cannot settle questions about the readiness condition of our armed forces. The devil may be in the details. But they do raise questions about the nature and extent of any purported readiness "crisis." A closer look is due -- and the case of the Air Force presents an excellent focus of inquiry.

1.2 Why focus on the Air Force?

Changes in budget and force levels since 1990 have been greater for the Army-Air Force team than for the Navy-Marine Corps, and the Army-Air Force team has carried most of the responsibility for those deployments that have drawn congressional attention. Between the Air Force and the Army, the USAF has suffered a somewhat greater percentage growth in the numbers of personnel on temporary duty assignment. And the Air Force also has been a more consistent source of complaints about Optempo. Finally, the prima facie evidence for readiness problems seems strongest or, at least, most accessible in the Air Force case. Both overall readiness ratings -- "C-ratings" -- and Mission Capable Rates for the service have declined notably, especially for the Air Combat Command. All these factors make the Air Force a good place to initiate our inquiry. However, in focusing on the USAF as a case study, this report cannot render any final judgment on the Air Force's condition relative to that of the other services.²

1.3 Alternative hypotheses on readiness problems

A common way of approaching readiness problems is to treat a service as if it were analytically a "black box" into which resources flow and out of which comes operational activity. In this view, readiness indicators are like external gauges on the box. A decline in these indicators simply expresses a mismatch between the flow of resources and activity. Problems can be resolved directly by either increasing resources, reducing activity, or both.

Another approach is to "open the black box" and assess both the manner in which the services gauge readiness and the ways in which they manage their resources and activity. Only by "opening the box" can policy-makers hope to distinguish between two types of problems: those requiring adjustment in overarching defense policy and those requiring change in how the services structure themselves and choose to operate.

A second and obvious distinction is between those problems that are deep, long-term, or systemic in nature and those whose source is discrete or transitory. Complicating the effort to distinguish between the two is the lens of partisan politics, which tends to either magnify or minimize problems. Also confounding the effort to distinguish between chronic and transitory problems is the process of post-Cold War military restructuring, which has slowed in recent years, but is not yet complete. It would be remarkable for any military to undergo such a transformation without some degradation in readiness. This might be short-lived -- but it would likely last throughout the transition period.

In line with the above observations, our review of the Air Force's readiness condition will make two types of distinctions:

• First, it will distinguish between those problems that seem rooted in policies external to the Air Force and those whose origins seem internal -- that is, rooted in how the service structures itself and manages its resources and activities; and



• Second, it will distinguish between those problems that are long-term, chronic, or systemic in nature and those that seem to be transitory or related to discrete and recent policy decisions.

1.4 A framework for analyzing the readiness condition of the USAF

In light of the previous discussion we will analyze the Air Force's present readiness condition in terms of three overlapping periods of "problem development." These reach successively further back from the present and involve different types of causes. The first period pertains to a set of problems that have their origin in Air Force budget decisions and calculations made during 1996. These decisions have led to real and even serious problems, but not chronic ones. The causes are discrete and the problems are relatively easy to correct -- although it will take time for the effects of remedial action to be felt.

The second "problem development" period pertains specifically to today's putative pilot shortage and the problem of pilot retention. These difficulties have a longer gestation cycle, reaching back to the early 1990s, and they involve some extrinsic causes -- some factors strictly beyond the control of either the Air Force or national authorities. The third "problem development" cycle is distinguished by casual factors rooted in the general transformation from a Cold War to a post-Cold War military establishment. This category engages the full gambit of readiness issues, and it raises the most fundamental and far-reaching questions about the USAF's adaptation to the new era.

2. Parameters of the Recent Crisis

The Air Force's readiness condition can be analyzed in terms of four elements: equipment, people, units, and infrastructure. The first three form a cluster with immediate implications for operational capability. The principal "hard" evidence of problems in this cluster is a recent decline in Mission Capable Rates (MCRs) and the service's growing difficulty in meeting its stated requirement for pilots. The reported decline in MCRs has been sufficient to pull overall unit readiness ratings down -- by 18% in three years for the force as a whole and a staggering 56% for state-side active component units, according to Air Force Chief of Staff Michael Ryan.³

Concern has also focused on a possible decline in unit training readiness, although in this case the evidence is sketchy and often anecdotal. There is stronger evidence that some inputs to Air Force training regimes -- such as training time and home-based training activity -- have declined; and this could imply broader problems in the future. Turning to infrastructure: one indication of problems in this area is the growth in property maintenance backlogs -- now estimated to total over four billion dollars.⁴

The supposed cause of these problems is a combination of limited O&M resources and the punishing pace of contingency operations since the Cold War's end.⁵ Resource constraints could impinge directly on mission capable rates by pinching the supply of spare parts and curtailing the service's repair and maintenance capacity. Indeed, the recent decline in MCRs has been associated with a growing equipment repair and maintenance backlog -- reportedly $323 million for FY1999 -- although the Air Force has managed this backlog in a way to minimize the grounding of aircraft.⁶ The USAF primarily ascribes the equipment maintenance backlog to a lack of spare parts. This shortage also has led to increased cannibalization of aircraft and increased "dipping" into Mobility Readiness Spare Packages (MRSPs).

The pace of operational deployments (Deptempo) would interact with resource constraints in an obvious way: it contributes to the work load of the maintenance system and, thus, can contribute to declining MCRs when the supply of maintenance does not meet the demand. The Air Force's recent Deptempo level also is supposed to affect negatively the retention of experienced personnel, such as pilots, by weighing too heavily on them and pushing them toward separation. Finally, by some analyses, current Deptempo is thought to depress training readiness by distracting units from their training routines. Involvement in peace operations is often criticized as especially deleterious to training readiness because the activity itself is thought to add little to vital operational skills.

In some sense, infrastructure problems are the worst ones faced by the USAF. Prior to receiving FY1999 supplemental funding, the USAF had identified shortfalls totaling $1.2 billion in base operating support, property maintenance, and military construction accounts. This compares to a reported shortfall of $750 million for spare parts, engines, depot maintenance, and training.⁷ Infrastructure problems tend to have an indirect and delayed affect on operational capability, which partly explains why the USAF has treated infrastructure as a bill payer. In 1998, the Air Force funded real property maintenance at a level sufficient to complete only emergency repair and critical work. The most immediate impact of infrastructure neglect on operational readiness is via a reduction in workforce efficiency and an increase in personnel dissatisfaction. Another possible effect is increased damage to aircraft due to deteriorating airfield conditions. At the very least, increased airfield monitoring and emergency fixes draw personnel away from other activities.

Key to the challenge of infrastructure support is the fact that the USAF retains excess infrastructure -- and by no choice of its own. It now routinely vents its frustration on this issue in its official posture statements and congressional issue papers:

The Air Force is spending scarce resources on unneeded facilities and spreading its airmen too thin....To enhance readiness, the Service must be allowed to right size its infrastructure so that it matches strategy and force structure.⁸

Summarizing the shortfall in infrastructure support before the House National Security Committee last year, ACC commander General Richard Hawley pointed out that, in terms of dollars allocated per square foot, funding for military construction and property maintenance has declined steadily since the mid-1980s.⁹ However, to properly appreciate this trend, another one must be considered: During the same period, the property holdings of the Air Force have increased steadily on a per capita basis. This reflects a lag between personnel reductions and infrastructure reductions.¹⁰ Thus, relative to its size and the size of its budget, the Air Force has more property than ever before -- and more than it needs. In choosing to fund base support at minimum levels the Air Force is struggling to prevent excess infrastructure from pulling its budget entirely out of balance.

Between 1989 and 1996, the per capita floor space owned by the service has grown by more than 15%.¹¹ This suggests significant excess capacity among the 172 installations operated by the Air Force worldwide. The Air Force expects that any future base closures would eventually entail annual savings comparable to those projected for past rounds: an approximate average of $56 million saved annually for each major installation closed.¹² Such savings could easily exceed the current annual shortfall in base operating, property maintenance, and military construction accounts.

The burden of carrying excess infrastructure has implications for readiness that go beyond the problems of base support. Although recent Air Force budget requests have kept base support and property maintenance to a minimum, this does not preclude the "migration" of funds from other areas into these accounts. The exigencies felt at the very top echelons of the service may be different than those felt by the commanders of wings and bases. Responding to a readiness questionnaire from the office of Senator John McCain, Air Force officials acknowledged that wing commanders do in some instances "plus up base operating support at the expense of other vital programs" in order to balance funding between mission and support areas.¹³

A 1996 Government Accounting Office (GAO) report that reviewed Air Force use of O&M funds during 1993-1995 also found that the service obligated significantly more funds for base support activities than it had sought from Congress.¹⁴ In 1995, for instance, the Air Force spent almost $1.2 billion more for base maintenance and repairs than it had requested. GAO concluded that these funds were more than likely reprogrammed from other Air Force accounts. By this mechanism, the effects of excess infrastructure can cut into aspects of readiness other than base support, for instance: home-based training and exercises.

Figure 2 and 3: Recent USAF Mission Capable Rate Charts

Measured against the left axis the above chart shows a "53%" growth in Not Mission Capable Rates (NMCRS> between 1991 and 1999. This actually reflects a decline from an 84% to 74% Mission Capable Rate -- a ten percentage point change. Note the sharp decline in 1997 and 1998. This preceded by a "plateau" of several years, 1993-1996, during which MCRs are close to 80%. Prior to this, 1991 and 1992 show historically high rates. The charts below depicts a portion of the same years but in positive terms -- MCRs rather than NMCRs.

Sources: Top chart from Gen. Michael E. Ryan, Testimony Before the Senate Armed Services Committe (Washington, DC: United States Senate Committee on Armed Services, 5 January 1999); Bottom chart from Air Force section in Sen. John McCain, Going Hollow: America's Military Returns to the 1970's (Washington, DC: U.S. Government Printing Office, October 1998).

The remainder of this report will focus exclusively on aspects of the readiness picture other than infrastructure; that is, it will focus on people, units, and equipment. To evaluate the perceived problems in these areas it will analyze both the supposed symptoms -- declining MCRs, reduced training readiness, and pilot shortages -- and a range of possible causes: resource constraints, high Optempo, and involvement in peace operations. Finally, it will assess a variety of remedial measures, both proposed and underway. In all this, however, we cannot forget the likely negative effect of excess infrastructure. The extent to which this problem conditions all the rest remains an unanswered question in the readiness debate -- not only for the Air Force, but for all the services. Suffice to say that for years the service Chiefs have consistently and ardently pleaded with congress to let them close more bases.

2.1 Mission capable rates and unit readiness

The recent decline in USAF mission capable rates has been reported widely and has figured centrally in recent Air Force testimony before House and Senate committees.¹⁵ This, and related reports of a decline in unit readiness, constitute the strongest evidence of a serious USAF readiness problem. In January 1999, Air Force Chief of Staff General Michael Ryan reported that the MCRs for the service's aircraft had fallen 10 percentage points since 1991 to a current level of 74% -- with almost one-third of that decline occurring in 1998 alone. As noted above, this decline is associated with an even more remarkable drop in overall unit readiness scores, which combine MCRs with other readiness measures.

Declines as steep and sudden as these point less to systemic problems than to some acute and exceptional cause -- in this case, a convergence of unexpected demand for spare parts in 1996 and an understatement of the requirements for parts in 1997 that left insufficient parts in the pipeline for 1998. As the Air Force stated in its response to a readiness questionnaire from the office of Senator John McCain:

Spare parts shortages have been driven by unanticipated requirements growth during the year of execution that exceed the levels funded in the Air Force Working Capital Fund (WCF) budget. The net result in the FY1996 WCF was that spares were funded at 90%. The requirements computation for FY1997 WCF inadvertently understated spares requirements, which resulted in an 82% funding level. These levels kept the spare parts pipeline from being filled at a rate to support actual demand and led to parts availability problems.¹⁶

Particularly hard hit by the spare parts shortage was Air Combat Command (ACC), which controls most of the Air Force's state-side combat aircraft -- fighters, bombers, and special mission. In September 1998, the ACC commander, General Hawley, presented data to the House National Security Committee showing that MCRs for the command's operational fighters had fallen from 88.4% in 1991 to 74.8% in 1998.¹⁷ ACC's exposure to this problem resulted partly from official USAF policy:

Supply policy states that engaged forces and forward-deployed forces receive priority on parts requisitions, so they will receive parts before forces remaining in CONUS [Continental U.S.]. The cost to the non-apportioned forces is an increased likelihood of parts shortages... For example, ACC CONUS forces received this lower supply priority which was compounded by the reduction in spares funding and inventories.¹⁸

The prominence afforded ACC in the recent readiness uproar has opened this policy to debate. Indeed, a characteristic of the recent readiness complaints issuing from all the services (except the Marine Corps) is a focus on non-deployed, home-based, or "second-to-fight" units.

2.2 The delayed effects of budgetary action

Prior to the crisis testimony of Fall and Winter 1998-1999, and before the Clinton administration added $112 billion dollars to the six-year defense budget, the Air Force was reporting that its spare parts problem had been contained and that a gradual restoration of MCRs was expected. The service had corrected its earlier underestimate of requirements and had funded spare parts in its FY1998 and FY1999 budgets at 95% and 100% of requirements, respectively.

In addition to improved funding levels, implementation of total engine life management planning and continued emphasis on supply chain management improvements should stabilize the decline of mission capable rates and begin recovery during FY1999.¹⁹

But it takes time for budget changes to affect the production and delivery of spare parts. Thus, both the spare parts shortage and its correction lag behind budgetary actions. In July 1998, acting Air Force Secretary F. Whitten Peters reflected that, "It has taken until recently for the underfunding in 1997 to work its way through the depot pipeline, so it will take at least several more months before the increased funding in 1998 and 1999 will take effect."²⁰ This was an optimistic estimate. Depending on the type of component in question, estimates of the lag between funding, delivery, and utilization of parts ranges between 12 and 36 months.²¹ More realistically, the remedial action taken in 1997 and 1998 would not be fully felt until the end of 1999. In the meantime, shortages have repercussions throughout the system. Among these are the cannibalization of aircraft and the draw-down of stockpiles and "packages" set aside to support operational missions. The capacity of the force (and the nation) to "ride out" these effects depends both on the existing level of threats and the size of the nation's material reserves -- stockpiles and spare aircraft. These issues are examined in Section 4.2 of this report.

2.3 Overall unit readiness ratings

Equipment readiness is only one variable in the readiness equation, but recently it has led the decline in reported levels of overall USAF readiness. That the decline in MCRs, itself a product of ephemeral circumstances, could produce the image of an Air Force plunged into crisis is partly an artifact of how readiness is calculated. Unit readiness ratings are composites that reflect the lowest value of their constituent parts -- equipment, personnel, and training readiness. One of these variables can sink the rest. Looking specifically at the "equipment readiness" category: this is based partly on MCRs, but does not simply reflect them. Indeed, the equipment readiness measure -- or "C-rating" -- will tend to either exaggerate or minimize change in MCRs.

Figure 4: USAF and ACC Mission Capable Rate History as of 5 January 1999

The report of an 18 percentage point decline in overall USAF unit readiness (and 56 percentage points for U.S. -based units) refers to a decline in the portion of units with readiness scores of C-1 or C-2. Unlike MCRs, these scores are discrete, not continuous. A small change in MCRs across the force as a whole could be sufficient to push a large number of units over the threshold from C-2 status to C-3. Likewise, modest remedial action might be all that is needed to push them back.²²

The recent testimony of General Ryan before the House and Senate armed services committees illustrates the susceptibility of the C-rating system to sudden swings.²³

• In September 1998, General Ryan had reported to the Senate a 15% decline since January 1996 in overall USAF readiness and a 47% decline for state-side units. This was based on mid-summer estimates.



• Before the Senate on 5 January 1999, he reported some improvement based on calculations made in November 1998: the declines had repaired to 14% and 40%.



• Two weeks later, however, he reported the current 18% and 56% figures to the House Armed Services Committee, based on January 1999 calculations.

The fluctuations in the reported decline for units belonging to Air Combat Command -- first a 47% drop, then 40, then 56 -- suggest an especially erratic cycle of progress and retreat in the war against unreadiness. Behind this appearance, however, are more modest changes, and among their causes might be both cyclical events, such as the resource shortages that often occur near the end of a fiscal year, and exceptional ones, such as the recent operations against Iraq. Unfortunately, the crisis mentality surrounding the readiness issue has Congress taking the Pentagon's pulse more frequently than is useful for sober analysis.

The absolute decline in the overall readiness ratings of Air Combat Command also deserves attention for what it may say about the peculiarities of both the current crisis and the method of readiness assessment. In the course of 1998 the percentage of ACC units reporting C-1 or C-2 levels of readiness fell to below 50% and have remained there. This puts ACC's readiness assessments uniquely close to levels reminiscent of the "hollow force" years. No other major command in any service of the U.S. military has reported decline on this scale.

Also remarkable is the discrepancy in the trend lines for ACC's mission capable rates and in those for its overall readiness scores. As noted above, the method for calculating readiness can either mask or magnify minor changes in MCRs -- but the experience of ACC is a limiting case: a three-year decline in overall readiness of between 40% and 56% is associated with a decline in mission capable rates during the same period of only 9% -- from 82.2% to 74.8%.

Figure 5: Discrepancy in Decline of ACC Mission Capable and Unit Readiness Rates, Indexed to 1995 Note: Data points are indexed to values for end of 1995 so that subsequent years appear as fractions of 1995 value. MCR curves reflect data points for each year. Unit Readiness "curve" reflects only two data points: end of 1995 and end of 1998, hence yearly variations are not represented. Sources: Director of Logistics ACC, Ten Year Look Back: Standards and Performance FY89-FY98 (Langley AFB: ACC, November 1998); General Michael E. Ryan, Testimony before the U.S. Senate Committee on Armed Services, 5 January 1999.

Moreover, the decline in MCRs had eased before the readiness alarms sounded in Fall 1998. According to the Air Combat Command Director of Logistics,

FY1997 was a tough year. ACC operational fighters began the year about 80% MC. Beginning in April-May 1997, the rate dropped to 74-75% MC. It has remained at his level in FY1998, essentially halting the MC rate decline. It was hoped that with better FY1998 Air Force Spares funding that some measurable recovery would occur. But this has not yet happened. FY1999 also has improved funding over FY1997, and measured improvement is still hoped for. At this point we are "cautiously optimistic" about measuring improved readiness in FY1999.²⁴

Of course, other factors figure in the determination of unit readiness, and some of these are partly subjective. Commanders must make judgment about whether the problems they see can be fixed in time to allow their units to deploy, and these judgments play a role in deciding overall readiness scores. In such judgments, service conventions and general guidance can make a big difference. In 1998, General Hawley, the head of Air Combat Command, cautioned subordinate commanders to be careful to avoid the inflation of readiness reports.²⁵ This may have prompted a re-evaluation of standards throughout ACC. At any rate, shifting standards are a frequent source of sudden change in readiness scores.²⁶

The plunge in the reported readiness of home-based units has pulled the overall Air Force average down, and has captured center stage in the readiness debate, which now focuses on the status of home-based and non-deployed units. Despite the expression of "cautious optimism" by the ACC Director of Logistics, the alarm generated by ACC's experience may inadvertently compel a more fundamental change in how America manages its military. Historically, DoD has informally "tiered" the readiness of military units by granting differential access to resources, with non-deployed and state-side forces coming last in line. In 1996, the idea of formally tiering the active force in order to reduce readiness costs and fund modernization gained some popularity.²⁷ However, the focus and tenor of the recent readiness flap is pushing policy discussion in the opposite direction.

The second major focus of recent Air Force readiness concerns is a purported shortage of pilots, supposedly resulting from pilot retention problems. Like the declines in MCRs and overall unit readiness, the change in the USAF pilot inventory is symptomatic of problems in the functioning of a complex system. A prerequisite to choosing an appropriate policy response, or even to understanding the real scale and vector of the problem, is a closer look at the workings of the system. Unlike the recent declines in MCRs and unit readiness scores -- which trace back to budget decisions and estimates made in 1995 and 1996 -- the gestation period for problems (and corrections) in pilot supply is five or more years. For this reason, the pilot shortage issue falls midway in our analysis between short-term or transient problems and those issues that concern the overall adaptation of the defense policy to the new era, which we address in Section 4.

3. Mid-term Problems: The USAF Pilot Shortage

Testifying before the Senate Armed Forces Committee in January 1999, Air Force Chief of Staff General Michael Ryan declared the pilot retention problem to be "one of the most serious pilot force challenges in Air Force history." Ryan predicted that "if pilot retention remains as it is today, we will be short 2000 pilots by 2002." Although this assessment may evoke the image of hundreds or even thousands of aircraft without pilots to fly them, a closer look at the issue reveals a less worrisome situation.

3.1 The pilot inventory: a tour in time

In January 1999, the active component of the Air Force had 13,146 "rated " pilots.²⁸ Only 7,862 of these pilots -- about 60% of the total -- were assigned to billets for operational force cockpits. Allocated to staff positions were 2,709 pilots with flying status, who keep current with requisite flying hours in anticipation of likely reassignment to cockpit duty. Finally, there were 2,483 "non-flying" officers -- a category of trained pilots assigned to certain staff positions as well as pilots allocated to the so-called "rated supplement." Pilots in this supplement are assigned to a variety of non-rated jobs in the Air Force, or to advanced Professional Military Education (PME), or other special billets. In the event of a large, protracted war, this supplement can serve as a "bench" pool of pilots available to the operational units.

All told, 60% of the Air Force's pilots carry the principal responsibility for flying the active Air Force's 3470 primary inventory aircraft;²⁹ 32% hold staff positions and 8% are in the supplement.³⁰

The Air Force calculates pilot shortages against its total rated pilot requirement. Thus, the Air Force has more than enough aviators to fulfill its flying missions, even when there is a formal "shortage." The negative effects of shortages are felt primarily in the ranks of more senior pilots, who sometimes must take extra flying tours in order to bring up the experience level in the operational squadrons, and who consequently may suffer some detours in their intended career path.

Figure 6: Air Force Active Component Pilots by Functional Assignment, January 1999 Sources: "Current Active Air Force Officers, 31 January 1999, Rated Officers, Grades Lt.-Lt. Col."

Over the last fifty years the total number of Air Force pilots has frequently deviated from official requirements, swinging between surpluses and shortages of 15% or more. In the post-World War II era, the first shortage (25%) appeared at the onset of the Korean War when the pilot requirement more than doubled in a single year, FY1951. As new pilots were trained, this shortage gradually abated, yet it persisted until FY1958, when requirements began to decline again and a surplus appeared, peaking at 15% in FY1965. With the Vietnam War buildup this seven-year surplus rapidly became a shortage of 16% in FY1967. By FY1972 there was again a surplus, peaking at 18% in FY1976. This surplus disappeared and became a shortage of 5% by FY1979.³¹

The decade of the 1980s was atypical in being characterized by a relatively stable relationship between pilot inventory and pilot requirements. With the beginning of the post-Cold War draw-down, however, a surplus developed, peaking at 9% in FY1993. This prompted a curtailment in training. But the relatively few pilots being trained in the middle years of the decade contributed to a shortage developing in FY1998, which Air Force officials now estimate will reach 14% in FY2001.³²

Figure 7: Percentage Surplus and Deficit of Air Force Active Component Pilot Inventory Relative to Requirements. Historical Data and Projection: 1950 to 2001 Sources: "Historical data from 1950 through 1992 from Thie, Taylor, et al, Total Force Pilot Requirements and Managements (Santa Monica, CA: National Defense Research Institute - RAND, 1995) p.13; historical data from 1993 through 1997 from Assistant Secretary of the Air Force, United States Air Force Statistical Digest FY 1997 (Washington, DC: SAF/FMC, Air Force, Pentagon, April 1998); data and projections 1998 through 2001 from various sources including Inside the Pentagon, 23 April 1998; John Pulley. "Plan: Pay Pilot Bonuses Through Retirement. Officials Hope Proposal Will Ease Aviator Shortage," Air Force Times, 1 March 1999; "Military Pilots Say Airline Salaries Aren't the Only Reason to Leave," Associated Press, 4 March 1999.

3.2 A shortage in the making

Effective management of pilot inventory is a perennial challenge for the Air Force. Various factors combine to push the inventory back and forth between surpluses and shortages. And because some of these factors are relatively unpredictable or beyond the service's control, some degree of unplanned variation in the inventory is unavoidable. The goal of management policy should be to have a sufficiently large array of personnel policy options to allow both an agile response to changing conditions and a dampening of the tendency of the system to precipitous swings.

Typically, when Air Force requirements rise, more pilots are trained and, as this cohort of pilots moves through their nine to twenty-five year career cycle, they create a bulge in the "years-of- service" inventory profile. Likewise, a downward revision of requirements means fewer pilots are trained, and this creates a trough in the profile. When these demographic bulges and troughs interact with relatively unpredictable or uncontrollable factors, such as war-related requirements and the cyclical fortunes of the civilian airline industry, they can amplify into worrisome surpluses or shortages.

Much, but not all, of this pattern can be forecast with reasonable accuracy by studying the years-of-service profiles and other key trends, such as pilot demand in the civilian sector and the changing percentage of pilots who opt for "continuation." For instance, in 1995, soon after the peak of the post-Cold War pilot surplus, analysts at the RAND Corporation predicted an 11% shortage of pilots in FY2001.³³ That is, they predicted almost 80% of the Air Force's 1998 revised pilot shortage estimate for FY2001. RAND found that most of the anticipated pilot shortage would result from the mid-decade recession in pilot training, which had reflected "severe management" measures and the Air Force's preference for retaining senior and experienced aviators when surpluses develop.³⁴

In 1989, the Air Force had trained 1,583 new pilots; the stated requirement at that time was 21,750. By 1994, the service had cut training of new pilots to a level below that needed for steady-state sustainment. The training level bottomed out in 1995 at 480 new pilots -- less than half the sustainment rate.³⁵ Since 1995, the Air Force has steadily increased the number of pilots trained, but yearly production did not rise to near sustainment levels until last year. Although the causes of projected pilot shortages are more complex than the following calculation suggests, a simple arithmetic summing of annual training deficits (sustainment level less actual) for the years 1994 through 1998 yields a total deficit of more than 1,900, a very close approximation of projected shortages. This confirms the earlier RAND finding that most of the shortage can be attributed to deficit training levels in the mid-1990s.³⁶

3.3 The pilot retention problem

The other part of the pilot shortage story is the increasing difficulty in keeping pilots in the service after their initial active duty service commitment. There are special bonuses known as Aviator Continuation Pay (ACP) offered to pilots who sign up for additional years of active duty. The Air Force has reported that by 1998 the so-called "take rate" of ACP had dropped 54% from its peak in 1994. There can be little doubt that the fall-off in continuation contributes to the near term pilot shortage. Questions remain: How much does decreased retention rates contribute to the pilot shortage and, How indicative is this of a lasting problem?

In order to answer these questions we must first consider that we are addressing that part of the shortage that is neither attributable to the mid-decade deficit in new pilot training nor to normal losses of pilots after their active duty service commitment. The rate of continuation has been quite variable historically (ranging in this decade from a low of 27% in 1998 to a high of 81% in 1994), so it can not be established with any confidence what a normal loss rate is. However, the 1995 RAND study projected pilot losses of about 580 in FY1998.³⁷ Actual pilot losses in FY1998 turned out to be 1,032.³⁸ Thus we may provisionally attribute 452 or 44% of the pilot losses in FY1998 to changes in the "continuation take" behavior of pilots. Several years of these increased rates of loss could build a sizable shortage in pilot inventory.

Air Force surveys of pilots leaving the service have identified four leading reasons for their decisions: 19% of respondent reported high operational activity (Optempo) as the leading cause, 14% - quality of life problems; 11% - airline hiring, and 10% - excessive time billeted in staff positions. Air Force personnel managers believe that many of the quality of life problems relate to absence from home due to high Optempo. Therefore they combine the leading two reported reasons for separation and suggest that higher Optempo is causing 33% of the pilot losses.³⁹ Regarding airline hiring: The industry is hiring more pilots today than analysts predicted earlier in the decade, however this pull on military pilots is market determined and certain to change. The "pull" of airline hiring probably combines with the "push" of Optempo, quality of life, and other issues to have a significant effect on retention. A change in any of these elements could shrink that effect.

Turning to pilot complaints about too much time spent in staff positions -- these point to a straight-forward way to mitigate the shortage problem: The USAF should use a greater proportion of its pilots in meeting cockpit, rather than desk chair, requirements. This would imply simultaneously a lowering of "desk pilot" requirements and a decrease in separations: a double benefit.

The Air Force has emphasized increased Optempo as a reason for its pilot retention problems. In Section 4 we examine Optempo in detail and review a variety of initiatives and proposals that can reduce its pressures significantly. But it is important to recognize that the Air Force's fixation on Optempo as a key source of its pilot shortage problem construes the problem too narrowly and restricts the range of "thinkable" solutions. A re-structuring of requirements or greater flexibility in how the service employs its considerable pilot assets would give the service a greater capacity to "ride-out" periods of relative shortage.

3.4 Ending the pilot shortage/surplus cycle

The Air Force estimates that it costs $6 million to train a pilot to full operational competence, making it the most costly specialty skill in the armed services.⁴⁰ The high cost of training pilots is one reason the Air Force stresses the goal of retaining pilots for an extended career. The high cost of training also provides a good reason to be wary of reacting to shortages by training too many new pilots -- some of whom may become a costly surplus later on.

So far, the current pilot shortage has caused some inconvenient reallocation of higher ranking and more experienced pilots into flying billets usually occupied by younger officers, and it has led to some less than optimal experience mixes in operational units. The focus of official concern is now shifting to the period beginning early in the next decade when the small cohorts of the 1990s will end their initial active duty service commitments. The Air Force worries that insufficient numbers of these pilots, who are needed to fill the requirement for senior command and staff positions, will opt for continuation.⁴¹

Prominent current initiatives to increase pilot retention include (i) an increase in the bonuses awarded pilots who opt to continue service and (ii) a partial revocation of the 1986 Military Retirement Reform Act, known as REDUX. The latter will substantially increase retirement benefits for career aviators. These initiatives may be sufficient to ease pilot shortages. Already there is some evidence that retention is improving -- although it is premature to attribute the change to any particular policy action.⁴² At any rate, the Congressional Budget Office (CBO) has raised serious doubt about whether REDUX has had much negative impact on service retention at all. Regarding Air Force officers, CBO's analysis found that "being a fighter pilot had a much larger negative impact on retention than being under REDUX did."⁴³

In light of the CBO data (and the dynamics of the pilot production system), legislating a major increase in retention incentives may not be a cost-effective way to bring pilot inventories into line with requirements. Today's pilot shortage is essentially a problem with a five-year duration, associated with the mid-decade period of training deficits. Rather than taking remedial steps that have the character of permanent financial commitments, it would be more efficient and wiser for the Air Force to focus narrowly on that portion of the next decade when shortages are forecast. Theoretically, pilot bonuses can be adjusted up or down freely in accord with changes in pilot supply and demand. However, it is difficult politically to adjust them downward because reducing bonuses is perceived as short-changing younger pilots. In addition, large pilot bonuses have the negative effect of creating resentment among service members who do not have access to this special category of compensation.⁴⁴

The proposed change in retirement benefits will have an even greater budget consequence. Concern about the "crisis of pilot retention" has been the prime motivator for changing the retirement formula. However, substantial pension enhancements will accrue not only to career pilots, but also to all career service people. Thus, the chosen solution to a five-year shortfall in pilot training will cost taxpayers many billions of dollars annually for an indefinite period.

There are ways to balance pilot requirements and pilot inventory that are less expensive, more efficient, and ultimately more stabilizing. In anticipation of today's shortages, a variety of such options were explored by RAND Corporation analysts as early as 1994 and 1995. These involve adopting more flexible personnel management policies and better utilizing the Air Force's substantial inventory of trained aviators. One positive step along this path would be the reform of pilot assignments to staff positions.⁴⁵

Currently 32% of rated pilots hold staff positions. In light of the current pilot shortage, the Air Force has been transferring some of these pilots to cockpits and replacing them with non-rated officers (which in some cases leaves lower priority staff positions temporarily unfilled). This policy, called "Prioritizing Assignments for Fill", is a fairly standard procedure during periods of shortage. But there is substantial room for the Air Force to routinely employ more nonrated officers (and, in some cases, civilians) in a range of staff positions, and thereby reduce the pilot requirement.⁴⁶ Given that about 4,100 pilot billets are staff positions and that the Air Force faces a 1,950 pilot shortage in FY2001, a substantial portion of this shortage can be addressed by placing rated aviators in cockpits instead of staff jobs.⁴⁷

In general the Air Force should seek ways to better utilize its corps of trained aviators in flying positions after their first decade in the service. For instance, there presently is a shortage of company grade officers (lieutenants and captains). In response, the Air Force could increase the number of field grade officers (majors and lieutenant-colonels) who serve in operational flying squadrons (and make it an expectation that some field grade officers will have longer or more frequent flying tours). Conversely, in periods with a relative surplus of company grade officers, the USAF could decrease the number of field grade officers in squadrons. With over 280 squadrons in the active Air Force, this sort of flexible staffing could provide a swing group of more than 500 cockpit pilots.⁴⁸ Together with reforming the assignment of pilots to staff positions, this could yield a significant reduction in pilot requirements.

The Air Force has considered a "fly-only" career path for aviators that would help keep requirements down in the long run.⁴⁹ It has resisted this option, in part because of an ethos that all Air Force officers should be generalists and potential senior leaders. However, there is no essential reason that a "fly-only" career path need be limited to a narrow specialty track. Most such pilots might fly for the Air Force for between fifteen and twenty years, while rising no higher than major or lieutenant-colonel before retiring from the military and embarking on a second, civilian career. Others, who show particular leadership talent, could be recruited to a higher command track after nine or ten years of flying -- but this recruitment would happen only when there were anticipated shortages of experienced higher-ranked officers.⁵⁰

Another method, used with much success by the Navy, is to employ active-duty reserve officers to help balance inventories and requirements. Over the past thirty years the proportion of Navy pilots that were reservists on active duty has varied between 13% and 35%.⁵¹ If we take these percentages as indicative of the range in which the Navy can employ reservists to augment its active pilot inventory, we can see that the Navy has, through the flexible use of its reserve component, the means to absorb pilot shortages of more than 20%. With some changes the Air Force should be able to achieve similar flexibility. In a complementary move, the Air Force could help balance the experience mix of its active component and reduce Optempo strains by assigning more senior active duty aviators to reserve squadrons.⁵² Air Force pilot requirements also include training assignments, which constitute about 16% of the total. More reserve pilots (and some civilians) could be employed to fill a portion of this requirement.⁵³ In anticipation of periods of pilot shortage, the Air Force could activate pilots for use in training and other roles.

Some combination of the above personnel management reforms and the increased production of new pilots already in the works should allow the Air Force to weather the coming shortage without extraordinary new investments now being planned. More important, reform along these lines would allow the Air Force to reduce its pilot requirement and dampen the tendency of the pilot inventory to swing between surpluses and shortages.

Notes

1. Jason Sherman, "Easy Riders: Joint Chiefs Walk Away With Embarrassment of Riches," Armed Forces Journal International (December 1998); James Kitfield, "The Hollow Force Myth," National Journal, 12 December 1998; Paul Mann, "Partisan Sniping Mars Readiness Debate," Aviation Week and Space Technology, 30 March 1998; and Pat Towell, "Military's 'Can Do' Budget Stance Heightens Hawks' Frustrations," Congressional Quarterly, 14 March 1998.

2. A second part of this study, to be published separately, will take a closer look at the reported readiness condition of America's other armed services.

3. Gen. Michael E. Ryan, testimony before the U.S. Senate Armed Services Committee, Washington, DC, 5 January 1999.

4. Gen. Michael E. Ryan, testimony before the U.S. House National Security Committee, Washington, DC, 20 January 1999

5. Peter Grier, "Stressed Systems," Air Force Magazine (June 1997); Julie Bird, "Washington, We Have a Problem: Doing too much with too little is taking its toll," Air Force Times, 24 March 1997; Art Pine, "Deployments take toll on US Military," Los Angeles Times, 19 March 1995; James Kitfield , "The Last Super Power," Government Executive (December 1994).

6. Air Force section in McCain, Going Hollow - An Update, p. 24

7. Air Force section in McCain, Going Hollow - An Update, p. 31.

8. U.S. Air Force, Posture Statement 1999 (Washington, DC: United States Department of the Air Force, 1999), p. 27.

9. Gen. Richard E. Hawley, testimony before the Readiness, Personnel, and Military Construction Subcommittees of the House National Security Committee, Washington, DC, 25 September 1998, pp. 11-12.

10. For a more detailed discussion of changes in Pentagon property holdings see: U.S. Congressional Budget Office, Paying for Military Readiness and Upkeep: Trends in Operation and Maintenance Spending (Washington, DC: United States CBO, September 1997), pp. 44-46 and pp. 58-59.

11.U.S. Congressional Budget Office, Paying for Military Readiness and Upkeep (Washington, DC: United States CBO, September 1997), p. 45.

12.Air Force section in McCain, Going Hollow - An Update, p. 14.

13. Air Force section in McCain, Going Hollow - An Update, p. 37.

14. U.S. Government Accounting Office, Operation and Maintenance Funding: Trends in Army and Air Force Use of Funds for Combat Forces and Infrastructure (Washington, DC: United States GAO, June 1996), pp. 6-8.

15. Elaine Grossman, "Strategic Airlift Shows Greatest Decline in Readiness; Air Force Moves to Correct Slipping Mission Capable Rate," Inside the Pentagon, 30 October 1997; Peter Grier, "Readiness in a Downdraft," Air Force Magazine (July 1998); Frank Wolf, "Readiness Not On Scale of 'Hollow Force,' Official Says," Defense Daily, 10 August 1998; Bradley Graham and Eric Pianin, "Military Readiness, Morale Show Strain; Budgets Contract; Deployments Expand," Washington Post, 13 August 1998.

16. Air Force section in McCain, Going Hollow - An Update, p. 28.

17. Gen. Richard E. Hawley, "U.S. Military Services' Ability to Provide the Necessary Forces and Their Capability to Carry Out the Nation's Military Strategy," statement before the House National Security Committee, Washington, DC, 25 September 1998.

18. Air Force section in McCain, Going Hollow - An Update, p. 29.

19. Air Force section in McCain, Going Hollow - An Update, p. 29. For an earlier review of remedial steps see Elaine Grossman, "Air Force Moves to Correct Slipping Mission Capable Rate," Inside the Pentagon, 30 October 1997. See also Frank Wolfe, "Pacific Air Forces Lack Needed Spare Parts, General Says," Defense Daily, 29 July 1998; and Director of Logistics, Ten Year Lookback: Standards and Performance FY89-FY98 (Langley: ACC November 1998), p. i.

20. Quoted in Peter Grier, "Readiness in a Downdraft," Air Force Magazine (July 1998)..

21. Lt. Col. Daniel L. Cuda, "USAF, The Hollow Force That Was," Air Force Magazine (April 1994), p. 73.

22. In Military Readiness: Concepts, Choices, Consequences, Richard Betts discusses the case of Army divisions in 1980 that achieved significant improvements in their readiness ratings within six months of having scored "unready." Writes Betts, "Critics suspected slight of hand, but army leaders cited the quirks of the rating system. Marginal differences in personnel fill accounted for the overall change." (p. 93). For a fuller discussion of the peculiar dynamics of the C-rating system see Betts, Military Readiness (Washington DC: Brookings Institute, 1995), pp. 87-96. General discussions of the C-rating and readiness reporting system can be found in Lt. Gen. Thomas Burnette, Jr., testimony before the U.S. House Subcommittee on Military Readiness, 18 March 1998; and, Trends in Selected Indicators of Military Readiness, 1980 Through 1993 (Washington, DC: Congressional Budget Office, March 1994) pp. 7-11.

23. Gen. Michael E. Ryan, testimony before the U.S. House Armed Services Committee, 20 January 1999; Ryan, testimony before the Senate Armed Service Committee, 5 January 1999; Ryan, testimony before the Senate Armed Services Committee, 29 September 1998.

24. Director of Logistics, Ten Year Lookback: Standards and Performance FY89-FY98 (Langley AFB: ACC November 1998), p. i.

25. Bradley Graham, "Strains on Many Fronts Drive Pilots From the Sky, Airmen Say," Washington Post, 13 August 1998.

26. And these changes can have powerful political ramifications. During the 1982 elections, Democrats criticized the Reagan administration for neglecting readiness. Despite a steep climb in O&M funding, service readiness ratings had slipped. This, according to Former Assistant Defense Secretary Lawrence Korb, was due to the services' raising standards "without even telling us." David C. Morrison, "Ready for What?," National Journal, 20 May 1995. Also see examples in Betts', Military Readiness: Concepts, Choices, Consequences, pp. 87-114.

27.James Kitfield, "Fit to Fight?," National Journal, 16 March 1996.

28. "Rated" is Air Force terminology for qualifying to crew aircraft and, among other things, requires keeping "current" with a requisite number of flying hours annually.

29. The total USAF Primary Aircraft Inventory (PAI) designates those aircraft available to units to fulfill various functions. However, only a subset of these form the authorized fighting structure of the Air Force. Hence, in 1997, the active-component Air Force held 1354 PAI attack and fighter aircraft -- but only 936 combat-coded "primary mission aircraft" in its fighter wings.

30. From "Current Active Air Force Officers, as of 31 January 1999, Rated Officers, Grades Lt.-Lt. Col."

31. This shortage coincided with the first episode of dire warnings in Congress and the media about "hollowness" in America's armed forces. Notably, the 1979 shortage was predictable as a consequence of the post-Vietnam War draw-down earlier in the decade.

32.In his statement before the House National Security Committee (25 September 1998), Gen. Richard E. Hawley, predicted a pilot shortfall of 1951 (14%) in October 2001. At the beginning of FY1998 the Air Force's projected pilot shortfall was 835 pilots and by the end of the year the actual shortfall was 648. Some of the reduction can be attributed to Air Force lowering its pilot requirement and some to apparent improvements in the rates of retention. If this pattern of improvement persists for the next several years the shortage in FY2001 may be closer to 12% than to 14%.

33. Harry J. Thie, et al., Total Force Pilot Requirements and Management (Santa Monica, CA: National Defense Research Institute-RAND, 1995).

34.Harry J. Thie, et al., A Critical Assessment of Total Force Pilot Requirements, Management, and Training (Santa Monica, CA: National Defense Research Institute-RAND, 1994), p. 32.

35. The 1995 pilot training level is cited in Michael C. Ryan, Military Pilot Retention: Issues and Options (Washington, DC: Congressional Research Service, 11 September 1998), p. 2.

36. In an interview Air Force Chief of Staff General Michael E. Ryan stated: "We made a terrible mistake six years ago when we reduced our pilot training to such a low level." Armed Forces Journal International (November 1998), p. 31.

37. Harry J. Thie, et al, A Critical Assessment of Total Force Pilot Requirements, Management and Training, p. 64.

38. HQAF/DPFFF, documented briefing, Headquarters Air Force, Personnel, Rated Force Policy, 7 Dec 1998.

39. HQAF/DPFFF, documented briefing, Headquarters Air Force, Personnel, Rated Force Policy, 7 Dec 1998. It should be noted that survey data is necessarily subjective in nature. Respondents may be motivated to weigh one factor in their decisions higher than others for a variety of reasons. For instance, some pilots may find it easier in their conscience or self-image to report their frustrations with Optempo rather than draw attention to the enticement of a well paying airline job. Other surveys and interviews with pilots find distinct preferences for operational duty in cockpits over desk jobs.

40. Michael C. Ryan, Military Pilot Retention: Issues and Options, p. 2.

41. Air Force efforts to increase the retention of pilots will contribute toward balancing overall inventory with requirements. But by simultaneously increasing retention in those "Cold War" years-of-service cohorts that remain in surplus, it will have the effect of exacerbating maldistribution of the overall experience and grade mix of the pilot inventory. Maldistribution of experience and grade caused by the large cohorts of new pilots in the 1980s and early 1990s and by the very small cohorts of new pilots in the mid-1990s will continue until around 2005. See charts in Ryan, Military Pilot Retention (11 September 1998), p. 8. It should also be noted that increasing pilot retention in the active force has the negative effect of reducing the size of the pool that feeds recruitment into the Air Reserve Component, itself a source of pilots to meet active component requirements. See Harry J. Thie, et al., A Critical Assessment of Total Force Pilot Requirements, Management, and Training, p. 62.

42. John Pulley, "Service Hopes Increase in Pilot Bonus 'Take Rate' Continues," Air Force Times, 22 February 1999, p. 12.

43. Christopher Jehn, statement before the House Armed Services Committee, Washington, DC, 25 February 1999.

44.Dave Brackett, "How the Air Force Can Keep Pilots," Air Force Times, 30 June 1997, p. 27.

45.In A Critical Assessment of Total Force Pilot Requirements, Management, and Training, Harry J. Thie, et al., concluded that "The Air Force will likely face overall shortfalls of pilots by FY02 and, perhaps, more serious, a significant maldistribution of experience. Changing retention does not offer much promise for changing either of these two outcomes, but reducing requirements and increasing training do. Reducing requirements even beyond the current cuts planned by the Air Force, particularly in the staff category, is crucial to reducing future shortfalls...," p. 23.

46.Complementing this initiative would be the use of more senior pilots to meet training requirements. See Michael C. Ryan, Military Pilot Retention: Issues and Options, p. 29.

47. See discussion of potential reductions in staffing in Harry J. Thie, et al., A Critical Assessment of Total Force Pilot Requirements, Management, and Training, p. 21. Separately, Air Force Lt. Col. Dave Brackett suggests that, "We do not need air-liaison officers below division level in armored and mechanized units, and we do not need many of the pilots we have in staff positions." Furthermore, he proposes the Air Force "privatize at least 50% of most nondeployable flying positions such as depot acceptance, test and specialized undergraduate pilot training." Dave Brackett, "How the Air Force Can Keep Pilots," p. 27.

48.See discussion of field grade manning initiatives in Harry J. Thie, et al., A Critical Assessment of Total Force Pilot Requirements, Management, and Training, p. 66.

49. Michael C. Ryan, Military Pilot Retention: Issues and Options, p. 27; and, Dave Brackett, "How the Air Force Can Keep Pilots," p. 27.

50.The principal consequence of this change in practice would be that those leaders selected from the "fly-only" pool would be three or four years older than aviators who ride the leadership track from the beginning. However, the benefits of reducing the overall pilot requirement should more than outweigh the modest personnel issue implied by the timing of officer advancement.

51. Harry J. Thie, et al., A Critical Assessment of Total Force Pilot Requirements, Management, and Training, p. 9.

52. Michael C. Ryan, Military Pilot Retention, p. 30; Harry J. Thie, et al., Total Force Pilot Requirements and Management, p. 29.

53. There is substantial precedent for civilian-contract pilot training in the Air Force. See Harry J. Thie, et al., A Critical Assessment, p. 56. The Air Force began a reserve pilot instructor program in 1997 and estimates that by middle of FY2000 some 500 reserve pilots (88 full-time reserve and 413 part-time reserve) will free up 225 active component pilots (about 10% of the training requirement) from training billets. See Darcia R. Harris, "USAF Reserve Effort Gets Nod for Reducing Active-Duty Pilot Shortage," Inside the Pentagon, 17 September 1998.

[Table of Contents] [Executive Summary] [Section 4] [Conclusion] [Bibliography]

Citation: Carl Conetta and Charles Knight, The Readiness Crisis of the U.S. Air Force: A Review and Diagnosis, Project on Defense Alternatives Briefing Report #10. Cambridge, MA: Commonwealth Institute, April 1999.
http://www.comw.org/pda/afread01.html

 Cover of this report as published by
 Taiwanese military translation service.

E-mail This Article