U.S. national security policy faces a double paradox. On the one hand, as a nation and as individual Americans we are remarkably safe and secure at the moment, at least as measured by long accepted, conventional standards of peril. But we don’t feel so safe or secure. Thanks to the media saturation of a 24-hour news cycle that places a high premium on shock value, we are constantly barraged with headlines about the threat du jour. Our deserved repose is thus artificially disturbed.

Beyond that, but certainly no less important, our sense of security is hard to separate psychologically from our general sense of well-being, which is informed by more than just national security factors. The Iraq War fiasco overlapped with the botched response to Hurricane Katrina, and these in turn bled seamlessly into the post-Lehman Brothers economic crisis and the disastrous rollout of Obamacare. It also doesn’t help us maintain appropriate perspective when our leaders speak as if nobody knows the troubles they’ve seen. Hardly for the first time in our history, declinism is back, this time in the form of a trend dating to around 2006.

This free-floating anxiety provides a general level of comfort for the fact that U.S. defense spending levels are at near-historic highs, even accounting for recent and projected cuts. It doesn’t matter that the reasons for these levels predate 2006, and it matters even less that no strong connection exists between our amorphous sense of insecurity and what spending money on defense might actually do about it. It’s a little like allowing oneself to feel better on account of having a lot of money in the bank despite having no clue what to do with it.

Unfortunately, U.S. defense expenditures have been poorly allocated in relation to U.S. strategy for more than twenty years. Worse, institutional dysfunction in and all around the uniformed military suggests strongly that this inefficiency will persist and may even grow. High spending during a period of low threat is therefore buying less relevant defense for situations, not far in the future, that could usher in once again a high-threat security environment. 

In other words, our present apparent strength is mostly superfluous under current circumstances in which U.S. security is overdetermined; but its inefficient reduction, relative to any strategic theory of what defense strength is for, may create weaknesses that will be critically relevant in all the wrong ways to future circumstances. We are putting ourselves in a situation where we might have to bring a very strongly worded memo to a knife fight.

National Security and Strategy

Before anyone can worry properly, we need a better grasp of what we’re really talking about when we invoke “national security.” Like many stock phrases that populate our cognitive universe, this one obfuscates as much as it reveals for all the obsolete assumptions it hauls with it.

The phrase “national security” entered the American public vocabulary between 1938 and 1941. President Franklin Roosevelt started using it to inveigh on subjects ranging from defense appropriations to Social Security. He urged Americans to look beyond traditional views of military preparedness to see the necessity for a more thorough, even total if necessary, mobilization of American society. In a sense he was merely encouraging Americans to understand statecraft as strategists had for centuries: the integration of all assets, domestic as well as foreign, in service to national goals. Intellectuals like Princeton’s Edward Mead Earle and Harvard’s Pendleton Herring promoted the President’s more encompassing notion of what constituted American safety in 1940 and 1941. As war approached, most Americans came to understand that how much steel we could produce, how innovative our scientists were, and how much morale a just, democratic society could generate were as important to securing victory as the number of planes, ships and soldiers we could muster.

The U.S. government does not talk anymore about the total mobilization of American society against enemies, foreign or domestic. So the term “national security” has outlived the circumstances that created it and arguably sustained it through at least the first half of the Cold War. The draft ended forty years ago, and mandatory male registration for it lies at the outer periphery of civic consciousness. Relatively fewer Americans serve in the armed forces today than at any time since before World War II. Deaths in American wars are lower, absolutely and proportionately, than experienced by any prior generation since the 1890s.

Replacing the older sensibility about “national security” as total mobilization is an extensive penetration of professional security institutions, public and private, into many facets of American life. This includes “homeland security.” American society exhibits an enduring sense of low-level, inchoate fear, mainly about enemies, citizens and foreign nationals alike, already within our borders. At its worst, this ambient fear has produced a kind of bureaucratized paranoia that is as expensive as it is counterproductive. (Not entirely by coincidence, too, the trans-galactic space invaders of 1950s popular culture have been replaced by zombies or vampires among us.) 

What does “national security” mean now? At a minimum it means protecting the country and its inhabitants from literal threat, whether from without or within. It still means trying to revive our economy’s faltering dynamism despite the apparent obsolescence of ever needing to invoke a military mobilization or industrial surge. Beyond that, consensus is elusive. To some it means privileging cybersecurity, to others heading off cataclysmic climate change and environmental despoliation. To some, it means not losing the zeal to create a more just society, while others fear that ever more intrusive government will make us less secure as we forfeit the liberty that is the nation’s moral raison d’être.

The point is that, beyond elementary needs, we do not agree on an inclusive definition of national security. Worse, perhaps, since the end of the Cold War the American political class has lost track of what U.S. strategy has been and, for lack of any explicit debate about it, should be today. In short, our grand strategy will be successful if it preserves an open, civilized world, in which America and Americans can thrive.¹ The core military requirements in this grand strategy are to prevent the emergence of a menacing hegemon in peninsular Europe, East Asia or Southwest Asia and to prevent the destructive contagion of sick states—such as violent extremists and criminal networks—from endangering the United States. These two kinds of military requirements drive needs for two kinds of military forces: one set, more conventional, available fast and forward, and a very different set, less conventional, available more deliberately. Since 1991, the United States has fought at least eight wars, three of them lasting nine years or longer.² In none of them did the United States have ready on the scene, or eventually develop, the capabilities it really needed for the kind of war it was in. 

As straightforward as this expression of U.S strategic guidance might be, getting members of the U.S. political class to actually say it, or something even vaguely like it, is very difficult. The Defense Strategic Guidance published in January 2012 was a promising start, but with very modest follow-through during the past two years from leaders in the White House, Congress or the Pentagon.

If we can’t agree on what national security means, and if our leaders cannot articulate the constant strategy we have followed now for generations, how can we possibly know how much and what kind of military capability we should seek? We try hard not to think about it. When we are feeling confident, as we have for clutches of years here and there, the very motive for devoting attention to the question melts away. When we are feeling pressed and anxious, we are too scared to engage in the wholesale, creative disruption that accompanies major innovation. The result is that despite a series of Quadrennial Defense Reviews since the end of the Cold War, we remain tethered to the legacy of the Cold War-era force. And that’s not even the biggest problem we face. Enveloping the intellectual deficiencies are the institutional ones.

Entropy and the Problem of Power Conversion

In classical physics, the entropy of a system is the measure of the energy no longer available to do physical work. For instance, thermal energy naturally tends to dissipate into useless heat. The universe naturally tends toward complete entropy. Industrious humans counteract this tendency by “organizing” the energy available to them, to conserve and direct it before it is lost to use. Possibly stimulated by his conversations with Albert Einstein, Sigmund Freud added a corollary suited to his own professional interests: “In considering the conversion of psychical energy no less than of physical, we must make use of the concept of an entropy, which opposes the undoing of what has already occurred.”³

No defense management consultant has ever said it better. If we approach national security not from a theoretical or strategic point of view, but from that of human institutions, then our most serious problem becomes apparent. The institutions that implement U.S. defense policy compose a giant and, for the most part, highly entropic system. The Department of Defense alone (never mind for now the Homeland Security Department, the intelligence community, and the relevant parts of the Department of Energy) is the largest single employer in the world with about three million people, plus millions more employed in the form of private contractors and their minions. (The next largest employer, at about 2.3 million, is the Chinese defense establishment; then comes Walmart.) Its energy inputs each year can be symbolized by their cost in money: about two-thirds of a trillion dollars. 

Now, in theory, as much of this energy as possible would be harnessed to do work that contributes to the current strategic requirements of national defense (assuming for the sake of argument that we know what these are) measured in terms of relevant, usable military capability. But, as we have seen, this depends on motivation, which depends in turn on leaders’ perceptions of the country’s national security circumstances. In comparable periods of relative security calm, U.S. institutions have produced sharply divergent records. We can see this if we compare the 1920s and the 1990s. 

In the 1920s, the U.S. government reverted after World War I to its usual peacetime habit of relying on a reasonably strong Navy, while maintaining only the embryo of an Army. But U.S. planners made some vital judgments about strategy and force posture during this decade. Noticing the promise of initial British experiments, the United States (and Japan) led the way in pioneering modern aircraft carriers and naval airpower. The U.S. Marine Corps also pioneered the development of amphibious warfare. (Alas, the Army did not advance as much.) Amid noisy controversy, the U.S. military also began developing an autonomous Air Corps. Ideas, training, technology and institutional culture coalesced around new ways of conducting warfare. Serious contingency plans, culminating in a series known as Rainbow, vitally foreshadowed the basic choices to come.

The 1990s were very different. After the Cold War settlements and the Gulf War, the decade was marked by complacency and drift. Between 1991 and about 1997, the Iraqi threat seemed well contained. China, Iran and North Korea were all militarily weaker than they are today. By 1998, however, the danger of catastrophic terrorist attack had become evident. Ashton Carter, John Deutch and I spotlighted the threat with a report and a 1998 Foreign Affairs essay on “Catastrophic Terrorism”, a sure sign that awareness of the danger was becoming conventional wisdom. 

Yet during the 1990s neither the U.S. armed forces nor the intelligence community made any great shift toward understanding and preparing for new patterns of warfare. The U.S. government thought it had plenty of superior capability from choices and acquisitions begun in the 1970s and 1980s. So the job was to cut back in scale, not to transform the character of its basic capabilities. There were no Rainbows. One critic at the time had seen the process close-up as a Vice Chairman of the Joint Chiefs of Staff and summarized the result: “The armed forces have been reduced in size by 40 percent in the last decade, but are still largely structured as they were during the Cold War.”⁴ 

The problem with this, as he and others saw it, was that the global conditions that had spawned the U.S. national security state designed between 1940 and 1960 substantially dissolved. But as they did so, the established equilibrium for distributing defense resources proved formidably durable. The result is that we were inappropriately dressed for the season, sort of like wearing shoulder pads to a baseball game. The surrealism was one more factor prompting Richard Betts to ask at the end of the 1990s: “Is strategy an illusion?”⁵

The 1990s were the great test of President Eisenhower’s famous January 1961 prediction of the coming military-industrial complex. He (and Freud) turned out to be spot on. Interest groups inside and out of the government clustered around every aspect of the system, including benefits and entitlements. These became immovable. So when it was clear that fewer overall resources were to be made available (the period between 1990 and 1995 was also marked by intense struggles over the size of the Federal budget) the path of least resistance was to cut around existing allocation structures. That is how the 1990s became the years that “the locust hath eaten.” 

To one of Washington’s then most seasoned observers of defense matters, George Wilson, “the fundamental mistake civilian and military leaders made . . . was failing to make and impose hard choices about the deployment, composition, and weapons of U.S. military forces. As I write this in the fall of 1999, those choices still have not been made.” Wilson described then-Defense Secretary William Cohen as “a consensus mechanic—and a good one. But consensus in national defense means settling for the lowest common political denominator.”⁶ Wilson seemed to suggest that Cohen should have spent less time writing mediocre poetry and more time rethinking the basic posture of America’s defenses.

Examples of what Wilson was talking about were not hard to find. After the Gulf War of 1991 the U.S. Army purchased 8,000 more M-1 tanks—only about 2,500 of which could be assigned to active units even then—before the main job turned out to require fighting insurgents, not conventional standing armies, in Afghanistan and Iraq. Throughout the 1990s, too, the Air Force was concentrating on investments in air superiority via the F-22 fighter. As of January 2014, the F-22 has yet to be deployed in combat. 

The Aspen Strategy Group, a longstanding collection of former officials and national security intellectuals, assembled in August 2000 to discuss these trends. The bottom line of that conclave was that, as Ashton Carter put it, “The organizational and managerial malaise of government constitutes a ‘threat within’ that over time will have consequences as serious as external threats.” That malaise, more than anything else, helps to explain, as I put it in my written summary of that 2000 meeting, the “great contrast [that] has emerged between the forces we have and the forces we will need. America’s armed forces have not been able to adapt because they have no real strategy to adopt.”⁷

More than thirteen years later, nothing has changed—for the better, at least. We still face systemic problems of deficient power conversion thanks to dysfunctional, corrupted institutions, and we still seem unable to articulate much of anything sensible about strategy in times of peace and uncertainty.

U.S. defense organization tends toward entropy for six interwoven reasons. First, stakeholders in the incumbent system will habitually claim resources, both in compensation and benefits and to support whatever they are currently doing. Second, the American system of governing the defense establishment, being very long on “courts and parties” relative to Executive Branch prerogative, gives stakeholders strong capacities to defend these claims against change.⁸ Third, if the requirements for American national defense change—as they rather radically did—the proportion of entropy in the system will increase as formerly functional uses of resources become less relevant and usable in relation to the new requirements.

Fourth, to reduce growing entropy, the system would need powerful capacities for central management and redesign to conserve and redirect available inputs of money and effort. But such powers of central redesign, always limited in the American system, fluctuate with the ups and downs partly correlating with the extent of concerted belief about threats and how to address them. And sixth, Americans are not, at the moment, seriously threatened, and consequently have an even more diffuse sense about the character of future threats. Thus the system’s capacity for central redesign should be expected to be very low indeed.

If these points remain valid, it is impossible to resist the conclusion that, however much money we spend on defense, it will have ever less relevant and usable effect.⁹ So, after 9/11, the system tended to get urgently needed new capabilities by bolting them on top of legacy capabilities. If my theory of defense entropy holds water, genuinely strategic investments will only be marginal investments. 

Note that this argument transcends well-justified complaints about the Budget Control Act and sequestration. Regardless of the approved topline defense-spending target, if entropy guts rational uses of inputs, the topline will not matter very much. More and more spending will buy less and less relevant capability. 

This is not “waste” in the usual way that term is used. Most of the irrelevant spending may be buying real equipment or soldiers, but the relevance of the inputs are so fractional to new requirements that colossal inputs are needed to achieve desired effects. What scale are we talking about? Let’s start with a summary of energy inputs over time, in constant Fiscal Year 2014 dollars.

With the onset of the Korean War in 1950, U.S. defense spending doubled from its post-1945 levels (measured in budget authority and including money for fighting wars) of about $200 billion. Between FY 1950 and FY 2003, spending fluctuated between about $400 and $600 billion. It hovered near or below the $400 billion mark between FYs 1954–1961, 1973–1980 and 1994–2000; it broke above about $600 billion only twice, in FYs 1952 and 1985. But it has been above the $600 billion mark every year from FY 2004–2014, the high-water mark exceeding $750 billion in FY 2008. That level of spending began to approach World War II levels. (Again, all of these and the following numbers are in constant FY 2014 dollars.¹⁰)

The Department of Defense currently plans for budget authority of about $622 billion in FY 2014, with a planned drop to about $570-575 billion by FY 2018, including funds for overseas contingency operations (OCO). If budget plans are revised to conform with current caps mandated in the 2011 Budget Control Act, then all these numbers might drop by about $50 billion, to FY 2014 authority in the $570–580 billion range, eventually dropping to the $520 range, if we assume about $35 billion for OCO. 

All of these currently planned spending levels—in the high $500 billions—represent very large commitments of resources, comparable in real terms to levels of resources committed at the height of the Korean War and the height of the Vietnam War. Even if Budget Control Act targets are met, the numbers are still relatively high. Inputs in the low $500 billion range are comparable to real-dollar commitments in periods of large buildups and high tensions like those of the Berlin crisis years of the early 1960s.

So much for overall numbers. What about “conversion” of these resources into relevant military capabilities and overseas presence? If there have been no major changes in conversion efficiencies over the years, a reasonable manager might ask whether, in a period of high relative security, the United States could afford to reduce budgets to, or even below, the $400 billion range.¹¹ To answer that question, a reasonable manager might want to dig a little deeper into budget categories. He might exclude military and civilian pay, and even operations and maintenance, and examine only those resources committed to capital goods: procurement; research and development; and testing, evaluation and construction. What would that comparison over time look like?

In FY 2014, the U.S. government plans to commit nearly $170 billion to such investments. Since 1945, measured in constant dollars, that level of investment has only occurred at the height of the Korean War and mobilization for World War III (FYs 1951–53), at the height of the Vietnam War (FYs 1966–68), the Reagan-era Cold War buildup (FYs 1982–91), and every year since 2003. The historic high point, nearly $300 billion, was in FY 2008.¹² Clearly, we are buying more stuff, but that stuff does not necessarily indicate a great deal of technological or doctrinal innovation. More important, is the increase in capital accounts relatively higher than the increases in military and civilian pay? The answer is no.

In addition to providing benefits of many kinds to millions of retirees and military family members, the Defense Department now funds very large civilian work forces, some of them highly specialized to maintain anachronistic systems like those in some of the 40-year-old aircraft carriers the U.S. Navy operates. Because of the long lead times in building an aircraft carrier, each one is idiosyncratic, a kind of floating time capsule of American technology for its era, with specialized maintenance workforces to match. Across the Department, civilian employment between 2001 and 2012 grew at five times the rate of growth in military employment (17 percent versus 3.4 percent).

Purely in the way it does business, before even getting into analysis of ideal force structure or strategic adaptation, every independent defense think tank finds a significant amount of self-defeating entropy in the Pentagon’s processes and business practices alone. This gets closer to falling into the category of “waste” as more traditionally defined. It appears that at least 8–10 percent of resource inputs are lost for these reasons.¹³ 

The obvious and less obvious inefficiencies in the use of defense resources do not change because the interest groups and institutions responsible for such waste are by now well accustomed to their critics and know how to dispatch them. This rentier elite mans well-settled, comfortable political fortresses with cleared and interlocking fields of fire, potential targets identified, ranges dialed in, and the surrounding terrain under constant observation. The weight of these institutions is not new, but after more than fifty years of established buildup, they are vastly larger and stronger. And they are now causing vastly more entropy. 

The last period in which U.S. defense spending as a whole was highly functional, with a relatively low proportion of entropy, was during the 1980s. At that time the United States had been building capabilities for a long time against a relatively stable set of requirements. My old professor, William Kaufmann, estimated at the time that a maximally relevant, usable capability, defined as able to perform all the major functions then postulated for the armed forces, would cost about 20 percent less than the somewhat less capable force then in train.¹⁴ Though no comparable analysis has been undertaken since then, I believe the entropy factor has grown significantly. 

To redo the Kaufmann analysis would require a model of desired military effects (not the same as force structure) in space and time. Military requirements have changed significantly, as has technology, as have the harvests of triumphant improvisation in wartime. Outstanding examples, immortalized now in established acronyms, are in the realms of Signals Intelligence (SIGINT), Unmanned Arial Vehicles (UAVs), and Mine-Resistant Ambush Protected vehicles (MRAPs). The counterinsurgency story is also interesting, if oversold. Nevertheless, a large part of the armed forces and civilian defense establishment layered the post-9/11 requirements atop their established patterns of life. Some of those established structures are no more relevant now than they were in the late 1990s. And some of the post-9/11 additions (like the numbers of MRAPs acquired) are not likely to be as valuable in the future either. My suspicion is that, if it could be quantified, the proportion of entropy in the American defense system has gone from around 20 percent in the 1980s to something like 30–50 percent today.¹⁵

For example, at the moment (early 2014) the U.S. military has ten aircraft carriers in active service in order to yield two aircraft carriers on station at any given time, deploying about 140 aircraft of varying kinds. Each of these ten aircraft carriers, including escort and support vessels and their work forces at sea or on shore, carry the costs of a decent-sized city.¹⁶ An idealized comparison would examine the desired capabilities—strike, antisubmarine warfare, air defense, air superiority, surveillance—and then examine what it would cost, in different configurations, to have comparable or greater relevant capabilities rapidly available via bases in the Persian Gulf or the East Asian littoral.

The U.S. Air Force musters, mainly through its Air Combat Command, about another ten Air and Space Expeditionary Forces (AEF). As with the aircraft carriers, of these only two are on station at any given time, deployable or deployed for immediate action. These two AEFs represent perhaps another 200 combat aircraft, roughly a tenth of the relevant combat aircraft in the regular and reserve forces (aside from bombers and missiles in the Global Strike Command).

The U.S. Army still has very large inventories of armored fighting vehicles. These include nearly 6,000 Abrams tanks. Of these, only about 2,000, thoroughly rebuilt so as to be practically new, can be assigned to active units, though none were or are being used by the Army in Afghanistan. (The Marines used about a dozen in Helmand province.) But Congress has forced the Army to keep the principal tank plant (in Ohio) open to produce or rebuild more and more Abrams tanks. The ground forces are also sized for a large-scale invasion and long-term occupation of a country in Eurasia with active, heavy units granted months to prepare, stage and deploy without much interference, scenarios now as realistic as planning for another Operation Desert Storm (1991).

Back in 1986, Kaufmann commented that it “may be stretching a point to suggest that defense contracts have become the Works Project Administration of the 1980s.”¹⁷ Amid the plutocratic political stew of the 2010s, no knowledgeable person thinks this a stretch. Former Pentagon Comptroller Dov Zakheim has argued, for instance, that the United States may not be able to sustain a meaningful overseas presence either in the Persian Gulf or in East Asia if we have to make do with resource inputs still well above $500 billion per year.¹⁸ In other words, we need more than $500 billion per year, in a period of relative security calm, to do what we did on much larger scales between 1950 and 2003 for around $400 billion (again we’re using constant dollars here). If this assertion is even close to being true, it would be vivid testimony to the tenacity of entropy. 

Management, Structure and Leadership 

Hard as it is to believe, it may indeed be true. Weak motive, strategic amnesia and the “executive-lite” nature of the American decision structure in the face of stubbornly entrenched rentier elites add to a persuasive case. What might change this constellation of constraints?

Motives for innovation and efficiency might arise from a threat to vital interests vivid to all, and that might lead to the empowerment of the central designers and their supporters. That happened during the two world wars, leading to significant organizational innovation during the wars and in their immediate wake (the National Defense Act of 1920 and National Security Act of 1947). 

But vivid threats, real or hyped for political reasons, historically have not always been boons to military innovation. Both in 1950 and in 2001, to take two examples, the first impulse was to mobilize everything the United States already had and try to make existing capabilities fit the new challenges. When really scared, as after the Chinese entry into the Korean War, the U.S. government reverted to the total mobilization mantra of FDR. This is not how strategic reform happens. Urgent business usually drives out anything that takes time, making a major shift of resources from near-term to longer-term use almost impossible. (The astounding record of U.S. innovation during World War II is of course an exception, with an exceptional history enabled by starting from such a low base of preexisting force structure.) 

Real anti-entropy reform is usually not driven by threats or by threats alone; it is driven by enlightened leadership. Even in the 1950s, much of the durable remolding of American defenses to considered strategic purposes had to await the Eisenhower Administration. And its companion was budget-cutting, not budget growth. So rather than look at defense budget constraints today as impediments to a strong national defense, we would be wiser to look upon them as a precondition—necessary but not sufficient—for real progress.¹⁹ Indeed, three key ingredients for a sustained effort at central redesign are now present.

The first is intense downward pressure on defense budgets. Barring a catastrophe, this seems likely to persist at least through the remainder of the Obama Administration and probably into the term of his successor. The second is the increasingly obvious dissonance between defense posture and the relevant security environment. The sluggishness of reform during the 1990s era of cuts had to do with the lingering belief that a Cold War-structured force would remain relevant long into the future. In contrast, no one now believes that U.S. armed forces should be called on during the next decade to perform missions akin to those that have dominated their work during the past decade. Especially important here are attitudes within the services, since they will have to do much of the heavy lifting in order to flesh out new objectives and figure out how to get there from the force posture they have now.

And third is a significant bipartisan consensus on retaining reasonably strong American defense capabilities far from our shores. This is a relatively recent development, and admittedly a debatable one. Although many criticize the supposed parochialism of Americans and their ignorance of the outside world, the truth is that the American public is more globalized in its outlook than past generations and is more willing to accept—in principle, anyway—the extension of American defense responsibilities to faraway places.

But it takes leadership to transform these three propitious conditions into an entropy-fighting redesign plan, and that leadership has been lacking. For all of the President’s initial anti-K Street, anti-special interest rhetoric, the sincerity of which need not be questioned, he has not been successful at reshaping the old national security state. The President devoted much of his energy to a quite serious civil-military struggle over the conduct of the wars in Iraq and in Afghanistan, and his White House, as an institution, does little (like most White Houses since Eisenhower’s) to master or guide the substance of defense posture. The Administration has not disturbed the K-Street transactional culture.

 In the defense area, the Obama Administration has stressed a few bastions it wants to protect, such as adequate forces in the Asia-Pacific region, strong special operations capabilities and a robust effort to develop and field capabilities in emerging areas like cyber. It is hard to argue with these choices, but their à la carte character indicates the lack of a broader vision clear enough to rally all of the allies and potential policy engineers it will need, including key congressional leaders. And again: It’s the vision, the intellectual coherence, that counts most in the end, not the size of the defense budget.

What might that vision be, if not to be taken up in the remainder of this Administration, then perhaps in the next? I have one to offer for discussion; it will be coming soon to a familiar magazine near you.