Is the U.S. Military’s Plan to Keep its Edge Fatally Flawed?


As 2015 came to a close, U.S. Deputy Secretary of Defense Robert Work described the Pentagon’s emerging third offset strategy at an event sponsored by the Center for a New American Security (War on the Rocks and CNAS have collaborated on the Beyond Offset series at this website). In a sharp break from the past 15 years, Work focused on the frightening prospect of conventional war against major near-peer competitors such as China and Russia. “This talk is all about conventional deterrence,” Work explained, a condition he believed is weakening due to the diffusion of advanced military technology around the world.

As with the previous two offset strategies (the Eisenhower administration’s focus on nuclear weapons in the 1950s and the precision munitions revolution of the 1970s), the third offset aims to achieve a new era of U.S. military-technical dominance. The need for such U.S. dominance should be indisputable — it would be foolish to simply spend more on legacy and increasingly uncompetitive systems and concepts when competitors like China will be able to match that spending and with lower unit costs.

Secretary Work should be commended for pushing this project. However, his speech (or, more precisely, what he did not discuss) revealed at least four concerns that raise doubts about the third offset initiative. The project is still in its childhood, with much presumably still under development. Yet these concerns are foundational and overshadow the technological points that were the center of Work’s remarks.

It’s the Access, Stupid

The third offset’s description doesn’t appear to address the central problem facing U.S. military forces. Work’s speech focused on the third offset’s five technology-related initiatives:

  1. “Learning machines” that take advantage of “big data” and iterative processes to improve performance,
  2. Human–machine collaboration, whereby equipment and software improve the presentation of data so humans can make better decisions,
  3. Assisted human operations, where machines will directly boost the physical and mental performance of humans,
  4. Human–machine combat teaming, where manned and unmanned systems will work in groups, exploiting the advantages of both, and
  5. Autonomous weapons that will be resistant to adversary cyber and electronic warfare effects.

These are interesting initiatives (although not particularly novel to U.S. military research, as I discuss below) and focus primarily on computer software development. What remains to be explained is how they relate to the principal military problem near-peer and continental competitors such as China and Russia will pose to the United States. That problem is how U.S. expeditionary forces will achieve access to Eurasia and its bordering sea lanes in the face of powerful, distributed adversary air, undersea, and land-based sensor and missile networks.

U.S. forces and doctrine are equipped, structured, and trained to fight for air and naval superiority from centralized hubs, either large forward air bases or aircraft carrier strike groups. As Work and many others in the U.S. defense establishment have long recognized, China, Russia, and other potential adversaries are designing and building sensor and missile forces to neutralize these hubs and thus U.S. power projection capability into and around Eurasia.

U.S. forces need to be redesigned to operate effectively and persistently without theater-located centralized hubs that will be magnets for adversary saturation missile attacks. Future U.S. power projection platforms and munitions will thus need longer-range, greater and more rapid strategic mobility, lower logistic support requirements, and most crucially, the ability to successfully operate in dispersed and distributed arrays. These requirements imply technology initiatives that give as much emphasis to engine efficiency, improved propellants, communication networks, and manufacturing efficiency as they do to software and human–machine interfacing.

Perhaps Work’s five-item software-focused list is designed to win a prospective missile salvo competition with the hope of keeping the existing forward hub concept viable against future threats. If so, the connection between these proposed software initiatives and vastly improved missile defenses over forward hubs needs to be explained. Until that happens, it is hard to see the connection between the third offset as Work described it and the principal problem facing U.S. expeditionary forces attempting to operate near and over Eurasia.

Missing an Economic Theory of Success

The first two offset strategies had clear economic bases for their success. President Eisenhower’s “New Look” strategy substituted a strategic and tactical nuclear weapons arsenal, along with large investments in U.S. strategic forces, for a large peacetime standing army that otherwise would have been required to deter massive Soviet conventional forces in Eastern Europe. The investments in the U.S. nuclear weapons complex, the Air Force’s Strategic Air Command, and the Navy’s ballistic missile submarine fleets were very expensive. But they were less expensive, and less disruptive to the U.S. economy and culture, than would have been an ongoing manpower mobilization at Korean War or even World War II levels, which effective deterrence might otherwise have required.

The economics of the second offset’s precision weapons strategy were equally clear. If one precise munition could replace a hundred or a thousand “dumb” munitions, the savings in logistics, supporting manpower, training, and organizations would be obvious. Once again, the United States would be able to sustain conventional deterrence without having to suffer economic and cultural costs the Soviet Union and its allies were imposing on themselves.

Is there a similarly clear economic case for the five technology initiatives Work discussed at the CNAS event? In his remarks, Work cited the continuing centrality of “deterrence by denial” to U.S. planning, the convincing ability to defeat adversary military forces, denying them their campaign objectives. For the clash of major powers that was the focus of Work’s speech, maintaining deterrence by denial implies a convincing U.S. dominance in a prospective missile salvo competition. Are the five projects Work discussed expected to yield orders-of-magnitude improvements in the accuracy or hit probability of U.S. guided munitions, resulting in a convincing U.S. dominance in the missile salvo competition? Will the projects result in a steep relative decline in the costs of U.S. munitions compared to those of adversaries? Third offset strategists have an obligation to explain their economic theory of success, and in a manner that will be as convincing to current competitors as it was to Soviet decision-makers in the 1950s and 1970s.

Chasing the Elusive Technological Lead

In his remarks, Work reminded his audience that technological leads are fleeting. Indeed, the disappearance of U.S. leads in nuclear weapons and precision munitions is the very reason for the current initiative. At several points Work mentioned that current U.S. leads would be similarly transitory.

We can wonder whether the U.S. leads in the five mainly software initiatives Work described actually exist, or if they do, whether they might be so short-lived that they shouldn’t be the basis for a fundamental Defense Department strategy. Work himself discussed the high Chinese and Russian interest and investment in military-related software and autonomous systems. For example, China and Russia routinely execute high-end cyber operations and both possess world-class computer programming talent.

During the first two offsets, Soviet leaders understood how far they were behind the United States, a condition that made deterrence a success. There is ample reason to wonder whether current Russian and Chinese decision-makers harbor such doubts about the technologies third offset strategists have selected for the competition. If they don’t, deterrence won’t be very strong. Third offset strategists need to explain why the strategy is based on a strong U.S. competitive advantage, why they expect potential adversaries to agree with that view, and how long they expect that advantage to last, if it even exists.

The Third Offset Might Stifle Innovation

In his CNAS remarks, Work strongly defended the notion of human–machine teaming. Work had disparaging words for a proposal by Gen. Valery Gerasimov, currently the chief of the Russian General Staff, to build and field fully “roboticized” units, capable of independently conducting military operations. Work asserted that such a vision reflected the view of an authoritarian regime, which believes people are a source of weakness. Work countered by declaring, “We believe the advantage we have as we start this competition is our people, the tech-savvy people who’ve grown up in a democracy, in the iWorld, will kick the crap out of people who grow up in the iWorld in an authoritarian regime.”

Work’s view implies that although all-machine combat units might make faster decisions than units that include humans (and thus win the so-called OODA Loop [Observe, Orient, Decide, Act] competition), all-machine units will be less adaptable since it will be impossible to program them to anticipate an endless variety of circumstances (though we should note that the first item on Work’s list — “learning machines” and their ability to self-improve with experience — seems to conflict with this assumption). In Work’s view, human adaptability and intuition, teamed with machines, will yield a decisive advantage.

Work’s assumption that human adaptability and intuition will trump machine speed may very well be correct. But that crucial assumption should be verified through experimentation. Such an experiment would require the Defense Department to build an all-machine test-bed combat unit to face off against a human–machine team. Writing recently at War on the Rocks, Work and Gen. Paul Selva, vice chairman of the Joint Chiefs of Staff, argued for wargaming and experimentation to test the validity of new ideas. In his CNAS speech, Work also expressed his strong views on what he believes to be the correct answer to this question before any testing has occurred. Will the Department allow the construction of a high-quality all-machine test-bed combat unit? We should hope that the department will permit an unbiased examination of all views — especially since it is likely that U.S. forces will some day have to fight some version of Gerasimov’s all-machine military force.

Answering the Third Offset’s Open Questions

Work is certainly correct that the Defense Department should vigorously pursue the five technology programs he discussed, if only because America’s top geostrategic competitors are doing the same, in pursuit of their own military-technical advantage. Indeed, various agencies inside the department have already been working on these and other related technology initiatives for many years. In 2010, the U.S. Air Force released its Technology Horizons report, a roadmap for its future science and technology investments. That report discussed in depth several items on Work’s list, including advanced data processing, autonomous combat systems, human–machine collaboration, and cyber and electronic warfare resilience. The U.S. Army’s Natick Soldier Research Development and Engineering Center has a long-standing human augmentation research and engineering program. The Navy’s research and engineering of air, surface, and undersea autonomous unmanned vehicles goes back many years. Work’s list of initiatives was thus mostly restating research efforts that have been underway throughout the department for some time already.

What will be more difficult, and what will certainly require strenuous effort from the department’s chief operating officer, is championing new operational concepts and the organizational and cultural disruptions within the services these concepts will require. Future U.S. military forces will need to be organized so that disparate elements seamlessly share information and coordinate their actions. They will have to transition from industrial-age organizations to information-age processes. Unfortunately, even the U.S. military’s best-trained units are far from achieving this standard.

Transitioning to a long-range and truly distributed power projection force structure would greatly undercut the investments potential adversaries like China and Russia have made in their access-denial sensor and missile networks. Senior Navy officials are developing a “distributed lethality” concept, while the Air Force’s global strike capability is inherently long-range and distributed. These could be fashioned to catalyze a transition to a seamless, integrated, and post-industrial future military force. Developing these concepts to the level needed will require greater funding, the creation of new organizational structures, and leadership attention. However, the majority of the department’s resources still go to propping up the legacy forward hub concept, a habit that will require Work’s full attention to break.

The Army and Marine Corps could organize for stepped-up security force assistance around Eurasia and help build friendly land-based access-denial networks. Senior Pentagon leaders could direct Special Operations Command to step up its preparations for unconventional warfare, in order to provide persuasive options for policymakers. Implementing these changes will be controversial and will require leadership from the top.

The third offset is just beginning and will presumably extend into the next administration and for many years thereafter. There is much work to do. From the start, the project has been about more than technology. Even so, the third offset’s not-so-new technological initiatives will not spare the Pentagon’s leaders from difficult and unpopular decisions affecting their department’s organizations and cultures.


Robert Haddick is a Visiting Senior Fellow at the Mitchell Institute for Aerospace Studies and the author of Fire on the Water: China, America, and the Future of the Pacific.


Photo credit: John F. Williams, U.S. Navy