Since I returned to government 6 years ago, I have been working with the acquisition workforce and defense industry to improve defense acquisition performance. There is a lot of evidence that we are moving in the right direction. We have also effectively partnered with Congress on some initiatives, and we are in the midst of a new cycle of congressionally led efforts to improve defense acquisition—as in other cases with the label of “acquisition reform.”
I would like to share some thoughts with you about the limitations of legislative tools, and also explain why I believe that lasting improvements must come from within the Department of Defense (DoD)—from our own efforts. Legislation can make our job easier or harder, but it can’t do this job for us. I recently was asked by Chairman Mac Thornberry to attend a roundtable on acquisition reform with the House Armed Services Committee. This article is based in part on the thoughts I communicated to the committee.
First of all, what it takes to be successful at defense acquisition isn’t all that complicated—to first order at least. It consists of just these four items: (1) set reasonable requirements, (2) put professionals in charge, (3) give them the resources they need, and (4) provide strong incentives for success. Unfortunately, there is a world of nuance and complexity in each of these phrases and words. They also apply to both government and industry organizations, but not always in the same way. The fact is that none of this is easy.
Reasonable requirements are not all that simple to create, professionals don’t exist by chance, resources are subject to budget vagaries and other constraints—including a predisposition toward optimism—and incentives are complicated and often have unintended consequences.
The work of making each of these four imperatives real for a given program is not easily accomplished, even with strong hands-on leadership. It is even harder to influence through legislation. I have some sympathy—and even empathy—for the difficulty that the Congress and our oversight committees face when they try to “reform” defense acquisition. Congress has two major challenges as it tries to improve acquisition results. The first is the structure of the defense acquisition enterprise itself. The second is the inherent limitation on the set of tools they have to work with to effect change.
One way to imagine the defense acquisition enterprise is as a layered construct. At the base of this tiered structure are the organizations and people that do the actual work of delivering products and services. These people and organization are almost all defense contractors. (I’m oversimplifying a little here—some services and products are provided within government, but this is an exception.) The next layer consists of the government people who actually supervise the defense contractors. This second layer is also the layer at which requirements—a critical input to the acquisition structure I’m describing—directly impact the work. There is a huge variety of contracted services and product acquisitions, and the government people who plan, issue and administer contracts cover a broad spectrum of roles and professional expertise. These two layers are where the action occurs in terms of delivering products and services. Everything else in the acquisition structure is about making these two layers function as effectively as possible.
Above these layers there are chains of command and direct stakeholders of many types, most but not all of whom are located in the organization (military department or component) acquiring the service or product. Next there is a layer of what we like to call “oversight” within the DoD, some of it in the Office of the Secretary of Defense but also a great deal of it distributed in the military departments and agencies. My own position as Under Secretary is a mix of acquisition chain of command responsibilities and policy or oversight.
Finally, at the top of the whole structure, and furthest from where the work is done, there is the Congress, which has statutory authority over the DoD and the entire Executive Branch and conducts its constitutional oversight role.
In order to achieve its objective of improving acquisition, Congress has to penetrate through all the other layers to get to those where the work is done. This isn’t an easy task. The DoD’s relationship with our contractors is defined primarily by contracts, so one route available to the Congress to improve acquisition is to write laws governing defense contracts. These laws then are turned into regulations in our Defense Federal Acquisition Regulation Supplement (DFARS) by people in the oversight and policy layer and implemented by the management layers that are in more direct contact with defense contractors.
As a practical matter, Congress tends to react to events as they occur by passing additional statutory provisions. Congress also tends to make changes or additions whenever committee leadership, members and staff change. Of course, lobbyists for industry and other interests play a role in this process. The result over time is a frequently changing, but usually increasingly complex compendium of almost 2,000 pages of DFARS regulations governing how the DoD contracts for work. A serious effort at acquisition reform would include a complete review of everything in both the Federal Acquisition Regulations (FAR) and DFARS with the first-order goal of simplification and rationalization and the second-order goal of eliminating as much content as possible.
This task would take a good-sized, knowledgeable team up to a year to complete and it would take at least a year more for review and modification to the resulting product. The DFARS is based on the FAR, of course, so this would need to be a federal government, not just a defense, endeavor. I believe this task is worth undertaking, but no one should expect it to achieve miracles; almost everything in the FAR and DFARS is there for a reason—usually as an expression of policy goals that are considered worthwhile. The tough questions have to do with whether the costs of all these provisions in terms of inefficiency, higher barriers to entry for industry, and taxpayer expense are outweighed by the benefits achieved. We may only be able to eliminate a subset of existing provisions, but what we could do for certain is have a more consistent, coherent and easily applicable body of regulations. Over time, I have no doubt that Congress would continue to add legislation that would take us down the same path of increasing complexity; a “reset” every decade or so would be necessary, but I still believe the effort would be of value.
In addition to influencing how the DoD contracts with industry, Congress also attempts to improve acquisition by legislating rules that affect the government oversight layers and the people in them. This indirect approach is based on the premise that oversight and supervisory bodies can have a positive or negative impact on acquisition performance and that laws can in turn improve the performance of those layers. The Weapon Systems Acquisition Reform Act (WSARA) was of this nature. It addressed the systems engineering and developmental test and evaluation offices and it created the Performance Assessment and Root Cause Analysis organization (all within the Office of Acquisition, Technology, and Logistics), for example. Congress also has taken some steps to improve professionalism of the government management team by mandating tenure for program managers and selection rates for acquisition corps officers. Many of the steps Congress has taken, like these, have in fact been helpful.
The more indirect approach to improving acquisition by redesigning oversight structures and processes also suffers from the problem that it only impacts what happens in the top layers of the structure—not the layers where the work is done. Many outside observers seem to confuse the efficiency of the defense acquisition system, (i.e., the process by which program plans are approved and program oversight is executed), with the fact of cost and schedule overruns on particular programs. I sometimes make the point that the DoD only has two kinds of acquisition problems—planning and execution. The burden on the military department or component of preparing a plan and getting it approved is an overhead cost we should seek to reduce, but that burden shouldn’t be confused with the failure to deliver a product or service on time and within cost. Where the DoD’s oversight structure falls short is when it approves an unrealistic plan and thereby fails to prevent overruns and schedule slips. The oversight mechanisms succeed when they produce a more affordable and executable plan. I think we are fairly successful in this regard. Execution itself is where we most often have problems—and that is squarely the responsibility of contractors we hire and the government people who supervise them—in the bottom two layers I described. Changing the oversight layer’s structure and processes can improve our planning, but it doesn’t lead to better execution.
In my experience, some of Congress’ efforts to improve acquisition have been problematic in three ways. In order of significance they are: (1) imposing too much rigidity, (2) adding unnecessary complexity and bureaucracy, (3) failing to learn from experience.
A lot of the work we have done over the past several years has been to identify and promulgate best practices, but a point I have made repeatedly is that the DoD conducts such a huge array of contracted work that it is counterproductive to impose a one-size-fits-all solution or way of doing business on everything that we do. Imposing rigid rules and universal practices is counterproductive. Overly proscribing behaviors also has the unintended impact of relieving our professionals of the core responsibility to think critically and creatively about the best solution to the specific problems they face.
One thing the DoD is very good at is creating bureaucracy. New procurement laws lead to the creation of more bureaucracy. Last year we provided Congress with a number of recommendations to remove reporting requirements and bureaucracy in the acquisition milestone decision-making process that our program managers go through. Many of these recommendations were included in the Fiscal Year (FY) 2016 National Defense Authorization Act (NDAA). Unfortunately, while some requirements were removed more were added. As indicated above, the overhead we impose on our managers does not directly impact the cost or schedule to complete a program or deliver a service, but it does have the secondary impact of distracting our managers from their job of getting the most out of our resources, and it does increase overhead costs. Frankly, I think we have enough rules; we need fewer rules—not more.
I’ve also been in this business long enough to have seen multiple cycles of acquisition reform. I tell a story sometimes about the first congressional hearing I ever attended. It was in 1980. I vividly remember someone on the committee holding up a program schedule and ranting about the presence or absence of concurrency between development and production. He was very passionate, but I don’t recall if he was for or against having more concurrency. We’ve been both for and against high degrees of concurrency several times over the years. Concurrency is one of the many judgments best left to professionals who understand the risks in a particular new product design and the urgency of the need. I also spent several years cleaning up the messes left behind in the late 1980s by an early round of self-imposed fixed price development contracting, which at one time was a presumed panacea to overruns in development. It was a disastrous policy that we swore we would never try again.
The sign outside my door, “In God we trust, all others bring data,” isn’t there as a joke. We need to learn from our experience, and the data tell us very clearly that fixed price development is usually, but like everything in acquisition, not always, a bad idea. We should not be making arbitrary acquisition policy changes under the guise of reform just because we are not fully happy with the results we’ve seen recently. Doing something different ought to reflect a factual basis for thinking that change will make things better. At the very least, novel ideas should be tried on a small scale in pilot programs before they are mandated more broadly. We need to learn from our experience, and, in general, passing laws that force us to repeat unsuccessful experiments is not wise.
Let me come back to where I started, with a description of what it takes to succeed in acquisition. Requirements drive what we acquire and they are set by our customers—the war- fighters and the organizations that use the services or products we procure. Setting reasonable requirements that meet user needs operationally but are still achievable within a specified timeframe, consistent with the need at an affordable cost is a matter of good professional judgment. These judgments can’t be legislated. They occur when operators, intelligence experts, acquisition professionals and technologists work together.
Creating complex new defense products that provide technological superiority is a job for true professionals, in industry and government. It is very hard to write a law that makes someone a better engineer or program manager. We have to develop these professionals over their careers in industry or government. Adequate resources are a concern of Congress, but they are authorized and appropriated in the context of the budgets the DoD submits. Historically, our greatest failing in building those budgets has been to be too optimistic about the resources we needed to deliver a product or service successfully, or about what we expected we could afford in the future.
Sound cost estimating, rational affordability constraints and leadership that insists on the use of realistic costs also are hard to legislate. Incentives for acquisition success in government come from the dedication of our workforce members and how they are encouraged and rewarded by the chain of command and their institutions. Again, this is about leadership, not legislative rules. For industry, it is a matter of aligning financial incentives with the government’s objectives in a way that successfully improves contractor behaviors. And this requires professional judgment that must be tailored to the individual situation—not something that can be directed in legislation with broad applicability.
The bottom line of all this is that there won’t be meaningful acquisition improvement except by our efforts. Congress can make things easier or harder, but this is still our job. We should be encouraged by the fact that we have made a great deal of progress over the last several years. The data support both that we are making progress and that there is still room to improve. As an example, we recently calculated the net Major Defense Acquisition Program overrun penalty for the Services that the FY 2016 NDAA directed. As of today, because of the savings we have achieved, we have built up a “credit” of more than $25 billion in underruns across the DoD. We also have some programs that have come in above their predicted costs, but the number of programs in which we are beating our original projections for Program Acquisition Unit Cost outnumbers the programs where we are seeing overruns by about 2 to 1. We need to stay on course; keep up the good work.