I was recently at an interesting round-table discussion where one gentleman was talking about a notional plan to cut $1 trillion of defense spending over ten years. He said there was interest on this subject from both the progressive left and the deficit hawk right. But how would you do it?
I see three basic strategies for attacking the defense budget, which I will label (1) Requirements, (2) Optimization, and (3) Ceiling.
(1) Requirements.
The requirements approach is the most obvious, and was indeed the default approach of the gentleman.
The requirements strategy reviews the specific project line-items in the budget and debates the merits of each project. Perhaps we don’t need a fancy new stealth bomber, or a 300 ship Navy will suffice over 355 ships. For each program, you count the dollars in the acquisition’s planned budget and estimate the impact on O&S costs. You find the programs least valuable to national security, and cut them out until you reach your savings quota.
This seems like a rational approach to finding savings. Target the superfluous items, manage the force structure for tomorrow’s needs based on affordability.
Indeed, this kind of requirements management by civilian leaders through control of funds was exactly the purpose of the program budget. It was put forth in 1911-1912 by the Taft Commission as a way for politicians, supposedly accountable to the public, to control the goals and activities of the vast bureaucratic machines they were nominally in charge of, but didn’t have the knowledge or power to direct.
The program budget was then installed in the DOD in 1949, but with the urgency of the Korean War, it didn’t take much effect until Robert McNamara became SecDef in 1961.
The problem with program budgeting is that DOD insiders have far more information on requirements and program costs than outsiders. If you suggest one program could do with less funding or be cut all-together, then the military can bring up the dire consequences, that they were already underfunded to the requirement, that our adversaries are catching up, and so forth. You don’t want to get into a program-by-program fight with people who will always have more information than you.
Frederick Mosher recognized this issue in 1954 after the first attempt to make the DOD move toward program budgeting:
…the military departments have been encouraged to “build up a case” for what they think they may need rather than to work from what they reasonably expect they may be granted… If carried to its dangerous extreme, this attitude might result in completely irresponsible behavior within the service. It is an attitude which might be expressed: “This is what I need, even though I know it is impossible for you (Secretary of Defense, Bureau of the Budget, President, or Congress) to give it to me. However, it will not be possible to do my job without all of it. If you make any cuts, you assume full responsibility for any dire consequences which may result.
So the most obvious strategy to cutting military budgets is the most dangerous. As you go through projects deciding what to trim, you will meet resistance from more knowledgeable stakeholders. In an effort to understand, you might even come out thinking they should get more funding!
(2) Optimization.
The optimization approaches budget cutting a bit differently. Instead of looking at military requirements and connecting those to program budgets, it looks at the acquisition process itself and determines where efficiencies can be gained.
This is a strategy more congenial to those who wish to see the Pentagon run in a more business-like manner. It focuses on consolidating IT systems and utilizing big data. It focuses on streamlining the contracting process and shortening cycle times. It focuses on business case realism and transparency.
In general, it looks something like the Defense Business Board’s plan to save $125 billion over 5 years which I discussed a few days ago. While the optimization approach sounds nice, the defense “business” is far larger and more diverse than any private business in the world. As former DCMO Peter Levine said, the DOD is more of an “economy” than a business. I am skeptical about traditional ideas of business efficiency because the DOD ultimately depends on innovation and scale, which requires a certain redundancy and flexibility not found in traditional cost conscious management practices.
Rational social design inevitably meets and conflicts with the motivations of the actors who make that system work. Contract regulations are so long because people continuously figure ways around the strict letter or the law. IT systems grow in complexity because the operational environments are nuanced and diverse; it is hard to force people into cookie-cutter processes that stymie their efforts at innovation.
Standard patterns of behavior are precisely what good business administrators intend to accomplish; but as administrative theorist Lyndall F. Urwick described, the paradox of a leader is to protect “the originals, the inventors, the crazy people to whom order is anathema… because it is from this lunatic fringe that he is most likely to derive something original.”
At any rate, this optimization approach has been tried many times before. In fact, optimizations are precisely what DOD officials are working towards on a day-to-day basis. But it often requires up-front investment for promised gains that may never materialize. Or it loosens regulations only until the next incident which ushers in another era of oversight. Or it creates new layers of approvals and processes that bog people down in unexpected ways.
(3) Ceiling.
The Ceiling approach is rather simple in this scenario: give the DOD a ceiling that is $100 billion less in each year for ten years, and tell them, “I expect the same end performance.”
Now, this might seem like the most ridiculous approach. I’ll explain why it is perhaps the best approach we have.
Back in the Eisenhower years and before, defense was given a budget ceiling from Congress. In the years before 1949, Congress appropriated funds directly to Army technical services and Navy bureaus, whose chiefs largely had freedom to allocate funds to projects they saw fit. Unlike today, Congress did not weigh in on allocation by project line-item. In the Eisenhower years, the services received a budget ceiling and the Service Secretaries generally handled acquisition matters (the bureaus became special staff organs).
Since the McNamara years, defense budgets have been controlled on a program basis, not by organization. DOD organizations only receive funding indirectly through program decisions. A “case” based on requirements was to be made that justified the budget level. Even in the Laird/Packard years, the supposed budget ceilings were programmatically justified.
Organizational budgeting has many advantages. It allows managers to take advantage of real options as they present themselves. They don’t need to justify project changes to many layers of bureaucracy, and develop a execution plan specifying what will be done over decades. It creates an alignment between acquisition and budgeting processes that has been broken. In effect, it avoids dangerous lock-in problems where decisions based on incomplete information are hard to redirect without a Government-wide consensus.
Moreover, when organizations receive a fixed budget, their attention is on getting the best value for that budget rather than growing their requirements and budget at the expense of others. Here is Pierre Sprey testifying to Congress in 1971:
If it is possible to arrange funding in such a way that you work under fixed pots, as I call them, for each service or perhaps even better for each mission that the Congress and the administration wishes to see accomplished, then I think you have created the atmosphere in which a service can properly judge the extent to which they want to trade simplicity and low cost versus complexity and small numbers.
Well said Mr. Sprey.
The ceiling plan I recommend has a three-step process. (1) Trace current program funding to the executing organizations (i.e., how much funding is overseen by each of the technology labs, Program Executive Offices, and Combatant Commands). (2) Make those organizations organic appropriations in the budget — they should have approximately the same size budget as they had under the previous pattern of appropriation. (3) Let defense and civilian leadership get together to decide whose taking the necessary cuts.
The cuts will be appropriate when moved into an organizational budget because you have simultaneously removed layers of bureaucracy. Former Comptroller General Elmer Staats said in 1969, “Estimates as to documentation costs range from 20 to over 50 percent of development costs, but reliable information is not available.” Professor Robert Judson said in 1975:
If you want to calculate the costs of the way we do business, and this is a conservative estimate, we spend an amount equal to 50 percent of the total dollars involved. So, if the total dollars involved, are a trillion dollars, we are spending $500 billion on trying to achieve various forms of accountability, and we do not get very much for that expenditure.
That percentage has probably remained relatively steady over the years. And that’s where the money will come from. And defense managers might even be more willing to take it under an organizational ceiling. As as Samuel Huntington said in his 1959 classic, The Soldier and the State:
The subordinate, if forced to choose, normally prefers fewer resources and greater freedom to allocate them as he sees fit than more resources less subject to his control. The result is a balance in which the subordinate acquiesces in the authority of the superior to limit resources while the superior leaves to the subordinate a relatively free hand in how he uses them.
The current defense management framework of the PPBS and program budgets is a major obstacle, which relies on centralized decision-making based on shaky predictions of requirements and costs. It treats the uncertain as certain, and thus pays a heavy price for its inflexibility. Accountability, however, should still be handled through programmatic accounting and contractor auditing.
The benefit of the ceiling approach over the requirements approach is that it forces decision-makers towards more austere, single-mission, designs that are more affordable and effective than overly complex, multi-mission, designs. The benefit over the optimization approach is that it doesn’t create a single solution for all organizations. They can create their own lawful structures for decision-making, contracting, and so forth, that will genuinely compete.
There’s a lot more to this story, but it is well acknowledged that defense acquisition has many unproductive processes. The organizational budget ceiling is the best way of rooting them out. The requirements and optimization approaches, on the other hand, will lead you into a quagmire. You would have to justify your claims — at a disadvantage — against “realistic” baseline plans.
Leave a Reply