Back in March, Matt MacGregor and I wrote an article with the provocative title, A bridge fund won’t solve the Pentagon’s emerging tech problem. While it won’t solve fundamental problems, I strongly believe in incremental reform and the importance of error correction. An innovation fund is therefore a crucial first step towards flexibilities in budgeting and requirements.
The basic idea of an innovation fund it to provide the department a funding account which is not tied to specific program activities. This is a problem because tying funds to programs requires a lot of bureaucratic decision time to agree on long-term objectives. Decision time for defense programs is one-to-three years for requirements, and realistically another three years for funding to be released.
It is increasingly absurd to assume that defense planners can predict the course of technology or enemy threats from multiple years out. New ideas from the defense labs or from startups and commercial industry cannot wait for a minimum of two years before funds can be allocated. DoD must keep pace with commercial cycle times. Even DoD’s own Prototyping Handbook finds the budget process to be the “biggest challenge” associated with of the dreaded “valley of death.”
Innovation funds provide a source of funding so that defense officials can pounce on new opportunities and then work to line up more stable funding in the budget. While flexibility is key to the very purpose of an innovation fund, it also gives critics reason to push back.
Slush Fund
The term “slush fund” is sometimes used for such innovation funds. However, Investopedia defines a slush fund the following way:
A slush fund is a sum of money that is set aside as a reserve. In accounting, a slush fund is a general ledger account of commingled funds that does not have a designated purpose. In more sinister cases, a slush fund may be used as a “black fund,” which is unaccounted for and kept off the books.
When people invoke “slush fund,” often the idea of “black fund” comes to mind where people are unaccountable for how they spend money. But a “reserve fund” sounds a little bit better. The funds go to legitimate purposes, there is transparency into how funds are spent, and decision-makers are held accountable for the outcomes of that spending.
Here is Dan Ward’s take on slush fund from our recent podcast episode: “There’s this sort of folklore about slush funds that we’re objecting to the flexibility. What we’re really objecting to is a lack of transparency.” He continues:
Let’s be precise with what we mean by slush fund. If we’re using slush fund to refer to just a bucket of funding that does not have a designated purpose — that’s one definition of a slush fund, as opposed to a bucket of money that’s off the books and there’s no accountability, there’s no insight into, it just disappears into a black hole. That’s the bad version of the slush fund. But something that doesn’t have a designated purpose, maybe we add the word “yet,” a bucket of money that is available to give us the flexibility to respond to emerging changes with appropriate insight and oversight, control, and reportability, this could still technically fit the definition of a slush fund, but it might be a good thing to have.
I think that’s exactly right. Indeed, by focusing so much attention on where money will go in the future, oversight loses track of evaluating the progress and value of what programs have been delivering. Execution to a plan becomes a substitute for seeing tests in person, talking with users about their satisfaction, tracing funds obligations, and assigning merit to decisions. Worse, the presence of uncertainty almost certainly means any plan has errors, and so execution to the plan may not indicate that useful capability was delivered.
Accountability
Accountability is increased when oversight is applied outcomes rather than plans, people rather than process. A person can only be held accountable if they had the authority to approve decisions, and had the responsibility to get it done. Similarly, a person can only be held accountable if oversight is able to determine whether their actions were praiseworthy or blameworthy.
DoD’s reporting needs to improve. The fact that the services don’t even have a list of their ACAT II and III programs is a shame. All appropriated dollars should be readily traceable to a contract obligation or expenditure. That traceability should allow oversight to see documentation or product to verify there was no waste, fraud, or abuse. This verification should be easy and routine, allowing oversight to ask the real questions about value: What were alternative courses not taken? How does this fit into an enterprise architecture? Is the requirement still relevant? Are users more productive with the outcomes?
How it used to be done
In the years before the Planning-Programming-Budgeting System, a different style of programmatic oversight predominated. For example, Harry Truman led an oversight committee during World War II vaunted for its effectiveness. Truman at the time was a senator on the Military Subcommittee on Appropriations. His committee staged 432 hearings that interviewed 1,798 witnesses between 1941 and 1948.
The Truman Committee did not control military plans before programs started. Instead, the committee fact checked and observed the consequences of program decisions. Truman himself was cognizant of the need to move fast and not unduly interfere with the executive branch. Yet he was able to save perhaps seven or eight times the entire cost of the Manhattan Project by exposing faulty weapons production. The investigations had knock on effects throughout the system. Fear of investigations created a deterrent and promoted an untold amount of honest dealings.
The Innovation Fund
In order for an innovation fund to succeed, the best path is to assign it directly to an organization like the Defense Innovation Unit, Joint Artificial Intelligence Center, Rapid Capabilities Office, or a Program Executive Office. Flexibilities in fund allocation should provide an opportunity to experiment with new methods of reporting and oversight. Universal questions of cost growth to APB are less important than contextual questions of user satisfaction.
Indeed, the current multi-year program process has some of the worse aspects of a slush fund. Program lifecycles are devised without regard to change and adaptation. The program is basically entitled to funds through the FYDP. The first major test isn’t scheduled for several years. Users provided input at the very front end, but are not involved until decisions need to be made about committing to production. At that point it’s too late. Oversight finds more problems then they can shake a stick at, but the program is now too far along, too big to fail, and legacy equipment is too old to continue maintaining.
Compare that to a with a well run innovation fund. There is direct insight into where funding goes. Test articles and product release cycles have something to show every six months or year. If something isn’t working, then how much trouble could they get in over a six-month period of performance? Pull the plug early. If it has potential, then pivot based on user feedback. In this way, the innovation fund is continuously being directed by the warfighter’s feedback. It is continuously guided to highest-valued efforts. Accountability to the user and transparency with oversight is what turns a so-called slush fund into a disciplined innovation fund.
Students in Hacking 4 Defense classes have routinely shown how this process of problem curation and lean methodology can create working solutions that fit users needs. Out of more than 450 problems, only two groups ended with the same problem that they started. “As the students are learning,” Steve Blank told me, “the sponsors are also getting schooled.” These development processes can be successfully scaled to solve large defense problems. An innovation fund is the first step to providing DoD leaders flexibility to experiment and demonstrate results.
Leave a Reply