Jerry McGinn: Welcome everyone I see everyone starting to join we’re really excited today to have assembled a very senior panel on a very niche issue: the DoD budget process. So when people talk about acquisition reform and generally talk about a specific either contracting authority or a way to approach program development or oversight, but a fellow for our center Mr Eric Lofgren — we’ll introduce in a moment — has written an ambitious paper around how the reforming DoD budget process will help the acquisition process and that is in the white paper that was sent to you all and we’ll be summarized in a video he’s going to show you shortly and then we’ll have the discussion with our subject matter experts. We’re very pleased to have two honorables and a very senior staffer, the former director of CAPE Mr Bob Daigle, the former assistant secretary of defense for acquisition the honourable Katharina McFarland, as well as Bill Greenwalt former senior senate staffer and former deputy and secretary for industrial policy. But I’m Jerry McGinn and I’m hosting you all for this discussion and our center for government contracting at george mason university within the school of business. But I want to turn it over to Eric. Eric Lofgren is our fellow and he’s come over from another part of George Mason last year where he was doing a doing a research on the history of defense acquisition and he’s also spent a lot of time in doing defense pricing where he was an analyst at OSD CAPE and he’s written this paper that you’re gonna see this video on. It’s very thought-provoking and ambitious paper. So i’ll let Eric take over for the balance of the conversation so welcome everyone.
Eric Lofgren: great, thanks Jerry. I wanted to say a few words here and then we’ll show a quick video and then the rest of the time we’re going to be focusing on this amazing panel that we have here. So I just want to frame the discussion here a little bit today and talk about again acquisition reform. Many of you are familiar with the concept of this pendulum of acquisition reform that seems to have been going back and forth between control and oversight versus speed and innovation, right, and in the last few years we’ve seen it shifting towards speed and innovation with of course middle tier authorities, with delegation to the services, other transactions and CSOs, a lot of innovation going on in SBIR, and trying to bring in new entrants, software factories… there’s a lot of things going on but one of the worries is whether we are peaking and starting to go back on the pendulum towards the other direction. One of the ways that we can think about this pendulum is having waterfall on one side and agile on the other. When we try to give organizations more freedom to do agile development processes, are they able to exercise those freedoms when you have waterfall processes that are institutionally surrounding the whole process? so a lot of times you’ll hear people talk about agile fall or water scrum fall, right, where you have an agile process that’s in the middle of larger institutional and business processes that are waterfall. The question is: will we be able to continue out pushing towards more reform or not? when we look back in history over the past 50 years, we’ve had plenty of very thoughtful intelligent leadership like Bill Perry in the 90s, David Packard in the 70s, even Frank Carlucci in the 80s, who had been trying to do rapid prototyping, competition, delegation down to the services, and we seem to always be coming back in these pendulums the other way. So we have to think about when we’re doing acquisition reform today, why did they fail in the past? I would suggest one of the reasons we’re here talking today is that progress in adopting new technologies and new concepts of operations really requires us to start tackling the budget process. That’s something that none of those reformers and pentagon leadership in the past have really addressed.
a lot of people say today that there’s a different feeling, right? it feels different today. There’s a lot of new things going on maybe we’re not going to swing back on that pendulum. One of the things that I think is actually pretty new is the innovation hubs such as defense innovation unit, AFWERX, and NavalX. There’s several of these defense acquisition innovation hubs in the ecosystem and they’re getting a lot of attention in the media and in discussions. I think that’s a pretty good place to think about it because they’re like a bellwether for the progress of acquisition reform because if they can really bring in new companies and new technologies and get some dynamism in the defense industry, well that’s a good reflection of the overall health of the acquisition system itself. One of the primary problems is what we call the quote-unquote valley of death, scaling the new technologies and getting them into a program of record. One of the major issues there, of course, is the budget process — or the planning programming budgeting execution system, where you have this two-year planning process. I can’t even start anything new in less than two years except for rare expedited means. I need the two years for programming to accommodate the one year for budgeting and it takes a long time to line up the funding. Then once you get that it’s actually locked in through the future years defense program for five or more years and so that’s one of the major contributors. We’ll never solve the value of death of course, that’s a major problem that we even see in the commercial side of industry. You’re not solving the value of death, but one of the big contributors to the problem that makes it even harder in the department of defense is the budget process and we have to ask, why does it take so long why is it so slow?
when we look at the budget process it’s not just a ceiling of money for organizations, like I have an income and then that translates into my budget which I can allocate. We’re not providing organizations and a certain amount of money to go after a broad mission or to achieve a portfolio of capabilities. Rather, the budget identifies specific projects. So like 20 million dollars to project x and 30 million dollars to projects y and so on. The issue with that is that you’re creating a process that’s very reliant on long-range prediction and control of future technologies, environments, future threats, user preferences, and even the economy is something that has to be baked into those projections of the future. When we do a new start for the most part we’re not controlling it in how much money do they need in the year of the budget, it’s what’s the life cycle cost and it’s a what you could call “teletic” versus incremental. It’s looking at the life cycle of what the whole program will be over the future for many years rather than an incremental decision process.
When we think about the essential element of the PPB — the planning programming budgeting system — what it really did when it was implemented by Robert McNamara in 1961 was change the classification of the budget from one focused on inputs to one focused on outputs. This is a philosophical point but I think it has real implications for people on the ground in the system as they work it. In the past before the PPBS, you had organic line items and appropriations. For example, in the army the department of ordnance and the signal corps. And these were organizations that had a general mission that they could go accomplish. The budget did not control the activities, the programs, the objectives to be accomplished by those organizations. In the navy, of course, you had the bureau of ships, the bureau of aeronautics, bureau of navigation, and these guys had more flexibility within the year of execution to start something new that they didn’t see coming from a couple years before, but also to pivot, ramp down, change direction. Then what we had with the PPBS when that started in 1961 was more reliance on the predictions of the future because now the budget controls what will be done and not just what is the cap of money provided to certain organizations.
When we’re looking to accomplish future technologies or bring new technologies and new concepts of operations into the present, there’s a general philosophical question that we can ask. Should we start with the output in mind and then work our way back on how to get there? or should we focus on the inputs such as people, training, enabling tools, and infrastructure, and allow the culture and through the process of interaction generate what those outputs will be?
I would suggest that a useful way of looking at this is when you actually know a lot about causality and prediction is not so much of a problem — When you know the relevant parameters of the problem and there’s not going to be a lot of uncertainty in the threats, in the technology readiness, and the like — then you actually do want to focus on the outputs and do something that looks like an optimization. But when you’re actually in an environment of extreme uncertainty, I would argue that you want to focus on the inputs and allow the outputs to result from that process.
Of course, McNamara and his whiz kids like Charles Hitch, Alain Enthoven, David Novick, and all these guys, when they’re looking at defense weapon systems and the choices made they really had the belief at that time that quantitative analyses could make uncertainty less of an issue — that they could put their hands around the problem and they could more accurately predict what the next five, 10, 20 years will hold. I would suggest that in this age — the way that we’ve seen commercial industry work — there’s a lot of uncertainty in technology and everything else and we have to move incrementally. That’s the foundations of agile and the ppbs is actually fragile to this constant change of information. So with that introduction I would like to show you a quick video and then we’ll introduce our panel and get their thoughts.
Video narration: the pentagon’s budget is the master controller of virtually everything that happens in the department of defense including how it buys aircraft, ships, launch vehicles, helicopters, and everything else. The army budget justification document for fiscal year 21 is documented over 39 volumes totaling a staggering 11,755 pages, and the army represents just one quarter of the pentagon’s total request. Of the RDT&E program elements, 18% are under 5 million dollars and more than half of all program elements are under 29 million and the mean program element is just 93 million. Compare that to the average venture capital fund which allocates 207 million dollars a year, and large VCs allocate more than 1 billion annually. It’s not just the number of budget line items. In order for a lab or firm to scale a new technology it takes at least two years and in many cases more to get it programmed into the budget. This is called the valley of death, and that’s assuming that over 50 offices approve it. Many defense offices have tried to bridge the valley of death including the defense innovation unit, SOFWERX, naval x, and many others. However, without budget reform their efforts may falter. The valley of death is just one issue. There are several other problems of equal importance that can’t be fixed by tweaking the current system. Do you think china waits two or more years to find money for new military technologies? their ability to move fast supports civil military fusion and has even resulted in their ability to access united states technology firms. In the 1950s and before budgets were classified by organization, such as the transportation service, the signal service, the medical and hospital department, and the engineering service. The question is for research and development, should we return to congressional heritage of financial control?
Here’s what we’re proposing for RDT&E. Today, the army has 182 program elements that average just 69 million dollars each. We’re proposing something like 24 program elements averaging just over 500 million dollars each. That’s a reduction in pes by a factor of 7.6, providing major organizations like the program executive offices and the laboratories the ability to exercise real portfolio management. The navy, air force, and DoD wide accounts can similarly be reorganized. Each major organization — usually led by a senate approved flag officer — will likely control between 250 million and perhaps two billion-dollar portfolios allowing them to be agile and accelerate innovation. These portfolios are not extravagant. For example, in 1951 the army ordnance service received an appropriation of 40 billion in today’s dollars.
The budgets will still maintain programmatic insight. The primary appropriations will remain the same. Within RDT&E, program executive offices will be the major unit of control. underneath the organizations, major thrusts will outline programs, and below that, lower level insight. With greater flexibility congress requires new tools for transparency and accountability. This should include quarterly reports for funding movements, real-time dashboard insights, and more in-person reviews and classified briefings. In return, reprogramming and new start thresholds should be raised to 70 million or 20 percent. As Shawn Barnes, an executive in the air force, said “… in today’s information environment there’s no reason why we couldn’t give congress habitual routine access to information.” the point is a win-win proposition for both the pentagon and the congress. Please join us in making budget reform a reality.
Eric Lofgren: so with that video, i’d just like to remind everyone that the video and the comments I was making are not necessarily the views of george mason nor are they the views of the panelists. Those were my personal views. And so we have a great panel to talk about these things and i’d like to introduce the panelists and allow them to make some observations. I would like to start with Katharina McFarland, who holds a number of roles and she’s currently a commissioner on the national security commission on AI, she’s a director for the procurement roundtable, and previously — as Jerry said — she served as assistant secretary of defense for acquisition. So Katharina, thanks for joining.
Katharina McFarland: thank you Eric. This is a great discussion. It’s actually fairly prevalent out there and I’m really just gonna sort of tease people’s minds before we get into the conversation because I think this will be a very rich one. I will offer that the system that we operate in is over five decades old that allows for us to do our acquisition. It’s post world war ii founded and inside of that was the environment that this system was built around. The environment has changed. I will just use a couple things that I found of interest from the draft NDAA languages both from the SASC and the HASC. One of which was collection of data. We have an ability now to inform ourselves much better than we’ve been able to in the past if we start leveraging it. I think that the tensions between the systems as they’re currently designed are built around the fact that we’ve had tons of people attempt to change the system without letting it mature and give us a full unveiling. So I think the opportunity space is there to take and gather information and start running pilots and measuring success to make change, especially given where we’re going in the digital world where we need to be able to make better informed decisions quicker. So I’m looking forward to seeing a change I’m looking forward to understanding how we could change the system and allow for the three branches of the government to have their ability to perform what their administration or their charter is and make balanced decisions and i’ll leave it at that.
Eric Lofgren: thanks Katharina, short and sweet. So next up is Bill Greenwalt. He’s a visiting fellow at the American enterprise institute, and previously he was the deputy under secretary of defense for industrial policy and he served also as a professional staff member on the senate armed services committee and there he helped formulate some of the recent authorities like middle tier. So Bill Greenwalt, over to you.
Bill Greenwalt: thank you Eric. First of all thank you for this really provocative paper and I think timely paper and i’d urge everybody in the audience to read it, but also to go back into your sources because I think it’s so important to start looking at the sources for this system. In other words going back to the wiz kids and going back to the RAND reports of the 50s and 60s. We created a paradigm, a way of thinking, a way of management, that started off in the late 50s and early 60s that we still live with, as you say in your report and the world, as Katharina says, has massively changed. We need to look at the reasons for that and I think that’s the next thing that this discussion is to go and look at the reasons why haven’t we reformed. I think the first thing, to start off, is to say ‘why do we even have the system’? well it was the best practice, the best thinking, the best management thinking of the 1950s. When mcnamara brought this system in, it was the guiding principles of ford motor. Now, we probably should have thought about that after the auto industry went into the tank 15 years later but it was the best practice and we brought that in and it grew out of this whole concept of scientific management and frankly a fear of the soviet union. This was post-sputnik, and the five-year plans and centralized planning were essentially what was going to take over the world. Americans were scared at that time and this was the way of getting some control. So the system was created and moved forward. Now, why haven’t we decided, ‘oh my gosh, this is ridiculous! why do we have a soviet style system that’s running the department of defense for 60 years’? well, probably two reasons. The first is power. This system encouraged and put the services and the OSD relationship into a different framework and I think we have to understand that. That’s going to be something that’s going to be hard to change. The second thing was this system gave the keys to the kingdom to the house and senate appropriations committees to essentially do what they’ve always wanted to do, which is to have control, have the ability to influence where this money is going to be spent. That system — those power structures — are still in place today and need to be addressed.
One more thing that probably has been stifling change is that this system was the system that won the cold war. Okay we won the cold war with the system, with the acquisition system, with all the other things. So why do we need to change the bureaucracy? well I think the focus is we won despite the processes and management procedures we had. When we competed against the soviets they had two centralized planning systems: one for their defense and one for the economy. They were both pretty much the same. Ours were entirely different and we won because of that. We need to understand that and start making the various changes. So I think when we look at this, we need to look at that history, try to figure out what the assumptions and rationale for it are — the barriers to change — and then come up with new criteria for success. I think your idea on looking at inputs and outputs is hugely important. I think the criteria we should start looking at is time. Time was something that we did differently in world war ii and the 50s. Time did not become relevant or valuable anymore. The value of outcomes was thrown out of out of the system for 60 years those are the type of criteria that needs to come back. So with that I will let my next colleague mr daigle think some provocative thoughts as well.
Eric Lofgren: yeah thanks for that bill, that was great. I’d just like to introduce Bob here real quick. Bob daigle is the director for strategy at rebellion defense, which is a technology startup, and in his previous role he was the director of the pentagon’s cost assessment and program evaluation office — so that’s actually charged with overseeing the programming aspects of the budget process. So Bob daigle is a real expert here and before that he was also a professional staff member on the house on services committee and held many other roles as well. So Bob take it over.
Bob Daigle: thank you and thanks for the opportunity to speak today. This is going to be a fun conversation. I want to pick up on something the bill just said because it’s right in line with what I was thinking and it has to do with time. I’m probably not allowed to say this as a former director of cape, but I will anyway. I think it’s important to recognize that we live in a world where none of us can know what tomorrow holds. We’re living in really interesting times right now, but even before that when bill and Katharina and I were working together and thinking about acquisition reform, I have no idea what technology is going to be in two years. I would challenge anybody to suggest that they know with any amount of certainty what innovation is going to come out of Silicon Valley or out of Austin or out of Boston and I think to your point, Eric, and to the point of this paper and the point of this discussion today, trying to project the future with certainty either through a requirements process or through a budget process does more harm than good these days. We restrict today the choices of tomorrow based on our current understanding without a recognition that that understanding is probably wrong and I’m just going to leave it at that.
I would suggest to Eric as I did in our precursor to this to this conversation that that central thought needs to go beyond the budget process. We need to take a really hard look at the requirements process. We need to look at — to bill’s point — the way congress does its work and the amount of work that is done today to both perfectly specify what a requirement is going to be five years from now, what a budget should be five years from now, and then lock that in so that it’s hard to change tomorrow, is a very big problem. So I’m a huge fan Eric of your central thesis of budgeting at lower levels of specificity — larger pots of money without as much detail around it. I think that’s true for the services and I think the oversight layers of OSD and the joint staff and congressional oversight would be better served getting the outcomes that they really want, which is faster adoption of innovation by taking some of the controls in the current system away.
Eric Lofgren: thanks Bob. Yeah, I hear what you’re saying. One of the things that it seems from reading some of the history was that when I look at the defense acquisition system and that chart that everyone sees the defense acquisition guide where you have these three rungs that are co-equal, you have the requirements JCIDS process, the little “a” acquisition process, and then the PPBE. When I was reading the history that wasn’t true. There was one that was above the rest. The system was the PPBS and its twin concept of systems analysis, and through that structure you actually had requirements as ancillary — something that fed in to those analyses to define the parameters required. And so the requirements-pull process as opposed to technology-push, I see it as a consequence and a necessary input to the functioning of the PPB system. Did you have anything to say on that bob?
Bob Daigle: yeah I don’t know if it has to work that way. During my time in cape, if we take the space development agency for example, just one tactical example we know that we need new space sensors and we know that those space sensors need to be driven to lower power requirements and smaller form factors. Where we settled on in terms of funding was to create a PE [Program Element] that provided broad discretion with regard to how are we going to pursue sensor technology in space and then allocated a certain amount of money to it so if you think about it, if you think about that as a small portfolio, I think portfolio management within the current budget process — current ppbs process — is a completely legitimate outcome. If you think about a portfolio as just a larger collection of money without as much specificity, that’s completely consistent with the ppbs process. I do think at the end of the day you’re gonna have — I don’t think you can get the genie so far back inside the bottle that washington isn’t going to ask, ‘hey what are the outputs that I’m getting for this amount of money’? I think that those questions are going to have to be answered. The question is at what level of specific specificity do they need to be answered.
Eric Lofgren: yeah I think that’s a really good point and we see these program elements that engender portfolio-based management throughout the department to some degree. We have JIDO, the joint improvised threat defeat organization, the space development agency as you said, the joint artificial intelligence center is also structured as its own program element, and strategic capabilities office is another one. So we see these interspersed throughout but it’s not necessarily the general way about going about defense budgeting. I think to the degree that we have more of that, I think a lot of the structure makes sense. But the PPBS is an instantiation of the program budget, and now it’s like are we focusing not on programs but capability area portfolios, or mission-driven organizations, right? so is it really a PPB process or not? i’d like to get some of your insights there.
Katharina McFarland: so you’re hitting upon something that as an acquisition person has always troubled me. If we’re really trying to meet the national defense strategy, we would be focused on outcomes — that is effect-based outcomes. What do we need to achieve in order to achieve over-match and meet our mission objectives? Instead, we’ve traced or followed the money. And let me give you a classic example of that. So in the army, we’ve come to the conclusion that we have to change what we have for combat vehicles, and we decided based on the totality of available funding that we would no longer want to procure tanks because necessarily tanks aren’t at highest of a priority. However, that was overturned because of the equities. Another area is if you’re really focused on national security in terms of priority, some of the lower cost items actually have a huge contribution to the effectivity of a service. Let’s call out command and control. You see with the peer threat that we have right now there’s a huge impetus to solve this joint air-defense command and control problem, but that’s because it’s languished. It’s not a high budget visibility item because the budget had become — and has become over time — more focused on as you talked in your presentation. The spend what we should be thinking about is the attributes of capability and how that translates into meeting the national defense strategy. I think if we did that we’d find more of these little pop-ups and more of these trade spaces where we know we have to have the equities and the visibility into the larger spend, but we would pay attention and husband through things like enabling technologies, command and control, the lesser spend, which are more important in terms of the actual effectivity and effect efficacy of our capability sets. Just thinking out loud.
Eric Lofgren: anything to add there bill?
Bill Greenwalt: the last thing I want to do is sit here and try to do incremental reform over PPBE instead of blowing it up. But this actually when you go back to the original intent as far as I can tell. There was a desire to do greater cost accounting, cost analysis, to drive the system into performance budgeting and performance-based on outcomes that never happened. If you look at — and I think this is really critical to look at how we do accounting which is really boring but I think this is significant — our accounting is to address out points. In other words where did the money go? just as Katharina was saying, that’s our financial system, and then congress tried to drive a financial system to add up what are our assets and liabilities and we’ll do the CFO act, which is almost more than worthless because we’re spending a lot of money to get nothing. The thing we didn’t do was cost analysis and cost accounting and we started to get there with the private sector, but rickover pushed this to ‘let’s use cost analysis to drive down profit margins’ and essentially drive getting the data throughout the department not just from the primes to managed to better outcomes.
Eric Lofgren: thanks bill. I think that’s a really important point. In the rickover hearings back in the 60s — where he’s trying to get the cost accounting standards implemented — just hilarious and really insightful. I would recommend people read some of those, but I think this also goes back to what Bob was talking about in terms of, well, what are you justifying? what are those outcomes? how do you show to the higher levels that have a real job to do in terms of oversight, how do you bring them in on the process and have them be comfortable? Brittany Clayton here has a nice question about this where she’s saying basically, the president’s budget documents they have all this historical cost and programmatic data and these go back quite a ways, and so she asked in the proposing framework with far fewer program elements would a similar level of detailed cost reporting be required? if not do we risk losing the transparency of that cost data? and real quick before I turn it over to you one of my thoughts here is that when we look to how the structure was in the 50s, 40s, and before, right, you didn’t tie the programmatic aspect to the budget necessarily, but there was always in the comptroller a program analysis shop which eventually became the office of systems analysis and then grew into CAPE, but what this shop used to do was actually do just that, what bill was talking about, collect cost data on what actually happened where did the money go, what did we get for it, and then tie that in with effectiveness. So instead of being a forward-looking plan based on predictions — instead of justifying these budgets to specific end items based on predictions of performance and costs — you’re actually collecting the real costs in real time and showing what actually did happen and provide that ecosystem of information that can inform decisions not necessarily through the budget, but you still had managers and oversight agencies able to focus on ‘what did I get and what what are the plans going forward’? and so through those appropriations hearings you heard a lot about what the services wanted to do. It’s just that you didn’t have a very detailed line items that tethered you to those plans defined a couple years before. So you want to want to jump in there and talk about, should we still have detailed cost reporting and where’s the transparency coming from?
Bob Daigle: i’ll take a shot at it. I think the oversight community would like to have — and I think the management would like to have — the ability to have the types of information that they can make better decisions on. My concern is I’m not certain whether the US government has the done the work to identify what the best practices are in that area. I think we should be looking at not just private but public sector foreign state and local governments who can actually do the best job of, what are the best practices in cost accounting, managerial cost accounting, and could those be brought in to the federal government to allow decision makers to make better decisions based on what we’re spending. I think that is that’s a financial accounting discussion that we just never had, because again we have two systems: what are we spending on, and let’s go count bullets and blankets and figure out what our what our assets and liabilities are.
Eric Lofgren: Katharina do you want to jump in?
Katharina McFarland: so I have a parochial view in this regard. I believe that at certain points in the system of acquisition there’s more fidelity associated, or knowledge associated, with the cost, and I believe at the beginning it’s obviously very nebulous and as you graduate towards fielding it’s much more detailed. I think we have right now in our grasp, if we could adopt it, systems that could help us have visibility into that information and not make it a reporting ritual where human royalties come into play with errors. I think that we could graduate even within the system as it’s currently constructed to be far more effective in this regard. I’ll hesitate to state that cost account standards should be totally revamped but i’ll offer to you as a nation we have many cost account standards that we impose upon our population, whether it’s industry people or government, and each of them seem to have reasons for having narrowly defined areas. So i’ll take the IRS and i’ll take our preferred use of cost account standings for programs and so when I’m an industry partner I have to report two different methodologies and I don’t quite understand that in today’s environment why I could not find one standard to apply for all of us. But I defer to those that are financial wizards, but going back to the premise of this discussion, which is where we are today, I think that we could indeed accelerate the application of the science and I’m very heartened to see what the authorization act drafts look like because there’s an implication on it and I totally am supportive of advancing our methodology.
Bob Daigle: I mean, I’m just a huge fan of data so my default position is collect it all collect everything what you do with it afterwards. Somebody smarter than me is going to come along sometime and crawl through all this data and realize that the meaning of life is 42 or whatever, but this is a fight that I had the last couple years inside the pentagon. Since Bill Greenwalt devolved all acquisition authority down to the services, the services don’t have to report any data on acquisition costs or results into a central system and we just drew the line and said that’s absolutely not correct, we’re going to need this data going forward for analysis. Centralization of data is should be one of the key lines of effort inside the department of defense. Since being in the private sector we’ve had working with a number of companies on the outside and Netflix and folks like that and we’ve spoken about how they think about data, and the idea that there is ownership of data and your organization owns this data and my organization owns this data is a foreign concept to most major data companies in America. Collect it all. Storage is cheap. Guaranteed we’ll figure out a use for it tomorrow.
Eric Lofgren: Yeah, I hear you. When I was a consultant in the CAPE my project was the cost assessment data enterprise, so shout out to the cade folks, and it’s really a rabbit hole this cost accounting discussion on it because every OEM [original equipment manufacturer] has their own different system and then how how deep do you go. We’ll have conversations with some where they just have a couple cost account codes and others where they’re going down to like tens of thousands of work packages.
Bob Daigle: … to Katharina’s point, most of the nuisance that comes out of industry is less about the transmission of data because my system can talk to your system and we can just transfer information automatically. It has to do with the bespoke formats in which the government requests data that requires manual manipulation or spreadsheet after spreadsheet after spreadsheet to transform the same data multiple ways to provide it to multiple offices. We should be past that, but I think what’s really important though is how you use the data and when and most importantly if there’s a time constraint. In other words, when we were collecting data in world war ii and operations research was the thing, it was time sensitive. Let’s do analysis and then we can keep doing the data for later analysis beyond, and the problem I think we have now is we collect data to collect data and then we wait and let the process become three or four years before we can make a decision. That’s what’s driving this long-term budget process, this paralysis by analysis. If you’re gonna get the data and you need to use it but you also have to be time limited and don’t in fact negatively impact things that are moving down the road that we need to go fast on. So I have to absolutely support what bill just was talking about in particular areas, and I’m not referring obviously to the building out of a ship but rather IT. The system is not agile. It’s not adaptable as it’s currently defined and we try to parse it out and make it fit into this cost and to this programming and budgeting world and if anything in this environment needs to be adjusted quickly given the threat of what’s going on in the world today we should closely look at this type of a system to see if there’s other agile budgetary methods that could be used that still provide adequate visibility in given where we are today because we’re falling behind rapidly.
Eric Lofgren: I have a question here from Jamie, he’s basically saying that engineering and program management functions seem to be employing agile pretty fully and so his question is really how can we posture for agile from the financial management perspective, and do you guys actually feel that it’s more difficult to implement agile in this functional area of financial management.
Katharina McFarland: I believe that we have a challenge that we have to find a way to work with congress and all the branches of government in this area, and that is not overnight — it is not agile by design. It can be made agile. I’ve seen excursions, the defense acquisition workforce initiative for example. We have a huge amount of cleanup at the end of the year that we just don’t get expended and it goes back into treasury and congress felt very strongly that the workforce needed to be improved so they gave us flexibility in utilizing expired funds to apply to the workforce. Over time the comptroller, who hated something out of family, basically killed that and went to appropriators so we have people who, if you look at it, the system can adjust, it’s just very — if I was to use material analysis the rc would be in excess of 90 — which is very very very brittle. So you don’t seem to have the ability to be flexible in this, and we need to start thinking about that because — and I believe with the use of data will help people understand — it is okay to have differences inside of the system.
Bill Greenwalt: I agree with Katharina 100 percent. I don’t think we we’re fearful enough, and the threat has to drive this. We are falling behind, and yes there are pockets of agility now in engineering, and yes there’s pockets of agility now in contracting, but it’s not the whole system, and from the budget standpoint we still have people again who still believe that this system is going to make it and don’t understand that they are now part of the problem and they’re in comptroller and we all have to get together and say ‘who’s the enemy here’ and what do we need to do to actually be competitive? Again, we’re not ahead anymore. We are we are falling behind, if we’ve not already fallen behind, and that requires drastic changes. Budgeting and finance is definitely an area — it’s not really sexy, but that’s going to be one of those things that’s going to potentially recreate our competitiveness in the future.
Bob Daigle: bill, I was going to congratulate you for not including CAPE as part of the problems. You said congress and comptroller, but then you said that budgeting and programming isn’t sexy, so I have to take your compliment away. Bill and I worked really well together when we were on the hill doing a lot of the acquisition reform stuff one of the conversations that we had often that is useful for this conversation as well is, don’t necessarily try to fix the current system. It’s too hard. Everywhere you look in the current system there is a stakeholder and that stakeholder is going to defend their turf. It is often better to just create new pathways through which activity can run outside of the current system. So Eric, you mentioned early on, hey the middle tier of acquisition or OTAs or some of those activities, I think the same is true as it relates to budgeting. When we start thinking about AI when we start thinking about IT programs — when we start thinking about agile — I would suggest to the system, I think Undersecretary Lord received a lot of credit for going after colorless money for software acquisition. I think she should go further than that and start thinking about different acquisition processes for software and things like that. There’s an opportunity when you’re working in a clean slate environment like that, when you’re when you’re just pulling up a blank piece of paper and saying hey let’s create a whole new thing from a blank piece of paper there’s no existing stakeholders, so there’s no existing stakeholders to defend the status quo, it might be a way forward in areas like this.
Bill Greenwalt: I agree 100 percent, pathways are a way to go and we do have acquisition pathways now we have pathways around the requirements process. The problem we’ve had and Katharina was right on target there. Things like the DAWDF [Defense Acquisition Workforce Development Fund] which were a pathway to actually get the acquisition workforce up to snuff to be able to use these authorities, was killed by the budget people. It was killed, by let’s go back to the soviet process and control money the way we’ve always controlled it. Budget pathways are going to be so much harder, and now we’ve done in the past, you can probably argue that OCO was one, and counter IED was one, and others that we’ve done for missile defense in the past, and so on, but we need several other different pathways in budget and they’re just going to be so hard to get because the antibodies not just are in congress but they’re in the department themselves, and that’s where I think we have to start.
Katharina McFarland: just think about the huge ecosystem that you’re talking about. The ecosystem involves our ERP, all of our data systems that we’ve invested so much money in which then ties us to industry because they support and work within our system as it’s currently defined. On top of that we have process. You’ve been talking about it, and I’m really thrilled to hear that because a lot of people can confuse process with practice. If we were to clean slate a system and really look at just what it takes to go build something and field it, if you would, but that is anything from the requirements down to getting in the hands of the user you would not see some of the oddities that that we’re dealing with. In terms of ‘let’s just pull software to the front,’ the way that the process is involves a lot of people who haven’t actually and aren’t actually ownership of the responsibility, or the outcome being involved in ensuring equities are met somehow within the design of that activity, so has in terms of things that we could do that don’t involve congress they involve ourselves we should be looking at the process and not the practice, but use the practice to help us define a better process. How many people really need to be there? can we use automation to confirm or deny that people the issues have been met like CMMC or cyber protection or fill in the blank, live fire test and evaluation? you could do it in any domain but we should be thinking about how can we remove the burden and the cost that candidly comes from that time out that you take in order to get the next approval step forward. I’ll give you a classic example of this. The army had been very frustrated under Heidi’s Shu about the timelines it took to them for them to get to ‘yes’. I think Heidi had said it’s like being on a school bus, and everybody in all of the seats in the school bus were trying to drive it for you even though you had the steering wheel. So if you take a look at all of those entities that become part of your ecosystem — the time that it takes to brief them, answer their questions, and brief everyone to get to the point where you can have a yes — takes years out of your program timeline and that cost is burdened in there. Now take that to the budgetary system. Have you watched it? my favorite chart is how many lines are on that chart that says where you are in the budgetary process. You’re developing, you’re executing, you’re reviewing. Oh my god! Right? think of the manpower that’s involved in that. So let’s pull the process out, put it over here, and take a real look at the fundamentals — what does it take to do the financial process? you might find your process has hindered your progress with your practice.
Bob Daigle: I would love to jump on that because I completely agree with you and it goes back to one of the points I raised earlier. You started that conversation that was, ‘what does it take to get a capability into a user’s hand’? so that means you have to start the process at the beginning of that — at the beginning of that thought which is often in the service requirements systems — and to me, we can, Eric, to your point about your paper, I’m a huge fan of pulling the budget process up a couple of layers to doing capability-based budgeting and offering the same freedoms in that world that we did in in parts of the acquisition world. I’m a huge fan of just burning the entire requirements process to the ground. I don’t see and I have not seen for years where it adds a lot of value into the process of what actually gets into the hands of the users.
Eric Lofgren: yeah, I tend to agree. When I look at the requirements process, it’s not that we don’t need requirements of course, we need an interaction between requirements and then technical feasibility, but just having this monolithic structure up front and then you get all — and rickover of course he talks about this a lot the layering of bureaucracy, to what Katharina was talking about — it’s not just that it takes a lot of time to get people and their approval — and rickover had this great story of a guy who was trying to get signatures from these people around the pentagon and then by the time he made his round the individuals had left their office and he was just too distraught to even go back out and start the process again for several months — but it’s not just the time of getting the approvals, and it’s not just the gold plating where everyone has their own little thing that they want to throw on and it’s not any one thing but it’s that marginal next thing that breaks the camel’s back, it’s also that — and george shares said this who was the one the chief designer of the b-52 for boeing, he said you couldn’t get 10 reasonably knowledgeable people into a room and look at technologies like nuclear power, the radar, the jet engine before they were created and have them agree on that that is a feasible and good thing to pursue. So when I look at the requirements process and then how it feeds into what actually gets funded through the budget and you can’t execute on the next new technology unless you get those approvals it seems that that’s where some of that layering of bureaucracy the non-consensual things are often the things that are most likely to prove revolutionary. Bob?
Bob Daigle: from an academic perspective, from a data management measurement perspective, careful when you start trying to measure decision-making time because we’ve had this conversation with the joint staff for years around JCIDS and they will rightfully point out that, hey, the amount of JCIDS time has declined precipitously over the last few years because of some reforms they’ve done internally and we’ve pulled the amount of time that paper is sitting in the joint staff inbox down from something like 240 days to 70 days. That’s great. Then when you talk to the services they’ll say, yeah but that’s only because they push all of that work to us and we’re spending all of that delta and more because they’ve jacked up the requirements for what is necessary to hit the the send button to the joint staff. So when you look at this, you need to look at it end to end and figure out what’s really going on. Where are people spending their time? And to your point, is that time adding value to the provision of capabilities to the warfighter?
Eric Lofgren: yeah, someone told me once that metrics are great, we love metrics, they’re actually very important but you need to have like a group of metrics and anytime you take one thing or a couple of things and then incentivize all performance and outcomes based on those metrics then it becomes a bad metric. So it’s almost like when you formalize a metric like speed to do this, well, they’re going to just figure out ways to do things on the periphery or that are outside that you didn’t intend but then turn that metric into something bad.
Bob Daigle: yep that’s called a perverse incentive.
Eric Lofgren: so we have a bunch of questions here. A couple of them, one from Douglas Berenson and an anonymous person, I would say that there’s two questions here. So first is this consolidation and this reform of the budget process an ideal thing something that we want to work for in the abstract? and then there’s the question of well if it’s good, what are the realistic ways to get there? and we talked a little bit about carving out some new areas and letting those on this side grow in a different way but the questions here are really, does congress have a desire to move to a different system? do they even want a change in that system? and then one of the other questions here is will the election in this upcoming year affect the likelihood of it getting any budget reform? so bill did you want to take that on?
Bill Greenwalt: sure. Bob, you were there as well so please pile on. I think congress and the members and the staff you really want to try to do the right thing. Okay? and so yeah, we have the system that they’re stuck in. They love it because it allows them to ear mark better, or at least to follow their ear marks because there are no ear marks but there are ear marks — but the reality is that this has to come from the department. Has to come from the budget shop in a way that explains to the appropriators that this is in our national security interest, and I truly believe based on all the time i’ve been up there, that the congress will eventually go there. But right now they don’t have any reason to. There’s no need to change. The department isn’t really complaining. People in the department complain, but the people they [the appropriators] deal with are the are the comptroller. The comptroller loves the system and loves all the ways of controlling the services. They’re trying and doing all this process, and so this has to come internally first and then try to sell it to congress.
Now, there will be some pushback. There’s gonna be need for greater transparency in exchange. But ultimately I think you there is the possibility to achieve reform here and I’m more optimistic than I would normally be in many reforms.
Jerry McGinn: I was going to ask you a bit about this because a lot of these ideas, portfolio management, have been around for a while. I mean the 809 panel certainly shed a lot of ink on it. But what you’re arguing is really that we just really need to come up — with the departments come with — a better idea on how to do that, and that they would be willing that the appropriators and the others on the hill would be willing to accept that approach?
Bill Greenwalt: I think you will have start having a discussion. But frankly, what we do is we have these little marginal discussions — I want this little pot here and then we immediately go down to well, ‘I’m not sure if there’s enough transparency and do you really need that pot? you really know what you’re doing’? and back and forth. I think the department needs to come up with comprehensive 1961 reform and say, the chinese are going to win this conflict if we don’t get our act together, and we need to work with you, congress, to bring this forward. Frankly I don’t think the appropriators or congress want to have the blame for the fact that we will not be able to compete with our great power competitors because they sit there and want to earmark something in the state of ‘X’. I think the department is not giving itself enough credit because I do think that there are people in the department who like this system and do not want to give it up. They blame the appropriators and of course the appropriators are going to say well I don’t need to do anything because the department didn’t know what’s it wants.
Katharina McFarland: so i’d like to add a little color to this. I would offer to you that it relates to what your underlying thesis is, which is the threat — the amount of visibility by joe public onto the threat — has really been diluted by all the other activities that are going on with COVID, you name it, upcoming election, the visibility into how right now the heightened environment it is just remarkable. If you were just to extract a few issues that are in the open press about what’s going on in the world today and have those other distractors removed, you would find public would be very interested, I believe, in trying to find out how to change the dynamic in terms of where our position in the world is relative to our ability to execute the national defense strategy, or our means to preserve our countries values, morals, etc. I find that a bit disturbing, I think, congress at least I know the staff is aware and are definitely trying to make the adjustments as best as they can, but they’re somewhat strapped by the traditional exchange that occurs between OSD and themselves and the services and trying to develop something that they can institute that their members would be able to focus on. I think this type of a dialogue with some meaningful attributes tied to data would definitely help them make changes. Can they make an overnight change? i’ll be possibly the only pessimist with a half glass. I don’t think you could do it overnight unless we had a significant emotional event, but I do trust in the nature of our society and our people that if we had a significant emotional event we’d move very quickly. The real question is, is it too late? so we do need to make these incremental changes and we need to make them informed so the members can still retain their needs and have their issues met, because they obviously it’s all a balance issue that industry can adjust and that, of course, the other two branches, legislative and executive, have the ability to manage and retain their abilities to oversee and execute their platform. I think there are things that we can do within the structure of the government to make change and perhaps it’s lesser than the perfect. It might be a grand compromise, but I think we’ve had a conversation here that talked about elements of change where we focus in on something that’s different — that we focus in on data to remove people from the equation. I mean we can go through all the little pieces commodity focus on national defense strategy not just on the top of the spend and we can improve the outcomes of the system I guess i’ll leave it right there.
Bill Greenwalt: Let me add one thing because there’s one part of this that we’re not focusing on that congress listens to a lot, and that’s what I’m going to call the oversight industrial complex. In other words, GAO, DoDIG, the audit agencies, the various stove pipes they’ve created to essentially be the checkers of the system to ensure compliance with the oversight and transparency process that’s been created over the last 60 years. This complex needs to come to a realization — and I think that has to come with a focus by senior management working with them — that they are now a part of the problem. The criteria that they are using is not the criteria that’s going to get us to deliver a capability faster to counter the Chinese or the Russians. The criteria is everything to them we have to understand what we’re evaluating. Time and outcomes have to be a greater criteria that the oversight complex measures to versus just compliance with process, because that’s what the appropriators are going to hear and that’s what the authorizer hear that DoD is not complying, therefore we better not give them any more flexibility.
Bob Daigle: Katharina I’m going to be far more pessimistic than you were. I think that there are pockets and areas generally new pockets in which a clean sheet approach might work, so if we think the joint artificial intelligence center, or SCO, or things like that, my experience over the last few years is that there are a few very thoughtful members, and some of the staff on the hill that is interested in systemic reform in order to make the process move more quickly so that we can get innovation across a validated death and into the hands of the warfighters more quickly. Where i’ve seen that general direction fail over and over again is on the Aegis and a couple of examples I will argue that there is the only thing more important from a capability development standpoint in the department of defense other than nuclear modernization is the JEDI cloud contract — the adoption of centralized cloud-based technology unlocks so much potential inside DoD and in creates a pathway to start to change how DoD manages data. It creates a platform for artificial intelligence et cetera. If nothing else it starts to create a pathway by which DoD can create a set of standards on which platforms should be built to allow for better data transfer better communications better cyber security et cetera. I can go on and on and on. The history of jedi has been a nightmare so far a capability that is widely accepted as being transformational in the broader society became to Bill’s point of political football inside the government we attempted another example is we attempted to put — Bill you and I talked all the time about acquisition speed — we need to be able to take risk and everyone around us was saying ‘yes fail fast, fail often, try it, it’s all good,’ except when it comes to unmanned naval vessels and hey we’re gonna put a lot of money against this and we’re gonna try it and we’re gonna fail fast and we’re gonna fail often. No. Go through the traditional acquisition system slow down do one prototype this year do another prototype seven years from now 15 years from now we might know whether or not this capability works, and I think that that’s where I’m more pessimistic is there’s a there’s a theory of reform but then on both sides of washington there is a the crushing oversight of the Aegis.
Bill Greenwalt: I just agree with that. The problem that you’re going to see here is some of these things are because of not the oversight industrial complex but the military-industrial complex that the lobbying for some of these things to keep these things from happening are great and how you can counter that? the only thing like I can really see that is frankly, if perhaps maybe some of our allies pick up the con and actually go off and have unmanned systems that are 20 times better than us and shame us into actually doing the right thing. But that may be the only alternative there.
Eric Lofgren: thanks bill. This is a great discussion we have a bunch of open questions here from the audience unfortunately I don’t think we’re going to be able to get to all of them but I wanted I wanted to get to one question here from Stephen Rodriguez. He basically says that nation states have struggled managing that necessary change from legacy systems and processes in order to create new systems, and then he points to the Wehrmacht and the PLA as being successful because they were able to more or less start from a blank slate. I would add the interwar navy in there where they had major reductions in capital ship building which allowed them to experiment through the general board and that pluralistic organization to move into carriers where you’d expect them not to. So I would just like for you guys to comment on the budget process in this interaction because it seems like for me that the PPBS is one of these institutions that’s locking in some of these legacy programs and it’s making it really hard to shift away from those things and find a new little wedge just to go do something new, and as Bob was saying, it can take a long time just to get spun up on something new whereas it felt like in the 40s and 50s we were able to pounce on new things relatively quickly right? rickover was able to build the very first working reactor for atomic power in less than five years and integrated it on the submarine in that five year time span. So can you guys just comment about the budget process and then how that how that interacts with shifting away from legacy systems and adopting new things?
Bill Greenwalt: well I’ll just say that mid-tier section 804 was designed based on two periods of US history. The first one was the experimentation that aircraft were under underwent in the 20s using almost other transaction-like contracting authorities, and then the that early period in the 50s. The idea was to make time and experimentation and start looking at blank slates as the way to go that’s the acquisition side of the house and it gets around some of the it’s a different pathway. The key thing is going to be is will the department ask for the type the chunks of money that are necessary for experimentation and have the arguments to the appropriators that yes, we need to experiment like we did and in the 20s and the 50s and achieve those new capabilities for the future. There have been a number of 804s started — number of mid tiers — that’s good. But they still have the two year, three year, lag time to get going because of getting into the budget process. Bob and I, when we were in the on the hill, tried to create a rapid prototyping fund to do this, and that just doesn’t exist right now, and that will be the key to see if we’re really serious — if we start funding that and actually put them toward these mid-tier type programs.
Eric Lofgren: one of the ideas is you can use these middle tier and then jump them back into the major capability acquisition, or do multiple of those for different subsystems, but it seems like it takes two or three years to line up money so you have to know ahead of time when you were going to be successful and that what that success would enable you to do. If you don’t know what you didn’t know two years ago, you’re not gonna have the money available. When you break it down, the SASC was talking about more prototyping of subsystems for ships, you’re still going you’re going to create these little miniature valleys of deaths between these various stages.
Bob Daigle: Eric think about that differently. So during our time period compare it to the traditional JCIDS leading into a program of record process. What we were able to do in a number of programs inside DoD is say, ‘hey, don’t start it as a program of record.’ you are going to have to wait two years because of the budget process no matter what you do because that’s the life cycle of the budget process. You have a choice. You can either go through a typical program of record process which forces you to start with the JCIDS process which forces you to spend three years coming up with the perfect set of requirements before you ever come to the table for the budget, or you can go right into a middle tier 804 program, skip the entire JCIDS process. You still have to wait for two years, but it’s not five years. So that was the real benefit of 804. Again didn’t have anything to do with the budget process, it had everything to do with the requirements process.
Katharina McFarland: there’s so much richness here. I would offer to you that a couple programs that didn’t have 804 utilize the same discussion — long range strike run bomber — the air force had invested in s&t for years before they decided to drop into the program of record process. I’m talking process for oversight, so naturally there’s tricks to the trade. The challenge is essentially the balance between public good and national security and that is when we need to, like MRAP, we accelerate fast, but when we need to have the people have their equities met inside of these programs, they slow it down so they can get access to them. Going back to the question that you asked, the whole history of the PLA for me, personally, is scary because they do have a fresh slate and they cleaned it out. What’s interesting is a different place is the Wehrmacht started by basically cleaning out the incumbents to include the process and just created their new slate. I don’t see either of those entities being able to be re-visioned here. It’s just not practical. But I do see the heightened nature of the environment, to our earlier conversation, setting us up at least in certain dimensions differently. If we recast our under secretary for R&E into someone who actually had the authorities by virtue of the original NDAA, which is to mature science and technology, it would help us carve off time because reaching that place where there is a valley of death, would be taken off the table because there would be a firmer relationship between those two entities. I think there’s ways that we can expedite some of this, and again, I would love to clean slate the process slice the practice so that people would actually recognize that the practice isn’t the problem it’s the people.
Bob Daigle: yeah, just to follow up on that Katharina real quick, because I completely agree with you and this goes all the way all the way back to Stephen’s original question. The PPBS process does that. You start off with asking what do we need to do, okay, that’s handed to us by the national defense strategy, then we answer the question, well how are we going to go about doing that, that’s the planning process, and then we turn around and we say, well what people, equipment, things do we need to do those things, that’s the programming process, and then you have the budgeting process that deals with OMB and Congress in the language in which they’re used to speaking. To Katharina’s point about practice, those three fundamental questions of what are we going to do how are we going to do it, and what do we need to do to accomplish that, we’re not going to be able to get away from those questions. There’s no organization in the world that doesn’t ask those questions of itself on a regular basis. To Katharina’s point, the problem that we’re grappling with is the way that we’ve chosen to implement the mechanisms to answer those questions are fundamentally cumbersome, way too complex, and way too detailed. If we could pull it all up a couple of levels and worry less about the individual specificity of those answers we could move a lot faster.
Eric Lofgren: i’d just like to jump in. Exactly what Bob was saying is right, and I would just recommend looking to some of the predecessors of the PPBS for how that was done in the 40s and 50s we had the R&D board and the munitions board where the leaders from those organizations were dual hatted, sitting on those boards jointly coming up with how they’re gonna — from an enterprise view — do programming, but that wasn’t tethered to the budget necessarily, even though it informed the budget, and then you also had — again i’ll bring it back to the interwar navy, read john Keuhn’s book, and there’s several other good books on it — how did the inter the inner war navy organized itself and that was another interesting model. So we’re coming up on time here. Bill, did you want to just round us out with a quick thought?
Bill Greenwalt: I think that you you’ve really just started to scratch the surface. I think Bob is absolutely right those are the questions need to be answered. I think we need to look at other alternatives and I think there are things out there, capital budgeting, or there there are ways — I know the Australians do things differently we may want to look at. Finally, I think in the in the interim we probably need to need some other accounts that are a little more flexible that deal with prototyping, that deal with technological maturation. I appreciate you having me here thanks.
Bob Daigle: Eric, Jerry keep up the good work.
Katharina McFarland: Yeah thank you so much folks!
Leave a Reply