Q & A – Admiral Mike Mullen: Armed with Analytics
Adm. Mike Mullen, Chairman of the Joint Chiefs of Staff, earned a master’s degree in operations research from the Naval Postgraduate School in Monterey, Calif. Photo by John Dean
Interview with Adm. Mike Mullen, whose master’s degree in operations research continues to serve him today as Chairman of the Joint Chiefs of Staff.
During the course of his illustrious career as a naval officer and now as Chairman officer and now as Chairman of
Joint Chiefs of Staff, Adm. Mike Mullen has been bombarded with military intelligence and information – some of it good, some of it not so good – but perhaps the worst advice he ever received came more than 25 years ago when he was anxious to pursue a master’s degree in operations research at the Naval Postgraduate School in Monterey, Calif.
At the time, Mullen had already achieved the rank of commander, and his assignment officer advised him not to take the tour in Monterey because it would “end his career.” Despite the advice, Mullen decided to go for the O.R. degree anyway, a decision he says continues to pay dividends to this day as he tackles myriad complex problems as the top uniformed advisor to the president.
Starting in the 1980s, Mullen commanded several ships at sea, including the guided-missile cruiser USS Yorktown.
Mullen’s last command at sea was as commander, U.S. Second Fleet, NATO Striking Fleet Atlantic. He also served as company officer and executive assistant to the Commandant of Midshipmen at the U.S. Naval Academy and as director of Surface Warfare and as Deputy Chief of Naval Operations for Resources, Requirements and Assessments on the Chief of Naval Operations’ staff. During the last decade, his four-star assignments included Commander, U.S. Naval Forces Europe, and Chief of Naval Operations.
President George W. Bush nominated – and the Senate confirmed – Adm. Mullen as Chairman of the Joint Chiefs of Staff in 2007. President Barack Obama re-nominated Mullen as Chairman in 2009, a move that was unanimously approved by the Senate.
Peter Horner, editor of Analytics and OR/MS Today magazines, and INFORMS Director of Communications Barry List interviewed Adm. Mullen on July 13 in The Joint Staff Flag Room adjacent to the chairman’s office at the Pentagon. The 30-minute, two-on-one session covered a range of topics centered on the military, operations research and Mullen’s career, from his grad school days at NPS to his take on the state of military intelligence coming out of Afghanistan today. Following are excerpts from the interview.
List: How did you come to study operations research?
I was inclined toward math in my undergraduate time [and] had always done fairly well in it … yet, it was 15 years between when I graduated from the Naval Academy and [when I] went to the Postgraduate School. I hadn’t cracked a book in 15 years.
I went over all the math I’d ever seen in my life in the first six weeks of prep training, even before the curriculum
started, and started there. What I found about the curriculum, and what I really did enjoy about it, was the application of it, the pragmatic side of it. Not every problem I solved or attempted to solve or missed on an exam was practical, but a lot of them were, so what I was attracted to was the underpinning of what I would call applied math in operations research, the modeling and simulation.
List: When you were doing your graduate work, ops research was used extensively in force-on-force analysis. What are the areas where ops researchers are making the greatest contribution to our nation’s defense today?
From my perspective, the operations research world and the problem set has gone from much broader force-on-force analysis to much more specific, I think pretty well-defined problems, although very, very difficult. And I just use that as an example. So where will the disturbance take place in the field, whether it’s under the water, on the water, on the land or in the air or in space, and then what do we do about it?
We don’t have systems, we don’t have enough money, to be able to create universal situational awareness with sensors. The world’s too big. An area that is really emerging in warfighting is space and cyberspace. So I would translate the Cold War force-on-force piece into the kind of both digitized world that moves so rapidly and moves us into areas to try to solve problems that seem almost intractable at this point in cyberspace, in space and even in the kind of warfare that we find ourselves in, which isn’t very often force-on-force anymore.
Horner: How did your O.R. education at NPS contribute to and help shape your career as a naval officer, and how does your O.R. background continue to serve you now, as Chairman of the Joint Chiefs of Staff?
I’ll take the second part first, which is, how does it help me now. One of the great things that the graduate education in O.R. taught me was how to think much more critically than I had before, and really, to frame a problem. Where that really helps me in this job is being able to ‘still frame’ a problem in my mind and to look at it differently than many of the people [who] bring those problems to me.
And then I have an opportunity to ask the right questions, or try to ask the right questions … It’s become a pretty natural part of how I do business … the ability to frame a problem – and this is the systems engineering, the applied math – that I try to bound a problem and then ask hard questions that … would push the system in a direction of an answer that clearly wasn’t forthcoming by the time it got to me.
I can apply this to areas that I’ve already talked about, but I can [also] apply it to just about every single difficult challenge, including policy challenges, that we have. What’s our policy long-term? … How do I think about policy with respect to China and the warfighting pieces that are a part of that? How do I think about it in my dealings with non-government organizations? How do I think about dealing with the challenges that I see for USAID … as we all need to be resource-constrained? At my level, decisions are oftentimes about apportioning resources – where is the best, or, to use an O.R. term – where is the optimum place to apply your resources? We need to do that.
We’ve had increasing budgets for almost the last decade. Those days are over, and we need to be able to think more critically, prioritize and optimize in ways that we haven’t been forced to do – and analyze, as well – in ways that we just haven’t been forced to do in the last 10 years because we’ve had plenty of resources. We need resources to fight these wars, but a whole different set of problems have surfaced that we hadn’t thought about before these wars started.
Horner: Many INFORMS members are current or retired military who would be interested in how your O.R. degree impacted your military career. You moved up rapidly …
I actually got in very senior. I was a commander [when I entered the NPS in Monterey]. I was very anxious to get a master’s degree. … In fact, I was specifically advised [by my assignment officer] against taking the tour in Monterey because it would end my career. Well, he didn’t quite have that right. …
I guess the advice I have with respect to that is we know ourselves pretty well. We certainly have to take the advice of those who assign us, but we also have some pretty good instincts about ourselves. And from that standpoint, it’s worked very well. …
The two years of study [at NPS] really disciplined me and it gave me an ability to think as I’ve described [and] practically apply it down the road. I’m a fairly pragmatic guy. It isn’t like I don’t like theory, but in the end, I’m a pretty pragmatic guy. The other thing that it taught me – Monterey – is, I didn’t want a Ph.D. in it. I didn’t want to know that much about any subject. People need to do that, and I have great admiration for them, but for my own practical application, a master’s was where I wanted to be.
Horner: O.R. is built on a foundation of mathematical modeling and fact-based, data-driven decision-making, yet U.S. presidents inevitably take into consideration the political ramifications of their crucial decisions. In your position as Chairman and given your O.R. background, how do you reconcile these two very different decision-making processes?
Well, I stay out of the politics. Again, part of it is my background. But I’m not unique in this regard. I see senior military officers all the time trying to make recommendations and come to their positions by understanding what the facts are. Not all of them, including myself, would do extensive modeling and simulation, although I push pretty hard in this job, specifically, to look at war-gaming.
And I don’t mean just war-gaming a military-against-military kind of thing, but in a very tough situation, how do we war-game out several steps into what policy might look like if we adopt a certain policy? Underpinning it, I think, with good models, good data – I think that is something we have to do more and more of so that the best decisions can be made.
My recommendation goes forward based on that. And then the president has to make – he’s the head politician in the country – that’s where politics kicks in. And I feel very comfortable with that: a best military advice approach, and then the politics kicks in, and I understand. This is a political town and I understand that, but I work very hard to not let the politics in any way, shape or form affect my recommendations to the president.
List: Turning to the influence of operations research in the military on the other sectors, I was a little surprised to see that you were on the cover of Fast Company. We expect you to be on the cover of the daily paper. Are there lessons that ops researchers and the military can teach the private world?
I’m sure there are. [During] another very brief period of time in my life I went to something called the Advanced Management program up at Harvard. It was an 11-week executive course focused on business. I was fascinated with this, in terms of how business looks at various problems, and I also found there to be a lot more overlap in problem-solving than I had realized.
That said, … many business leaders I met then were also very focused on fact-based, data-driven solutions. So I would offer – although they didn’t talk much about operations research – I would offer that there is plenty of opportunity on the business side for the kind of things that operations research gets at.
Some of [the] easily transferable skills or areas would be in the whole logistics world – the whole supply-chain piece, which, quite frankly, many businesses have done better than the military. So it’s almost like learning in the other direction, but clearly, the operations research skills and techniques would apply there, and would apply, again, in the optimization world. I don’t know that there is a – and I’ll probably get in trouble for saying this – but there is an area of study that focuses more on optimization than ops research, or can, when you apply it correctly. And business leaders I’ve talked to – they want optimum solutions, specifically.
Now they get into, very specifically, the whole P&L, bottom-line piece. Our bottom line in the military is different, but we both have bottom lines. So I think there is a great deal that could be used, specifically from the military perspective, in the business world if we knew more about each other. One of the reasons I pay attention to what’s going on in business and in economies is that security – or lack of security – is in great part driven by healthy economies, or lack thereof.
We’re fighting in places, for the last several years, where the economies weren’t very healthy. I’m worried about instability in places where the economy isn’t very healthy. … I guess my overall thought would be, we need to share a lot more than we do because I think there is a lot of overlap and a lot where we could help each other. And this is but one area, again, because I feel so comfortable and so strong about what operations research can do to underpin good decisions, which business leaders want to make just as well as military leaders want to make. It’s very, very applicable.
Horner: The military has been asked to take on more and more noncombat missions, such as disaster relief, infrastructure building and police training. How will these types of missions transform force structure to what some are now calling “resource structure,” going forward, and do you see a role for O.R. in that transformation?
If we go back to that, the signature event that really got my attention in the last several years was the tsunami in ’04. I was stationed overseas in Italy, on the other side of the world, when that occurred. But I marveled at that time at the ability of the United States, and in particular, the United States Navy and Marine Corps, to be able to build a city at sea where there was no infrastructure in a devastated country where the infrastructure was wiped out, and, in building that city, be able to provide food and water and medical help to bridge to a situation that was stable enough for the country to pick up that humanitarian assistance and disaster-relief role.
I would then extend that to our own country, the following year, when Katrina hit in July of ’05, and what we were able to do in the military. It was, as it oftentimes is … led by the Coast Guard that punches way above its weight. But it was all the military services during Katrina that came together. We’re very good at organizing.
We’re very good at what we call command-and-control. We’re very good in terms of providing structure or presence. And then extend that to Haiti, where all services went down there very rapidly, again led by the Coast Guard, but all the services provided a huge boost and a presence.
As I think about the problem, this is really a systems problem. It’s a flow problem. It’s a prioritization problem. So we were in Haiti, just to give you an example. There’s one airstrip in Haiti. There were 20 sorties a day. We needed to – and eventually got – the number of sorties per day upwards of 140, 150, 160 in a very small airport. That’s an optimization problem, and it’s a constrained problem.
I haven’t mentioned my favorite two subjects [at grad school at NPS], linear and non-linear programming, but I can remember the practicality of the problems that linear and non-linear programming [could address]. The optimization issue, the constraints issue, which any major problem, like disaster relief or humanitarian assistance has, you could apply that very specifically.
I actually brought up and wanted to know, in the Gulf – the oil leak we have right now, which has been such a disaster – are we, the military, and the Navy in particular, which has some incredible computing power – are we modeling this out? Are we using the kinds of power that we have in our technology to be able to understand where this is going to flow, how long it’s going to last? Have we done that kind of modeling and simulation work?
Horner: A report issued earlier this year by Maj. Gen. Michael Flynn maintained that senior leaders, from the president on down, “are not getting the right information to make decisions” [in regards to Afghanistan]. You’re certainly one of those top leaders. What’s your view of the report and are you getting the intel information that you need?
General Flynn is the senior intel officer for our forces in Afghanistan. He also was my J-2, so I know him well, and I agree with him completely.
In 1997, I deployed to the Persian Gulf with my carrier battle group, and I was assigned to liaison with the British aircraft carrier that was out there at the same time. There were four carriers in the Gulf at the time. And this was just prior to what could have been a war in Iraq in that timeframe – late ’97, early ’98.
I walked into the command information center – CIC – on this carrier – the British carrier – and there was this scope in the middle. And the scope in the middle, written in pencil, said, “This is knowledge.” So whatever information was on that scope was fact. And if I were to talk about – and I’ve talked about this a lot over the years – for many of us in many different fields, the intelligence and information field being a huge part of it – I need the right information, the right knowledge. I don’t need information; I’ve got tons of it – I need knowledge at the right time.
So how do I optimize flowing the right information to a decision-maker? It varies. If you’re a captain in the trenches out there in a village in Afghanistan, the information you need is obviously significantly different from the information I need. How do I have systems which support that, that pore through this glut of information that’s out there to get the president and that captain in the trenches the right information at the right time? I would call it knowledge, not information.
And the phenomena that has changed in warfare because of these wars is the ability to feed that information and intelligence into operations, conduct an operation, generate more intelligence and information and feed it – in very much an integrated, rapid way – that outpaces the enemy. So I think what Flynn laid out there, that’s how I think about the problem.