 Ladies and gentlemen, good morning. Welcome to the Center for Strategic International Studies. I think it's a wonderful sign that defense spending is still a topic of some interest, notwithstanding all of the reports of its imminent demise. Some of us in the think tank business have been saying for years that the time will come when the rise in defense spending will cease and the reductions will begin. And we all have known that sooner or later we would be right. We may be on the verge of that. I have a couple of administrative notes. I'd like to welcome those who have joined us on the web this morning as well. There should be a way for you to access the charts that we're going to be showing here this morning so that you can follow along at home. And also you can send us your questions when we get to the question and answer period. We'll put my email up as the last chart. That'll give you an opportunity to send us an email. And I'll have my blackberry out and be able to read them. That also gives me the opportunity to remind you here in the audience in the room to silence your cell phone and other devices so that when somebody calls you with an important message you won't get it. So if you would please do that as well. We've been here at CSIS running the Defense Industrial Initiatives Group now for the better part of a decade. And over that period of time we've been spending a great deal of effort trying to understand where the federal government spends its contract dollars. Many of you have tracked for years our annual reports on federal spending on services, professional services, industrial base as we call it. About a year ago we began to realize that we should go beyond the services business to look at all contract spending and also to focus on the key entities or the key components across the federal government. Obviously the Defense Department is the biggest chunk of this. Two-thirds of government contract spending is done by the Defense Department. Back in January we issued a preliminary version of this report for defense contract spending and that includes products and services across the board. This is an update from that report. We had a number of issues that came up and we've gotten new data. So in a sense what we're doing today is revisiting what we discussed with many of you back in January. But here's what's really new. Number one is we now only have fiscal year 10 spending data. At that point we only went up through FY09 and given that it was already 2011, obviously that was a bit of a lag. There's about six months between the end of the fiscal year and the time that the federal procurement data system has a pretty good set of data up. It doesn't get uploaded at all. It wants to get uploaded sort of one contract at a time and so we have to wait until we've got a cumulative balance that looks good enough that we think we can draw some conclusions from it. In addition, all of the numbers in this report are in FY10 constant dollars. So every single number that you got in January is now different. So throw that chart away and use the report here that we've issued today. We do also have a number of key issues that we've updated. One of the questions that we spent a lot of time looking at has to do with the expenditures in research and development and with expenditures on classified data. There were some disconnects between the overall appropriated dollars in DOD for R&D and the amount that we could track through contracts. Much of that discrepancy is related back to the fact that the federal procurement data system basically doesn't have classified contracts in it. And so we've done a good bit of work reconciling that. You'll see that in our charts in our discussion this morning. The second issue that came up that we began to wrestle with also starting in January is that there's a disconnect sometimes in the way the government classifies whether a contract is a product or a service and the way the company itself would look at those data. For instance, and we'll cover this when we get to some of the charts on there, but for instance, the company may categorize that contract as a service contract because in its mind it's logistic support and that's a service. The government may categorize it as a product because it's tied to a particular weapon system and they put it in that weapon system category. And the third is that we've done a good bit of additional work in breaking down each of the DOD components in the categories of products and services and R&D. So that's kind of what's changed between then and now. We're going to go through about 30 charts here this morning. We'll go through them pretty quickly. We'll then throw the floor open for questions and if there aren't any, we'll let you all go to lunch early. Guy? Yes, thanks David. Let's go to the next slide. Just a few quick comments in addition to what David said on the data that we use. All charts that you'll see here are primarily based on federal procurement data system numbers. They will all be in 2010, FY 2010 dollars and the most recent year will be 2010. The other thing to mention here is that supplementals are not classified separately. So we have no ability to distinguish between what came out of the base budget and what was a supplemental spend. Other than that, let's get to the actual charts. This chart really portrays two different sets of data. The bars that go start at the bottom and go up is total defense outlays any given year. The source of the data is OMB. And you can see the reductions during the 90s as we start and this is all in FY 10 constant dollars at about 450 billion in spending in 1990 dipping down to about 330 billion by the mid 90s and then climbing back up to 664 billion last year. Inside that bar, the lower part, the dark brown, is spending on contracts. The upper part of that bar is all non-contract spending. So that would basically be pay and benefits and in-house operations, if you will. The intriguing thing is the trend over the last three or four years in terms of total DoD contract spending and that is the numbers. They're kind of hard for you to see. You can see them in your paper copy a little bit better. From a peak of 390 billion dollars in FY08, a drop of about 7 billion in FY09 and then a drop of almost 20 billion in FY10. So you're beginning to see in terms of total contract dollars spending by DoD, a decline that actually started two years ago. When we sat here in late January, we said we had one year of data. It was only down a few billion. We didn't know if that was an anomaly or a trend. We can now see that it's not only a trend, it's accelerating. And it's accelerating at a time that outlays were still going up. Of course, part of the reason outlays are still going up, even though the budget is rather flat, is there is a lag time for a lot of these dollars in terms of when they actually spend. Just now, at this point, I'll give you a brief outline of how we'll break down the rest of the charts. The main areas will be, again, as we had in January, the product services and R&D breakdown. Overall, for DoD, for the particular military departments and then other DoD, we'll look at contract characteristics, specifically competition, levels of competition, the funding mechanisms that DoD uses for its contract awards, and the vehicles that it uses for those. And then the last few charts we'll look at the industrial base. For those of you who remember the January brief, we'll have the top 20 companies overall broken down into product services and R&D. And then we'll look at the breakdown of the DoD market into small, medium, and large companies doing sort of a comparison of 99 and 2009 to see what's been going on there. So our next chart is contract spending and DoD for products. And the line across the top is the percentage of total contract spending. So you can see back in the early 90s, it was a little bit over 55, 56% of total contract dollars were going towards products. By the mid-90s, it dipped down into the low 40s in terms of the percentage, because in fact we were in the procurement holiday, and the outsourcing and the reinventing government initiative had led to a lot more spending on services. The post-911, there was a little bit of a bump in percentage, but interestingly enough, and you'll see this later as we compared the charts of products and services, the increase in services contract spending actually outpaced the increase in product spending, except for a couple of years. And so products never got back up to that 56% level that they were at the end of the Cold War. You've got a little bit of a bump on the right in 2007 and 2008. That's kind of the MRAP effect, if you will, in terms of overall product dollars. And you can see here is where essentially almost 100% of your dollar decline in overall DoD contract spending comes in products, as we've dropped 30 billion between FY08 and FY10. Next is services, and this has been the fastest growing sector of the three in the last 21 years that we're looking at in this report. Just over 6% annual growth over 21 years, and again, R&D in products not in that same, not with that same growth rate. But what's interesting to see is that despite being relatively flat over the last two years in terms of the overall spend, the share of services has actually gone up to a level, probably its highest level in this 20-year period. Just in the sort of 45%, 47% of overall spend on DoD is now services despite, again, the beginnings of the downturn that we're beginning to see. We've mentioned this before, but let's just mention it again here because it happens a lot in services. The issue of how FPDS classifies the services, how some companies might classify services, this accounts for a little bit of differences in absolute numbers here. And it's worth, I think, saying again as we look more into how we can improve the data and how we improve policy that's based on the data. Right, and specifically, we believe that the services dollar value here is understated because the military departments classify as a product contract things that are in effect logistic services. And so the products dollars are overstated, the services dollars are understated. We don't know by how much. You'd have to go through contract by contract to figure that out. The Federal Procurement Data System, FPDS, classifies R&D as a service. And so whenever OMB or the Office of Federal Procurement Policy issues its own summaries of spending across the government and they show services dollars as high and products as lower, you'll know that R&D is classified as a service. Now, a lot of those R&D dollars actually deliver a product, as you well know. There's about $3 billion in R&D in the F-35 program. That's actually delivering airplanes. There's about $1 billion of R&D in the DDG-1000 a year per year. And so, you know, there are some numbers here. We break it out separately because actually for defense tracking R&D separately is pretty critical because that's your investment in the future. Here you'll see that it's been relatively flat, roughly around the $40 billion level for the last several years. But here is where I think it's important for us to emphasize to you this does not include classified contracts. And using other data that is outside of the Federal Procurement Data System, and we are in part beholden to the Center for Strategic and Budgetary Assessment for its reports on this, that $41 billion in R&D for FY10 in the Defense Department in R&D should have added to it the following numbers. There's roughly $20 billion in classified R&D contracts that does not show up in here. That gets you up to about $61 billion. There's probably something on the order of $11 or $12 or $13 billion that is spent internally. That is on government personnel and on government operations and buildings and so on. And that leaves us somewhere around $6 or $7 billion that we cannot track to one of those three categories. That's not inconsistent with the discrepancies that we see between appropriations data, agency data, and the Federal Procurement Data System. So by and large, this understates the total R&D spending by about 50%. It understates the R&D contract spending by about a third. This next chart is basically the last three charts as trend lines thrown together just to give you the effect of how each of these sectors shares has changed over time. What's interesting for us here is, I think, two things. One, if you consider the left-hand side of the chart, the early 90s, as where we witnessed our last big drawdown in defense spending, where, as David mentioned, products began dropping and we started spending more on services as we cut back on military personnel, that gives you some indication of potentially where the next drawdown might be heading that we started in about 2008. Again, products dropping significantly, services going up. Only this time, the lines might meet and the trends might continue with services beating products out in the next year or two. So if the last drawdown on the far left is anything to go by, the drawdown that we're beginning to experience on the far right, that might give us an indication of what this drawdown might look like. We were kind of thinking that when we got the FY10 row of this that we would have some visible indication of where this was going, because the FY09 data said, okay, it looks like one's going down and the other's going up in terms of products and services. The reality, though, is that they both veered enough that we have no clue what FY11's going to look like. And so, you know, we've got a big question mark sitting up there. I would also like to put in a plug here. All of you who were invited to this session today will be invited to an event that we're having on June the 8th. It's our Global Security Forum across all of CSIS, but one of the particularly interesting panels of that will be the lessons from the last drawdown and how they would apply to what we think is coming down the future. It'll be a very exciting and interesting panel as we go forward there. Now let's look at spending by component. So we've looked at products and services in R&D across DOD. So the top part here, the sort of yellow-orange-ish khaki colored, is the other DOD. That's everything that's not Army, Navy, Air Force, Military Department. The light blue is the Air Force, the dark blue is the Navy, and the green at the bottom is the Army. This is all updated from before. This is in FY10 data for, and you can see actually what a significant increase in the Army that really drove an awful lot of this total expenditure and also a very significant increase in the other DOD, which back in FY2000 was $23 billion. It tripled to $74 billion last year. This, of course, is everything from DLA, Missile Defense Agency, Special Operations Command, et cetera. Let's look at each one of these in turn. So for products, as David mentioned, the big growth really has been in the Army and in other DOD. Other DOD, again, primarily driven by DLA and MDA here, Defense Logistics Agency, Missile Defense Agency. What's interesting, again, is we might be getting some indication in the data from the last two years that the drawdown is beginning. FY10 alone, compared to FY09, sees a drop in about $3 billion in each of the services except for the Air Force. So, yeah, I mean, products, again, in the last drawdown was the area hardest, and it might be looking like that's happening again. I should note that at the bottom of either of these charts, we give you the combined average annual growth rate for each of these. Of course, that's taken over the whole 21-year period, and really there's sort of two separate decades in that period. There's the 90s, where things were sort of going down, and then there's the aughts where everything was sort of going up. In our report, we've got more detail on each of these, and as we continue down this road in the future, I think we'll continue to expand that out. This is services spending by components, and again, over the last decade, what you see here is that the Army has basically tripled. The other DOD component has basically tripled. The Navy and the Air Force have gone up by roughly 50%. So the distribution of growth is dramatically different between the ground forces and the support for those ground forces that come out of other DOD accounts, and the sea and air forces, which have gone up by substantially smaller percentage. On R&D, and again, I would caution you to remember David's earlier comment about these numbers being quite low in terms of overall spending. We know that DOD spends about 80 billion a year last year on RDT&E, so the total here of 43% is about 50% of that. What's interesting here is that of the DOD components with spend on R&D, the one that's underrepresented most here as a result of classified programs not being counted in FPDS is the Air Force. About 50% of the Air Force's RDT&E dollars are in classified programs, so that light blue second bar from the top in each of these years is about 50% higher in reality. We just can't capture those classified programs, those classified R&D programs. The same is true for the other DOD category, which is where you have a lot of your other classified intelligence contracts. So the 6 billion in R&D reflected here for other DOD, there's almost 5 billion dollars in classified contracts in that same category, so it's understated by about 50%. Now let's look at the Army. This is contract spending with the Army. The bottom, the bar graph at the bottom shows you the dollar value, and you can see there's been a bit of a decline since the peak in 08. The top line shows you the percentage, and while there was a little dip between 08 and 09, that's flattened out. Essentially what's happening here, and we'll be very interested to watch this trend as we get the FY11 data, is that the Army's dollars are declining, but their percentage share of DOD appears to be stabilizing. And it's stabilizing, of course, at a level that's roughly double what it was in the 1990s. Whether this is sustainable over the long run, I kind of doubt it, but whether it is sustained up and through the time that we actually have substantial drawdown from Afghanistan, which is sometime between now and 2014, that's a much more likely outcome. So I think the future picture of the Army here is really two different questions. One is the next year or two. The other is the longer term 2014 and beyond. Let's look now at the other services. Looking at the Navy, probably an unnecessary comment for our audience here today, but I'll make it anyway that Navy's budget includes the United States Marine Corps budget. So that's factored in here. We, again, have no ability to break that out because FPDS doesn't categorize the Marine Corps differently. And so despite the Navy's budget, including the Marine Corps budget, and we know the Marine Corps have had an increase over the last 10 years at least, given the operational tempo that they found themselves in, the Navy's share of overall DOD contract spending has dropped from about 37% in 1990 to about 24% in 2010. Well, we'll be interested to see how that trend continues, especially given that in the next few years we've got LCS coming up, we've got the second Virginia class submarine coming up. We'll probably drive this trend in a little different direction than it's been going the last few years. And with Air Force, you see the similar decline in terms of percentage from its historical average in the 30s down to actually below 20%. It would be above 20% if you had classified contracts in here that would add another $12 billion basically to the Air Force's total. But what's intriguing to us is that that decline, has been pretty steady now for about 15 years, seems to have leveled off in terms of percentage over the last couple of years. And as far as other DOD components go, again, we've mentioned this already, but the share in terms of their overall DOD spending has almost tripled. Well, sorry, the value has actually tripled in the last 10 years, which is what has been driving up the trend line on the top of this chart. Again driven primarily by DLA and MDA in the last 10 years. This is a chart that combines Army, Navy, Air Force, and other DOD together. At the time we briefed this back in January, I noted that there was a signature event here that occurred when the Air Force dropped below other DOD back in 2008. An event worthy of a press release, except we actually couldn't notice it until two years later when the data become available. Now, if you actually adjusted this to add back in the classified contracts, the effect would still be the same, but the difference between the other DOD line and the Air Force line would be so small that on this chart with the scale that's here they would actually look pretty much like the same line. Nonetheless, it's a dramatic redistribution of the way the DOD spends its money. The fact that, in fact, the military departments or that no military department dominates to the extent it did in the past and the other. Now, once we actually reduce our fuel consumption, and you tell me when that will be, it will show up in one of these years out here, others are going to drop back down because roughly $12-$15 billion a year of those contracts is for fuel, and at some point that goes down. We'll look now at the breakdown of product services in R&D for each of the departments or each of the DOD components starting with the Army. Overall, services has been the largest, has seen the largest growth in the Army over the 21-year period. Of course, if you just look at the last 10 years, it's been about even between products and services, given operations in Iraq, Afghanistan, other. One other thing to note about the Army's R&D spending is that there's actually very little classified information or classified programs there. So this number has actually fairly real in terms of actual spending overall. Navy spending by category R&D has been relatively flat over the last six or eight years. Procurement dollars or products are up considerably because of growth in ship deliveries in the aughts. Services dollars relatively constant. And for the Air Force, the service category has actually, again, been the one with high growth here. Almost doubled from 1990 to 2010. One other thing to note is that though products have been relatively flat and seen very little growth in this 20-year time frame, we've got some big ones coming up in the next few years, like tanker, joint strike fighter, next-generation bomber maybe. So that lower bar, the dark blue bar, might be the next one to see significant growth in the coming years. And here we have the spending product services R&D for the other DOD category. Keep in mind, of course, there's about five billion in classified R&D that doesn't show up in this chart. Now we're going to look a bit at a different way of slicing it. We're going to step aside from the DOD components, Army, Navy, Air Force, other DOD, and take a look at the different types of contracts themselves. And the first contract characteristic we'll look at is competition. Interestingly, almost consistently over, and by the way, for these categories, we're looking at the last 10 years or 11 years, as opposed to 2021 years, just because that's the period for which we have what we feel is data that we can be confident in. Earlier years just don't have the level of granularity of the data that we'd like to see in order to be confident. So over this last 10, 11-year period, what's been interesting is that the split between competed and uncompeted contracts has been about 50-50. And I say this because for us, even though the middle bar here of competition with single offer is technically a competed contract, the fact that there was only one offer for us makes it a no-competition type of contract. And so if you add the pinkish-red bar at the top to the middle bar, the purple bar in the middle, that for us really is the no-competition category, where either there really was no competition, the contract wasn't competed, or it was, and the government thought it was going to receive multiple offers. It actually only received one. Sadly for us, this was the category, the competition with single offer, this was the category which saw the highest level of growth in these last 10, 11 years. Here on this chart, and for those of you following at home, this is chart 22 on your website, you've got basically 20% per year increase in competition with a single offer. One of the things that we have to do that we have not done yet is break this data down by product services and R&D as well. Because one of the most important factors in our minds is that growth in competition with a single offer in the category of products, where it would make sense in many cases that you would like to have competition and you want to have competitive categories, but there's just not a lot of bidders out there, or is it in the services arena where, at least in theory, there should be multiple offers for almost every service that's put out there. We're going to have to look at the data and break that down. We also looked by what we call funding mechanism, that is the type of fund structure of the contract. This is chart 23. The bottom, about $240 billion in DOD contract spending in FY10 is fixed price contracts, and that has had about a 10% per year growth rate. The next part of each bar graph, the cost reimbursable, $104 billion in FY10 has grown at an average rate of about 7%. That means that we've had a higher emphasis on fixed price contracting, and that that emphasis has, in fact, yielded results. The next category up, the kind of oranges category there, is time and materials. It's had a pretty good growth rate, about 9% per year, but there's only about $15 billion. But the biggest factor that's a change from when we were looking at this before with FY09 data is our concern over the category called combination, which was basically in the FPDS guidance was where you had a mix of fixed price contract line item numbers and deliverables and cost-based or time and materials-based, and we couldn't therefore distinguish what type of contract funding mechanism you had in place. Our concern was that it had had such an enormous growth rate going from about $4 billion in FY06 to $45 billion in FY09. Somebody woke up because it's down to only $6 billion in FY10. Now, we have no proof that this is a direct result of CSIS's analysis, but we would love to take credit for this result. So we're going to go out and we eagerly await to see what happens in FY11 and whether or not this continues. But obviously, from the point of view of, if OMB and the Office of Federal Procurement Policy is attempting to track compliance with the guidance that says increase the use of fixed price contracts, increase the use of competition, and they are using as their guidance memory in the state the federal procurement data system as their way of measuring, this is major progress, because actually now you've got some correlation between the data you're trying to judge and the data you're managing by. We're going to turn to the... We're now going to turn to the... You skipped one? Oh, I skipped a chart. That's because I didn't turn over my page. Spending by contract vehicle. One of the things that we started tracking with Vigor in our services contracts data is the use of definitive contracts, indefinite delivery vehicles, purchase orders, single-award indefinite and multiple-award indefinite. And we've broken down the DOD contract spending for the last 12 years by that category here. What this chart shows you is in fact a pretty remarkable stability by type of vehicle, notwithstanding the fact that there's been much more of an emphasis on indefinite delivery vehicles over that time. Again, what we need to do is to break this down by product services and R&D, because we think there's some substantial differences across those categories here. And in the next update of this report, when we're rolling the FY11 data next year, we'll certainly have that in play. Now let's turn to the industry rankings. This chart shows you the top 20, and rather than track year by year, as all the other charts do, we just used 99 in the start point and 2009. We haven't updated for 2010 yet because the granularity of the data available from the Federal Procurement Data System didn't arrive in time for us to be able to break it down. We'll probably issue an FY10 update of this in advance of next year's reported to be available on our website. I think it's important that we point out that these numbers are not necessarily equal to what the firms themselves show in their data. This is what we can track against the Federal Procurement Data System for DOD. There's roughly, for instance, $5 billion of DOD contract spending that goes through a GSA contract. Many of the firms who know that it's DOD money coming to them would show this as defense revenue. We don't capture it that way because the way we rate the Federal Procurement Data System number, those show as GSA dollars. So there's going to be disconnects between here. In addition, of course, agencies, firms often capture subcontract dollars as show that as defense revenue as well. And here, all we have is prime dollars. And so we're allocating 100% of a prime contract to whoever is the prime contract awardee, even though they may actually only recognize from a revenue point of view 30 or 40 or 50 or 60% of those dollars. There are no good data yet available for breaking this down by subcontractor. Companies know that, the government does not. And so there's potentially some substantial disconnects here. Nonetheless, you won't actually be surprised by the names up here. The top five, Lockheed, Boeing, Northrop Grumman, General Dynamics, and Raytheon are essentially the same top five as in 1999. There's substantial change in the next 15, if you will. It's also worth noting that these numbers in this ranking does not reflect the separation of Huntington Ingalls Industries incorporated from Northrop Grumman. That'll be a pretty dramatic change when we get to FY 11 and 12 data in terms of the rankings here. HII essentially will show up as somewhere in that 5 to 10, 6 to 10 category there, and Northrop will probably drop down to number 5, although there's obviously a lot of other things that can happen between now and the time that that comes up. So we're just showing the government data here. This is what the rankings are. Two interesting things that we'll point out. Let's look at them now by products, services, and R&D. So for products, looking at these two years, 1999, 2009, you really see the effects of operations in Iraq and Afghanistan. The new companies in the top 20 list in 2009 are really the companies that are either, A, in the business of selling fuel to the Department of Defense, Shell Oil, BP, Bahrain Petroleum, sort of on the lower end of those top 20, but newcomers since 1999, or B, it's the companies that specialize in ground systems. That's your BAE systems, that's your McAndrews and Forbes holdings who own AM General, it's Oshkosh, and all these companies just didn't exist or weren't on the, sorry, didn't exist, weren't on the 1999 list at all. Of course, some of that, for the land systems companies, some of that growth was achieved via mergers and acquisitions, and in fact, that's true for the other companies as well. Northrop Grumman bought Lytton, bought Avondale, and bought Newport News in the 1999-2001 timeframe doing wonders for its placing, taking it from six to five, but if you look at the overall contract value, that's a significant growth. Same for BAE systems, United Defense and Stuart Stevenson were bought, and those helped propel it to the position that it was in in 2009. Looking at services, and particularly focusing on the 2009 category, the single most dramatic thing that leaps out at you is that when the Defense Department says the fastest growing part of the budget is healthcare, it's clearly reflected in these numbers. You have HealthNet, you have TriWest Healthcare, if you actually added up the services dollars in DoD healthcare and combined them into a single category, they'd be either number one or number two on this chart. That's where the growth has occurred immensely over time. The other thing that you see, of course, is there is a substantial, in addition to healthcare, substantial changes in the elements of the companies that make up this category. We think this will probably be the most volatile area going forward out of these three. And by comparison, in R&D, very, very little change, especially if you consider the share of the top five and the top 20 of the overall DoD, R&D, T&E spending, again, this is just the unclassified spending, bear in mind. But the top five account for about 50% of that and the top 20 account for about 75% of that. So you really have a relatively small number of players taking the sizable chunk of the DoD, R&D, T&E budget. Another interesting thing to note on this list is the primacy of non-private sector elements. In 1999, there were four entities that are either FFRDCs or UARCs, University Affiliated Research Centers, or non-profit research organizations. And in 2009, there were five, MIT, Aerospace Core, APL, MITRE, and BATEL. And all of these have significant chunks of R&D contracts and are not part of really the, sort of what we would consider the traditional or the classic defense industrial base. The final set of charts that we'll look at breaks our industry down into three categories. We call them large, medium, and small. This is not actually by size of company. It's by the total amount of dollars, contract dollars that they get from DoD annually, if you will. So large contractors are those that receive $3 billion or more in FY $10 or... No, it's annual revenue. ...annual revenue, all right. And medium or between above the small business set-aside threshold and below $3 billion, and small, of course, is those that meet the government's tests of small business. What the chart shows you is the difference in terms of the percentage and in terms of the dollars between 1999 and 2009. The top bar, obviously, the shorter bar is 1999. Fewer dollars being spent. The larger bar is the, the bottom bar is, and the larger bar is FY09. And then you also have, essentially, an ability to compare the percentages. What you see is that overall in DoD contract spending, large contractors have grown at an annual rate in terms of their revenue of 10.8%, medium at 7%, and small contractors at 9.6%. Those who are in this medium category think of this as being squeezed. Now, obviously, growing at a rate of 7% per year is not exactly squeezed. If you were on Jenny Craig, that would not be called a success. But nonetheless, they're not keeping up with the same growth rate. Let's look then at each of these, at this by category, products and services, et cetera. So for products, really, the trend that we saw in the previous chart is really highlighted the most. The share of the overall market for large companies between 1999 and 2009 grew from 57% to 62% at an annual rate of 11.4%. And that's really been a sort of the driver of the mid-tier squeeze, so to speak. And it's been a squeeze from the top and not from the bottom. The small companies, and here the good news is that the small business set-aside seem to be working. And almost all the areas, the small businesses have maintained their market share over that 10-year period. And so the growth of the large companies has really come at the expense of the market share of the medium-sized company. In the services arena, this effect is less so. Where you have medium contractors, it's probably been the least squeezing here. Nonetheless, medium-sized contractors in 1999 had roughly 30% more dollar value business in services contracts from DoD than large contractors did by FY09. Large contractors were up about 8% over medium-sized contractors. Small businesses increased at a 9.3% per year rate in that same time period in services. It's actually interesting, I'm going to go back one. Small businesses' annual growth rate in products is actually higher, 11.6% per year growth rate than it is in services, 9.3%. We didn't expect that. We actually expected that small businesses would have a higher growth rate over that period of time in the services business than in the products business. And this is one where we don't believe this is because of DoD categorizing services as a product because a lot of small businesses are not in the pure weapons system service arena. Sorry, just another quick comment on the services slide. The share of the overall market going to small companies is actually largest in services, where small companies account for about 23% of the market in products and R&D. They're at about 13%. So really a significant difference in the service sector. R&D, again, the story is sort of the lion's share of the market being held by the large companies. That share has grown from 65% in 1999 to 76% in 2009. Again, those top five, top 20 companies that we saw in the previous chart dominating this market. And again, the squeeze from the top affecting the medium-sized companies that fell from 22% of the market share in R&D to about 14% with the small companies being relatively protected and remaining relatively stable at 12%, 13% over this 10-year period. And keep in mind, this is where 50% of your contract dollars aren't reflected because they're classified programs. You know, anecdotal history says a lot of those classified contracts are going to medium-sized companies, but it doesn't take very many big satellites to sort of offset that in terms of the distribution of dollars. So from that. That essentially ends our briefing. You've got to copy this book. There's a company in text on each of the charts in the book. There may be places where our words here this morning are different than the words in the text. In every one of those cases, the text is right and we're not. And where the text is wrong, it will later be updated so that it will be right. This is kind of a living process for us. We do maintain these documents on our website. As time goes by, we get better data. We update them. And so the hard copy that you have will have a half-life. The document on the website will have a much more robust constancy to it. And the beauty of the federal procurement data system is not only does it update to add next year when it gets here. It's constantly going back and updating the past as well so that there's a continual change in the reality which it reflects. We often don't understand what those changes mean, but we track them diligently nonetheless because we have to live on the assumption that the government's entering its data accurately, even though we know that's not true. But where we find discrepancies and disconnects, we take some precision in pointing that out and making sure that the FPDS custodians are aware of that. And so over time it constantly continues to improve. But the highest amount of change year over year that we've seen in terms of correcting the data back in FPDS has been about $10 billion, which depending on where you are... And that's in the aggregate, not individual $10 billion entries, as you will. Right. That's the one where you stand. That's either a lot of money or not a lot. Right. Well, out of $400 billion, that's probably... Governor White, out of almost $600 billion in contract spending, that doesn't probably affect the quality of the analysis. I'm going to open up the floor now for questions and comments. Those who are on the web should have access to this by email. If you have a question, you can send it by email and I'll pick it up. We've got microphones in the audience, so I'll recognize you. And then you wait for the mic. If you would identify yourself and your affiliation and then ask your question or make your comment. We'll start up here and then we'll go to... Howie Lin with Floor Corporation. Could you go to the Army contracts spending by category slide? In the book it's page 20, but I think it was different numbering on the slides. Yeah. Okay. So you made comment earlier about the differentiation between services and products. Does log cap fit under the services? I assume so, but that's the question. I think all of the log cap contract dollars that we found here are in the services category. Okay. So then looking at the far right column there for fiscal year 10, do you have any idea of the $11 so far that are being spent? No. That I'm afraid there's been almost no data entered yet for 11 on the federal procurement data system. Typically there are data starting to come in by now. Our assumption is that the impact of seven continuing resolutions has decreased the enthusiasm of the government to put the data in until they actually knew what their total dollars were and where they were going to go. Let me go to the back and then come up to the front here. Bob Bill Howard with Boeing. David, in your last chart you have here where you have products by contract or size and this is prime contracts only so the amount of subcontracting that the primes do to the mid tiers and the mid tiers to the small companies is not reflected, right? That's correct. Yeah, I think the, have you tried to correlate like the size of the firm in terms of employees as another way of kind of getting it, how much work, I realize getting supplier dollars to make an adjustment there on how much money is flowing to mid and small tier from the larger companies is hard to do but certainly the size of the firm itself in terms of number of employees would give you a feel for how much of the money would flow. Has any kind of correlation been done on that on the size of the firm versus the amount of money that couldn't be done by their own employees? We haven't looked at that correlation Bob and I think there's a parallel correlation or parallel data element that comes into play there that would have to accommodate as well and that is that we define size of firm by total revenue not by government contracts revenue so those firms that are purely, totally dependent on the federal government for revenue would have a different distribution obviously than those firms that have outside commercial or non-government business. I think the other element of that is that clearly from the point of view of small business dollars we're only capturing here small business prime dollars and as we know from being in that business a lot of those small business prime dollars flow to a large company as a subcontractor. Similarly and in reverse a lot of large business prime contract dollars end up flowing to small businesses because every DOD prime contractor and every government prime contract has a small business set aside a threshold that they're trying to achieve so being able to sort this out and what kind of changes are has proved to be too daunting a task analytically for us to do. There is indication that the federal government is beginning to collect subcontractor data and array subcontractor data. It has not yet been made publicly available to us we have no idea what the fidelity of those data will be when they come but we intend to fully exploit them as soon as they become available because I think those are both critical contracting questions they're critical questions in terms of execution and they're critical questions in terms of policy because ultimately we've got a whole set of policies out there that believe they're driving us towards a set of objectives and we have no idea really whether they're achieving them or not. I think there's some questions down in front here. Let's do the right hand side first and my right and then my left yours. Hi Michelle Jomresco from Bloomberg I wanted to ask about the funding mechanism category. Specifically the combination and other category it's my understanding that OMB and DOD agreed to eliminate that category for 2010 and yours following and that's why you see that residual is that part of your analysis? We're talking about I'm trying to look at my chart number here I'm calling it chart 23 although I don't actually see a chart up there and your question Michelle is in the category which we call unlabeled. No the combination category it's sometimes called combination and other combined but you have combination and I understand the residual for 2010 is 6 billion Right, for reasons of OMB and DOD deciding to eliminate that category for future contracts is that your analysis? That's our assumption and we're looking forward to see whether FY11 actually reflects that or not and of course we applaud that because actually we think that the use of that category is both misleading and eliminates the ability to do analysis. You just had no transparency on how much of the contract was fixed price versus cost plus Having said that, let's look at what the reality of the result is you do have complex contracts which have fixed price cleanse and cost reimbursable cleanse and typically the rule is go with the one that has the majority and that's the category you put it in that doesn't eliminate the fact that you could have a $2 billion contract 55% of which is fixed price and 45% of which is cost reimbursable it will now show up as fixed price since there's no longer a category called combination or it may now show up as fixed price depending on the judgment exercised by the person entering the data so there's still some analytical challenges underlying that and one needs to be careful about drawing conclusions of trends especially you look at the difference year to year here it's not that dramatic it's really hard to look at these data and say clearly the drive to increase the use of fixed price contracts is a successful policy effort you really can't be sure based on the data here because the trends are not dramatic enough our conclusion in many ways is that there's a level of stability here that's probably driven by something other than policy guidance ultimately what do we use fixed price contracts for when the requirements are stable enough and the dimensions of what it is you need to bid against or well known enough it makes sense to use a fixed price contract the government knows what it's trying to buy the vendors know what it is they're trying to sell and how they line up nobody benefits from forcing use of a fixed price contract where either the requirements are too ambiguous or the performance is too hard to measure and you're much better off in a cost reimbursable basis so ultimately the relationship between the data and the policy isn't just one way it's a two-way feedback and it's going to be very hard to tell how that's going that's a longer answer than you wanted I know but I think we would love to have more precision than we've got and it's not coming yet next question here Dave Eisenberg, PMSE observer with regard to top 20 contractors for services which is page 29 of your report you raise an interesting question or caveat regarding the way policy makers categorize the data which is are they throwing in O&M as service contracts you raise that question I was wondering I understand you can't speak definitively but as of right now I mean what is your view or answer as to what policy makers are doing because if what you write is correct it would seem that they are undervaluing or minimizing particularly things like log cap within services I think we're approaching a circumstance where within DOD and across the federal government there's an increasing awareness that entering data correctly into the federal procurement data system is actually important and needs to be paid attention to we have been there before there was a time in the mid to late 90s where there was a very serious attempt inside DOD to actually match the military department's own contract data with what was in the federal procurement data system my memory of watching that exercise is it took a huge amount of time and it produced a result that was barely discernible in terms of improved data results but that was then and this is now and I think there is a greater emphasis nonetheless I will tell you that almost every entry we find visible inconsistencies fields left blank that should be entered fields where the category that's entered is not one of the categories available based on the guidance of what you can put in that field the amount of confusion I don't know if you have an estimate of percentage of accuracy of data entry but let's say that I'm a college professor and I'm grading on the scale of 90 to 100 is an A, 80 to 90 is a B at the overall contract level we're probably at a B at the total number of data entries we're probably at a C but I think we're getting better and I think that the impact of the discrepancies is diminishing and I think that over time we all benefit from that now ultimately what you'd like is single entry data I won't live long enough for us to get to that point but boy what a great result that would be both for management and obviously for analysis you have anything to add to that guy? I think part of it is just that the deeper into the data you go the more you find both in terms of how you can use it and how it's being misused or mis-entered and a lot of that is just really it's just human error you're entering into this database millions of rows a year and so there's got to be some accounting for just human error but I think the real test is that when we do find these issues where there are data entry issues that impact policy that really impact policy substantially that impact trends in the database as a whole and can raise them I think that's us doing our job and hopefully there's somebody here listening and doing something about it and on major disconnects we'll actually go to the government forms the DD350 contract forms and do a comparison now here we make a heroic assumption we assume the DD350 is correct and FPDS is wrong that's probably not 100% of the case but I think that's more likely the reality because that's somebody who's actually putting their signature on the line for contract purposes so that's kind of our assumption I think we have a question over here one in the front here and then we'll call see if we've got any from coming on the web Cameron Luthi Booz Allen Hamilton if we could go back to the defense contract spending by funding mechanism Mike, a question is more it's only breaking out between combination time and materials cost or reimbursement fixed price couple more back I think it looks like our back button has been defunded it's alright the question is about what's not shown right, we actually merge a number of categories together here in fact if you look at our core analysis in the services contract we actually have 11 different categories of what we call funding mechanism here and we break them down and in the services contract business those tend to make more sense I think we merge them here by and large just for ease of visualization we've got data that breaks it down by a much more definitive level of category and I'd certainly be happy to talk with you about that at some point and to that point broken out between services R&D and products that part is research still to be done okay thank you and here in the Larry Ferreiro from the Defense Acquisition University well the back button worked but on chart 4.1 4-1 spending by competition can you clarify the difference between the awards with the single offer which was labeled no competition that's the first question there is a field in the federal procurement data system of was this contract competed or not and it's pretty much a binary entry it's either yes or no of course there's a third option which is to leave it blank which does also happen that's our category called unlabeled that's not in the instructions by the way but some people seem to come back and that part is declining there's another category which lists a number of offers that were received and so this is one that basically fall in the category of the data interiors labeled this as a competitive contract so in our own sampling we've determined that they're generally when they say yes on competition it generally was competed there were multiple offers solicited but only one received and that tends to fall into that category and my second question is whether the data has enough fidelity or expanse that you could determine what some of the differences are in the cost performance of contracts awarded where there were multiple competitors multiple offers, single offers and sole source what a great question Larry we've actually had a research project underway for two years now attempting to tie contract to performance and I would say that we are we have made some progress in the answer to the question of how we're going to measure performance in terms of this but we're a long ways away from being able to achieve that I think it's a very valid question I think it's probably a question that's different in the services business than it is in the hardware business we've focused on the hardware side of this on the major defense acquisition programs as our touchstone there and unless there's been a breakthrough that I'm not yet aware of we're a long ways away from being able to correlate performance in terms of results of product delivered with the contract type and the funding mechanism just a very technical note on the competition slide the unlabeled category includes both what was originally unlabeled in FPDS but also where we found an error in the labeling or in the categorizing so for example if a contract was defined as no competition but there were five bids we will mark that as unlabeled because it's clearly an error somewhere either it was competed and there were multiple bids or it wasn't and there was just one so that category includes both FPDS entry errors as well as just plain unlabeled originally in FPDS was there another question there's one in the back in front here Dave Drabkin Northrop Grumman did I understand you to say that you agree that if a competition is labeled competition and they receive only one bid you're not counting that as a competitive bid we are counting as a competitive bid we separate the category out as competitive bid with a single offer and the reason would be we find that it's interesting to see the difference between competitive bids with multiple offers and competitive bid with single offers much of the data that shows the benefit of competition if you break it down by competition with single offer versus competition by multiple offers the benefits from competition clearly show up better in terms of reduced price and impact on the dollar spent where you have multiple offers that's the historic analysis shows that I think that at one point we actually had a different name for this category and we moved it to what we call it now because we think it illustrates better it's clearly though I think an area that needs further analysis because there are multiple good reasons why you would have a competitive solicitation and a single offer and a lot of those reasons are both consistent with policy and desirable in terms of the contracting entity there are other places where in fact people were probably caught by surprise now we all know that the flip side of this is pretty easy because all of you who are in the contracting business know that sooner or later the contracting officer is going to say Joe I really need you to bid on this and you'll say I don't want to and they'll say I need you to we have to have competition here I'm sure that's not happening to anybody in this audience specifically but it does occur and so you can mask a lot of things here and woe onto you if you happen to not want to win and you end up winning anyway so then you actually have to perform that did happen to me a couple of times so I think there's still a fair amount of work to be done here I think the thing though that we find as we look at this data is it really calls into question one of the fundamental political premises that's at work in the field of competition and the political premise is that if we only worked harder at it we could have a lot more competition than we have today that somehow the government's falling short and not competing enough and I think our analysis shows that by and large there's not a lot of additional room available for competition that isn't already occurring and we think that's an important policy point here you can't double the amount of dollars awarded under competitive contracts it's just not physically possible and there's a point at which you may have a level of diminishing returns from a policy point of view and it will be useful for more analysis to discover where to put that emphasis and where you have the potential for the best payoff and I think that's a different response of what you got but this would be one of the areas I would say would lend itself most visibly to that kind of effort would be areas where you think you're going to receive competition from the government side and there's only a single offer those would be the kinds of things where you'd really like to ask yourself was this a surprise to those who put out the solicitation that they only got one offer back and if so why and what can they do about it so that's why I think more analysis is needed in that area especially since this is the fastest growing area and it's to the tune of almost 55 billion dollars a year now in DOD alone right in DOD alone now we haven't rolled out dollars or similar reports yet for other agencies we are in the final stages of wrapping up a report similar to this for the Department of Homeland Security obviously the total dollars don't come anywhere close to the 367 billion dollars that DOD spends under contract each year but many of the dynamics appear to be very similar and we'll be rolling that report out sometime in the next few weeks and I think many of you will be interested in the results of that plan for similar reports over the next year on the other agencies with significant contract dollar expenditures NASA the Department of Energy GSA and I think it will be interesting to see the degree to which the kinds of dynamics that we see operating in DOD are in fact present for other agencies as well or whether there's some unique attributes to the Defense Department that won't show up with the other agencies we're looking forward to seeing the results of that and I think being in this job where you get excited about whether or not DHS has the same percentage of competition with single offers as DOD does and then you wonder why or why not I mean it's an exciting life if you had told me when I was a kid that I would grow up being excited about this I would have shot you but none the less we do find it to be quite interesting and I'm sorry we can't bring the charts but you've got your books so you can go at it here another question in the front here nothing on the web nothing on the web we may not even have anybody on the web Paul Sullivan Usec just to pursue the performance question is there anything in the database that gives contract closeout prices it's a simplistic measure of performance granted multiple years growth changes in scope etc but over the law of large numbers could you say comparing the contract award prices contract closeout prices measure of performance I don't think we've looked at that and I think that's actually a very interesting set of data that would be worth looking at Paul sorry for that and of course you also have the lag time of getting it closed out but yeah we'll have our FY09 closeouts in 2016 but it's still that's a very worthy set of data to look at there a quick question here back to the chart on page 4-1 the characteristics the pink for no competition maybe I missed it why is that so high historically since 99 it's been a huge part of the whole picture there there are several different categories lumped together here there's a whole chunk of contracts awarded by DOD that have an exception to SICA for no competition unique qualifications many of the contracts awarded for example to the federally funded R&D centers fall into that category there's quite a few billion dollars a year that are under an exception to SICA or the competition contracting act and there are several different exceptions that call into play there there's another category of where it's essentially follow on to a competed contract but there's only one potential follow on and so it's awarded without competition again back to the question you asked about breaking it down into multiple categories I think three or four different categories together into this area if you break it out none of them dominate to the $130 billion level but we'll look at reflecting that data as we refine this and update it so you can see the greater revisions it doesn't work very well in a bar graph format but it would work fine in a table form with the dollar numbers in it and we'll look at putting that into here because I think it's useful to know those things that are follow on the manufacturer of this particular helicopter is obviously the only one who's going to get the follow on work and there's no use to pretend that that's competitive it's one of the reasons in fact why we think there's a limit to how far you can carry the competition you could run a competition but we know who's going to win it so why bother wasting the time there's only one viable provider was there another question here in the middle or I think we're close to being done here I want to thank you all for your attendance your attention I want to thank you for your support on behalf of Guy Benari who's my deputy here at CSIS we welcome your reactions and feedback all the time on this virtually everything we do that's better now than it was before is as a result of something you've pointed out to us and so we kind of rely on our you guys as our customers to make sure that we continue to get better I want to thank those who joined us on the web today and with that we'll call the meeting adjourned thank you