NASA has embarked on a path toward a total overhaul of its accounting system. While a lot of progress has been made, much more lies ahead. Of course, not everyone is happy with these changes.
At his confirmation hearing in December 2001, Sean O'Keefe was asked about the fact that NASA raised its surprising $4 billion ISS cost overrun by yet another $800 million only days after O'Keefe testified on the subject on Capitol Hill in early 2001. O'Keefe noted with some irony that it was referred to by some of his colleagues as "the O'Keefe bump".
O'Keefe said "this reminds us that we should not have any more confidence in these numbers than any that proceeded us. This concerns me. There is nothing that says that we can take this to the bank. There is no interest whatsoever in offering NASA a blank check -indeed that would be irresponsible."
There is an urban legend of sorts that Sean O'Keefe showed up at NASA shortly after being nominated and asked to see NASA's books. So the story goes, he was less than pleased to learn that you could not simply sit down at a computer and see the current status of NASA's balance sheet.
This is not the sort of situation you want to put a bean counter in.
Whether or not the story is true, O'Keefe acts as if some sort of rude fiscal awakening confronted him. Within a short period of time he brought in several people to overhaul NASA's accounting system Patrick Ciganer was brought in from the private sector placed at the head of this effort as Integrated Financial Management Program Executive.
Ciganer sat down with reporters last week to discuss the IFMP.
Getting it Right the Third Time?
When Ciganer arrived NASA had already been trying to get its financial act together for a number of years under the auspices of the "Integrated Financial Management Program (IFMP)". Two previous attempts at reforming NASA's Byzantine accounting system had already failed.
According to Ciganer "this is the third time we are trying this. The first time was in the late 1980's. This attempt lasted 4 years and cost $35-40 million." The second attempt was in the mid 1990's under Dan Goldin. According to Ciganer "The logic behind this effort was 'if we can't do this internally we'll go outside.' NASA hired an organization to cut code from scratch instead of using off the shelf software." This effort eventually flopped after costing $130 million.
Ciganer cast the current effort as being "mandated to use COTS (commercial off the shelf) vendors on the GSA Schedule." SAP and Accenture eventually won the contract for the current incarnation of IFMP implementation. Looking back at – and learning from - the causes of the previous two failures, Ciganer now oversees an effort that is modular. This approach was taken since earlier efforts were seen as having been too large and incapable of meeting everyone's requirements all at once. In addition to a modular approach, these modules will be introduced in a phased plan so as to allow lessons to be learned along the way and incorporated into existing and future systems.
In the implementation of the new IFMP, there was also the problem of how to deal with NASA's existing 'legacy' accounting systems. In essence, each NASA Field Center had its own unique accounting system - 10 all together - each of which had to be converted to the new agency-wide system. Some of these Field Center accounting systems involved a bit of what Ciganer called "archaeology" i.e. there were several systems that used COBOL. According to Ciganer it was getting to be rather difficult to find people who knew how to program in COBOL - a language which first came out in 1965. Another tool, called NOMAD" was used in private industry in the mid 1970's.
While Ciganer suggested that it might have been ideal to bring the new system online in parallel while the legacy systems continued to operate, there were simply not enough people to allow that to be done. Instead, implementation was done in three waves. The first wave involved several pilot centers (MSFC and GRC). Wave 2 brought KSC, JSC, HQ, and JPL online. Wave 3, initiated last week (and the cause for this session with reporters), brought the remainder of NASA online i.e. LaRC, GSFC, SSC, and DFRC.
Ciganer spoke of lessons learned from the first two waves of software rollouts noting that this far (after only 24 hours) that Wave 3 is "very quiet". He also added that while such accounting overhauls had been performed within large corporations in the private sector, that NASA is the 'pathfinder' for the Federal government. "Everyone has big plans - we seem to be at the forefront" he said.
The Latest Fad?
Just because you trot out fancy new software doesn't mean that it will work right off the bat. Moreover, there will be some institutional resistance to change. Ciganer addressed these issues by asking "how do you upgrade? How do you get people to use it?"
Referring back to the old way of doing business, Ciganer noted that there was no way to "drill down" into the data so as to get greater and greater specificity or granularity. The data was provided from each center, having been developed using that center's own unique accounting system. This information was then rolled into the agency's books. Now, according to Ciganer, within this unified system, you can drill down into the data, eventually to a great level of detail - and this will eventually be possible in near real time.
Of course, just because you provide people with more information doesn't mean that it will change any thing. Ciganer spoke of a need to "get managers to really understand what they need". After all, the culture within the agency is to not rely upon real-time accounting data - since there was none to rely upon. Estimates and projections were made with periodic snapshots - with data that was months old. As such managers often used other methods, albeit imprecise ones, to gauge program cost and performance.
One Piece at A Time
Not all of the special features of NASA's new IFMP software are now available. What has been implemented thus far is the so called "Core Financial Software" which sits atop an accounting database. Remaining to be brought online are the Budget Formulation, Asset Management, Human Resources Management, and Contract Administration portions of the system. It will not be until 2006 that these 4 remaining components are fully operational.
In addition to the Core Financial Software, some other portions of the IFMP are online and operational. "NASA Stars" is NASA's automated Staffing and Recruiting System. This service allows job seekers to post their resumes online. The goal was to reduce the time required to match job openings with qualified candidates from months down to weeks - or less. According to Mike Mann, over 50,000 people have used the service. According to subsequent information from NASA, of these 50,000 users, 1,525 people were actually placed into jobs at NASA. The jobs listed were created using the Position Description management system.
It is one thing to pay your bills. It is quite another to understand how a program is actually doing. As mentioned before, NASA was plagued with financial information that was rather old by the time it was in the hands of people managing a program. To expedite the availability of this information, NASA has created "Erasmus", the IFMP's Program/Project reporting system. While access to this system is forbidden outside of NASA, some charts were shown to reporters which show it to be a system where cost trends can be tracked and problems highlighted as they start to make themselves known.
Where's My Travel Voucher !?
While cost overruns are of prime interest to Congress and the media, in the work-a-day world at NASA, the most vexing issue tackled by IFMP has been travel reimbursement. Indeed, it is the most common thing NASA Watch readers complain to me about. I am not at all surprised. Speaking from personal experience as a former NASA civil servant, the system never worked.
Having once had to put up with this, I always felt that asking someone to have credit card - one that they had the responsibility to pay off each month - while disinterested accountants at NASA took their time writing checks was unfair. When you were issued the card, you were told that the credit card company would not contact you if payment on the card was late (due to NASA sluggishness) and that the cardholder's personal credit rating would not be affected. Neither are entirely true.
I asked Ciganer if the policy whereby individual employees would still be responsible for paying their travel bills (after being reimbursed by NASA) would still be followed. Ciganer said yes. The issue, according to Ciganer, is one of policy - a broader, overall government policy that covers much more than just NASA.
Ciganer did acknowledge some transition issues with travel reimbursement. Knowing that there'd be a delay in processing vouchers Ciganer said "we asked people to try and submit their vouchers early." Ciganer then said that he personally contacted Bank of America (who issued the government credit cards in question) and said "we want you to contact us - not employees" if bills were not paid on time. According to Ciganer this was mostly successful "B of A being B of A, that worked 80% of the time with 20% of the employees contacted about their unpaid balance."
Ciganer added "we as an agency committed to reimburse people within 30 days during the transition period." The issue of prompt travel reimbursements became an unexpected challenge after the Columbia accident. Due to this large spike in travel "we had to manually process vouchers. We were not going to wait for the system. People at JSC had to be reimbursed."
As to whether the current system will always be the standard Ciganer said that the IFMP software can accommodate either direct reimbursement of employees or payment by NASA to the credit card company.
When fully implemented, Ciganer says that the IFMP will allow great specificity when it comes to seeing who is working on what. This capability will be available when the Human Resources Management portion of the system comes online. I recalled my experience as a civil servant when all I wrote at the end of each day on my time card was "7.5 hours". I was not asked to say what I worked on - simply that I was at work for 7.5 hours.
In the era before full cost accounting (where all costs are tied to a particular NASA project and consist of direct costs, service costs, and G&A costs), such granularity was more or less irrelevant since labor hours came out of an utterly different pot than did program funds - the bulk of which went to contractor labor and materials. Now, with all costs to be accounted for, NASA has a major cultural shift to contend with i.e. accounting for everyone's actual time spent and what it is spent on.
While the overwhelming majority of folks at NASA work hard on their tasks every day, a number, to be quite honest, don't. Of course, this is a fact of life in almost every work environment. In the private sector, a variety of charge codes allow management to track employees' time better. Not only does this allow them to see how much effort is being devoted to a project, but it also allows management to see when people are overtaxed on some things and not able to perform others - or when they are underutilized. Again, without such tools for civil servants, managers need to resort to less precise means akin to keeping their ear to the ground to listen for trouble ahead.
A lot of people at NASA will likely embrace a system that allows them to actually see how a project is progressing - and who is doing what. Some will want to do so as to show their management what it is they actually do every day in the hope that they will get some additional help. Others will see an opportunity to show that there are cheaper and more efficient ways to do things than is currently the case.
Just as many people are likely to shun these new tools - either because they represent an anathema to the way they have done things for years - or they put pressure on them to perform in a way they are not used to. I can recall older managers at NASA moaning and groaning about being forced to use email in the late 1980's and early 1990's. Later, they complained about being forced to do their own charts and surf the web. Now they are going to have to track projects using IFMP tools. In each case, each of these new tools reduced the amount of time required for information to go from one person to another, yet each required a bit of a paradigm shift.
I asked Ciganer how he was going to get people to adopt these tools - since they are useless if they remain un-used. Moreover a process that requires their use can be crippled when just a few people refuse to use them. Ciganer replied, with a smile, that people need to be trained on how to use these new tools. Moreover, they need to be shown just what these tool s can do for them.
Ciganer noted that he had been exposed to this issue when he was involved with the Young Commission's examination of ISS costs. "Managers were skilled, but they did not have access to information" he said. "The NASA culture adapted to the fact that that information wasn't available. The bottom line is, a lot of folks are actually quite interested in being able to make the right decision."
Ciganer expressed optimism that people would adopt these tools and that he did not think that there would be "too much push back". He added that he saw the workforce as being genuinely concerned with getting information out noting the plethora of emails that surfaced shortly after the investigation into the Columbia accident.
He added that there "will be a natural selection". If information does not show up in a manager's reports then that manager will need to talk with their employees and say that this information has to show up. If the only way that this information can be gathered is via IFMP tools, then this will serve as an impetus for employees and managers to use those tools.
How Much Does it Really Cost?
Several reporters touched on what the cost of Return to Flight (RTF) would be and whether this system was going to be able to track that. Ciganer and Mann gave replies that did not address the issue directly. When asked if specific codes had been - or could be established to track RTF costs, Mike Mann said that temporary codes could indeed be added. Of course, there needs to be some distillation and differentiation between the costs that would be accrued anyways - regardless of whether there was an accident, and those that specifically relate to the accident and recovery from it.
I asked Mann if NASA would ever be able to give a firm answer as to just what a Space Shuttle launch actually costs. While $500 million is the general estimate NASA will not shy away from, they have never placed a firm, official cost on a launch. There are two ways to look at this: what is the overall shuttle budget and how many missions it was designed to support and what is the nonrecurring cost of adding another flight - usually expressed as the cost of an External Tank, SRBs, etc. The other is to wait until the end of the fiscal year and simply divide the Shuttle budget by the number of launches in a given fiscal year. Both are valid approaches Both are also misleading.
Mike Mann started to give a standard NASA answer and I cut him off noting that it has been a game when I worked at NASA to toss things back and for between Space Shuttle and payload providers, and between the Shuttle program and Space Station program. Depending on what your assumptions were you could easily come up with numbers which, when added up and compared against NASA's actual budget, would not agree. I asked if there couldn't be some what to make standard assumptions about what percentage of which cost went where.
Mann replied that the IFMP system did allow for the use of such rules and that they were "spending time on these rules right now" as they related to shared resources. I noted that this system is supposedly more than just a way for NASA to keep its books, but also a way to assure that accurate cost estimates are passed onto Congress. Ciganer replied that IFMP was indeed more than just an accounting system- that it was intended to be a "Decision Support System".
Playing the System
I then asked Ciganer about "grantsmanship" - the way that NASA 'intramural' researchers submitted their proposals. I once ran peer review panels at a non-profit organization for NASA. I can recall one rather astute peer review panelist asking me about how NASA employees went about submitting proposals. In particular, they referred to a batch of proposals submitted by one center - all of which covered the activities of the same group of several people.
"If all of these proposals are funded - and we pay for 100%of these individual's time - several times over, do they get a big pay raise?" he asked rhetorically. Of course, the answer is ‘no' - the money goes somewhere else. "If none of these proposals are funded will they be laid off?" he then asked. Again the answer was ‘no' since civil servant salaries came from a different pot of money altogether.
In other words there were hidden money sources at work which did – or did not factor into the situation. You could never figure out what something really cost.
I asked Ciganer if IFMP would be designed such that it stopped this 'funny money' approach to grant proposals from going on within the agency. Ciganer said firmly that the IFMP system, would be alert for this and that it was being designed "to adhere to best business practices".
During the course of this session with reporters (only 4 of us were there as opposed to the dozen(s) that can be present when Sean O'Keefe is present) representatives from SAP and Accenture were present and had a few words to say. Suffice it to say SAP and Accenture tried to put a happy face on any possible downside to the IFMP that may have been identified - either in this press opportunity - or elsewhere. This was not surprising, of course. I hardly expected them to be critical of its progress thus far.
Skepticism is abundant throughout the agency with regard to IFMP. To be certain, some of it is to be expected given that this is the third attempt to do what the private sector does on a regular basis with NASA's two previous attempts universally regarded as flops. Ciganer expects that the final cost to fully implement will be $497.5 million with approximately $200 million spent thus far.
Recently, the General Accounting Office issued a report about NASA's progress on the IFMP. They summarized their findings as follows:
"The core financial module, if implemented as planned, may provide some improvement to NASA's accounting system environment. However, NASA is not following key best practices for acquiring and implementing IFMP. In acquiring IFMP components, NASA is facing risks in understanding dependencies among commercial components. NASA has not analyzed the interdependencies among selected and proposed IFMP components, and it does not have a methodology for doing so. For programs like IFMP, which involve building a system from commercial components, it is essential to understand the characteristics and credentials of each component to select ones that are compatible and can be integrated without having to build and maintain expensive interfaces. By acquiring IFMP components without first understanding system component relationships, NASA has increased its risks of implementing a system that will not optimize mission performance and will cost more and take longer to implement than necessary."
NASA concurred in a general fashion with what the GAO had found.
Smoke and Mirrors
Having worked at NASA, I can honestly say that not having a fully detailed outline of what things truly cost was a benefit. It allowed you to not have to worry about things that would never fly in the private sector. You never had to worry about civil servant labor costs since no one you knew actually paid for them. If you needed help you simply got someone to move onto your project.
Many projects benefited from being able to o below the radar screen where costs were hard to truly gauge. Look at the X-38. It started with George Abbey's discretionary funds, borrowed people from other projects, and hummed away very efficiently without attracting too much notice- until such time as it needed a dramatic infusion of funds and then voila, there it was almost out of no where.
There are a lot of people who have spent their entire careers at NASA behind a murky screen. This is not to imply that good projects cannot emerge, but rather that the agency lacks the ability to understand what it is actually doing at any given point and be able to explain that with a straight face to -and the public.
Of course, there is an increasing number of folks at the agency who realize that this simply has to change. The $4.8+ billion space station cost overrun which simply emerged out of nowhere only weeks after NASA management told Congress that costs were under control is a prime example of something that can topple a program. With NASA burdened with the cost of recovering from the Columbia accident - and looking to build a wholly new human space transportation system (Orbital Space Plane) cost credibility ahs never been more important.
NASA civil servants: to be certain, the IFMP has a long, long way to go to fix what everyone agrees was broken. Sure, the process will trip on itself. It already has - and it will do so again. But, from my perspective having been inside and now outside the agency, there does seem to be an interest in trying to get it right this time.
When something is broke, point it out and try and to help fix it. But don't just sit at your desk and point out problems and then refuse to use new tools because you don't want to change. To do so moves you from the column of NASA human assets to the column of NASA human liabilities. Put another way, you can be part of the problem - or part of the solution.