Q&A: Mark Forman and Jeff Steinhoff on the future of federal financial management
Federal financial management is poised to undergo a new wave of transformation, say former information technology and financial management federal officials. With the closing of the Financial Systems Integration Office comes an opportunity to integrate disparate systems with cloud computing, say Mark Forman and Jeff Steinhoff, both former federal officials now with consulting firm KPMG.
FierceGovernmentIT recently spoke in depth with Forman and Steinhoff; below is a full transcript of the interview.
Forman headed the office of e-government and information technology at the Office of Management and Budget from mid 2001 through mid 2003. Steinhoff ended his government career in 2008 after lastly serving as the Government Accountability Office's managing director for financial management and assurance, and assistant comptroller general for accounting and information management.
Forman is a partner at KPMG and Steinhoff is executive director of KPMG's government institute.
FierceGovernmentIT heartily recommends you read the entire transcript--and not just because we like page views. Despite the length of the Q&A, there's insight a plenty. However, for highlights from the conversation, read this recap.
FGIT: So what happened at the FSIO (Federal Systems Integration Office) conference last week?
FORMAN: Basically, OMB explained the justification for shutting down FSIO, and laid out their vision for the future on federal financial management reform, and then explained , through speakers from OMB and Treasury and others, what's going to change in financial management, where they're going to invest resources and what the implications of moving to cloud computing means for federal financial management, which is a fascinating discussion. Jeff's got some more specifics.
STEINHOFF: Folks acknowledge there's been a lot of progress in the last couple of decades in financial systems, but they expressed a view that a lot more can be done, and we have to do a little better job of getting a return on the funds invested in systems. There was the common view from the various speakers, from OMB and Treasury, that we're spending over $2 billion a year in financial systems and development projects, and that these projects oftentimes take 8 to 10 years. Many times they're related to complex, ERP-driven projects. Folks feel there has not been enough successful proven results from this, and that data quality suffers, because data is managed in multiple locations through multiple systems. There was an expressed frustration that despite good work and good intentions, we have not gotten the result that is needed.
Also, they focused on the factor of multiple processes that are common across the federal government, but that people handle with different systems in different ways. They focused a little bit on the open government directive, and the fact that we have to do a little better job of ensuring the quality of spending information, and ensuring that what we have is consistent. And, that we've got the emergence and growing maturity of new Web-based data exchange technologies that are providing an opportunity to take a more innovative view of this. If one is looking at the open government directive and accountability and transparency, one is looking at the integration of information from a variety of systems. In looking at how the financial systems requirements were developed, they were largely predicated on accounting rules and regulations, and the ability to prepare a financial statement. Whereas now, we're trying to push out information on spending--recovery.gov and other forums. And, that we have to have more of a collaborative approach where we're looking at information.
They announced the establishment of the Office of Financial Innovation and Transformation. This will be a collaborative partnership between Treasury and OMB, and its goal is to set what they called a new course in financial management. They're looking to partner, not only within Treasury and OMB, but to receive collaborative support from the federal agencies and tie together not just the CFOs, but the CIOs, CAOs and other communities.
In looking at this particular office, they're looking at 10 to 12 people. Adam Goldberg, who now is at OMB, I think he starts his new job at April 1. [Goldberg is currently OMB's financial analysis and systems branch chief.] They want to be a canvas for change, they want to fundamentally work change on how financial management is viewed.
They have a number of goals. They want to reduce costs--there was a wide belief that we're spending too much to get too little. [They want to] improve the quality of data, create efficiencies.
I think what we're really seeing is that they want to go beyond individual agency solutions and move to centralized solutions for common business practices and processes. Examples would be invoicing, intergovernmental transactions, create standard ways of doing business, business rules, implement standards that support standard processes. No longer will everybody be able to say ‘I know that a payment is a payment, but I'm different. I know a loan's a loan, but I'm different, I need to have a different system or a different way of doing it.'
FGIT: Weren't the Financial Management Line of Business shared service centers meant to solve a lot of the problems you just mentioned?
STEINHOFF: Yes, and I think what this is the next stage of evolution of that concept.
FORMAN: If you think about the whole cloud computing concept, let's just bucket it as software as a service, or buying IT as service. That's what the line of business initiative was. Our [KPMG's] studies on Sarbanes-Oxley implementation show that an awful lot of companies got to achieve the quality of financial management required under Sarbanes Oxley by moving to a shared services model and by consolidating and integrating their diverse financial platforms. That was the intent of the FMLoB. Now a number of the financial management services are moving into the cloud. That's the same model--shared services--it's just rather than buying it in bulk, you're buying it on a utility model, you're buying it on a per-use basis.
If you went to the shared services organizations that exist in the government under the Government Management Reform Act, these franchise organizations, like National Business Center, National Finance Center, are supposed to allocate their cost over the customer base and charge them back during the financial year. That's a little different model than a per transaction financing basis. But the whole notion under the line of business of buy once, use many--that shared service concept, that notion that we're going to get it done once or twice and we're not going to pay for 50 different organizations to figure it out--I think that's pretty much what OMB is saying.
STEINHOFF: One of the concerns that OMB raised was the fact that while we have the standards, the system requirements, that the FSIO was responsible for, and in fact potential core systems were tested against those requirements, [but] there was so much customization and everybody claimed they were different, that at the end of the day, we really weren't buying a commercial off-the-shelf package. There really were very few entities that went that way.
And, then when people started going to huge ERP solutions, for which they were dumping hundreds of millions of dollars, in some cases in excess of $1 billion, they were over-engineered, not well designed, everybody invented their own wheel. There was just problem after problem. At the same time, there's a recognition--I think some of the speakers were saying this very well throughout the conference--that when we went to a recovery.gov approach, we found that a lot of these agency systems couldn't really accommodate that. They knew how much money went out in certain ways for accounting purposes, but not in terms of ability to use the information for this broader reporting that tracked where the money went and how the money was used. That triggered even greater focus.
If we can't really use these systems we're investing in and spending a lot of money on for something other than preparing a financial statement at year end and having it audited, and we're not able to compare, contrast, exchange and adapt information, then we're probably not going in the right direction.
When they spoke at this conference, they spoke in conceptual terms. Eventually, the devil will be in the details on how they roll out this new approach.
There were also a number of speakers, Jeffrey Zients [OMB deputy director for management and chief performance officer; click for a .pdf of his conference presentation] and others, who were making a point that the federal government has fallen further behind the private sector's ability to capitalize on technology.
You've got all these legacy standalone systems that can't readily talk to each other well, you've a lot a heroic efforts going on just to pull information together. We need to step back and say, 'Is there a better paradigm here?'
However, in life you have cultural transformation. And that doesn't happen quickly. Harvard has studied this--I think they said it takes seven to eight years. I think we're in that zone where people have been bating this around for seven to eight years, we have more and more service being provided today than ever envisioned before, and I think what you're going to see is through the budget process, people are going to be told, 'We're not going to give you $100 million to replace your so-and-so system.'
Some of the areas they want is to have information more accessible, to stakeholders and the public, they want to look for ways to eliminate the paper flow, to actually reduce the financial transactions, where things have to be recorded and re-recorded, which are very expensive. They want to have a strategy for assessing financial data, if someone needs financial data for something, through one source. As opposed to, one of the examples cited quite often was, was ‘We need to know what something cost or what happened,' and the person calls all around the country and reports back ‘We reported umpty-ump dollars, with one caveat--Bill is on sick leave today, so we don't know what that center spent, but the last time we called, Bill guessed it was so much.'
They want the systems to be where the information is residing, and they have the ability to pull that information out.
In characterizing where they are today, they referred to the information technology as outdated, the systems fragmented--these are their words--projects poorly executed. Information technology inhibits productivity and performance gains. I think they're looking, they made it very clear, to close the gap by going to cloud computing technologies, because they do want to improve not only the quality and the usability of the information--the value of it--but they want to reduce the cost. They feel they're spending far too much, not only to develop systems and to run those systems, but to actually process transactions.
FGIT: I wonder if there's two things going on. First is an expansion of the data that's going to come out of financial management systems--in order to track spending and not just put together audit reports--and secondly a change from service centers current model to cloud computing. Is that right?
STEINHOFF: I think cloud computing is a way to facilitate the organization of the information. I think they'll still have these service centers, and they'll have standard approaches for doing certain things.
FORMAN: I would look as the cloud approach as the next generation of federal financial management tools. Just to give you a feel for part of their vision, there's a company called NetSuite. And I think you see in Microsoft's Azure a very similar approach.
NetSuite is a software as a service model, so that's a cloud ERP. My last company went to NetSuite as our ERP. It took us a month to get operational. Now, we were a small company--NetSuite handles companies up to $1 billion...but it's a month to two months. That's a fraction of what it takes federal agencies, and some of them are not a lot bigger than that. Even if they are bigger, the types of transactions are not different. It's the same transactions, just with a lot more iterations.
FORMAN: That doesn't really change the accounting model. Now, what makes that different?
In the status quo for federal agencies, you have a finance system that's separated from the budgeting system, sometimes. You have the finance and the budget systems that are separated from purchasing or asset management, other supply management systems, HR systems. Whereas what industry is moving to is simplifying all those disparate systems. Those systems in the government and the commercial world grew up in the days of client-server applications. If you look at the federal IT budget, the vast majority of spending over the last 10 years, in fact, I'd say over the last 15 years, has been on client server systems. It doesn't matter if you're talking about financial management or program management. The government, like everybody else, was buying client-server systems, because that's the way the IT industry supplied technology. That's all shifting now. You buy technology on an utility model. And so whether it's an application that you buy as software as a service or hosting that you're buying virtualized--could be storage, could be network processing or data center processing--you're buying the technology differently now than you were 10 years ago.
That's going to have an impact not just on financial management, but on all those other disparate systems. It turns out in financial management, people are saying this is an area that we've got a lot of manual processes that should be automated, like Jeff's example, where an organization goes out for a data call to find what's the status, what's our current financial data. Nobody has it, because a lot of these systems don't talk to each other.
When you go to the cloud, you're buying the transaction processing as a service, so there's much more standardization. That data is going into a standardized database, so you don't' even have to do data warehousing with the complexity of what you used to do, because of the standardization of the data. OMB has started down this path with the standard financial information structure and the common government-wide classifications, the CGAC, and the Defense variant of that is called SFIS, Standard Financial Information Structure. So, that's all that's happening now--it's a natural evolution to the way you buy IT. But, it is a big shift from the client-server separation between the finance systems, the budgeting systems, the HR systems, the procurement systems. That's going to be more tightly coupled, and that's going to simplify people getting real time access to real-time financial data.
FGIT: You couldn't achieve that with policy rather than buying new technology?
FORMAN: Unfortunately, the reason that this is a policy issue is because, as Jeff said--based on the Harvard research--it takes seven to eight years for people to change what they've got. And that, in part, is because these systems lock in with the current organization structure. Now, that organization structure has to change away from the silos of information. I think you've got to expect to see a lot more integration between purchasing processes and financial management processes.
STEINHOFF: It gets down to, many times, to who owns the information. These agencies are very large for the most part. The CFO Act agencies have got quite a bit of autonomy. I go back to the CFO Act, when it was enacted back in 1990, and when one looks at that act, I've always felt that the three most important provisions in that act was not an audited financial statement at year's end. That was just one mechanism to get folks to focus on some of the blocking and tackling. The three most powerful provisions in that act were the call for the system management of performance, the development of cost information, and the integration of systems. Meaning financial, budget and program systems.
Again, folks have come a long, long, long way from where they were back in 1990. It's not even the same playing field today. It's like going from a Little League team to being on the Yankees. But at the same time, there's a recognition that we've spent a lot of money and we have to, as a government, look for a different solution. And there are a lot of opportunities here.
If one looks at the transformation objectives that were discussed by Treasury folks at this conference, one of the areas they're going to focus on first is to centralized vendor payment processing information. They want to have an integrated approach to it, where they do all invoice purchase and payments on a government-wide level. They want to have a central utility for intergovernmental transactions. That's one of the reasons Treasury can't pass a financial audit. One agency records a transaction one way, another agency might record half of that transaction a little different way. Both of them pass an audit test, but the government is out of balance by hundreds of millions of dollars if not by billions of dollars.
FORMAN: We used to talk about the feeder system issues, and the fact that one of the reasons why financial management systems failed so badly was that feeder systems were so complex, that they required customization of interfaces with the financial systems. And that tended to disrupt a number of things that led to audit problems. So where are the key audit problems that you ought to focus on if you're going to fix these issues? One is in what the government calls the procure-to-pay process. In the commercial world it's called order-to-cash. But it's basically the processes that relates the financial management, the purchasing and the inventory management all together, the receipt of goods. That's the feeder system issue that's got to get fixed. The government tried to fix that with the Financial Management Line of Business, and yet because of the difficulties and the fact that the end-to-end process integration was not generally done in agencies, it was not fixed.
Similarly, inter-agency transactions is another recurring audit problem. Although it's being done on a computer, it's an automated manual process. The commercial world has focused on this, they had to because of Sarbanes-Oxley. In moving to the cloud , I think again OMB is looking for the shared services approach, the line of business on steroid approach, where you're adopting that same buy once, use many model. But we know that when we buy once, it's done right, and the expectation is that each agency doesn't have to figure out how to do it right, but each agency will have to use that instrument that the government has invested in to get it right.
FGIT: In order to get that integration, must there be a policy change?
FORMAN: Management policies shouldn't have to change. The primary ones that I'm looking at are A-123 [.pdf] and A-127, on the financial and business management side. They already lay out the path for integration. I think that when you dig in, the interpretation of those policies are going to have to be refined to reflect today's technology.
If you were to look at A-127, as an example, A-127 is the one that covers the financial and business systems of government. OMB issued a revised version of that in January 2009, with an effective date of Oct. 1, 2009. I think it's fair to say that at the FSIO conference, a lot of speakers didn't believe that FSIO was able to do what they needed in order to actually get that policy implementable this fiscal year. And so, it was supposed to be working the first day of the fiscal year, and it hasn't been. I think that was some of the frustration that was relayed. So, OMB is saying ‘We can't do this with business as usual. We're going to move to take advantage of cloud computing concepts in order to implement this new A-127.'
A-123 similarly has the controls environment, and that was updated in 2008, to take into account specific the acquisition interfaces, the procure to pay process that we're talking about. The policies, I think, have been updated to do that.
You could argue that A-130, which is the systems policy, needs to be updated, because that was built for the client-server era. A-127, A-123, cut across the management elements of OMB. A-130 is owned my old office, the office e-government and IT. That should be updated, conceivably, but I don't know if there are any policy updates in that update.
STEINHOFF: The sense I got from the speakers is that OMB just didn't do this in a vacuum. They've been listening to and talking to CFOs and CAOs and others, and there was quite a bit of frustration that, again, going back to that central thesis that OMB and others laid out quite upfront - we're spending somewhere in the order of $2.3 billion a year on financial management systems that are taking us 5 to 10 years. Then, when they come up, they're not working as well as one expected, and we've had some very costly failures for systems that are running on life support, and they haven't provided to us the improvements we were looking for.
FORMAN: I don't know if I would call that so much a policy change so much as a management change.
STEINHOFF: I would agree.
FORMAN: I think the underlying policies are still pretty good. It's the difficulty in trying to implement those policies by continuing to invest in client-server financial management systems with all the other changes in the other client-server systems that they've had to interface with. What I think OMB is saying is ‘We know there's not a complete set of technologies out there to replace those old approaches, but we also know that our old approach doesn't work. So, why keep spending money trying to make it work with the old approach, when technology is now evolving?'
FGIT: I guess the success of this assumes that there's going to be some federal agency offering cloud computing services.
FORMAN: I think that's more than an assumption. I think at the conference it was a very clear notion that there is going to be a set of shared services available in a cloud computing framework, and Jeff laid out three of those. I'd say from a process standpoint, you're really looking at the procure-to-pay and the budgetary reporting processes, where you'll see heavy emphasis on the use of cloud.
STEINHOFF: They're looking to have government-wide government financial management repositories, with one standard based architecture for all payments, all collections, other government-wide transactions. And it'll have supporting details that'll be in this repository that users can come in and pull and they want to have an integrated budgetary and financial reporting process that is supported by the underlying transactions, where someone can just go to the system and array the transactions in the way they wish to array those transactions, as opposed to being locked into a certain format or a certain way for that pattern to be displayed. If one looks at what they're doing in recovery.gov, they are arraying the information in a variety of ways to show how that money is being spent. Compare that to what DoD faced, when it was told by Congress, please report to us on how much the war on terror is costing, what's the Iraqi conflict costing? The department did not have systems, despite having tons of information in various forms.
There's no lack of information in systems in DoD, and they're extremely smart people, but they couldn't capture this cost information in a way that they could reliably report it back, and they basically had to resort to the data call approach, and having rules. Here are the rules when we call you on the phone on how you report this information back to us. And then you had people trying to understand what the rules were, and trying to adapt whatever information they in fact had. So, they're talking about building these repositories.
Again, as I said before, this is not something that's going to happen immediately, and time will tell how this goes. But I see it as the next stage.
Changes coming to federal financial management