Here is the fundamental challenge: Each new grant program passed by Congress creates a tiny new fiefdom in a government agency. This fiefdom directly gets funding and has the ability in many cases to set up its own management processes and systems. New grants typically have some unique rules and process steps, so most program managers believe that their processes and systems must be created from scratch, leading to a colossal amount of reinventing the wheel, duplicate systems, and waste.
Since the Office of Management and Budget (OMB) oversees the activities of federal agencies, it is one of the most logical bodies to attempt to curb this rampant duplication. Two major initiatives from OMB are:
- The Federal Enterprise Architecture: The FEA is a framework for enterprise capital planning through which agencies and sub-agencies can organize process improvement and IT development initiatives. This includes identifying common services that can be implemented across programs and processes to share resources and reduce waste.
- The Lines of Business (LoB): These are sets of high-level processes that are common across many agencies (like HR, financial management, and grants management). The goal for each LoB is to develop common government-wide standards and identify service providers so that all agencies can combine these duplicate efforts.
I feel that both of these initiatives have shown great value (for example Grants.gov and the FEA Practice Guidance), and that groups and individuals working on them have demonstrated intelligence and drive to reduce waste. However, fundamental decisions by OMB in running these initiatives have put their overall success at great risk. Specifically:
Mistake 1: Centers of Excellence do not have consumer-driven standards and can inadvertently descend into self-centered mediocrity. Once a line of business reaches a level of maturity, OMB picks a small number of agencies to become Centers of Excellence (CoE) or government-wide service centers for that LoB. For example, grants management has three CoEs led by DoEd, HHS, and NSF. These service centers could potentially provide great service, however, since OMB never declared any requirements or gave potential customers any service rights, it is completely up to the service centers to provide quality. Instead OMB forces other agencies to pick a CoE and demonstrate movement toward being a CoE's customer in order to avoid penalties. This model forms a monopoly with all of the associated dangers of mediocrity and overblown cost. CoEs have no real impetus to advertise their capabilities, to entice customer agencies, or even adapt to address customer agency needs. Instead the CoE has earned the luxury to embellish their highly-specialized systems, leaving other agencies to pray that one of the CoE's can actually meet their needs. Due to the incredible efforts of individuals, some CoEs have provided demonstrable value, but it is only because of their efforts, not because of a reasonable market-based drive.
Mistake 2: Most Line of Business service centers are monolithic systems that are difficult for agencies to adopt. In OMB's FEA Practice Guidance, they suggest looking at defining segment architectures not only from vertical core mission processes (like grants management), but also from horizontal enterprise services (modular process steps, like authentication, that are common to multiple organizations or processes). OMB should recognize that adopting systems that implement whole vertical processes is difficult since a vertical system must completely support a customer's process to be used and requires great investment and stakeholder buy-in to develop. However, adopting a well built enterprise service (such as authentication, grant search, or grant apply) is a lot more feasible and cost effective. Therefore enterprise services like grants.gov are dramatically much easier to adopt than the grants CoE's provided by DoEd, HHS, and NSF.
Mistake 3: The FEA Practice Guidance skips a fundamental goal: creating an Enterprise Architecture. The FEA Practice Guidance looks like an incredible cost-saving tool that helps agencies define and implement enterprise-wide services and coordinate the IT efforts across multiple programs. The guidance covers defining segment architectures, conducting EA transition, and using performance metrics to analyze the efficacy of an EA program.Yet, it does not cover developing an EA since "there is already a broad body of knowledge on these topics". The closest compatible document is the Practical Guide to the FEA that was written in 2001, and most of which is superseded by the Practice Guidance. The absence of a compatible plan to develop an EA makes the Practice Guidance extremely difficult to implement.
All is not lost. Through LoBs and the FEA OMB has a brilliant infrastructure from which it can still radically reduce redundant systems. It has galvanized a great number of people driven to provide better services and reduce cost. What does it need to do?
Fix 1: Create a market-driven set of Line of Business Enterprise Services. Grants.gov is a wonderful example of a multi-agency enterprise service. However, usage is mandatory and is in some cases customer agencies are being driven to use it for formula grants, which is just plain silly.
- Instead of focusing on CoEs that implement a full LoB, OMB should encourage the development of enterprise services that are smaller, applicable to more organizations and processes, and are easier to adopt.
- OMB should open up the creation of LoB service initiatives to agencies, companies, and research institutions.
- LoB services should be allowed to operate as either software-as-a-service (SaaS) (as all current CoE's operate) or as off-the-shelf software, so that agencies that wish to retain ownership of their data can run the services locally.
- OMB should define requirements and standards for multi-agency initiatives, including marketing materials that LoB initiatives need to release to candidate customers.
- OMB should create and maintain a central directory of LoB service implementations.
- Agencies should be allowed to not use LoB services if they can provide evidence from enterprise planning analysis. Instead it should be the responsibility of the LoB service providers to provide such quality and marketing to attract the customer agencies (in other words, create a customer-based market).
Fix 2: OMB needs to create a single reference for FEA implementation in agencies. The FEA Practice Guidance needs to include a section that defines a light-weight process for creating an organization-wide enterprise architecture model. The Practical Guide to the FEA delves too far into organization-wide to-be architectures as well as implementation strategies, which are not needed when using the segment architecture approach of the FEA Practice Guidance. With one more section, the FEA Practice Guidance could be the go-to document for enterprise planning.
Fix 3: The OMB 300 needs to check for enterprise planning within an agency. Why should any grant program require a new system? What about services that can be shared across programs or processes? These questions should be directly asked by having applicants research nearest neighbors within an agency (and within LoB service implementations) and describe how a new IT initiative uses those nearest neighbors, or provide compelling reasons why they cannot use them. The application should include evidence of enterprise planning including a segment architecture diagram and an enterprise sequence plan. If enterprise planning is occurring at an agency, then these one-page diagrams would be easy to generate.
Right now, agencies are holding their breath to determine if Lines of Business will survive into the new administration. Formed with the best of intentions, LoBs have languished under its implementation as huge monopolistic SaaS deployments with customer agencies being dragged into their use. It is time for reform and allow LoBs and the FEA to truly save taxpayer money.
I wish to remind readers that the opinions expressed in this article are my own, and do not necessarily reflect those of TCG or its employees.
Robert W. Buccigrossi, Ph.D.