Talk to any state or local official about how the federal government approves and grants funding for IT projects, and you will likely hear a litany of complaints. In 2001, for example, John Cuddy and Jerry Friedman, then officials with the Human Services Departments of Oregon and Texas, respectively, pointed out just how outdated the entire process had become in their report, Re-engineering the Approach by Which the Federal Government Approves and Monitors the Creation of State Human Services Information Systems.
"The decisions that led to the current approach were made many years ago in an environment much different from that in which IT projects are conceived and managed today," the report states. What may have worked when mostly mainframe-based information systems were stable and "big-bang" systems development predominated does not apply in today's world -- where rapidly changing technology and new approaches allow for more flexible and improved system development.
Former Georgia CIO Larry Singer, in a 2002 testimony before the House Subcommittee on Technology and Procurement Policy, described the tension between the federal government and states even more strongly. "Of all federal grant processes, this APD [Advance Planning Document] process is the most cumbersome, expensive and provides the least value to the federal oversight process and to the states."
On one hand, the federal government believes it exercises expected and appropriate controls over the billions of dollars it gives states to implement IT projects. On the other hand, many states feel the federal grant process is onerous, hindering them from implementing quickly and efficiently. The prevailing question that arises can be expressed in several ways: Where do we hang the "trust net" between the federal and state governments? Does federal money have to come with strings, or can states be trusted to spend wisely? When does the accountability requirement for public funds interfere with good management practices and efficiencies that yield savings?
Addressing such a complex issue and set of questions is a difficult undertaking. States receive federal funding for IT projects covering health, homeland security and transportation, to name just a few key areas. Nothing is more problematic, however, than health and human services. The difficulties can be traced back to welfare reform legislation passed in the early 1990s that levied on states many more federal directives, including those addressing key programs, such as Temporary Aid to Needy Families (TANF), child support, food stamps and Medicaid.
As families became more independent and moved from assistance to self-sufficiency, states realized a more integrated IT infrastructure to more efficiently manage caseloads was necessary. But as Cuddy and Friedman pointed out, the federal funding approval process had not kept up with the changes taking place. Tension grew to the point that many states argued the federal IT grant process actually stymied their efforts to help those most in need.
Problems, Problems
Although most observers would agree that today states are much more knowledgeable in defining IT systems, obtaining good contractors, controlling scope, managing projects and getting good results, former Ohio CIO Greg Jackson argues that problems continue for the health functions in many states, for example. Notoriously underfunded, these departments received a surge of funding in the wake of 9/11 and the anthrax attacks that followed. Although welcomed, the funding presented a dilemma: States now had lots of money, but not enough resources and staff experience (including program management expertise) to manage the sums adequately. "States need program management services," Jackson said. "Anytime you have a large amount of money, it requires a large amount of program and project management."
Some state officials say the current IT approval process no longer reflects the reality of what is needed for good project oversight. In his statement before the House Subcommittee on Technology Procurement Policy, David McClure, director of information technology management issues for the Government Accountability Office (GAO), spoke about the complicated process of funding IT in human services.
In his statement he reported on a GAO study, in which approximately one-third of states studied said they had problems with the current process or wanted a more streamlined one. State officials commented that the process takes too long, it is difficult to meet the many requirements for receiving federal funds, compliance with the APD is costly, and communication with federal reviewers has declined. One observer noted the APDs and their updates require extensive and sometimes excessive work by both federal and state staff during a time when many federal employees are retiring and state staffing levels are shrinking.
Another problem is the delays inherent in the APD process. Federal agencies responsible for funding state human services projects must respond to an APD within 60 days. If the federal agency disapproves the document or requests clarification, however, the clock starts over. Late responses from federal agencies, the GAO study found, were typically due to lack of staff, complicated issues, multilayer review within the agency or difficulty in reaching agreement with another agency.
Although most states that experienced delayed federal responses said they were unaffected by the delay, a few requested retroactive approval for actions already taken -- something APDs were established to avoid. In fact, in one case, the work was in vain. By the time the federal government finally approved the state's APD, the state was barred from implementing the project when its legislature froze spending until the next fiscal year. In other cases, federal delays resulted in the implementation of obsolete equipment.
A third problem with the APD is the process strongly discourages use of federal programs to create a common IT infrastructure. When submitting an APD -- or in some cases, an APD update -- states must submit cost allocation plans, which are used to identify, measure and allocate expected project costs between the states and the federal program(s). According to Office of Management and Budget (OMB) Circular A-87, an agency may not expend appropriated funds for purposes other than those authorized by the appropriation.
"Although [a common IT infrastructure] frequently costs less to the original program than creating a separate, stand-alone program, it's precisely because it would benefit other programs that it's often termed unallowable," said Aldona Valicenti, former CIO of Kentucky.
Frank McDonough, vice president of Intergovernmental Solutions at Guerra, Kiviat, Flyzik and Associates -- and former federal official -- brought the point home recently when he described one state office with multiple computers on a single person's desk to access various independent applications.
Over the years, various efforts have been made to address these problems and the overall cumbersome nature of the APD process. To date, little substantive change has been achieved.
Over the Rainbow
So does a solution exist? Is there a credible way to improve federal funding of state IT projects that balances the need for accountability while maintaining innovation and efficiencies? Depending on which expert you ask, the solutions vary. Here are a few worth noting:
Work Within the System. McDonough, who is the former director of the General Services Administration's Office of Intergovernmental Solutions, takes a big-picture view. He noted that statistics on the many governing bodies in the United States are staggering. "There are six types of governments in the United States: federal, state, county, city, regional and tribal," he said, adding that there are 70,000 governing institutions among the various government levels. "When you layer this infrastructure with the numerous granting organizations of the federal government, you paint a very complex picture." Most program applications are already in place and run as independent stovepipes with their own processes and requirements, he said. Considering the size of the governance issue, good program management within the stovepipes provides a smart approach. "There are so many different governing bodies contributing to the magnitude of the situation that a more grandiose approach simply isn't feasible," McDonough said. He suggested programs should be performance-based wherever possible to let states retain flexibility in implementing systems while maintaining overall requirements for delivering results.
Former Ohio CIO Jackson took McDonough's approach one step further, noting that if the federal government were more prescriptive in its IT funding requirements, less waste would exist. "It would be beneficial if we all knew what architecture to build to, so we could all build to the same system," Jackson said. "The federal government never went the extra mile to give direction for architecture."
Block Grants. Block grants represent a middle ground between federal control and state freedom. The term generally refers to a funding package created by consolidating several existing programs into one, providing states more latitude with how money is spent.
In his testimony, former Georgia CIO Singer noted that states can now apply TANF funds to IT projects without APD review. "Since that time, there have been no major system failures or federal censures, procurement scandals or failed audits related to TANF IT expenditures. In that same period of time, there have been many such problems with the systems that have remained under the APD process."
State officials are split on the block grant issue. Proponents argue that block grants give states the flexibility to implement programs more cheaply, creatively and efficiently. Opponents say block grants obscure the focus of federal programs by giving states too much leeway and reducing accountability. For example, in one famous "Golden Fleece" award, $27,000 was used for a study on why prisoners wanted to escape from prison.
Hord Tipton, CIO of the Department of the Interior, notes positive experiences with block grants, though he indicated they are more effective when there is trust established between federal and state counterparts. One way to build this trust is on the longevity of a relationship. Another important component of the trust required by block grants is clearly defined requirements. "The grant giver needs reassurance," Tipton explained. "You can get into trouble quicker with a mismanaged grant than with anything else." Grant administrators come and go, so complete documentation is a must. Lastly, the more experience state and federal grant administrators have with a certain type of grant, the more comfortable they are operating and reporting on it.
The Enterprise Model. To South Dakota CIO Otto Doll, an enterprise model makes the most sense. Since 1996, South Dakota has had a centralized IT organization. Doll cited the state's E-Rate implementation as an example of his department's success with consolidated systems. E-Rate, a $2.25 billion per year program authorized by the Telecommunications Act of 1996, provides affordable telecommunications services to eligible schools and libraries. Through the program, qualifying schools and libraries receive up to 90 percent off the cost of Internet and telephone services and internal wiring. E-Rate was immensely complicated to implement in South Dakota, as it involved nine different contracts. But because South Dakota consolidated IT, one contract administrator at the state level handled all the contracts, covering virtually all of the telecommunications needs for 176 different school districts.
Doll also noted that South Dakota deployed solutions such as a single time accounting system that supports every state worker in the state's 66 counties and a statewide radio system throughout all of its towns. Doll said the struggle to convince towns to join the state radio system when they had previously invested in other systems was eased somewhat with the supply of new radios. "By spending money for the enterprise versus programs, all agencies benefit," Doll said.
While Doll admits centralization is significantly easier to achieve in a sparsely populated state like South Dakota, he is convinced a centralized IT function is achievable in all but the largest states (New York, Texas and California). Doll offers states a few pointers:
1. Begin by standardizing commodity technologies, such as data centers and hosting. This may be the highest level larger states can standardize to without running into technical problems, but it still offers significant benefits.
2. Dream big. "We don't spend enough time visioning and missioning. We accomplished IT implementation by thinking and working through the ability to do it for everyone -- citizens, businesses, etc. It has added a level of complexity, but it is worth it."
Best Practices: The Ties That Bind
Across the board -- from those who feel that the only realistic way to deal with the state IT funding issue is to work within stovepipes, to those who believe in an open enterprise model -- commonalities exist. These notions provide a meeting ground for those holding differing opinions and offer a springboard for launching new ideas and approaches.
Trust first. It is the public servant's mantra -- whether he or she works for a state legislature, the governor's office or federal agency. Honorable intentions must be assumed among all participants who share the same ultimate goal to promote public welfare.
Influence good behavior. The federal government plays a role in influencing good behaviors and decisions. Ohio's Jackson, for instance, suggested that the federal government should prescreen the three best state solutions for a particular IT application, then influence other states to take advantage of the earlier experience and success.
Standardize the grant process. Standardize federal requirements so fewer requirements are driven down to the states. If the federal government distributed grant money consistently -- and reporting rules were the same -- one system would be feasible. Denis Gusty, director of the GSA's Office of Intergovernmental Solutions, noted that the process states must endure to get a grant is very stringent. "Once you've satisfied that level, you've satisfied the criteria for getting the money, so you should be able to use it as you see fit," he said. "Although states shouldn't have to rejustify the need for grant money, they still need to show good stewardship of it."
Consider centralized governance. Joe Leo, former U.S. Department of Agriculture CIO and now a vice president at SAIC, said there should be one federal governing board and one pooled revenue pot for all social service-related IT to encourage statewide system development and management. "Allocation should be based on legislative standards," he said. "A governing board would ensure fair allocation."
Collaborate. Unite representative bodies to craft a single solution and achieve buy-in and commitment. Management experts suggest today is an "Age of Interoperability," where work is accomplished through networks, partnerships and collaborations, and all are valued for their wisdom, experiences and contributions.
The issues surrounding federal funding of state IT projects will not disappear anytime soon. The deep-rooted history of the problem is ingrained in the cultures and processes of thousands of governing bodies throughout the country. Congress desires results and accountability; federal agencies care about effective and successful implementations; states and localities want freedom to maximize investments. Regardless of perspective, most believe improving the process should be "institutionalized in our spirit." We, as public servants, should constantly strive for improvements in funding state projects at all levels of government. Public funds are a public trust. Strings or no strings, the dollars are the same shade of green and require the same obligation of duty -- to serve and maintain the welfare of this country's citizens.
