In ancient Greek mythology there was the story of Jason and the Argonauts. They were enticed to sail their boats toward treacherous rocks by the sweet song of sirens. Their adventure calls to mind the experience of human services information technology professionals drawn into expensive and high-risk projects, without adequate planning, by the appeal of "enhanced funding" from the federal government.
The latest example is the promise of 75 percent matching funds, for a 25 percent state government investment in state automated child welfare information systems known as SACWIS. In order to receive the matching funds, states and their contractors must complete the implementation of federally recommended functionality by October 1996. States are rewarded for what many might describe as "quick and dirty" implementations, where speed of installation is of the highest priority.
The SACWIS systems specifications, which call for much-needed new functionality to support the adoption, foster care and child protective services programs, are ripe with complex rules and high transaction rates. The funding formulas strongly encourage the adoption of new client/server implementations, with more immediate payment for items like personal computers that cost $5,000 or less, and delayed payments for more expensive mainframe and server platforms. The factors of complexity, size and the adoption of leading-edge technologies -- added to the fact that many of the case workers who will be the end users of these systems are not experienced with computing -- all add risk to the development and deployment of these multi-million dollar projects.
It may sound like I'm opposed to the SACWIS systems and the fact that their development is encouraged by Health and Human Services leadership from Washington, D.C. Actually, nothing could be further from the truth. I believe that one of the fundamental responsibilities of government is to provide infrastructure, like that represented by this key system architecture. I believe the functionality prescribed by the HHS team is necessary and sorely needed.
What I am opposed to is the deadline imposed, and the carelessness exercised in the breakneck pace encouraged by the lure of "other peoples money." Some states -- like Colorado -- have exercised great restraint and foresight by recognizing the need for SACWIS functionality, while not rushing headlong into a strategy that puts speed ahead of thoughtful planning, analysis and design. They have taken this strategy because they understand the risks and complexities of this system, and the necessity of its successful implementation.
Ken Muroya, a systems manager at the Colorado Department of Human Services said that his "department has always had a philosophy which says that they have to prove a good business case and demonstrate that they have fully analyzed all options and information needs before starting a large systems effort." He also said that "philosophy explains our success in the past and having all of our human services systems certified as functional and fulfilling their intended purpose."
Colorado and some other leading states have placed a higher premium on meeting the needs of their child welfare programs than on receiving enhanced funding. Other less enlightened states have issued requests for proposals (RFPs) that have unreasonable and unachievable delivery schedules as their key feature. These RFPs include massive financial penalties for contractors and integrators that do not achieve required time goals. The latter approach will certainly result in systems that will be difficult to maintain or enhance, even if they are successfully deployed, but more likely will result in failed efforts and lawsuits.
We have historical precedence showing that these large social services projects require very careful consideration and planning, and may require mid-project changes of direction. Strategic systems efforts that are large and complex should encourage conservatism and flexibility, not be focused on speed at all costs.
Check out your own state and see if the Child Support Enforcement systems that were begun over the last couple of years have overrun their budget, or are behind schedule, or have less functionality than originally planned, or are in serious risk of being declared failures. I believe that only the state of Montana has a new certified system in place, with virtually all of the others in trouble. Those child support systems were the last to utilize the "enhanced funding" breakneck speed approach.
One remedy for this problem is contained in new federal legislation that calls for federal participation in funding state administered social programs to be implemented through the use of so-called block grants. Block grants eliminate federal mandates that direct the use of federal funds by the states. This change in approach will allow state legislatures to use federal funds in the manner they see fit. There will be no mandate to use any portion of their new "block grant" for information systems. The recent adoption of a block grant formula by the U.S. House of Representatives is supported by a majority of senators, who will likely consider similar legislation this summer. If the president signs block grants into law it will kill the SACWIS enhanced funding before it is ever paid out.
The American Public Welfare Association believes that block grants will be implemented in the federal government's next fiscal year (Oct. 1, 1995). If it is put into place as projected, the advent of block grants eliminates the incentive for states to pull out all the stops to beat the October 96 deadline, meaning states should immediately reevaluate their approaches to child welfare and extend the analysis of options to include those that may take longer than previously would have been acceptable.
Unfortunately, block grants may also eliminate the incentive for states to take on the challenge of building a child welfare information system at all. That would be a tragedy.
Block grants might eliminate the input of federal technology professionals that have traditionally provided tremendous insight in their recommended functional specifications and technical architectures. The new federalism contends that by moving decisions and program development to states and local communities, the result will always be better programs. Colorado's Muroya said that this view might be wrong. "When you have local control, various interest groups and legislator's inputs may drive a level of complexity that exceeds federal requirements, especially without the benefit of seeing what has worked elsewhere."
The greatest risk presented by the new block grant formula is that there will be two classes of states, those who have a technically enlightened political leadership and so have access to innovative technologies, and those that don