The interactive kiosk was going to be government's answer to its image as an uncaring bureaucracy. Taxpayers would see just one caring "face" for government. Once-stovepiped programs at the federal, state and local level would appear as an integrated whole, allowing users to look for jobs, renew a car registration, apply for welfare benefits and pay for a building permit -- all at the touch of a screen. It all sounded great, and even looked good on prototype kiosks, with their multimedia presentations.

But after all these years, the multi-government kiosk has yet to catch on. Nor have other high-profile intergovernmental projects. Some, like shared software systems for child welfare, never got off the ground. Others, such as a national electronic benefits transfer (EBT) system have been in gestation for years.

Intergovernmental IT projects are supposed to be one of the great outcomes of today's computer-driven government. Networks, not bureaucracies, would not only deliver cost-effective services more efficiently, they would integrate once-separate programs into a seamless service as well. The larger the network, and the greater the number of services it delivers, the lower the unit cost. At the same time, intergovernmental projects would spread costs further as the variety of transactions increase.

Yet all these perceived benefits have added up to little in the way of concrete projects, especially when it comes to projects that work across federal, state or local boundaries. One problem became obvious from the start. The larger the system, the greater the need for coordination, said Jerry Mechling, director of the Program on Strategic Computing and Telecommunications in the Public Sector at Harvard University's Kennedy School of Government. "Coordination -- necessary to resolve conflicts -- increases if the government entities are further apart, which is the case with federal-state and state-local intergovernmental projects," he said.

"Cross-government projects have failed because they didn't try to rethink how government should deliver services," stated P.K. Agarwal, chief information officer for the California Franchise Tax Board, and a former president of the National Association of State Information Resource Executives. Without any reengineering, intergovernmental projects really don't change the way government does business, they just put old government programs into sleek new computers. "With intergovernmental projects, the feds were really looking for creative ways to get involved with their customers," he added.

Agarwal faults the federal government for not having a clearly defined mission when it comes to intergovernmental projects with the states. As a result, projects such as kiosks and even EBT face the specter of growing beyond their purpose, taking on more transactions than government can deliver reliably or that a taxpayer can comprehend.

The Postal Service is testing a national kiosk initiative with the idea that ordinary citizens can use touchscreen technology to search for jobs, apply for unemployment benefits and job training scholarships, grants and loans. In addition, individuals can locate the nearest homeless shelter, health clinic and, oh yes, apply for welfare and Medicaid benefits.

Federal-state EBT initiatives call for multi-state networks that distribute food stamps and welfare benefits. For several years now, state alliances have been trying to hash out everything from interoperability and joint procurement issues to how users can access benefits across state lines and whether other benefits such as Social Security or state and local aid programs should somehow be included.

While EBT coalitions are moving forward toward operation by 1999, there's concern in some states that these coalitions may collapse because of their unwieldy size.

Another big issue with intergovernmental IT projects is over control. Often, it is IT people who gather around the intergovernmental table. According to Agarwal, their first interest is building, implementing and running the project, not turning it over to somebody else. "What happened with the kiosk business is that everyone came to the table saying, 'let me see if I can sweet-talk everybody into being part of my enterprise,'" said Agarwal. "When you have established geo-political boundaries, you are not going to find people who are willing to give up turf. That's like a private business voluntarily giving up market share!"

Intergovernmental projects based on coalitions of federal and state agencies also unravel because of a lack of unanimity on where the design and development of the system will take place and who will be contracted to help with the project. Often these loosely structured projects end up with one state taking the lead, making the whole project less collaborative and intergovernmental than when it began.

Finding constituents who actually need intergovernmental assistance and will use these proposed systems is another question mark. Agarwal argues that "there is little meat behind the clich