social impact or their degree to which they are inappropriately ignored in most jurisdictions today.

Jerry Mechling is director of Strategic Computing and Telecommunications in the Public Sector, at Harvard University's John F. Kennedy School of Government. Victoria Sweeney is a research affiliate with the same program.

THE EASIEST SOLUTION

The two easiest IT budgeting improvements involve educating and training those who propose and evaluate IT investments and use capital budgeting vehicles for projects that create multi-year assets.

Education. Public-sector managers considering IT often know what business results they want but fall short when it comes to knowing how technology can help. This gap can and should be filled. Education should focus on the human and organizational dimensions of IT-enabled changes in work processes and on improving communication between the technology community and various program and political communities. Senior executives need to better understand how IT creates value -- for instance, through network-based work processes. Without this knowledge to guide design and implementation, IT projects are often unduly risky and not worthy of funding.

The primary lessons to be learned are not about technology per se but about what technology can do, particularly as related to organizational change and implementation. Technology can also help change how education itself occurs within the organization -- for instance, by offering distance education, groupware and "just-in-time" delivery of relevant information as teams form to work on specific projects.

Capital budgets. Many IT projects are necessarily broad in scope and scale. They can be costly, with high upfront costs that can be killers. Multi-year funding in such cases -- by spreading costs over the life of the investment -- can substantially improve the chances for project approval. Though some governments have moved to multi-year funding, many have not.

Capital funding is immediately attractive because it reduces the "sticker shock" of large-scale projects, but it also has other benefits. For example, the commonwealth of Massachusetts has twice issued bond bills for technology projects -- the second effort coming because of the overwhelming success of the first. After the initial Information Technology Bond Bill of 1992 went through -- with 21 projects -- many managers in the state government thought they would increase the probability of getting funded if they cooperated in producing another technology bond bill. This second IT bond bill, introduced and adopted in 1996, included 112 projects totaling over $300 million. Because it represented such a huge opportunity, it encouraged agencies to develop more comprehensive and strategic plans for IT.

THE MOST VALUABLE SOLUTIONS

Education and capital funding are valuable steps for most jurisdictions and among the easiest to implement. But infrastructure and cross-organizational projects typically offer even greater net value, along with typically greater implementation difficulty.

Infrastructure projects. Infrastructure investments -- in networks, standardized data, etc. -- offer a happy combination of exceptionally high value and only moderate implementation difficulty. Information infrastructure can improve service delivery processes in the near term while also creating shareable data and communications capacity for the future. While infrastructure requires long-term commitments, it often facilitates projects not feasible or even imaginable at the time of the initial investments.

Cross-organizational projects. Often, applications that integrate specific services across agencies and governments are extremely valuable. Certainly, as organizations become familiar with the Internet, the potential for cross-organizational links becomes more apparent, allowing workers in many settings to provide a broader range of services to the citizen. But the Internet isn't the only way. Information kiosks operating under other protocols can also allow citizens to access services without having to own their own computer.

Cross-organizational projects, however, raise issues that are often easier to resolve in traditional, inside-the-organization projects. For instance, agencies and governments may be using technologies that can't "communicate," just as 5.25-inch disk drives can't make any sense

Jerry Mechling  |  Contributing Writer