It's apparently not too early to start thinking up ideas for how the next administration can improve the government's management practices.

With the presidential campaign field taking shape, think tanks and public-service and business groups are preparing to develop policy proposals and ideas for the winning candidate.

The IBM Center for the Business of Government, for example, plans to launch an "interactive conversation" today to help find ideas for the next White House.

To start the conversation, the District-based IBM Center hopes to prompt an online discussion on issues raised in a new report, "Reflections on 21st Century Management," written by professors Donald F. Kettlof the University of Pennsylvania and Steven Kelmanof Harvard University. The Web address is

The Government Performance Coalition, a group of organizations interested in management issues, is sponsoring a Transitions in Governance 2008 Web site, in hopes of helping shape the agenda for 2008 and beyond.

The group's steering committee envisions the Web site (http://www.http://transitionsingovhttp://.org) as a way to get political campaigns, federal managers and the public involved in defining management issues for the next administration. A "transitions blog" is scheduled to go live today.

Jonathan D. Breul, a former Office of Management and Budget executive, will be heading up the Performance Coalition. In a bit of a coincidence, he takes over the position of executive director at the IBM Center today, replacing Mark. A. Abramson, who will continue as a consultant to the center and reopen his own company.

Other organizations, such as the Council for Excellence in Government, also are making plans for the next presidential transition. Patricia McGinnis, that group's president, said the council plans to take its "Prune Book," an analysis of the plum jobs in government that require experienced leadership, and move it to the Internet. "We want to push out and reach a lot of people across the country," she said.

The council plans to sponsor a series of discussions for career managers to prepare them for what to expect during a transition, when a crop of new political appointees arrive in Washington.

Thinking up new ideas for improving government could be challenge, after almost 16 years of "reinventing government" by the Clinton-Gore administration and the Bush-Cheney team's "president's management agenda." Those initiatives were met with some resistance from Congress and were greeted skeptically by many federal employees who perceived them as efforts to downsize and outsource government operations.

In recent months, the government's efforts to improve the performance of programs and the delivery of services has been overshadowed by the controversy over the Iraq war, allegations of contract waste and fraud, and the sluggish response by federal officials to Hurricane Katrina.

Still, the federal government has taken what Kelman calls a "performance turn," driven by a 1993 law, the Government Performance and Results Act, that has prodded agencies to show the public what it is getting for its tax dollars. Efforts to link budgets to program performance measures also have become more common, Kelman says.

The next president, Kelman suggests, will take steps to increase the use of performance management techniques, improve management of federal contracts, and look at public-private partnerships and interagency collaborations as ways of improving service to taxpayers.

But Kelman says more research is needed to show that some of these efforts are more than just a management fad and can make a difference in how agencies and programs perform.

Kettl believes the next president will face pressure to reshape the government so that it can more quickly respond to what he calls "non-routine problems," such as terrorism, natural disasters and health emergencies.

The government's most important problems "refuse to stay within the boundaries of the government agencies established to solve them," Kettl writes.

"New agencies created to tackle the new generation of problems, like the federal Department of Homeland Security, have struggled to find their footing." he argues. "Even more important, they tend to be backward looking, focused on solving the last set of problems rather than scanning the environment for the next set of problems that must be solved."

Stephen Barr's e-mail address is