within the research program, or out of the office altogether, can result in more lost time and declining public support than virtually any other mistakes. Leaders in ORD should consult regularly with, and be perceived to seek and consider advice from, the scientists and engineers within ORD. Research managers in ORD should be selected on the basis of scientific competence and the personnel skills needed to lead and nurture professional development of their staff scientists and engineers.

Our committee's vision for the future of EPA's research program requires leaders who have technical competence; managerial abilities; communication skills; knowledge and skills in research planning and administration and in the public decision process, including its political dimensions; and the ability to marshal constituencies for an effective research program.

The issue of scientific leadership is discussed further in the last chapter of this report.