Agricultural research in northeastern Thailand

initiated a project to establish a major multidisciplinary agricultural research center at Tha Phra in Northeastern Thailand. This report summarizes the project's setting and activities, describes its impacts, and presents lessons learned. Project experience was mixed. An excellent physical facility was built, 118 scientists received U.S. training, and a substantial research program (although not sufficiently linked to farmers' needs) was developed at the Center. However, since the project ended in 1975, research programs have been reduced and many Center employees have been absorbed into other units of Thailand's Ministry of Agriculture and Cooperatives. Present research is routine, laboratory- or station-focused. Bureaucratic conflict has created an atmosphere in which research done at the Center is often rejected out of hand by central Ministry units. On the other hand, a variety of creative outreach programs, deserving of modest A.I.D. support, have been undertaken by the Center since the project ended. These include television and radio programs, a mobile extension unit, and an agricultural information network. USAID/T officials were in a position to recognize the Center's bureaucratic impotence well before the project ended, but failed to take action to either strengthen the Center's bureaucratic position, support only promising (e.g., training) components, or withdraw support completely. Project experience shows that institutional development projects must fit the existing bureaucratic structure and be flexible enough to allow for major readjustments; that advisors and funds provided under technical assistance contracts may shield a project from direct bureaucratic opposition only as long as they are present; and that A.I.D. procedures and incentive structures do not promote investigation, analysis, and action to solve implementation problems. The following remedial measures are recommended: a detailed, retrospective review of A.I.D professional development work every 5 years; project review for mid-course correction every 2 years; and division of large complex projects into modules to provide more flexible options.