Discussion: Methodological Problems in Agricultural Policy Research

Abstract

In casting about for a handle by which to grasp topic of methodology as it applies to agricultural policy research, Dr. Brandow, like Neptune, has come up from sea bearing a three-pronged spear, or trident with which to attack methodological problems in agricultural policy research. He suggests that research problems in policy be arrayed in three orders. First order problems deal with choice of of several alternative solutions to a policy question. Second order problems deal with consequences of alternative courses of action with respect to agricultural policy, while third order problems deal primarily with questions of facts and relationships ancillary to policy questions. I have some difficulty in viewing this particular trident as a weapon with which to pursue methodological problems in agricultural policy research. Why? There are two basic reasons. First of all, I would argue that element upon which Brandow's division of policy problems into orders must stand is one of scope rather than method. As Dr. Brandow points out, this hierarchy of policy questions moves from the single complete question at top, [through] a few broad questions about consequences of alternative courses of action . .. [to] a base of numerous specific questions about facts and relationships in several fields of study. These are primarily separations according to scope of problem. Even if we waive objection that this division of policy problems into three categories is based upon scope, it is difficult to see basis for separation of questions of choice of among several alternative courses of action from an appraisal of consequences of alternative courses of action. If a decision as to what is best is to be made with respect to policy alternatives, this choice can be made only after consequences of different alternative courses of action have been appraised. In turn, choice of alternatives and their consequences calls for facts and relationships. Without such facts and relationships no evaluation of consequences of courses of action could be made. These observations lead to opinion of relative inseparability of policy problems on this basis for purposes of appraisal of method. My second point is that scientific method in research applied to any problem involves a specific sequence of elements regardless of problem's position in any hierarchy of problems. Since Dr. Brandow has recognized interrelation of these three orders of problems, one can conclude that separation stands on difference in method applied to problems at three levels. Scientific method usually requires, in addition to a sensing of existence of a problem situation, formulation of hypotheses about problem and its solution, collection of data for testing of these hypotheses and application of logic and statistical analysis to data and hypotheses to arrive at conclusions that will support or cast doubt upon hypotheses. It is evident that in any one of three orders of problems methodology is substantially identical as far as basic elements of method are concerned. Even in third order