Sugar Cane Irrigation: A Case Study in Capital Budgeting

A study of irrigation, economics was conducted in 1963 on a 10,000 acre sugar plantation in Hawaii and repeated in 1964 on three additional plantations covering 34,000 acres. The objective of the study was to recommend, on an agronomic and economic basis, a program for investing large amounts of capital in newly developed automatic and semi-automatic irrigation systems. Although basically a problem in capital budgeting, a number of complicating factors precluded the straightforward application of traditional investment analysis techniques. The more important of these factors were: 1. Many Alternatives. The wide range of physical conditions found on each plantation ruled out the selection of a single "best" irrigation system. As a result, a number of systems had to be evaluated for each of several hundred cane fields. 2. Combinatorial Aspects. Investments in individual cane fields were interdependent since groups of fields shared common water sources; thus, combinations of investments had to be analyzed, increasing the number of investment alternatives astronomically. 3. Opportunity Value of Scarce Resources. New irrigation systems offered substantial water savings. However, the value of these savings arises from the "opportunity" value of water in alternative uses. Alternative uses, in turn, depend on how much water is saved. Thus, no value for potential savings on the plantation could be established until the investment analysis was run, first for individual alternatives and then for combinations of alternatives. 4. Complicated Timing. Because of crop planting schedules, new systems would be installed on small increments of acreage over a period of 8 to 10 years. Thus, "nonconventional" investments (outlays interspersed with benefits such that the stream of cash flows had multiple sign changes) occurred in multiple time periods; furthermore, because of crop harvesting schedules, returns from these investments also were distributed over multiple time periods with varying amounts of delay between investments and returns. The approach used to solve this problem is described in the paper.