Optimal resource allocation for land settlement in Indonesia : a combined static linear programming and dynamic budgeting approach

Abstract

The integration of land settlement under Nucleus Estate-Smallholder (NES) schemes in Indonesia is aimed at increasing export crop production and improving the settlers' well-being. In spite of improvements in the schemes' implementation, profit-maximising farm plans have rarely been formulated and used as a basis for settlement development. As a result, the productivity of existing resources has not been optimised. Consequently, incomes accruing to the farms have not been a stimulus to the development of both settlements and regions. This study seeks to devise an optima] farm plan that maximises farm incomes within the framework imposed by limited resources and other restraints. The approach is a combination of static linear programming and multi-period budgeting. Although planning for long-lasting perennial crop farming systems can be handled by dynamic linear programming, this technique requires a main-frame computer. Such computational facilities are not available at a regional planning level. By contrast, the chosen approach relies on readily available microcomputer hardware and software. Thus, the technique can be adopted even at more remote regional planning offices. The optimal farm plan is formulated in three steps. First, static linear programming is employed to formulate an optimal combination of enterprises when the farm is assumed to be at 'maturity'. This is when the main perennial (rubber) reaches its yield plateau. Two perennial crops, rubber and coconut, are included in the optimal plan. The former is forced into the farm plan by the scheme management (nucleus estate), while the latter dominates the third of the three parcels of land that make up the settler's holding. Based on this solution, in the second step, a linear programming problem at 'immaturity' is specified. Here, the perennials are given equality constraints at their mature levels although their current returns are negative. Also, intercropping is allowed on the perennial blocks for the first several years. Static linear programming is used to derive the optimal farm plan at immaturity. In the third step, multi-period budgeting is employed to link the optimal static solutions and to span the total 30 year planning horizon. The empirical results show that rubber, with the establishment cost charged to the settler, does not pay. The rubber block is subsidised not only by the government through its credit scheme, but also by the other crops. Intercropping the rubber block for the first three years and the coconut block for five years, allows the second parcel of farm land to remain fallow. The delay in utilising this parcel has two advantages. Firstly, the nucleus estate will have sufficient time to clear the second parcel and allot it to the settler. Therefore, together with the settler, they can concentrate on the upkeep of the immature rubber. Secondly, the newly opened second parcel of farm land will give good yields, at least for several years, before the demanding efforts of soil management need to be undertaken. Without taking explicit account of the cost of family labour, the investment in the farm as a whole is highly remunerative as is shown by its benefit-cost ratio of 2.3. The internal rate of return is 37.8 per cent. Net income, as a return to family labour and management, is substantially higher than the poverty level that settlers are assumed to face in their places of origin. The sum of net present value is more sensitive to changes (over the whole planning horizon) in the gross revenue than to material costs. However, it is important to stress that the actual solution should not be used for any current policy conclusions owing to the inadequacy of the data used. What is far more important is that this study has demonstrated an innovative technique for optimal planning of long-lasting farming systems. The combination of static linear programming and dynamic budgeting provides a wider range of economic measures than does either technique in isolation.