Tariffs and economic development: A comment

The optimality criteria of linear programming transportation and spatial equilibrium models never ‘explain ‘ real world flow patterns. This paper provides reasons for the difference between an optimal solution and real world patterns. Data for the linear programming exercises are derived from the four stages of a rice marketing system in Sri Lanka at a time when the state had monopoly control over distribution. The examination of factors more important than transport costs in explaining residual flows sheds some light on policy and institutional problems associated with monopoly procurement.

Substantively, a comparison of the optimal solution with reality shows a fairly high degree of transportation efficiency throughout the system, except at the last stage, where rice changes hands between two parastatal orginisations (the Paddy Marketing Board and the Food Commission) to be distributed to final destinations. Inefficient store locations rather than commodity allocations generate the greatest waste of transport. Reasons for the difference between programming solutions and reality include uncertainty, congestion, policies and institutional structure conducive to a deterioration in quality of the commodity handled, problematic regional preferences for rice type, inadequate communications, unpredictable timing of rice imports and corruption.