A STATE-SPACE APPROACH TO PERENNIAL CROP SUPPLY ANALYSIS

In perennial crop supply analysis, separate estimation of the qualitatively different new planting and replanting decisions is desirable. Commonly, data paucity restricts estimation to a single reduced form equation. This study develops a dynamic unobserved components method where separate estimation of the structural equations is possible in the absence of detailed data on new plantings and replantings. The proposed method is empirically implemented in a case study of the Florida grapefruit industry. In terms of both statistical and forecasting performance, the proposed structural approach is found superior to a single equation reduced form specification.