2026 Volume 64 Issue 2 Pages 1-10
Fragmented farmland ownership impedes mechanization and productivity and thus has been a major challenge in Japanese agriculture. Farmland consolidation, aimed at improving land use efficiency, has been a cornerstone of Japan’s agricultural policy. In farmland improvement, as farmland plot size increases, small dispersed farm plots by each farmer are consolidated through replotting. In the replotting process, farmer assignment to a plot is determined mainly by farmers’ discussion in the region. However, achieving consensus is not easy compared with a market mechanism based on individual and decentralized decision-making of resource allocation due to conflicting farmers’ interests. Considering these problems, the study employed the mechanism design method of matching game theory and a multi-agent simulation to develop and propose an improved farmland replotting protocol to achieve replotting easily for farmland improvement. The simulation modeled an agricultural village comprising 25 farmers and 324 plots, reducing the latter to 81 plots after consolidation. Multiple scenarios involving varying farmer preferences and plot distributions were tested to ascertain each mechanism’s performance based on the consolidation rate, original land’s inheritance rate, and welfare (sum of the ranks of the farmers and plots in the final assignment). Simulation results showed that the deferred acceptance (DA) algorithm satisfies both stability and strategy-proofness for farmers. While it achieves a lower consolidation rate than the current mother farmland replotting method, it performs better in terms of the inheritance rate of original plots and sum of the ranks of the farmers and plots in the final assignment. Therefore, the DA algorithm can be regarded as an effective mechanism for the replotting method.
