<![CDATA[This study estimated production elasticities and partial allocative-efficiency indicators for rice inputs in technically irrigated and rainfed areas of Tangerang Regency, Indonesia. Cross-sectional data were obtained from 88 farm units (44 in Sukadiri and 44 in Solear) and analyzed using log-linear Cobb-Douglas models. In addition to conventional ordinary least squares, the revision reports HC3 robust standard errors, Wald tests of returns to scale, a pooled regime-by-input interaction model, per-hectare sensitivity models, and bootstrap intervals for value-of-marginal-product-to-input-price ratios (VMP/Px). Cultivated area was positively associated with total output in both areas; urea was also positively associated with output in the rainfed sample. HC3 estimates supported these results. The sums of elasticities were 1.090 (95% CI: 1.043–1.138) and 1.117 (95% CI: 1.092–1.141), each statistically above one. Joint slope interactions were significant (Wald statistic = 11.643, p = 0.040), driven principally by the cultivated-area interaction; the urea slope difference itself was not significant. Per-hectare sensitivity models had much lower explanatory power (adjusted R² = 0.247 and 0.107), confirming that the near-perfect fit of total-output models largely reflects farm scale. Bootstrap VMP/Px intervals supported a positive allocative signal for cultivated area in both samples and urea in the rainfed sample, but were inconclusive for most other inputs. Findings are associative and location-specific because irrigation regime is confounded with district, the sample comprises 44 farms per model, and several predictors show substantial collinearity.]]>
