<![CDATA[Objective: To support Indonesia’s progress toward the SDG Zero Hunger target, this study compares the performance of univariate panel data regression against the Panel Seemingly Unrelated Regression (Panel SUR) model in analyzing stunting, wasting, and underweight (2007–2023). Method: Using a two-way Feasible Generalized Least Squares (FGLS) approach, the study simultaneously estimates the malnutrition system to account for cross-equation error correlations. Results: The results demonstrate that Panel SUR outperforms the univariate model by correcting the theoretical inconsistency of the Low Birth Weight (LBW) coefficient—switching it from a counterintuitive negative to a positive sign while significantly increasing the explanatory power for the wasting equation and achieving superior estimation efficiency as evidenced by a lower Mean Square Error (MSE). Implications: These methodological improvements are substantively critical; correcting the LBW sign ensures that policy interventions accurately prioritize prenatal nutrition as a determinant of acute malnutrition, avoiding misleading inferences common in isolated modeling approaches. Conclusion: Consequently, Panel SUR offers a more robust empirical framework for formulating integrated nutrition policies than traditional univariate methods.]]>
Copyrights © 2025
