🤖 AI Summary
This paper addresses the challenge of accurately estimating the opportunity cost of caregiving at small-area levels when individual-level population data are unavailable. We propose a small-area estimation method that integrates calibration weighting with a mixed-effects random forest (MERF) model. Relying solely on aggregate-level auxiliary data—such as census summaries—our approach is the first to embed a calibration weight mechanism within the MERF framework, thereby eliminating dependence on microdata while enabling high-resolution point estimation and rigorous uncertainty quantification. Applied empirically in Germany, the method generates fine-grained spatial maps of caregiving opportunity costs. Monte Carlo simulations demonstrate that, compared to empirical best linear unbiased prediction (EBLUP) and standard random forests, our estimator reduces bias by 32% and achieves a 94% coverage rate for 95% confidence intervals—substantially improving the reliability and policy relevance of small-area estimates.
📝 Abstract
Evidence-based policy-making requires reliable, spatially disaggregated indicators. The framework of mixed effects random forests leverages the advantages of random forests and hierarchical data in small area estimation. These methods require typically access to auxiliary information on population-level, which is a strong limitation for practitioners. In contrast, our proposed method - for point and uncertainty estimation - abstains from access to unitlevel population data but adaptively incorporates aggregated auxiliary information through calibration-weights. We demonstrate its usage for estimating opportunity cost of care work for Germany from the Socio-Economic Panel and census aggregates. Simulation studies evaluate our proposed method.
