metafrontier: Unified metafrontier analysis for efficiency and productivity in R

Comparing the technical efficiency of firms that operate under different production technologies requires a metafrontier framework that envelops group-specific frontiers. Despite a substantial methodological literature and growing applied demand, no comprehensive R pack-age for metafrontier analysis has been available on CRAN. We introduce metafrontier, a package that provides a unified interface for three complementary approaches: the deterministic metafrontier of Battese, Rao, and O’Donnell (2004), the stochastic metafrontier of Huang, Huang, and Liu (2014), and DEA-based metafrontier models. The package estimates group-specific frontiers, constructs the metafrontier envelope, and decomposes efficiency into group technical efficiency and the technology gap ratio. Additional features include the metafrontier Malmquist productivity index with three-way decomposition, bootstrap confidence intervals for technology gap ratios, Murphy–Topel corrected standard errors, latent class metafrontier estimation via the EM algorithm, panel stochastic frontier models with time-varying inefficiency, and ggplot2 visualisation methods. An interoperability layer accepts pre-fitted models from sfaR, frontier, and Benchmarking. We illustrate the package with simulated examples and a Monte Carlo study demonstrating parameter recovery under both metafrontier estimators. All examples and simulations are fully reproducible and complete in under ten minutes on a standard laptop.