Frontier is a dataset for Maximum Likelihood Estimation of Stochastic Frontier production and cost functions. It implements two model specifications: the error components specification with time-varying efficiencies and a model where firm effects are influenced by explanatory variables. The dataset was created by authors Tim Coelli and Arne Henningsen, based on the methodologies of Battese and Coelli from 1992 and 1995.
Use Cases
- Estimate time-varying technical efficiency based on the error components specification.
- Model how firm-specific variables influence efficiency scores using the Battese and Coelli (1995) specification.
- Benchmark production or cost efficiency across firms or industries using stochastic frontier analysis.
- Conduct econometric research on productivity and efficiency frontiers.
Strengths
- Implements two established econometric specifications from peer-reviewed literature (Battese and Coelli, 1992 & 1995).
- Authored by recognized contributors in the field, Tim Coelli and Arne Henningsen.
Limitations
- Column-level documentation is absent; field semantics must be inferred after download.
- Row count and dataset scale are unknown, which may limit suitability assessment.
- Last update date is unknown; freshness unverified.
Provenance
- Source
- Tim Coelli, Arne Henningsen
- Collection Method
- Likely contains simulated or example data for demonstrating the Maximum Likelihood Estimation method.