Fast and user-friendly econometrics estimations: The R package fixest

fixest is an R package for fast and flexible econometric estimation, providing a comprehensive toolkit for applied researchers. The package particularly excels at fixed-effects estimation, supported by a novel fixed-point acceleration algorithm implemented in C++. This algorithm achieves rapid convergence across a broad class of data contexts and further enables estimation of complex models, including those with varying slopes, in a highly efficient manner. Beyond computational speed, fixest provides a unified syntax for a wide variety of models: ordinary least squares, instrumental variables, generalized linear models, maximum likelihood, and difference-in-differences estimators. An expressive formula interface enables multiple estimations, stepwise regressions, and variable interpolation in a single call, while users can make on-the-fly inference adjustments using a variety of built-in robust standard errors. Finally, fixest provides methods for publication-ready regression tables and coefficient plots. Benchmarks against leading alternatives in R, Python, and Julia demonstrate best-in-class performance, and the paper includes many worked examples illustrating the core functionality.

💡 Research Summary

The paper presents fixest, an R package that combines a novel C++‑based fixed‑point acceleration algorithm with a unified, expressive formula interface to deliver fast and user‑friendly econometric estimation. The authors begin by motivating the need for speed in fixed‑effects models, noting that traditional approaches (e.g., within‑transformation in lm or plm) become computationally burdensome as the number of observations and fixed‑effects dimensions grow.

The core of fixest is a fixed‑point algorithm that iteratively demeans variables with respect to one or more fixed‑effects groups until convergence. Implemented in highly optimized C++ and parallelized via OpenMP, the algorithm reduces both memory footprint and CPU time, achieving convergence in far fewer iterations than conventional methods. The authors provide a theoretical justification based on the Frisch‑Waugh‑Lovell theorem and prove convergence under broad conditions, extending the approach to generalized linear models and maximum‑likelihood settings.

From the user’s perspective, fixest offers a family of functions—feols, feglm, fepois, feiv, etc.—that all share a common syntax:

dependent ~ regressors | fixed_effects