Citation

Prudencio, D. (2023). Productivity in procurement auctions of pavement contracts in Mexico. Journal of Productivity Analysis. https://doi.org/10.1007/s11123-025-00733-3

BibTeX Entry

@article{prudencio2023procurement,
  author    = {Prudencio, Daniel},
  title     = {Productivity in procurement auctions of pavement contracts in {Mexico}},
  journal   = {Journal of Productivity Analysis},
  year      = {2023},
  publisher = {Springer},
  doi       = {10.1007/s11123-025-00733-3}
}

Methodological Code

The paper uses three separate computational environments, each for a different stage of the analysis:

Stage 1: Structural auction model (Julia)

The nonparametric GPV estimation, including kernel density estimation of bid distributions and recovery of pseudo-costs via the inversion formula, is implemented in Julia. Julia’s performance is well-suited to the computationally intensive kernel bandwidth selection and cross-validation steps.

Stage 2: Stochastic frontier analysis (Stata)

The SFA models (half-normal and truncated-normal) are estimated via maximum likelihood in Stata.

Stage 3: Figures in the paper (R)

All figures published in the Journal of Productivity Analysis article were produced in R using ggplot2.

About This Website

This interactive summary was built with Quarto, a scientific publishing system that renders Markdown and code into HTML. The charts on this website are reproductions of the paper’s figures, generated using Python and R, drawing on the same underlying data as the published analysis.

The source files for this website are available at github.com/danstad/claude-code-my-workflow.