Skip to contents

A balanced county-level panel of Walmart store openings and retail employment in the United States, 1977–1999. No county in the panel housed a Walmart before 1986, and openings are staggered thereafter, so that the data are a natural testbed for (non-binary) staggered difference-in-differences. This is the dataset used in the empirical application of Wooldridge (2026).

Usage

walmart

Format

A data frame with 29,624 rows (1,288 counties x 23 years) and 5 variables:

fips

County FIPS code.

year

Year (1977–1999).

cohort

Year in which the county's first Walmart opened (1986–1999), or 0 if the county did not have a Walmart by 1999 (never treated).

n_open

Number of Walmart stores open in the county in that year.

log_retail_emp

Log of county retail employment (rounded to four decimal places).

Source

Subset of the data constructed by Brown and Butts (2025), as distributed with the replication files for Wooldridge (2026). See the References below.

Details

The underlying data were constructed by Brown and Butts (2025), with the sample of counties following the selection criteria of Basker (2005), and were subsequently used by Wooldridge (2025, 2026). We are grateful to the original authors for permission to redistribute them here.

Please note that this is a subset of the original data. It retains only the five variables needed to illustrate the etwfe functionality, whereas the original contains many more (e.g., employment in other sectors and pre-treatment county characteristics). The outcome variable has also been rounded to reduce file size, although this does not affect results at the precision reported by Wooldridge (2026). Users interested in the full dataset should consult the original sources.

References

Basker, Emek (2005). Job Creation or Destruction? Labor Market Effects of Wal-Mart Expansion. Review of Economics and Statistics, 87(1), 174-183. Available: https://doi.org/10.1162/0034653053327568

Brown, Nicholas and Kyle Butts (2025). Dynamic Treatment Effect Estimation with Interactive Fixed Effects and Short Panels. Journal of Econometrics, 250, 106013. Available: https://doi.org/10.1016/j.jeconom.2025.106013

Wooldridge, Jeffrey M. (2025). Two-way fixed effects, the two-way Mundlak regression, and difference-in-differences estimators. Empirical Economics, 69, 2545-2587. Available: https://doi.org/10.1007/s00181-025-02807-z

Wooldridge, Jeffrey M. (2026). A Heterogeneous Slopes Approach to Difference-in-Differences with Non-Binary Treatments. Working paper. Available: https://doi.org/10.2139/ssrn.7530698

Examples

data("walmart", package = "etwfe")

# Effect of each additional Walmart store on (log) retail employment, cf.
# Wooldridge (2026, Table 7.1)
mod = etwfe(
  log_retail_emp ~ 0, tvar = year, gvar = cohort, data = walmart,
  ivar = fips, vcov = ~fips,
  wvar = n_open
  )
emfx(mod)
#> 
#>  .Dtreat Estimate Std. Error    z Pr(>|z|)    S  2.5 % 97.5 %
#>     TRUE   0.0386    0.00506 7.63   <0.001 45.3 0.0287 0.0485
#> 
#> Term: n_open
#> Type: response
#> Comparison: +1
#>