Skip to contents

Callaway-Sant'Anna staggered DiD (module-name entry point)

Source: R/causdidcs_native.R
morie_causdidcs.Rd

Entry point onto morie_cssant; see there for the full description of the group-time ATT(g, t) estimator, its clean comparison groups and its aggregations.

Usage

morie_causdidcs(y, D, unit, time, cohort = NULL, control = "notyet")

Arguments

y

Outcome, long format.

D

Binary 0/1 absorbing treatment, long format.

unit, time

Panel identifiers.

cohort

Optional per-unit adoption period, constant within unit, as an alternative to deriving it from D.

control

"notyet" (default) or "never".

Value

The list returned by morie_cssant.

References

Callaway, B. and Sant'Anna, P. H. C. (2021). Difference-in-differences with multiple time periods. Journal of Econometrics, 225(2), 200-230.

Examples

set.seed(1)
nu <- 6; T <- 5
unit <- rep(1:nu, each = T)
time <- rep(1:T, nu)
ft <- rep(c(Inf, Inf, Inf, 4, 4, 3), each = T)
D <- as.integer(time >= ft)
y <- rnorm(nu * T) + D * 1.5
morie_causdidcs(y, D, unit, time)
#> $estimate
#> [1] 0.9134894
#> 
#> $se
#> [1] 0.4239056
#> 
#> $ci
#> [1] 0.08264971 1.74432915
#> 
#> $att_gt
#> $att_gt$`2|0`
#> [1] 0.2772586
#> 
#> $att_gt$`2|2`
#> [1] 0.4904441
#> 
#> $att_gt$`2|3`
#> [1] 1.54583
#> 
#> $att_gt$`2|4`
#> [1] 2.044359
#> 
#> $att_gt$`3|0`
#> [1] -0.3333132
#> 
#> $att_gt$`3|1`
#> [1] -0.7193809
#> 
#> $att_gt$`3|3`
#> [1] 0.1817652
#> 
#> $att_gt$`3|4`
#> [1] 0.9751313
#> 
#> 
#> $n_by_cell
#> $n_by_cell$`2|0`
#> [1] 1 5
#> 
#> $n_by_cell$`2|2`
#> [1] 1 5
#> 
#> $n_by_cell$`2|3`
#> [1] 1 3
#> 
#> $n_by_cell$`2|4`
#> [1] 1 3
#> 
#> $n_by_cell$`3|0`
#> [1] 2 3
#> 
#> $n_by_cell$`3|1`
#> [1] 2 3
#> 
#> $n_by_cell$`3|3`
#> [1] 2 3
#> 
#> $n_by_cell$`3|4`
#> [1] 2 3
#> 
#> 
#> $event
#> $event$`0`
#> [1] 0.2846582 0.7396668
#> 
#> $event$`1`
#> [1] 1.1653642 0.4979689
#> 
#> $event$`2`
#> [1] 2.044359 0.363257
#> 
#> 
#> $cohort_att
#> $cohort_att$`2`
#> [1] 1.3602109 0.2856451
#> 
#> $cohort_att$`3`
#> [1] 0.5784483 0.6120612
#> 
#> 
#> $calendar
#> $calendar$`2`
#> [1] 0.4904441 0.3510050
#> 
#> $calendar$`3`
#> [1] 0.6364535 1.0534611
#> 
#> $calendar$`4`
#> [1] 1.3315405 0.6189579
#> 
#> 
#> $pretrend
#> $pretrend$`2|0`
#> [1] 0.2772586
#> 
#> $pretrend$`3|0`
#> [1] -0.3333132
#> 
#> $pretrend$`3|1`
#> [1] -0.7193809
#> 
#> 
#> $pretrend_max_abs
#> [1] 0.7193809
#> 
#> $pretrend_note
#> [1] "cells with t < g are not effects; they are the parallel-trends check and are reported separately"
#> 
#> $control_group
#> [1] "notyet"
#> 
#> $cohorts
#> [1] 2 3
#> 
#> $n_cells
#> [1] 8
#> 
#> $n_units
#> [1] 6
#> 
#> $n_periods
#> [1] 5
#> 
#> $clean_controls
#> [1] "no already-treated unit is ever used as a control, so no negative-weight comparison can enter the aggregate"
#> 
#> $method
#> [1] "Callaway-Sant'Anna (2021) group-time ATT(g,t)"
#>