Direct and spillover ATT under an exposure mapping

With a binary own-treatment D and a binary neighbourhood exposure E the three exposure conditions are direct (D = 1), spillover (D = 0, E = 1) and control (D = 0, E = 0). Each contrast against the shared control condition is estimated by the doubly robust moment of Sant'Anna and Zhao (2020), equation (2.6), restricted to the two conditions being compared, so the two contrasts are on the same scale.

Usage

Drspr(y, D, X = NULL, exposure = NULL)

Arguments

y

Outcome change, one entry per unit.

D

Own binary treatment.

X

Optional baseline covariates.

exposure

Neighbourhood exposure; any strictly positive value counts as exposed.

Value

List with estimate, att_direct, att_spillover, se_direct, se_spillover, total, n_direct, n_spill, n_control, n.

References

Aronow, P. M. and Samii, C. (2017). Annals of Applied Statistics 11(4), 1912-1947. Sant'Anna, P. H. C. and Zhao, J. (2020). Journal of Econometrics 219(1), 101-122.

Examples

V <- c(1, 2, 3, 4, 5, 6, 7, 8)
Drspr(V, V)
#> $estimate
#> [1] NaN
#> 
#> $att_direct
#> [1] NaN
#> 
#> $att_spillover
#> [1] NaN
#> 
#> $se_direct
#> [1] NaN
#> 
#> $se_spillover
#> [1] NaN
#> 
#> $total
#> [1] NaN
#> 
#> $n_direct
#> [1] 8
#> 
#> $n_spill
#> [1] 0
#> 
#> $n_control
#> [1] 0
#> 
#> $n
#> [1] 8
#> 
#> $method
#> [1] "DR-DiD with spillover"
#> 
#> attr(,"class")
#> [1] "morie_rich_result" "list"