Continuous Treatment DiD#

The continuous treatment DiD module provides methods for handling continuous treatment variables in difference-in-differences settings. This includes nonparametric instrumental variables (NPIV) estimation and spline-based methods. The implementation follows Callaway, Goodman-Bacon, and Sant’Anna (2024).

Main Functions#

cont_did

Compute difference-in-differences with a continuous treatment.

Non-Parametric Instrumental Variables#

Core NPIV Functions#

`npiv`	Estimate nonparametric instrumental variables model with uniform confidence bands.
`npiv_est`	Core sieve TSLS estimation for the nonparametric IV model.
`npiv_choose_j`	Select optimal B-spline dimensions.
`npiv_j`	Implement Lepski's method for optimal sieve dimension selection.
`npiv_jhat_max`	Determine the upper limit of the sieve dimension grid.

Uniform Confidence Bounds#

`compute_cck_ucb`	Compute honest and adaptive UCBs.
`compute_ucb`	Compute uniform confidence bands for nonparametric instrumental variables.

Spline Functions#

Product Splines#

prodspline

Create multivariate spline basis with B-spline components.

B-Spline Basis#

BSpline

B-spline basis functions.

Result Objects#

`NPIVResult`	Container for nonparametric instrumental variables estimation results.
`PTEResult`	Container for panel treatment effects results.
`PTEAggteResult`	Container for aggregated panel treatment effect parameters.
`GroupTimeATTResult`	Container for group-time average treatment effect results.
`DoseResult`	Container for continuous treatment dose-response results.