Honest DiD

The Honest DiD module provides sensitivity analysis tools for potential violations of the parallel trends assumption, following Rambachan and Roth (2023). This allows researchers to assess the robustness of their DiD estimates to various forms of pre-trend violations.

Main Functions

honest_did	Compute sensitivity analysis for event study estimates.
construct_original_cs	Construct original (non-robust) confidence set.
create_sensitivity_results_rm	Perform sensitivity analysis using relative magnitude bounds.
create_sensitivity_results_sm	Perform sensitivity analysis using smoothness restrictions.

Confidence Intervals

ARP Confidence Intervals

compute_arp_ci	Compute Andrews-Roth-Pakes (ARP) confidence interval with no nuisance parameters.
compute_arp_nuisance_ci	Compute Andrews-Roth-Pakes (ARP) confidence interval with nuisance parameters.
compute_least_favorable_cv	Compute least favorable critical value.
compute_vlo_vup_dual	Compute the truncation bounds for the test statistic using dual approach with bisection.
lp_conditional_test	Perform Andrews-Roth-Pakes (ARP) test of moment inequality with nuisance parameters.
test_in_identified_set	Test whether \(\bar{\theta}\) lies in the identified set using the ARP conditional approach.
test_in_identified_set_flci_hybrid	Hybrid test combining fixed-length confidence interval (FLCI) constraints with ARP conditional test.
test_in_identified_set_lf_hybrid	Conditional-least favorable (LF) hybrid test.

Fixed-Length Confidence Intervals

compute_flci	Compute fixed-length confidence intervals under smoothness restrictions.
folded_normal_quantile	Compute quantile of folded normal distribution \(cv_{\alpha}(t)\).
maximize_bias	Find worst-case bias subject to standard deviation constraint \(h\).
minimize_variance	Find the minimum achievable standard deviation \(h\).

Restriction Types

Relative Magnitude Restrictions

compute_conditional_cs_rm	Compute conditional confidence set for \(\Delta^{RM}(\bar{M})\).
compute_identified_set_rm	Compute identified set for \(\Delta^{RM}(\bar{M})\).
compute_conditional_cs_rmb	Compute conditional confidence set for \(\Delta^{RM}(\bar{M})\) with bias sign restrictions.
compute_identified_set_rmb	Compute identified set for \(\Delta^{RMB}(\bar{M})\).
compute_conditional_cs_rmm	Compute conditional confidence set for \(\Delta^{RMI}(\bar{M})\).
compute_identified_set_rmm	Compute identified set for \(\Delta^{RMI}(\bar{M})\).

Smoothness Restrictions

compute_conditional_cs_sd	Compute conditional confidence set for \(\Delta^{SD}(M)\).
compute_identified_set_sd	Compute identified set for \(\Delta^{SD}(M)\).
compute_conditional_cs_sdb	Compute conditional confidence set for \(\Delta^{SDB}(M)\).
compute_identified_set_sdb	Compute identified set for \(\Delta^{SDB}(M)\).
compute_conditional_cs_sdm	Compute conditional confidence set for \(\Delta^{SDI}(M)\).
compute_identified_set_sdm	Compute identified set for \(\Delta^{SDI}(M)\).

Combined Restrictions

compute_conditional_cs_sdrm	Compute conditional confidence set for \(\Delta^{SDRM}(\bar{M})\).
compute_identified_set_sdrm	Compute identified set for \(\Delta^{SDRM}(\bar{M})\).
compute_conditional_cs_sdrmb	Compute conditional confidence set for \(\Delta^{SDRMB}(\bar{M})\).
compute_identified_set_sdrmb	Compute identified set for \(\Delta^{SDRMB}(\bar{M})\).
compute_conditional_cs_sdrmm	Compute conditional confidence set for \(\Delta^{SDRMM}(\bar{M})\).
compute_identified_set_sdrmm	Compute identified set for \(\Delta^{SDRMM}(\bar{M})\).

Result Objects

HonestDiDResult	Container for honest_did sensitivity analysis results.
OriginalCSResult	Container for original confidence set assuming exact parallel trends.
SensitivityResult	Container for a single sensitivity analysis result.
FLCIResult	Container for fixed-length confidence interval results.
APRCIResult	Container for APR confidence interval results.
ARPNuisanceCIResult	Container for ARP confidence interval results with nuisance parameters.