DiD–ActoCAUSAL ANALYSIS

THE STATISTICAL TOOLKIT

Know the method.
Understand the evidence.

Modern difference-in-differences, synthetic comparisons and randomized designs—with named estimators, explicit eligibility, and tests you can interpret.

21registered estimators, with availability labeled below
9HonestDiD restriction families, subject to exact-profile support
1approved plan tying methods, assumptions and checks together

Primary is the main estimate in an eligible plan. Robustness is a comparison, subject to compatible targets. Exact profile means the design, options and inference must match a supported procedure. Expert MCP uses a separate request path. Python only / no planner role is not a normal guided MCP choice.

The Python registry contains 21 canonical estimators. That count includes experimental and research-only procedures. Start with the design and target: the planner then selects eligible methods, explains its choices and lists the checks included in your analysis.

DIFFERENT UNITS. DIFFERENT START DATES.

Modern staggered-adoption DiD.

Study an overall effect or its evolution after treatment. These procedures differ in how they build comparisons, aggregate effects and estimate uncertainty.

Primary / robustness

Callaway–Sant’Anna (CS)

Estimates cohort-by-period treatment effects, then aggregates to overall, event-time, cohort or calendar targets.

Design, availability & inference

Binary absorbing treatment with eligible never-treated or not-yet-treated comparisons. Conditional panel profiles support 1–5 finite, unit-constant baseline covariates; the exact profile must match.

What you receive: Influence-function covariance, pointwise intervals, and multiplier-bootstrap simultaneous bands for supported vectors.

callaway_santanna

Exact profile

Sun–Abraham

Interaction-weighted event study for effects that vary across adoption cohorts and time since treatment.

Design, availability & inference

Balanced, unweighted panel, never-treated controls, no covariates, reference event −1 and the pinned fixest 0.14.2 convention. Dynamic robustness by default; may be explicitly requested as the dynamic primary. Other variants are unpromoted.

What you receive: Unit-cluster t inference at 95%, cohort-event support and a joint pretrend test. Not an HonestDiD source.

sun_abraham

Exact profile

Borusyak–Jaravel–Spiess (BJS)

Imputes untreated outcomes from a model fitted to untreated observations; reports overall ATT or a post-treatment event path.

Design, availability & inference

Exact balanced, unweighted, unadjusted LOCO simple/dynamic profiles. Ask the planner to confirm the exact imputation and covariance options for your design.

What you receive: Unit-cluster leave-own-cluster-out smart-residual covariance; pretrend leads are a separate diagnostic.

borusyak_jaravel_spiess

Robustness / exact profile

Wooldridge extended TWFE (ETWFE)

Saturated cohort-by-calendar regression with a treated-row aggregate effect.

Design, availability & inference

Balanced, unweighted, zero-anticipation panel. Exact unadjusted profile, or exactly one unit-constant baseline covariate with never-treated controls. Uses the pinned fetwfe 1.10.0 treated-row convention.

What you receive: Unit-cluster covariance and normal-reference 95% intervals. Alternative aggregation choices need their own eligibility.

wooldridge_etwfe

Exact profile

Fused extended TWFE (FETWFE)

Regularizes the treatment-effect structure, then refits selected effects with native post-selection OLS.

Design, availability & inference

Exact native post_ols simple/dynamic profiles: balanced, unweighted, no covariates and never-treated controls. The CRAN cran_bridge route is withdrawn from governed use.

What you receive: Post-selection unit-cluster CR1 normal inference within the supported profile. A selection with no effect coordinates produces unavailable effects, not proof of a zero effect.

fused_etwfe

Robustness

Gardner two-stage DiD

Two-stage estimation of an overall ATT with first-stage estimation uncertainty carried into inference.

Design, availability & inference

Eligible unadjusted panel designs using the default analytical two-stage GMM procedure.

What you receive: Two-stage unit-clustered GMM covariance with a normal reference. Legacy bootstrap is a separate option, not the promoted default.

gardner_did2s

Robustness / exact profile

Efficient staggered GMM

Combines clean two-by-two comparisons through an efficient generalized method of moments procedure.

Design, availability & inference

Balanced, unweighted, no-covariate panel; converged pooled or cohort-specific Toeplitz weighting and the exact 95% profile. Unrestricted covariance weighting remains unpromoted.

What you receive: GMM covariance and normal intervals; comparison support, reweighting convergence and Hansen overidentification diagnostics.

staggered_gmm

MATCH THE DESIGN

Common adoption, donor comparisons and time series.

A common start date, baseline adjustment or a small treated group can call for a different approach.

Primary for common adoption; diagnostic if staggered

Mean DiD / two-way fixed effects (TWFE)

Conventional treated-versus-control change comparison with unit and time effects.

Design, availability & inference

The common-adoption default when its design fits. Under staggered adoption the coefficient is labeled a TWFE coefficient, not a heterogeneity-robust overall ATT.

What you receive: Cluster-robust inference; the implementation supports unit or unit-and-time clustering, subject to the approved plan.

twfe

Dynamic / pretrend diagnostic

TWFE event study

Relative-time coefficients show the outcome path around adoption and support joint lead checks.

Design, availability & inference

Eligible common-adoption dynamic analysis and pretrend diagnostics. Under staggered heterogeneity, plain TWFE event-study coefficients remain diagnostic.

What you receive: Unit-clustered covariance and a joint pretrend F test using eligible leads.

event_study

Primary for adjusted common adoption

Doubly robust DiD

Combines propensity-score weighting and an outcome model for an adjusted overall ATT.

Design, availability & inference

One common adoption date, controls and declared finite baseline covariates. Requires acceptable propensity overlap, effective sample size and weight concentration.

What you receive: Unit bootstrap with complete trajectories. Every required draw must succeed.

doubly_robust_did

Robustness for adjusted common adoption

Inverse-probability-weighted DiD (IPW)

Reweights control outcome changes using propensity odds.

Design, availability & inference

Common adoption and declared baseline covariates with supported propensity overlap. The propensity model is part of the identifying specification.

What you receive: Unit bootstrap; overlap, effective sample size and concentrated-weight checks.

ipw_did

Primary for eligible few-treated designs

Synthetic control

Builds a comparison path from a weighted pool of untreated donor units.

Design, availability & inference

Balanced common-adoption panel, usable pre/post paths and adequate donors for treated-size-preserving placebo groups.

What you receive: Pre-fit RMSPE, in-space placebo RMSPE-ratio p-values and donor diagnostics. Placebo-SD normal intervals are heuristic, not coverage-guaranteed sampling intervals.

synthetic_control

Robustness

Synthetic DiD (SDID)

Combines unit and time weights with a level correction for a common intervention.

Design, availability & inference

Eligible balanced donor design with at least two treated-size-preserving placebo assignments and remaining donors for each fit.

What you receive: Placebo RMSPE diagnostics and heuristic placebo-SD intervals; fit and solver checks.

synthetic_did

Primary for eligible time-series designs

Interrupted time series (ITS)

Models intervention-associated level and slope changes and the average post-intervention effect.

Design, availability & inference

Suitable single-intervention series. Interpretation depends on the continuation of the pre-intervention trend and the absence of competing changes.

What you receive: Newey–West heteroskedasticity-and-autocorrelation-consistent (HAC) inference.

interrupted_time_series

SEPARATE INPUT CONTRACTS

Repeated cross sections and randomized assignments.

These supported expert paths preserve what makes the design different. They are not ordinary panel data with a different label.

Separate expert MCP profile

Conditional repeated-cross-section CS

Cohort-time ATT and simple, dynamic, group or calendar aggregation when different respondents are observed each period.

Design, availability & inference

Exact unweighted profile with 1–5 baseline/pretreatment covariates, unique respondents, explicit stable composition and independent cross sections, and adequate four-arm support. Uses the conditional_rcs request path, outside the ordinary panel conversation.

What you receive: Respondent influence-function covariance and receipt-bound vector bands. Saved results and reports are available; this expert path keeps claim and export authorization false.

callaway_santanna_rcs

Separate randomized MCP workflow

Randomized difference in means

Sample intent-to-treat (ITT) for individual or cluster randomization, with declared strata where applicable.

Design, availability & inference

A valid assignment contract is required. Cluster assignment targets an equal-cluster-weight sample effect, not an individual-weighted population effect.

What you receive: Exact or Monte Carlo two-sided sharp-null randomization tests. Resolution depends on the allowed assignment space.

randomized_difference_in_means

Non-confirmatory randomized robustness

Lin covariate-adjusted ITT

Fully interacted baseline-covariate adjustment alongside the randomized difference-in-means result.

Design, availability & inference

Declared eligible baseline covariates and adequate design rank, including under reassignment. Not a confirmatory replacement for the primary procedure.

What you receive: Sharp-null randomization inference, assignment-rank checks and few-cluster/discrete-p-value limitations.

lin_covariate_adjusted_itt

CHALLENGE THE RESULT

Tests, diagnostics and sensitivity.

Ask what threatens the interpretation, how uncertain the estimate is, and how the conclusion changes under specified departures. Checks run only when the approved plan and data support them.

Joint pretrend tests

Test whether eligible pre-treatment coefficients are jointly zero. Source-matched Wald test for staggered CS; cluster F test for plain event study and the exact Sun–Abraham profile. Governed gating needs at least two supported leads. Read a non-rejection alongside the plotted leads and the test’s ability to detect a meaningful difference.

Placebo in time

Move the intervention into an eligible pre-treatment window as a falsification check. Available when predeclared and supported. A substantive rejection gates the evidence; a structurally unavailable check is not_assessed. Timing assumptions still matter.

Unit-permutation and in-space placebos

Compare the observed result with reassigned treatment or donor-placebo results. Procedure-specific and dependent on the assignment comparison being meaningful. Synthetic methods preserve the treated-group size and report RMSPE-ratio permutation diagnostics.

HonestDiD / Rambachan–Roth sensitivity

Calculate intervals under specified departures from parallel trends. Formal relative-magnitude and smoothness procedures plus exact supported directional/composite cells. Requires a certified event path, full covariance and sufficient stable support; see the detailed families below.

Dynamic heterogeneity test

Jointly test equality across post-treatment event-time effects. Uses supported coordinates and joint covariance. This concerns variation over event time; it is not automatically a test of differences between cohorts.

Cross-estimator equivalence and conflict

Compare eligible estimates of the same realized target using simultaneous difference intervals. Confirmatory equivalence needs a predeclared margin and eligible uncertainty. Automatic comparison uses a conservative standard-error bound and Bonferroni adjustment. Overlapping marginal intervals do not establish equivalence.

Hansen overidentification

Assess compatibility of the overidentifying moment restrictions used by efficient staggered GMM. Available on the eligible GMM path alongside clean-comparison support and convergence checks. Review the assignment assumptions separately from the moment-restriction test.

Propensity overlap and weight diagnostics

Check common support, effective sample size, separation and concentration of weights. Relevant to DR, IPW and conditional procedures. Invalid support can block fitting; poor overlap calls for a design or sample review before estimation.

Synthetic donor and fit diagnostics

Review pre-treatment RMSPE, donor concentration, effective donor count, solver convergence and placebo feasibility. Synthetic control also includes leave-one-donor and contamination/spillover diagnostics. Use these to identify fragile donor choices, then review possible spillovers using knowledge of the locations and intervention.

Panel, cohort and event support checks

Validate data structure and identify which comparisons support each requested effect. Checks include treatment timing, baseline availability, contributing cells and event-window support. Thin or unsupported coordinates may retain points while losing inferential eligibility.

Randomization design checks

Validate the assignment contract, admissible reassignments and p-value resolution. Separate randomized workflow. Lin adjustment adds covariate-rank checks; few clusters and a small assignment space limit what can be resolved.

Read the tests & inference guideExplore all nine HonestDiD families

LIBRARY SCOPE, CLEARLY LABELED

Implemented beyond the guided workflow.

These complete the 21-name registry. They are useful to understand the library, but the current planner does not offer them as general governed analysis choices.

Implemented; no current planner role

Goodman–Bacon decomposition

Reconstructs unweighted TWFE and reports the weight on problematic treated-versus-treated comparisons.

Design, availability & inference

Balanced panel; weighted decomposition is not implemented. Available in the Python catalog, but not selected by the current governed planner.

What you receive: Decomposition and reconstruction diagnostics; not a new causal ATT estimate.

goodman_bacon

Implemented; no current planner role

Counterfactual forecast

Fits a treated series to controls and optional trend before the intervention, then projects a counterfactual.

Design, availability & inference

Requires stable post-period relationships and an appropriate error process. Python procedure, not a current automatic MCP planning choice.

What you receive: Forecast uncertainty for the average post effect; no per-period uncertainty band.

counterfactual_forecast

Implemented; no current planner role

Changes-in-changes

Distributional counterfactual mapping with an overall ATT and quantile-treatment-effect provenance.

Design, availability & inference

Common adoption and untreated controls. Multi-period results map each post period from the final pre period. Python procedure, not a current automatic MCP planning choice.

What you receive: Unit-trajectory bootstrap and period-specific/quantile diagnostics.

changes_in_changes

Experimental Python API

Continuous-dose DiD

Dose-response ATT(d) on an explicitly supported dose grid.

Design, availability & inference

Separate balanced continuous-dose input contract. Unpromoted and unavailable through governed panel planning; covariates are rejected.

What you receive: Pointwise unit-clustered normal inference within the experimental implementation; no general governed claim.

continuous_did

Aliases: bjs, did_imputation and didimputation identify BJS; etwfe identifies Wooldridge ETWFE; fetwfe identifies fused ETWFE; gmm_staggered identifies staggered GMM. Aliases are not additional methods.

Open the full estimator reference

YOUR FIRST QUESTION STARTS HERE

Bring a question.
Build the evidence.

Start with a shared sample, learn the workflow, then bring your own data.

Read the getting-started guideOpen your workspace