From jpam-skills
Defends causal identification for JPAM manuscripts using RCTs, DiD/event study, RD/kink, IV, or synthetic control. Strengthens design assumptions without writing estimation code.
How this skill is triggered — by the user, by Claude, or both
Slash command
/jpam-skills:jpam-research-designThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Credible identification is JPAM's **core bar**. The journal evaluates the effects of real policies and
Credible identification is JPAM's core bar. The journal evaluates the effects of real policies and
programs, so the design must connect the theory of change (jpam-theory-building) to evidence a
policymaker can trust. State the estimand, the assumptions that license a causal reading, and
how each is defended — then rule out the single strongest rival explanation. Selection-on-
observables alone rarely clears the bar.
For the single strongest rival explanation (selection, anticipation, concurrent policy, mean reversion), write one sentence: "If the rival were driving the result, the data would look like ___; instead they look like ___." If you cannot, the design does not yet identify the policy effect.
A state raises a benefit eligibility threshold; the team uses an RD at the income cutoff. The design write-up states the estimand (effect at the threshold), defends continuity (covariates smooth across the cutoff, no manipulation by a density test), reports bandwidth and local-polynomial robustness, and adjudicates the strongest rival: "If families were sorting just under the cutoff to qualify, the running-variable density would spike there; it does not." It then flags that the estimate is local and discusses how it might differ away from the threshold. (Illustrative.)
【Design】RCT / DiD-event-study / RD-kink / IV / synthetic control
【Estimand + population】what is identified, for whom
【Key assumption(s)】and how each is defended
【Rival ruled out】the adjudication sentence
【Inference】clustering, multiple-testing, weak-IV plan
【Next】jpam-data-analysis
../../resources/code/ — runnable DiD / IV / RD / DML skeletons (Stata + Python)../../../shared-resources/empirical-methods/reviewer-objection-checklist.md — objections by identification strategy, each with a preemptionnpx claudepluginhub brycewang-stanford/awesome-journal-skills --plugin jpam-skillsStress-tests quasi-experimental policy-evaluation designs (DID/event study, IV, RDD/bunching, RCT) for AEJ: Economic Policy manuscripts.
Stress-tests causal identification for JHR empirical-micro manuscripts: RCT, DID, RDD, IV, event studies, decompositions, policy shocks, and reconciliation with prior estimates.
Stress-tests causal identification strategies for JPubE manuscripts against public-finance norms before drafting tables.