Working Paper: CEPR ID: DP18271
Authors: Atsushi Inoue; Oscar Jorda; Guido Kuersteiner
Abstract: An impulse response function describes the dynamic evolution of an outcome variable following a stimulus or treatment. A common hypothesis of interest is whether the treatment affects the outcome. We show that this hypothesis is best assessed using significance bands rather than relying on commonly displayed confidence bands. Under the null hypothesis, we show that significance bands are trivial to construct with standard statistical software using the LM principle, and should be reported as a matter of routine when displaying impulse responses graphically.
Keywords: local projections; impulse response; instrumental variables; significance bands; wild block bootstrap
JEL Codes: C11; C12; C22; C32; C44; E17
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| significance bands (C20) | assessment of hypothesis regarding the effect of treatment on outcome (y) (C12) |
| confidence bands (C46) | misinterpretations in assessing impulse response coefficients (C22) |
| joint hypotheses (C12) | construction of significance bands (C46) |
| significance bands (C20) | clearer picture of statistical significance of impulse responses (C22) |
| LM principle (L00) | reliable estimation of response of y to s (C51) |
| significance bands (C20) | tighter intervals than confidence bands (E32) |
| treatment variable (s) (C32) | outcome variable (y) (C29) |