Working Paper: NBER ID: w15690
Authors: Bennett T. McCallum
Abstract: So-called "spurious regression" relationships between random-walk (or strongly autoregressive) variables are generally accompanied by clear signs of severe autocorrelation in their residuals. A conscientious researcher would therefore not end an investigation with such a result, but would likely re-estimate with an autocorrelation correction. Simulations show, for several typical cases, that the test-rejection statistics for the re-estimated relationships are very close to the true values, so do not yield results of the spurious type.
Keywords: No keywords provided
JEL Codes: C22; C29
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| spurious regression relationships between random walk or strongly autoregressive variables (C22) | significant autocorrelation in their residuals (C22) |
| significant autocorrelation in their residuals (C22) | false significant relationship (C52) |
| autocorrelation (C22) | reestimate regression using iterated Cochrane-Orcutt procedure (C51) |
| correcting for autocorrelation (C22) | rejection frequencies of the null hypothesis approach true significance levels (C12) |
| integrated moving-average variables (C32) | simple autoregressive corrections may not suffice (C22) |