Working Paper: NBER ID: w2765
Authors: Jeffrey A. Miron; Stephen P. Zeldes
Abstract: We examine two measures of monthly manufacturing production. The first is the index of industrial production; the second is constructed from the accounting identity that output equals sales plus the change in inventories. We show that the means, variances, and serial correlation coefficients of the log growth races differ substantially between the two series, and the cross-correlations between the two seasonally adjusted series are in most cases less than .4. A model of classical measurement error indicates chat in 15 of 20 2-digit industries measurement error accounts for over 35% of the variation in the monthly growth rates of seasonally adjusted industrial production.
Keywords: Industrial Production; Measurement Error; Business Cycle
JEL Codes: E23; E32; C22
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| if (Y60) | y4 (Y60) |
| measurement error (C20) | monthly growth rates of if (O49) |
| measurement error (C20) | monthly growth rates of y4 (O42) |
| y4 (Y60) | variance of production based on y4 (C29) |
| if (Y60) | variance of production based on if (C29) |
| measurement error (C20) | differences in volatility and persistence between if and y4 (G17) |