Working Paper: NBER ID: w1371
Authors: Alan S. Blinder; Douglas Holtz-Eakin
Abstract: It has been known for a long time that inventory fluctuations are of great importance in business cycles. But inventory fluctuations are fundamentally a short-period phenomenon. Consequently, annual data may shed relatively little light on the nature of inventory fluctuations; most of the "action" may be played out within the year. For this reason, economists know precious little about inventory behavior before World War II. This paper seeks to lift this veil of ignorance in two ways. First,we create -- from some admittedly incomplete and imperfect data -- monthly time series on inventory holdings in manufacturing, durable manufacturing,and nondurable manufacturing. To our knowledge, these are the first such series ever made available.(The data are available on request.) Second,we apply to the prewar data certain statistical procedures and models that are in common use with postwar data. In this way, we can address the central issue of the paper: Has inventory behavior changed? While we do not wish to overstate the case, we were struck more by the similarities in inventory behavior between the prewar and postwar periods than by the differences. But the relevant stylized facts and regressing are displayed below, and each reader can make up his or her own mind.
Keywords: Inventory Fluctuations; Business Cycles; Economic Dynamics
JEL Codes: E32
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| inventory fluctuations (D25) | real GNP (E10) |
| inventory changes (G31) | peak-to-trough change in real GNP (F44) |
| inventory behavior (L81) | understanding dynamics of recessions (E32) |
| inventory swings (L81) | cyclical contractions (E32) |
| inventory fluctuations (D25) | business cycles (E32) |
| business cycles (E32) | inventory fluctuations (D25) |