Working Paper: CEPR ID: DP7069
Authors: Albrecht Ritschl; Samad Sarferaz; Martin Uebele
Abstract: This paper reexamines U.S. business cycle volatility since 1867. We employ dynamic factor analysis as an alternative to reconstructed national accounts. We find a remarkable volatility increase across World War I, which is reversed after World War II. While we can generate evidence of postwar moderation relative to pre-1914, this evidence is not robust to structural change, implemented by time-varying factor loadings. However, we find moderation in the nominal series. Moreover, we reproduce the standard moderation since the 1980s. Our estimates confirm the NIPA data also for the 1930s but support alternative estimates of Kuznets (1952) for World War II.
Keywords: dynamic factor analysis; US business cycle; volatility
JEL Codes: C43; E32; N11; N12
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| World War I (N44) | increase in business cycle volatility (E32) |
| post-World War II (N44) | reversal of business cycle volatility (E32) |
| pre-World War I era (N93) | business cycle volatility (E32) |
| post-World War II (N44) | moderation in nominal series volatility (C22) |
| time-varying factor loadings (C22) | no moderation in real economic activity volatility (E39) |
| 1930s NIPA data (E01) | support for estimates during World War II (H56) |