Oil price shocks and U.S. unemployment rate: a robustness check

ABSTRACT

This study investigates the sensitivity of the responses of the real price of oil and the U.S. unemployment rate to several real economic activity (REA) measures proposed in the literature. It then estimates the responses of the above macroeconomic variables to a set of alternative exogenous oil supply shock series driven by different modelling strategies. While the results are qualitatively robust across the REA measures, they are sensitive to the exogenous oil supply shock series selected.

KEYWORDS:

• Oil prices
• U.S. unemployment rate

JEL CLASSIFICATION:

• C32
• Q41
• Q43

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ The intuition behind it is driven by market fundamentals, i.e. escalations in the demand for shipping driven by growing global economic activity would increase the cost of shipping.

² I refer the reader to Kilian (2009) and Kilian and Zhou (2018) for a detailed description of the index, its benefits over other measures, and some possible limitations in predicting commodity prices.

³ Kilian and Murphy (2014) include changes in global oil inventories in their sign-identified structural VAR model.

⁴ Güntner (2014) uses sign restrictions in a SVAR model to identify the response of crude oil producers to oil demand shocks. His findings are consistent with the notion of a vertical short-run oil supply curve.

⁵ The updated index is now published, on a monthly basis, by the Federal Reserve Bank of Dallas at: https://www.dallasfed.org/research/igrea

⁶ The index is available at: https://sites.google.com/site/cjsbaumeister/datasets

⁷ While Baumeister and Hamilton (2019) use WIP in growth rates, Herrera and Rangaraju (2020) and Zhou (2020) use log-linearly detrend WIP when comparing with the KI.

⁸ For the complete list of variables, see Baumeister, Korobilis, and Lee (2020).

⁹ Note that WIP, GSP, GECON, and RCPF are available from Baumeister’s website from 1973.1 onward: https://sites.google.com/site/cjsbaumeister/research.

¹⁰ $OS{S}_{k09}$ is extracted in this paper.

¹¹ $OS{S}_{BH19}$ is available at Baumeister’s website: https://sites.google.com/site/cjsbaumeister/research.

¹² $OS{S}_{GH21}$ is available at Güntner’s website: https://sites.google.com/site/econjochen/research.

¹³ Kilian and Park (2009) build on Kilian (2009) structural VAR decomposition of the real price of crude oil, and relate the U.S. stock market variable to three structural demand and supply shocks that drive the real price of oil.:

¹⁴ See Kilian (2022) and Kilian and Zhou (2020) for a thorough justification on the assumption of a vertical short-run supply curve.

¹⁵ In line with Karaki (2018).

¹⁶ Note that GECON covers measures of expectations and uncertainty, and energy related indicators including long-run oil price uncertainty. That might explain why the responses using GECON are different from the other measures.

¹⁷ Note that I am not comparing only the different identifying strategies, but, in the cases of K09 and BH19, also different REA measures used in their model. In other words, $OS{S}_{K09}$ and $OS{S}_{BH19}$ are driven by different modelling strategies using KI and WIP, respectively. Hence, this cumulates the differences between Kilian (2009) and Baumeister and Hamilton (2019).