Measuring the degree of persistence in the U.S. economic policy uncertainty index

ABSTRACT

Researchers utilizing the U.S economic policy uncertainty index and its sub-categories need to be cognizant of the unique persistence profile of each index. We use fractional integration techniques to estimate the degree of persistence in the U.S. economic policy uncertainty index and its 11 sub-categories. The results indicate the estimated values of the differencing parameter, d, are in the interval (0, 1) supporting fractional integration and rejecting the classical models based on stationarity (d = 0) or unit roots (d = 1). Though there is a fair amount of heterogeneity across indices, shocks will be persistent, but mean reverting.

KEYWORDS:

• U.S. Economic policy uncertainty index
• persistence
• fractional integration
• long memory

JEL CLASSIFICATION:

• C22
• E60

Acknowledgments

Comments from the Editor and an anonymous reviewer are gratefully acknowledged.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ Baker, Bloom, and Davis (2012, 2016) provide a detailed analysis of the construction of the economic policy uncertainty index and its categorical data.

² The índices associated with Trade Policy, Financial Regulation, and Sovereign Debt, Currency Crisis contained zero values. In our analysis analysis with the log-transformed data (those data with non-zero values), the results were almost identical to those reported in Table 1.

³ Regarding the time trends, as d moves from 0 to 1, the trend will disappear in the long-run. For example, in Equation (1), if d = 1 and u_t is white noise, y_t becomes a random walk process with a constant and the time trend disappears.

⁴ The possibility of structural breaks, and thus potentially producing spurious long memory is a consideration for future investigation.

Additional information

Funding

This work was supported by the Ministerio de Economía y Competitividad [ECO2017-85503-R].