On time and frequency-varying Okun’s coefficient: a new approach based on ensemble empirical mode decomposition

Myeong Jun Kim, Stanley I.M. Ko, Sung Y. Park

Research output: Contribution to journalArticlepeer-review

4 Citations (Scopus)

Abstract

This study revisits the time-varying Okun’s law, using US data over the period 1948Q2–2015Q3. The estimated Okun’s coefficients are negative over most of the time horizon and the absolute values of the time-varying Okun’s coefficient is getting smaller. The short- and long-term fluctuations of the time-varying Okun’s law are reconstituted using the ensemble empirical mode decomposition (EEMD) method, and their determinants are analyzed. The empirical results show that the number of working hours and utilization are important factors affecting the long- and short-term fluctuations of the time-varying Okun’s coefficients. More specifically, the short-term fluctuations of the working hours and utilization have significant positive and negative effects, respectively, on the magnitude of short-term fluctuations of the time-varying Okun’s coefficients. It is also found that the long-term fluctuation of the estimated time-varying Okun’s coefficient has a very similar pattern to the detrended real GDP series. We also show the estimated regression estimates are very stable with respect to the considered EEMD method using a simple simulation.

Original languageEnglish
Pages (from-to)1151-1188
Number of pages38
JournalEmpirical Economics
Volume61
Issue number3
DOIs
Publication statusPublished - 2021 Sept
Externally publishedYes

Keywords

  • Determinant of Okun’s law
  • Ensemble empirical mode decomposition
  • Okun’s law
  • Time-varying coefficient

ASJC Scopus subject areas

  • Statistics and Probability
  • Mathematics (miscellaneous)
  • Social Sciences (miscellaneous)
  • Economics and Econometrics

Fingerprint

Dive into the research topics of 'On time and frequency-varying Okun’s coefficient: a new approach based on ensemble empirical mode decomposition'. Together they form a unique fingerprint.

Cite this