Skip to main navigation Skip to search Skip to main content

Employment Trend-Cycle Decomposition and Forecast

  • Kyu Ho Kang*
  • , Samil Oh
  • *Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

This study aims to decompose recent employment fluctuations in Korea into structural and cyclical components and forecast future employment trends. To achieve this, we develop and estimate a trend-cycle hidden factor model that incorporates Korea’s macroeconomic environment and demographic structure. Within the model, employment is modeled as the sum of a unit root process (trend) and a stationary process (cycle). The trend and cycle are each designed to have a dynamic correlation with key macroeconomic variables and demographic structure. The main results are threefold. First, the cyclical component of employment for all ages as of the fourth quarter of 2023 is estimated at 150,000 and 130,000 for those under 60. Second, the trend of employment for all ages and those under 60 is mainly determined by the population growth rate and aging rather than the potential growth rate. On the other hand, the cyclical component is closely related to the GDP gap and economic sentiment index. Fi-nally, employment for all ages is expected to increase by 228,000 by the end of 2024, of which 74,000 is cyclical. On the other hand, employment for those under 60 is expected to decrease by 115,000, but the cyclical component is 56,000, indicating that employment is expected to exceed the trend.

Original languageEnglish
Pages (from-to)1-31
Number of pages31
JournalJournal of Economic Theory and Econometrics
Volume35
Issue number4
Publication statusPublished - 2024 Dec

Bibliographical note

Publisher Copyright:
© 2024, Korean Econometric Society. All rights reserved.

Keywords

  • business cycles
  • Population aging
  • state-space model

ASJC Scopus subject areas

  • Economics and Econometrics

Fingerprint

Dive into the research topics of 'Employment Trend-Cycle Decomposition and Forecast'. Together they form a unique fingerprint.

Cite this