Common Breaks in Means for Cross-Correlated Fixed-T Panel Data

Research output: Contribution to journalArticle

Abstract

This note considers a panel data model in which the variable of interest has undergone a common structural break in the mean. The object of interest is the unknown breakpoint. The challenge is to device an estimator that is consistent when the data are cross‐correlated and the number of time periods T is fixed and cannot be increased without bound. The proposed solution involves taking an already existing estimator initially proposed for cross‐section uncorrelated panels and applying it to defactored data. Consistency is established as the number of cross‐section units N grows large, and is verified in small samples using Monte Carlo simulation.

Details

Authors
Organisations
External organisations
  • Deakin University
Research areas and keywords

Subject classification (UKÄ) – MANDATORY

  • Economics
Original languageEnglish
Pages (from-to)248-255
JournalJournal of Time Series Analysis
Volume40
Issue number2
Early online date2018 Jun 10
Publication statusPublished - 2019
Publication categoryResearch
Peer-reviewedYes