Is MORE LESS? The role of data augmentation in testing for structural breaks

- This paper addresses the issue of data augmentation in structural change testing.
- Theoretical and simulation analysis shows that increasing the sample size may decrease power.
- An empirical example conrms the findings.

In this paper, we examine the impact of increasing the size of a data set in detecting structural breaks. Based on an empirical application, supported by theoretical justification and a simulation experiment, we find that larger sample sizes may make it more rather than less difficult to determine the existence of a structural break.