Abstract
Annual estimates of productivity are reported for periods over 500 years for eight countries and for five other countries over shorter periods. One- and two-break time series models are used to investigate discontinuities in productivity growth. The results support two-break models of long-run productivity and they favour approaches to unified growth modelling with three epochs. However, the lessening of productivity gaps and the decisive shifts to higher productivity occurred in the twentieth century, chiefly in the years around the World War II. The timing of the breaks and the complexity of the historical record highlights a need for unified models to connect more closely with economic history.
Acknowledgements
An old version of this article was presented at UC-Davis, May 2010 and comments received from the seminar participants are gratefully acknowledged. Comments by a referee are also gratefully acknowledged. Jakob B. Madsen acknowledges financial support from the Australian Research Council (ARC Discovery Grant no. DPO877427 and DP110101871).
Notes
1 For full discussion, see Lee and Strazicich (Citation2003) and Lee and Chang (Citation2008), who utilize the test with long historical data.
2 For full discussion, see Harvey et al. (2009).
3 For full discussion, see Bai and Perron (Citation1998, Citation2003).
4 The result is not reported here but full details are available upon request.
5 Of course, the boundaries of modern Ireland were not established until 1921 and identifying its growth trends before that date is problematical.