In Discussion Paper No. 1087, Research Fellow Nicholas Crafts and Terence Mills aim to provide acceptable estimates of trend growth in output per person in order to obtain a clearer idea of the validity of the notion of a `Golden Age' of European growth. The results are also of interest in the context of the continuing debate in the time-series literature over the unit root hypothesis. The paper examines growth in output per person in 17 OECD countries from the late nineteenth century to 1989 considering the possibility of several breaks in trend. The results amount to a rejection of the unit root hypothesis in favour of a segmented trend stationary alternative, for all the countries investigated. This applies both to the case where 1973 is treated as an exogenous breakpoint and where the post-1950 break is found endogenously. These results are contrary to what might be expected were growth fully endogenous with constant returns to the accumulation of reproducible factors of production.

The authors also investigate the hypothesis of reconstruction followed by a `return to normal' by considering the validity of a number of restrictions on the fitted segmented trend models. Only for one European country, Denmark, do they fail to reject the hypothesis that recent growth is along a trend path extrapolated from before 1914; for all other European countries, output per person in the 1980s was above this path.

Europe's Golden Age: An Econometric Investigation of Changing Trend Rates of Growth
Nicholas F R Crafts and Terence C Mills

Discussion Paper No. 1087, January 1995 (HR)