Abstract
This paper analyses the integration order of the consumption/output and investment/output ratios for a group of the most important OECD countries. Results show that these ratios can exhibit some structural breaks. Thus, the use of those statistics which do not take into account the presence of breaks leads to the acceptance of the unit root null hypothesis. A clear increase in the number of rejections of the unit root null hypothesis is found when these structural breaks are included in the model specification.