There is a recurrent assertion that the elderly want more public resources to be spent on social protection and health, the young want more on education, and such preferences are reflected in the actual government spending policy. This study aims to empirically confirm whether the assertion is valid in OECD countries. For that goal, we propose an estimation method to exploit the comparison of the actual share of government expenditure and its theoretical share by using aggregate data. The empirical finding is consistent with the recurrent assertion in the sense that the fraction of the young has a significantly negative effect of the spending share of social protection and health but a positive effect on the spending share for education even though we can not find a significant effect of the elderly. In particular, ageing leads to a smaller fraction of the young and a larger fraction of the elderly. Hence, the empirical finding predicts that the ageing trend is likely to bring more public resources to the social protection and health areas, and less public resources to education.