I have a question about rebasing indices, hoping someone might be able to help.
I'm working with two sets of index data for residential property prices from the Australian Bureau of Statistics. the indexes are constructed from house prices for the 8 main cities, weighted for population.
one dataset is referenced to 1990, the other to 2012. There are 8 points of overlap between the two series.
the issue i am struggling with is that the ratio between the overlapping data points are not consistent as the table below shows.
I am guessing that there are revisions to the index which I am not aware of. Perhaps because the index is a weighted average of house prices from 8 cities, there may have been some changes to the weighting at different points in time? but i don't know why that would create such large changes in ratio between years.
Can anyone shed some light on 1. why this might be, and, 2. how to work around it when trying to rebase the old series? eg does it seem reasonable to take the average ratio of the 8 overlapping data points and use that to rebase the old indices?
any help would be really appreciated.