Also...

(following up on yesterday's thought)

It seems to me that for the inflation-adjustment of debt -- maybe for all 'real' calculations, but definitely for 'real' debt calculations -- you want to have the base year at the start of the dataset, not at the end of it.

The 'real' calculation removes the effects of inflation. Inflation is prices going up. If you remove the increase in prices, your numbers get lower. To show the "lower" you have to put the base year at the start of the series. So it just makes sense to do it. (See also mine of 5 April.

Coincidentally (or not), using the first year as the base year solves the starting-year problem that I discussed yesterday. Come to think of it, when I did inflation-adjusted debt calcs a while back I wrote about a "floating" base year.