@professor:

I've actually coded in both Fortran-77 and C++. The former, though, was a very long time ago--thanks for bringing back some old memories!

It's certainly true that pensions indexed to the CPI aren't growing as fast as the actual cost of living--anybody who runs a household can see that. My point in this article, though, was that a good bit of those changes (especially in the cost of food and fuel) are not inflation, but rather are declines in standard of living. A perfect inflation index (if there could be such a thing) might actually be producing much the same result just now, because a big chunk of the problem is that everyone is poorer now than a year ago.

Indexing for inflation isn't supposed protect you when people are getting poorer, just when the money is getting less valuable. Clearly, both thing are going on just now, and it's hard to see how much of any particular price change is due to one versus the other.