Tim wrote:

"Ronald Raygun" wrote
While it was easy enough for you to make up a set of value movements
which made the 2007/8 value proportions the same as the income ones, ...

It's not a matter of "making up" the value movements,

It is if I use "make up" in the sense of "posit ficticious
values for". Naturally it goes without saying that you can't
make up *arbitrary* values for f and g, the two years'
inflation factors. There is only one pair of values which
will work to achieve the effect you wanted, and of course it
must be calculated.

but rather *calculating* them from their yield.

Well, I suppose you can do it by working out the yields first, but
that seems a bit long-winded. How I would do it is to observe that
the two three-way ratios 100fg:60g:75 and 20:15:17 must be
equal, which boils down to solving the three equations

[1] 75k = 17
[2] 60gk = 15
[3] 100fgk = 20

for f and g, which *can* be done by calculating k first (which
happens to be one of the yields) from [1], then using that in [2]
to work out g, and finally using that in [3] to work out k, but it
can also be by eliminating k algebraically, calculating g directly
by dividing [2] by [1], and f by dividing [3] by [2].

60g/75 = 15/17 = g = 1.103
100f/60 = 20/15 = f = 0.8

"Ronald Raygun" wrote
... the underlying assumption that the value inflation rates are the same
in each of the three locations in any one year is a pretty unrealistic.

Of course, but in the absence of further information (the question
had no other data), it's got to be the best assumption, hasn't it?

It's the only possible assumption, which makes it simultaneously the
best and the worst. But since it is objectively bad, might it not be
better than solving the problem on the basis of it, to throw up one's
hands in warning and say that the answer will be silly so there is no
point in working it out?