> Doug Henwood wrote:
>
> >Wouldn't the revision down of the "income" side only make no difference
if
> >it were revised in such a way as to still be plausibly within the range
of
> >statistical error for income?
>
> I don't get what you mean.
>
Sorry, I meant to ask about the statistical error for the product side. Isn't the main way that they reconcile the difference between the income and product sides of the accounts by means of "statistical discrepancy" used in figuring national income from net national product (ie. Table 1.9, line 15)? I was wondering whether a "revision" of the income side would look fishy if this statistical discrepancy were exceptionally high.
It also surprises me that the product side of the accounts is considered more accurate, since it includes capital consumption/ depreciation, which is really hard to measure, right?
The section 5 tables are the ones I originally looked at. Personal saving for 2000 is about .68% of GNP. Gross private saving is about 13.1% of GNP, but in 2000, corporate capital consumption/ depreciation accounts for 55% of that. Over the period between 95 and 00, depreciation accounts for more and more of private saving (starting from about 40% in 95). I assume that this number is not derived from corporate tax accounts of capital consumption, and remains hard to measure. In which case, we might suspect that this relatively ok number for private saving isn't quite what it seems. No?
Christian