I think that this finding of Robert Blecker's is relevant to the present
discussion:
"...the net natinal saving rate excludes all depreciation allowances, for
corporate and non-corporate businesses as well as for households. What has
driven the net national saving rate to such pitiful levels is, aside from
the rising governent budget deficit, mainly a notable increase in
'consumption of fixed capital' as share of GDP. It is largely the rising
deprecitation rate, along with the government deficit, and not a sudden
decrease in private thriftiness that has made the net measures of the
saving rate sink so low.
"This raises the question of why the depreciation rate has risen and what,
if anything, should be done about it. Briefly, the depreciation rate has
risen for two main reasons. First, thanks to the slowdown in the growth of
current output and income...the depreciation on the existing stocks of
capital equipment and structures (which were created in the past, when
growth was higher) looms ever larger in proportion to current GDP. And
second, there has been a notable shift in the composition of investment
away from structures (including residential housing) to producers' durable
equipment, which has shorter economic lifetimes, and within the latter
category toward more short-lived types of equipment such as computers. In
other words, the rising depreciation rate that has depressed net savings
measures is partly a symptom of the growth slowdown, and otherwise is
mostly a reflection of changing technology.
"Theoretically, a low net saving or investment rate means that a country is
not adding much to its capital stock, beyond replacement of depreciating
buildings and equipment. But this can be misleading. Especially in the
case of producers' durable equipment, replacement investment almost
invariably takes the form of new 'vintages' of machinery embodying new
technology. For eample, no one replaces an early 1980s IBM personal
computer (PC) with an equivalent machine today. The new PCs being
purchased today have capacilities that still required much more expesnive
mainframes ten years ago--yet cost less than an original IBM PC. While
computers are an extreme case, they represent an increasing fraction of new
equipment expenditures, and the same principle holds (albeit in less
extreme fashion" for other types of equipment. For these reasons, the
hysteria based on low NET savings rate is simply misguided."
>From Understanding American Economic Decline, ed. Michael A Bernstein. CUP,
1994: 290-1