re: profit rates

["Devine, James" <jdevine@xxxxxxx>]

Daniel Davies writes:> The assets of WorldCom and Global Crossing are
worth exactly what they were worth before the meltdown, as stock market
movements don't mean much to cables in the ground.  The fact that the
stock market's assessment of the future excess returns to be earned from
renting out those cables no longer provide a viable basis for making
interest payments don't change the capital employed for the purpose of the
BEA numbers.<

yes. It's important to remember that how one calculates the "rate of return"
depends on what one's purposes are; not all measures of the rate of return
are good for all purposes.

The forward-looking rate of return would be the _internal_ rate of return,
the interest rate that sets the present discounted value of net profits
equal to zero. To some extent, this is the rate of return based on
stock-market valuation, since the stock market prices are based on guesses
of future net profits (not very good ones, of course). A firm will have a
different estimate, based on its projected costs, revenues, etc.

The other rates of return that people have been discussing have other
purposes. I find that the BEA estimates of the profit rate, for example, to
be pretty well correlated with other estimates (such as those of Dumenil and
Levy), while having something to say about the validity of theories such as
that falling profit rates lead to stagnation and/or inflation (or both).
That is, the data work in practice, though of course we can't put too much
faith in them.

One of the basic rules of empirical data is that they're never very good. If
you examine the calculation of the "real wage," for example, there are lots
of caveats. People like Boskin take these caveats and run with them, trying
to figure out ways to revise the data to make them look better (either
empirically or in terms of the orthodox economics ideology or both). But
that doesn't say we should give up on statistics. Rather, it says we have to
be humble in using them and to be very conscious of the econometrics of
"errors in variables."

A non-statistical or anti-statistical approach unfortunately eschews any
sense of context. Everything is a special case, with no connection or
comparison with other special cases. Everything ends up being a bunch of
anecdotes. And we know how bad anecdotal evidence is: my wife lost (and is
keeping off) 40 pounds by going to a hypnotist, but that doesn't mean I can
do so.

There are other ways of putting things in context, such as saying "all
countries in Africa have a similar social relationship to the rich countries
of the capitalist core" (the imperialist nations). But unless there's some
sort of statistical meaning to that relationship, it ends up being too
abstract.

Statistics shouldn't be rejected, but should instead be treated carefully,
and as complementary to other knds of evidence.
JD