Re: Vision of the three Rs: Regular, Reliable and Reusable

On Feb 18, 9:51 pm, Fred J. McCall <fmcc...@xxxxxxxxxxxxx> wrote:
Ian Woollard <ian.wooll...@xxxxxxxxx> wrote:
:Derek Lyons wrote:

:> In other words - pork and handouts.
:
:Yeah, it's a subsidy on water in LEO/GTO.
:
:I believe I'm right in saying that subsidies can be economically
:justified in situations where the long-term situation is one where the
:lower prices are going to be in line with the subsidy.

It's not quite that simple. It assumes there is a 'knee' in the cost
curves which must be overcome in order to get down to the subsidized
price AND that the 'knee' is related to scale of operations. In other
words, this only works depending on very specific curve shapes and
very specific economies of scale.

Fred, either you're stating the obvious or you're hinting at (without
explicating) something economically subtle. I'll assume the latter,
and ask: care to get very specific about what you mean by "very
specific"?

As for the shapes of cost/investment curves, there can be many "knees"
after the first few big ones. In fact, it's mostly knees on down to
optimal. (I was once told that IBM kept offering employees incentives
for improving the process for making Selectric typewriters, down to
the point where they were shaving 1 cent off the cost per unit. Given
the sales volume, it was actually worth an engineer's time to think
for a few weeks about cost savings that low.) Nathan Rosenberg, in
his Inside the Black Box: Technology and Economics, points out that
upwards of half of all gains in economies of scale in industrial
production come from steady, incremental improvements -- each of which
represents a capital investment decision, just not a very big one
compared to the initial ones. Back off far enough from the curve, you
don't see these later "knees". A process that appears superior at the
outset can even lose out to another that is more amenable to a
learning curve.

Can you make a case that the specifics of these supposed "very
specific" curves simply can't match what's needed to reduce the cost
of producing water from space resources? Can you make a case that the
costs will never be significantly lower than simply launching water
from Earth, because the "very specific" economies of scale will never
yield low enough costs?

If you can't, I'll just say you're hand-waving. And then we're hand-
waving at each other, aren't we? Takes one to know one, I guess.

Here, let me give you start on rebuttal. I am, after all, interested
in this question being pursued objectively.

Establishing and improving economies of scale in a new industry often
depends on supplies from other industries that have their own
established, growing markets and their own improving economies of
scale. The quantity purchases made by new industries (which may
develop in fits and starts, with many failures) can become a source of
new capital to the supplier industries, helping to further feed their
own production process improvements and raise their own capital based
on the new earnings. But those suppliers aren't totally dependent on
the new industry--they couldn't be, not nearly, or they wouldn't
already exist. So you get a virtuous cycle of capital formation, a
very resilient commerce web within which the new industry can try,
fail, try again, and eventually succeed on terms that have become
easier, both internally from what's been learned by failure in the new
industry, and externally from what's been learned through success in
its supplier industries.

In the case of space development, however, one could argue that so
many of the technologies are exotically space-specific that any large-
scale investment in off-Earth water production will depend on
suppliers who would have little more than that one single market. The
notional space-water-extraction industry would have to carry almost
the entire burden of capitalization; it would have to incubate and
establish economies of scale all the way down its supply chain, on its
own, without the benefit of other consumers for its suppliers, and of
financiers willing to advance capital to firms that would be serving
those non-existent consumers. Thus, it might not only be a non-
starter in the capital markets, it might bust the (politically
tolerable maximum) budget of any government trying to subsidize such
an industry into existence. It would be a little like trying to do
the whole Industrial Revolution in one go.

Maybe. But I'm just handwaving here.

-michael turner

.

Quantcast