Michael Turner has a piece in today's The Space Review arguing that Moore's Law won't apply to space development. His argument fails, at least to me, because it rests on a false premise (and a common myth)--that the reason access to space is expensive is because we don't have the "right" technology.

While I don't literally believe in a Moore's Law for space (in the sense that we can see seemingly never-ending halving of costs on some constant time period), I do expect to see dramatic reductions in cost in the next couple decades, but not because there are vast ranges for improvement in the technologies, but because there are is vast potential for improvement in the real problem--the heretofore lack of market.

Costs will come down dramatically when we start flying a lot more. It's that simple. Once we reach a plateau, in which the costs of propellant start to become significant in the overall costs of flight, then we should look to some new technological breakthroughs, but we're sufficiently far from that that some form of Moore's Law, at least in the short term, is actually quite likely to hold.

I'm disappointed that the article didn't even try to detail the price/performance reality for rockets. Fundamentally, Moore's Law is an observation, with the variables hashed out later.

If you plot the inflation-adjusted potential cost per pound to LEO (choose an early program versus SpaceX), then we are experiencing a drop in cost of a couple percent per annum. It's not sexy, but tortoises can win races.

The author totally ignores the effects of mass production on what is now an essentially hand made product and ignores the advantages of a truly reusable system such as overcoming the "infant mortality" syndrome.

I'm surprized that people fail to apply basic economic principles to new markets like spaceflight.

The reason there is a moore's law in silicon chips is that there is an analogous effect in ALL products and services. The cost to manufacture things (or to provide services) declines with the number of units manufactured -- this is as close to a universal economic law as there is.

There reason there is such a "law" is that the capitalist manufacturer is constantly striving to produce the product cheaper in order to maximize the return on investment AND to prevent any new entrants from claiming a share of the market with a cheaper manufacturing process.

THere is always room for improvement in manufacturing (cheaper, faster, etc) because new techniques and ideas can be applied.

This relationship has been measured in many diverse industries and (over the long haul) has shown to be true. It will be true in spaceflight -- one can see the curve already declining with chinese/russian/sea launch/EELV. (Of course, short term there can be price spikes and drops not indicative of the long term trend).

We have a long way to go on the curve before things are "cheap". We don't know how step the curve will be -- very, very steep like semi-conductors or rather shallow like aluminium production.

"Moore's Law" (more of a forecast, and Gordon Moore himself was surprised at how people responded to it) has been often and repeatedly misused in articles on the electronics industry itself. It isn't a surprise it could be misused outside the industry.

There were two major electronics/computer revolutions before microchips came along: vacuum tube and discrete transistor. Even AFTER the microchip came along there were many big changes - TTL versus CMOS, for instance. All useful technologies have a period of discovery, then rapid development leading to a plateau where development dramatically slows. (Incidentally, things may be slowing a bit now for chips - CMOS seems to be getting near a wall, there are big scaling problems now).

The unusual thing about microchips is that the technology was very useful fairly early in development and has had an extremely long period of rapid development. No, it isn't likely space travel will have decades of exponential technological improvement - but unless you plan on shipping megatons to Mars for pennies it doesn't matter.

Right now, we are in the very beginning of the non-government space market. We haven't even touched the technology development stage yet.

There's a graph of historical prices here:

http://www.theculture.org/rich/sharpblue/archives/000066.html

I did a quick interpolation assuming an exponential decay of the price since about 1988 on Ariane. I obtained a price reduction of about 6% per annum on average.

That's a halving of the price every 11 years.

Of course that assumes a Moores-law type relation has taken over at about 1988; it doesn't seem totally unreasonable to me that we may be in that regime now. IMHO I think that Moore's law is a function of how quickly costs come down, and turnover goes up, for a given investment in money. With semiconductors the relationship is quite direct and quick; with rocketry it seems rather slower- mostly because the market is inelastic right now.

Anyway, take it with 2200 lbs of salt.

Economics is part of it, but it also depends on the underlying technology. There wasn't as much room for evolutionary improvement in vacuum tube technology, for instance. There were three major revolutionary jumps before we got to microchips with their very long evolutionary exponential improvement.

I expect there will be revolutionary changes along the way with space travel, but there are physical limits with conventional chemical rockets. The point is that we don't NEED exponential technological development for the price of space travel to drop dramatically. The current hardware used by the government was designed for extremely low-volume high value payload. That's the reason costs are so high.

I think the prices are high, only because they can be high. The entire American military space program is working on a cost-plus basis; they pad the bill out to a reasonable degree. The Russians price their hardware to match the western programs and laugh all the way to the bank. Pretty much, that's it really.

Rand argues that my argument fails "because it rests on a false premise (and a common myth)--that the reason access to space is expensive is because we don't have the 'right' technology."

Rand, my argument is not that simple. Go back and read the piece. I also cite problems on the demand side. I am actually a big fan of John Walker's keeper essay, "A Rocket a Day Keeps the High Costs Away". This is, however, still a proposal that governments test the market by subsidizing even frivolous launches, if that's what it takes to keep launch rates high. And he admits it would be only a test.

You say, addressing an issue that I DID bring up in the essay, that there "is vast potential for improvement in the real problem--the heretofore lack of market."

You speak of (1) "dramatic" improvements in technology in the offing, (2) "vast" potential for a market. You say that (3) "in the short term," something like Moore's Law is (4) "likely to take hold." I admire your optimism. I would just like to see how you'd make the case for all four optimistic assessments, in concrete terms, citing present trends.

I started my career in VLSI CAD. For a few short years, I lived Moore's Law as a reality in the way Carver Mead spoke of it. It was a world in which I met people who spoke of fear. Fear of the sheer scale of the next chip design. I met people who made career choices based on that fear - choices that took them out of the path of the juggernaut that was delivering the exponential growth in IC component power. It had been barreling on for two decades when I joined it. It has continued its careening trajectory for two decades since I started moving out of its path myself.

In my admittedly-few brushes with rocket science, I have found nothing like the same atmosphere. If anything, it seems more like an inherently logarithmic improvement curve, with no cusp turning into exponentiality on the horizon.

Tell me why that perception is wrong. I would really like it to be wrong. So would a lot of people. Including some rocket scientists.