I did a thread on Twitter this morning.

It's worth noting that one of the reasons we never got space-based missile defense was that it was only recently that we've finally gotten launch costs down sufficiently to make it financially feasible, due to an almost demented policy failure for the past three decades. [1/n] https://t.co/ouaaIS9eUk

— Rand Simberg (@Rand_Simberg) January 18, 2019

The first serious proposal for space-based missile defense was Lowell Wood's concept of "Brilliant Pebbles": Kinetic interceptors in orbit. But in order to implement it, launch costs had to be reduced far below those of the Shuttle and conventional USAF expendables.

— Rand Simberg (@Rand_Simberg) January 18, 2019

The purpose of the DARPA DC-X program was to demonstrate the potential for reusable Single-Stage-To-Orbit, which many viewed as a requirement for low launch cost (SpaceX has since proven this to be mistaken).

— Rand Simberg (@Rand_Simberg) January 18, 2019

DC-X did demonstrate vertical take-off and landing of single vehicle in an atmosphere (the Apollo LEM was two stage in a vacuum). It also demonstrated relatively rapid turnaround of a LOX/LH2 propulsion system. But then NASA took it over.

— Rand Simberg (@Rand_Simberg) January 18, 2019

On one of the test flights of the NASA-modified vehicle, someone left a pneumatic hose off one of the legs, and it crashed and burned at White Sands, ending the program.

— Rand Simberg (@Rand_Simberg) January 18, 2019

Another thing that the DC-X program demonstrated before its demise was that traditional cost models for new concepts were utter crap. SpaceX has since validated that. NAFCON cost model has been shown to be worse than worthless for non-traditional activities.

— Rand Simberg (@Rand_Simberg) January 18, 2019

One of the biggest launch-policy errors of the 90s was to confine the military to expendables, and assign reusable space transports to NASA. It was nothing short of disastrous, setting us back over a decade.

— Rand Simberg (@Rand_Simberg) January 18, 2019

After the X-33 debacle, which no one saw coming except anyone who understood how to do X programs, the idiotic lesson (fallacy of hasty generalization) drawn from it by NASA was that reusable launch systems weren't practical. Tell it to SpaceX.

— Rand Simberg (@Rand_Simberg) January 18, 2019

X-33 should never have been awarded to Lockmart (their proposal wasn't compliant, in that the business plan was nonsense, but no one at MSFC would recognize a business plan if it kicked them in the ass). Also, should never have been a single award.

— Rand Simberg (@Rand_Simberg) January 18, 2019

A key rule of X programs is that a vehicle only tests one new technology, on a platform that is otherwise well understood. VentureStar was testing single-stage to Montana, with a linear aerospike engine, and a conformal composite hydrogen tank. Huge and obvious tech risk.

— Rand Simberg (@Rand_Simberg) January 18, 2019

X-33 was an example of NASA's Wile E. Coyote approach to technology development: Try some crazy thing, then when it doesn't work, don't try to figure out why and improve it, just assume it can't be done and go on to the next crazy thing.

— Rand Simberg (@Rand_Simberg) January 18, 2019

And so we entered the 21st century with no one, neither USAF or NASA, even attempting to get launch costs down. Former was focused on mission assurance of expendable EELVs, and latter had devolved into a jobs program for giant expendable rockets.

— Rand Simberg (@Rand_Simberg) January 18, 2019

But now, having done that, it's useful to go back and re-examine concepts for space-based missile defense that were financially infeasible with traditional launch costs of many thousands of dollars per pound. Cubesats are also a game changer.

— Rand Simberg (@Rand_Simberg) January 18, 2019

[Update a couple minutes later]

Trump’s missile-defense strategy.

As I noted above, if the space segment is now feasible, it’s despite, not because of government launch policy for the past three decades (except possibly for COTS).