Rand Simberg has an interesting (and plausible) proposal for getting the U.S. space industry out of its present rut.

I know that some people will not like this idea, but the fact is, we already have plenty of military activity in space, and have had almost from Day One (recall that the first man-made object to reach what is generally defined as "space" was an offensive weapon). Rand cites GPS as an example, and wisely so: like the internet, GPS technology has become enmeshed in so many civilian applications as to be inseparable -- imagine the economic impact of a suitably-motivated enemy detonating a small nuclear device on the fringe of the atmosphere, and destroying a good chunk of the GPS constellation via EMP. And whatever can take out the GPS constellation to any significant degree can even more easily take out the less-hardened commercial satellites on which global communications, resource management, and weather prediction depend. It isn't merely the need to protect and (if the need arises) replace space assets because the military depends on them, but because the civilian economy does as well.

He is also right in the significant impact an affordable, reliable, high-rate, and rapidly-deployable launch system would have on satellite design. By removing the need to make satellites simultaneously lightweight, powerful, ultra-capable, and heroically robust, the cost of satellites will drop dramatically -- and the same would apply to the construction and maintenance of manned facilities in orbit.

Building a military-civil space transportation system could be a suitable means for turning around the demographic erosion of the aerospace industry. And if the military takes the lead, it might actually stand a chance of coming to fruition.