What you described is one type of subsidy. It's not the only type.

From that wikipedia article you linked: "In standard supply and demand curve diagrams, a subsidy will shift either the demand curve up or the supply curve down. A subsidy that increases the production will tend to result in a lower price, while a subsidy that increases demand will tend to result in an increase in price."

Defense R&D increases the demand for engineers and thereby raises their "price". I don't know if it has a significant effect though.

Neither tlb nor cadlin are entirely right about this from an economic standpoint.

The government employing engineers is strictly speaking not a subsidy. It's just an increase in demand. Yes, one type of a subsidy does result in an increase in demand. But causing an increase in demand is not a necessary&sufficient condition for being a subsidy. Because now the government is demanding engineers. It does not decrease supply. Someone who is employed is still by definition part of the labor market. The only person who is not a part of the labor market is someone permanently not looking for work.

Defense spending goes to MRU's who train engineers who go into the private sector. E.g. MIT receives almost $1 billion in DOD funding each year, which benefits all the companies that hire MIT engineers.