Matt Yglesias hits the nail on the head:
"I think it’s unquestionably true that one strength of the United States vis-a-vis Europe is that our setup is friendly to a certain kind of startup firm and this is why we’re world leaders in much of the high tech industry. But this kind of rhetoric is annoying and inaccurate:
[Google CEO Eric] Schmidt said that in parts of Europe the venture capital industry was being held back. “It takes a very long time to get a proper venture capital industry going. And in Europe there are many countries where the failure of venture capital is in fact seen as criminal, Germany, for example.” [...]
“Jobs are created by the private sector not by the public sector. Wealth is created by the private sector not by the public sector. If you are going to have great companies you have to be willing to support and encourage the creation of real wealth-generating capitalists. That means you have to put up with them.”
You would think the CEO of an Internet firm wouldn’t be totally dismissive of the idea that CERN and DARPA are contributing to wealth creation. And of course education is important to wealth-creation, and much of it is state-financed. And transportation infrastructure—typically state-financed—creates wealth. Of course even if the state completely removed itself from education finance, it’s not like nobody would go to school. Presumably the private sector would step in to fill some of the void. But conversely in places where the state steps beyond what we’d consider appropriate in the United States it still creates wealth. State-owned firms like Areva are generating value.
I raise these points not to make the case for a bigger or smaller public sector (I would say bigger than the US but smaller than France is ideal, but both are clearly workable) but simply to underscore the point that the performance of public agencies matters a lot."
This is the Robert Higgs school of GDP, of course: just don't count the public sector. The reassuring thing is that when they redefine the very definition of words like "wealth" you know they're desperate - the problem is that I think they actually believe this stuff. It's one thing to say that the government in most spheres of economic activity is less efficient at wealth production than the private sector. That is unambiguously true and completely uncontroversial. But in certain fields they are better (that's why there are a few fields where we regularly see governments: evolutionary pressures practically dictate we will see these sorts of efficiencies capitalized on by successful societies). And even where they're not better you don't just pretend they don't exist.
Anyway - this is a pet peeve of mine too. Similarly, it bugs me when people only think of "crowding out" as public activity crowding out private activity. Clearly that is a problem and it exists (again - unambiguous and uncontroversial), but you never hear anyone talk about private activity crowding out public activity. They simply define it out of the discussion. They assume it away. This is what an externality is though, if you think about it: private activity crowding out social activity. That's one way to look at it at least.