I highly recommend William Voegeli's essay on conservatives and government-cutting. There's a lot to be said about it, but I'll confine myself to this passage:
One yardstick may help conservatives feel a little better about themselves. In 1981 federal spending was 22.2% of GDP; last year it was 20.3%. This measure hovered in a very narrow band for the whole era, never exceeding 23.5% or falling below 18.4%. Adding expenditures by states and localities confirms the picture of a rugby match between liberals and conservatives that is one interminable scrum in the middle of the field. Spending by all levels of government in America amounted to 31.6% of GDP in 1981, and 31.8% in 2006.
Conservatives, though, can't take much solace from fighting Big Government to a draw. Looking back, the dynamic growth of the American economy after 1982--real per capita GDP was two-thirds higher in 2006 than in 1981--offered a great opportunity to reduce the relative size of the public sector. This economic vigor meant that more people had more money to spend on their own health, education and welfare, presumably enabling the government to spend less for such purposes. It also meant that government spending could have grown robustly and still expanded more slowly than the economy, leaving the public sector to absorb a significantly smaller portion of GDP in 2006 than it did in 1981. Even this modest achievement eluded conservatives.
In conservatism's defense, though, I think it's worth considering Tyler Cowen's hypothesis (unfortunately I can't find the place where he advances it, but I'm pretty sure it's his) that as rich countries get richer, the demand for state services - and particularly middle-class entitlements - may naturally rise, because people increasingly feel like they can afford the pinch of taxation that comes with, say, lavishing more money on public schools or giving Grandma free prescription drugs. A richer country is, in this sense, a less tax-sensitive country, in which voters are more favorably disposed to liberalism's spiel that, in Voegeli's phrase, "we want the government to give things to you and do things for you." Which in turn suggests, perhaps, that the optimal moment for government cutting isn't a long period of economic expansion, but a period of stagnation or slight decline - when people still feel well-off enough that they don't need government to survive, but not so well-off that they're willing to throw their extra tax dollars after extra government services. This may be one reason why the small-government message was so resonant in the late 1970s, and again (albeit to a lesser extent) in the early 1990s - both were eras of belt-tightening and diminished expectations, and the public reacted to them by expecting less out of government itself.
This hypothesis also offers at least some reason to be a little more optimistic than Voegeli where Medicare and Social Security are concerned:
Looking forward, government spending as a percentage of GDP is about to rise dramatically. The oldest baby boomers, born in 1946, will be eligible for Social Security's early retirement benefits in 2008 when they turn 62, and become Medicare beneficiaries when they turn 65 in 2011. These two programs, along with Medicaid, accounted for 41% of federal spending in 2006, even before the baby boom cohort had started collecting benefits. All three will increase relentlessly due to the longevity and sheer numbers of Americans born between 1946 and 1964. The columnist Bruce Bartlett estimates that the magnitude of this growth will be "on the order of 10% of the gross domestic product over the next generation even if no new government programs are enacted or current ones expanded." This is the Swedenization of America on autopilot.
Except that it won't happen on autopilot; it will require enormous tax increases to sustain. This will force voters into hard choices in a way that the long Reagan-to-Clinton boom simply didn't - and if my gloss on Cowen's theory is correct, an appetite for government-cutting is more likely to arise from hard choices than from easy ones. Voegeli quotes, favorably, Paul Pierson's remark that "if conservatives could design their ideal welfare state, it would consist of nothing but means-tested programs." Well, the crisis of Social Security and Medicare will give conservatives their best chance yet to means-test both programs, since they'll be able to offer voters a choice between means-testing and tax hikes. If voters choose the former, which there's good reason to think they will, then we'll have a welfare state that's more expensive than today's (at least during the Boomer bulge), but that's also much closer to the right-wing understanding of what a welfare state should be. And that would be no small victory.