Jim Manzi argues that, as conservatives strive to claim a decisive voice in governance, we should see the welfare state not so much as a demolition project, but as remodeling, with a different end-goal in mind:

… it would be foolhardy, from a conservative perspective, to eliminate a system so central to day-to-day life and long-term planning — and especially to do so all at once, acting on an unproved theory.
While it is always possible that some future society will find a way to cultivate widespread wealth and stability without a welfare system, or that existing welfare systems will wither away, the welfare state appears to be concomitant with the growth that capitalism creates. As far as can be determined from history, the idea of an advanced capitalist society without a welfare system is misplaced nostalgia — or more accurately, an anachronism. It is like wishing for a commercial jet aircraft without wing stabilizers. …
… First, the primary purpose of the system should be to support capitalism, not to oppose it. Second, we should seek the system’s maximum alignment with the elements of human nature that make us want it in the first place. Together, these two criteria simply mean that we should be as informed as possible about the costs and benefits created by the welfare system as we seek the greatest possible benefit for each unit of theoretically forgone growth that we invest in it. Third, we should attempt to shoot ahead of the duck by modifying the welfare system in a fashion that anticipates foreseeable changes in society and technology while leaving us maximum flexibility to respond to unforeseeable changes.

This all begins to sound like the folly of central planning — only declaring that our experts will manage the beast better than their experts. To be sure, we should not forsake those who simply fall through the inevitable economic and charitable cracks of a freely operating society, and we should not pull the rug out from under those who have taken the central planners at their word without offering a path off of it. Still, if increased individual autonomy and shrinking government are not the goals of reform, then the same incentive structures that have pulled the welfare state to its current position will remain in place.
The distinction manifests in all aspects of government operation, including in education:

… For several decades, a goal of the libertarian Right has been to voucherize social programs so that the government provides the cash but allows private firms to compete in markets to provide the services. But this is not always as practical as it sounds. …
To return to the example of K–12 schools, the focus on true privatization has been both doctrinaire and artificial. If school choice ever grows beyond Tinkertoy demonstration projects, taxpayers will appropriately demand that a range of controls be imposed on the schools they are funding. Would we allow families to use vouchers to send children to schools that taught no reading or mathematics, but only bomb-making, or that offered lavish “support payments” to parents that were, in effect, bribes? No, we would inevitably — and justifiably — have a fairly detailed set of regulations, along with inspection, adjudication, and enforcement mechanisms. At that point, what would be the difference between such “private” schools and “public” schools that were allowed greater flexibility in hiring, curriculum, and student acceptance, and had to compete for students in order to capture funding? Little beyond the label.
Publicly funded private schools is an oxymoron, but greater flexibility to meet different needs and to improve general performance through market competition can nonetheless be found in a public-school system involving parental choice and the freedom of schools to operate outside of collective-bargaining agreements and other restrictions. The most basic institutional requirements of a market would be present: consumer choice and widely distributed buying power on the demand side, capacity and flexibility on the supply side.

Here, Manzi embarks on the same sort of argument from extremes that his entire promotion of temperance eschews. The conservative principle of trade-offs and individual assessment of costs should make it obvious that the problem that Manzi raises is a matter of the degree of vouchers. The core rationale behind including private schools in a voucher system even if, for example, they concentrate on religious education, is that the state is giving the money to the parents for the purpose of education, not “establishing” the religion that its money ultimately supports. That perspective draws obvious lines for vouchers.
The system could be one in which parents can allocate every penny earmarked for their children’s education to the full-fledged public school of their choice, including charter schools that are freed from some of the chains of the broader system. But parents could also receive a portion — say, only that money which they pay into education through taxes — if they opt for private schools. In that way, the vouchers are only relieving the parents of the burden of paying for schools that their children do not use. Even by that structure, the government could reasonably have limited requirements that the schools actually teach basics like mathematics, whatever else they include in the curriculum.
The idea, again, is not to change the guiding principle of the overwhelming state machine, but to increase the autonomy and authority of individual citizens.