Bryan Caplan writes,
1. Universal social programs that “help everyone” are folly. Regardless of your political philosophy, taxing everyone to help everyone makes no sense.
2. In the U.S. (along with virtually every other country), most government social spending is devoted to these indefensible universal programs – Social Security, Medicare, and K-12 public education for starters.
3. Social programs – universal or means-tested – give people perverse incentives, discouraging work, planning, and self-insurance. The programs give recipients very bad incentives; the taxes required to fund the programs give everyone moderately bad incentives. The more “generous” the programs, the worse the collateral damage. As a result, even programs carefully targeted to help the truly poor often fail a cost-benefit test. And while libertarians need not favor every government act that passes the cost-benefit test, they should at least oppose every government act that fails it.
Read the whole thing. However, these first few paragraphs made me worry about a nirvana fallacy. Nirvana would be a program that gives money or services to poor people without creating incentives that tend to discourage work. I do not believe that nirvana exists, so that in the real world we have to compromise. You either let poor people suffer or you provide a program that dilutes their incentive to work. Some further remarks:
1. Programs to help the poor do not necessarily have to be provided by the national government. They could come from local governments, or from private charity. I personally would like to see a mix. I can make a case for a universal basic income provided by the national government, with a marginal tax rate of 25 percent or less. For example, with a 25 percent tax rate, if a household of 4 with zero income gets $12,000. then when its income reaches $20,000 it gets $7000 and when its income reaches $48,000 it gets zero. Beyond the basic income, local governments and private charities could provide supplemental income and services to households with special needs, such as a child with expensive medical problems. If we are worried that households will not budget to meet their basic needs, we can give them money not in dollars but instead in the form of a flexible benefit that can only be spent on food, housing, medical care, and education.
2. I also can make a case for having generous welfare states, but with no involvement at the national level. Denmark and Sweden have less than 10 million people each. Why do we need to spread a welfare state over 300 million people?
3. I think that the case for abolishing or phasing out existing social welfare programs is very strong.
4. Those who favor a role for the national government in providing a welfare state should worry about a political nirvana fallacy. Take your ideal welfare state. Maybe it looks like (1) above). Maybe it looks like some idealized version of Scandinavia. In practice, how do you get from here to there?
Overall, I think that there is a powerful pragmatic utilitarian case for reducing the role of the national government in the U.S. in providing support for education, health care, housing, and income security. Not because we do not wish to be generous in helping people with those benefits. But because the set of national programs is so wasteful and inefficient.
For a more philosophical counter to Bryan, see Matt Zwolinksi. I do not like to discuss these issues solely at an abstract philosophical level. I am more focused on taking reality as it is and posing the question of what is the direction for improvement.