>Universal benefits often use of a lot government funds on middle class and
>wealthier citizens, often essentially redistributing from middle class tax
>payers to middle class recipients with little economic redistribution. Social
>security is a good program, but for all its universality, the rich pay very
>little into it proportionate to their income. The argument for universal
>programs is that they are politically bulletproof, but social
>security is partly
>bulletproof because the wealthy are not taxed to pay for it, so they don't
>mobilize against it. National health care in a number of countries is funded
>heavily by national VAT (sales-type) taxes that are not particularly
>progressive.
>
>The advantage of means tested programs is that the money spent is almost pure
>redistribution, largely at the federal level from progressive income
>taxes paid
>overwhelmingly by the wealthy directly to those most in need. For the same
>reason, those programs are under continual political assault.
Comparative studies show that countries with means-tested benefits pretty consistently offer lower benefits than those with universal schemes, so the targeting argument is spurious. (See the Luxembourg Income Study references at <http://www.panix.com/~dhenwood/LISBiblio.html> for more.) There's a lot of truth to the cliche that programs for the poor are poor programs. Also, in countries with universal benefits, competing private sector programs are unlikely to develop, so the upper part of the income distribution is less able to opt out of the public program. Finally, there's some evidence in these LIS studies that countries with more progressive tax structures have less redistribution overall.
Doug