This is the classic problem of universalism vs. targeting efficiency, but
I'm
not sure I come down on the same side you do. On the universalistic side,
money for the poor requires, as a kind of informal political blackmail,
money
for the rich (or at least the more affluent). Targeting focuses the
benefits,
but risks the stigma of welfare and has all the other problems that you and
Nathan have been debating. Under ideal circumstances, I would prefer
universalistic benefits that are taxed differentially, because I think they
encourage the formation of more durable political coalitions. What's not
clear to me is whether you are defending targeting as a political necessity
in these politically constrained times, or as something you would choose
even in the face of more robust options. The former I can understand; the
latter, I have more trouble with.
Joel Blau
>>>>>>>>>>>>
The scheme I'm working up now is universalistic,
but to be as redistributive as I want it still
requires something extra for the bottom, which
in turn requires some kind of phase-out, if only
a partial one.
For something extremely 'universalistic,' say
a straight grant of X per child for all children,
an adequate grant would require a more progressive
tax system than we have now. Otherwise you're
moving an enormous amount of money around to
little effect.
Elsewhere I've written that a big tax system that
is roughly proportional (relative to income) will
be more redistributive than a small progressive
system. So I'm definitely in the universal camp.
But we have the worst of both worlds -- a small
and not terribly progressive tax system (taking
everything together). So I conclude that some
targeting for an incremental change in a benefit
program is inescapable.
mbs
