Julian Sanchez has an excellent post up about the sophistic notion that there is no reason to draw a distinction between different types of government redistribution:
I've also been a little astonished to see progressives acting as though opposition to "redistribution" is just some bizarre incoherent notion, because duh, almost all government programs redistribute money in some way. At first, I thought this was just a cute bit of sophistry, on par with claiming that all government programs are "faith based initiatives," because Congress must have "faith" that the program will accomplish its goals. After all, the past half century of political philosophy has been occupied with debates over whether and to what extent it's appropriate for the state to redistribute income, and all the parties to that debate at least seemed to acknowledge that they had a real disagreement about some intelligible question. But the argument appears to be offered in all sincerity. So maybe it's helpful to consider a few different types of distribution.
At the very least, most government programs that require spending tax revenue will involve what I'm going to call incidental redistribution. Take the least controversial government functions, like national defense or courts. These are textbook cases of public goods requiring public provision--they're supposed to benefit everyone, but in such a way that people can't be individually excluded from the benefit provided. Hardcore libertarians will probably disagree, but a similar case can be made for public subsidy of general education in a democracy. Now, given that people vary widely in their ability to contribute tax revenue for such goods, even a flat tax means that people are going to kick in different amounts for these goods. So certainly there's a sense in which one might say provision of public goods involves "redistribution": People who can't afford to pay much, or anything at all, toward them at a given time nevertheless benefit from the funding provided by the better-off.
It's not especially helpful to talk about this as "redistribution" in the context of the historical debate over the idea. The justification for these programs is that they are a net benefit to everyone (or almost everyone) in society, including those who foot a disproportionate chunk of the bill for them. In terms of the Doctrine of Double Effect, most familiar from just war theory, we can say that redistribution is an inevitable consequence of the provision of public goods, but not the reason for which programs providing public goods are enacted. Redistribution of a sort occurs, but it is not the point.
Now there are two other kinds of redistribution, and here I think it may be helpful to reference an excellent paper by Joseph Raz on distributional equality, about which I hope to have more to say in a future post. First, we have what I'm going to call altruistic redistribution. What I'm talking about here is transfer programs aimed at helping the badly off, where the justification for the program is specifically the benefit to the worse-off, and not centrally any benefit to the people footing the bill. Obviously, both justifications may be in play with respect to a particular type of transfer. You may believe that all citizens in a democracy, including the very wealthy, benefit on net from public subsidy of education. But you may also believe that, quite apart from this rationale, we have a moral duty to ensure that the children of the poor have access to some threshold level of education, and that this would be the case, even if doing so were not a net benefit to the folks paying the bills.
Finally, we have what I'm going to call egalitarian redistribution, which is the view that resources should be transfered from the rich to the poor and middle class because economic equality is a good in itself. The idea here is not simply that we have a duty to provide folks who can't afford it with some basic minimum quality of life--say, by ensuring that they have things like health care or education or food--but that fairness independently requires a more equal distribution of resources, above and beyond whatever duty we might have to promote the welfare of the badly-off in absolute terms.
Now, I think Raz argues pretty persuasively that the argument for distributional equality as an intrinsic good is very weak. And indeed, I've argued on this very blog (though I'm not finding the post just now -- update: found it, link added) that most people who argue against income inequality don't really do so on the basis of the intrinsic-value egalitarian argument, even if some of them think they do. Still, this sort of argument does get made, so it makes sense to include it, because it's arguably the most pure rationale for redistribution--an argument for redistribution as such, independent of any particular benefit to which we might think the poor are entitled.
Within mainstream political discourse, just about everyone accepts the need for the first type of (incidental) redistribution. Most people are also on board with some level of altruistic redistribution, though there's obviously a pretty broad gradient here, especially as egalitarian redistributive arguments get thrown into the mix. You can probably make a "now we're just haggling price" move here against people who support only minimal altruistic redistribution, but there are probable plenty of points within the "altrustic" rubric where you can reasonably draw a line and say that redistribution beyond this threshold goes too far, even if political campaigns are poor fora to dive into the weeds and establish the precise location of that line.
So I'm just not terribly impressed with arguments that move from incidental redistribution, or even low-end altruistic redistribution, and proceed to the conclusion that it's somehow nutty or incoherent to deploy rhetoric that's skeptical of a "redistributive" philosophy of the role of government.
It seems to me that if you can see the difference between the Soviet leaders losing 10 million Russians in World War II, and the Nazis killing 11 million "undesireables", you ought to be able to tease out the distinction between redistribution as a goal, and redistribution as a side effect.