My last few posts have all been dedicated to an argument that redistribution is a misnomer: there is no such thing as redistribution, just different distributions. In making that argument, I have insisted that we adopt a blank slate vantage point when talking about distribution. I think doing so makes discussions clearer because it does not allow individuals to assume certain distributional consequences as somehow preceding distributional policy. You might think I insist on this because getting individuals to go along with that framing is the only way I can make my point. You would be wrong.

In my previous post, I responded to a post from Adam Ozimek of Modeled Behavior. Ozimek argues that if someone had a watch, and then someone else just took it from them, that would still constitute redistribution whether there were governments in place or not. This is supposed to be a common sense notion. Originally, I just pointed out that this is not the correct distributional vantage point because it smuggles in pre-existing property as if that property is not itself a consequence of distributional choices. But we can go along with it as well.

In that example, someone is being deprived of a watch that, in the status quo, they have a distributional entitlement to. We have other things like that too. For instance, right now tens of millions of Americans get food stamps. In the status quo baseline distribution, those food stamps belong to them. Many people — perhaps Ozimek even — want to take away those food stamps. If we are going to call taking the watch away a form of redistribution, then we must also call taking the food stamps away redistribution. Both entail a change in distribution relative to the status quo carried out by taking away something from someone. There is no functional difference between the two.

But that is not how certain people would describe it. They would say the food stamps are redistribution, and taking them away is unwinding redistribution. This is where we start wading through confusion. Such an individual — whether they state it or not — is actually endorsing a view that designates some sort of “market income” as the correct distribution. Deviations from that become redistribution. But designating market incomes as the correct distribution is a normative view. Such distributions are not commanded by the universe, not default, and are certainly not naturally existing absent government intervention: markets are constructed, contract law is constructed, property law is constructed, and so on.

In the absence of a natural, default distribution, we are left with three options for talking about “redistribution.” The first option is to reject it as a basically incoherent category. The second option is to use the term to refer to migrating from one distribution to another (e.g. taking away a watch is redistribution just as taking away food stamps is). The final option is to embrace the highly ideological construction of the word “redistribution” by treating it as meaning “distribution contrary to my normative beliefs about who deserves what.” In that way, a Rawlsian can call distributions that do not follow the maximin principle redistributive, and a libertarian can call all distributions that deviate from laissez-faire distributions redistributive.

Notably missing from this list is the way we currently talk about redistribution, in which we pretend we can objectively and non-ideologically figure out what is and is not redistribution by merely describing a given distributional policy. That is completely untenable, and nobody who wants to be taken seriously should talk that way.