If I tell you about my personal budget and talk about my income and my costs, and you tell me that some of my costs aren’t actually costs because they’re just shifting resources from me to someone else, you’re being irrelevant. That’s what you’re doing here, with one more level of indirection.

> No, I think it reflects your failure to understand what we’re talking about.

This is ironic, because this whole post is based on your failure to understand what Krugman and the politicians he’s referring to are talking about. They’re very specifically talking about the government budget, and costs and revenues relative to that. The “costs” you’re speaking of are something entirely different, and very much not what either the “deficit hawks” meant, nor what Krugman meant.

Playing semantic games would be one thing (”no, that’s not what the word ‘cost’ means!” even when you can tell what they’re talking about), but what you’re doing is subtler and much more obfuscatory – you *assume* that the terms they use mean what you want them to mean, and criticise as if that were the case.

Hmm, I suppose some people might think that was lacking in earnestness; sorry, Steve. More seriously, have you ever blogged about the existence of a “Nobel” in economics versus the lack thereof in mathematics?

Well, given this statement: That is, the classical model seems to imply there are greater social losses living with a world that has taxes, property rights enforcement and roads than living in a world without., I guess we can *also* conclude that you fail to understand the classical model.

Cute, cute.

But briefly, and earnestly, I’d be interested to hear people’s thoughts on the question of whether there is any theoretical basis to conclude that the deadweight social loss related to taxes is greater or lesser than the deadweight social loss related to a lack of transfer payments that the taxes might have financed. Are these just reciprocal graphs?

Of course, my analysis may simply reflect my failure to fully understand the classical model.

No, I think it reflects your failure to understand what we’re talking about.

You are pointing out that government expenditures can be worthwhile. Nobody disputes that.

I am pointing out that government expenditures can be non-worthwhile. Nobody disputes that either.

I made no argument about the value (or lack thereof) of any program. My only point is that programs should be evaluated on their merits, and that Krugman has not only failed to do that, but directed attention away from it.

P.S. Well, given this statement: That is, the classical model seems to imply there are greater social losses living with a world that has taxes, property rights enforcement and roads than living in a world without., I guess we can *also* conclude that you fail to understand the classical model.

But a loss relative to what? To a world in which there is no taxation – in which things like property rights magically enforce themselves and roads magically build themselves, etc. In short, it’s a fantasy world.

No. A loss relative to a world where you enforced property rights and built roads exactly as you do now, but offered fewer unemployment benefits.

Well, in the world in which I live, I pay for property rights enforcement and roads with income taxes (and, ok, some other taxes). And in the world in which I live, the classical model show a “deadweight social loss” with respect to income taxes used for ANY purpose – including the purpose of property rights enforcement and roads. That is, the classical model seems to imply there are greater social losses living with a world that has taxes, property rights enforcement and roads than living in a world without. Or, to put it more charitably, the classical model appears to assume that the state of property rights enforcement and roads is exogenous, unrelated to the level of taxation. This strikes me as a big, fat weakness of the classical model.

I suspect Landsburg means to bolster reliance on the classical economic model by using it to refer to incremental taxes for extraneous things like the social safety net, unrelated to important things such as property rights enforcement. But this begs the question I posed above: Are social safety net expenditures really unrelated to other variables? Or is it reasonable to expect that increasingly desperate people will turn to crime or revolution? Is it reasonable to think that increasingly stressed people will require additional public health expenditures? In other words, is it appropriate to assume that a social safety net is not a productive investment, offsetting other costs?

Again, the classical model’s analysis of the consequences of taxation seems to be useful, assuming you can conclude that the taxed resources are simply dumped in a pit. Absent such an assumption, the classical model seems pretty flawed. No economist—let me repeat that—NO economist, not even Steven Landsburg on the days when he’s being an economist—would count a the costs of a policy while excluding its benefits for purposes of policy analysis.

Of course, my analysis may simply reflect my failure to fully understand the classical model. Yes, taxes cause parties to refrain from mutually-beneficial transactions that otherwise would have occurred. But don’t transfer payments also permit the recipients to engage in mutually beneficial transactions that otherwise would NOT have occurred? That is, is there a theoretical basis for saying that the deadweight social loss that results from taxation is greater than the deadweight social loss that would result from the LACK of a transfer payment?