Ladies and gentlemen, we have a deal

Fair is fair, Brad DeLong shall now name the forbidden candidate. Brad and I have reached agreement on sundry matters concerning immigration, liberalism, and the welfare state. Here is Brad's response to my challenge:

"In response to my evangel, Tyler Cowen throws down the gauntlet:

Marginal Revolution: Brad DeLong tries to convert me: Finally, I'll offer Brad a deal. I will refuse to vote for the Presidential candidate he specifies (guess who that might be), if he will write in his blog, with no subsequent irony or repudiation the following: "The classical liberal recipe of increased immigration is superior to strengthening the welfare state. I just don't think it will or can happen, so I will advocate the next best thing." As a pure freebie, I will in advance volunteer the concession that most tax systems should be mildly progressive rather than flat or regressive. [That's Brad quoting me, if you are confused, now back to Brad.]

I can deal with the no irony or repudiation part, but there do have to be three clarifying footnotes:

Footnote 1: As Robert Waldmann has pointed out, you can't be an economist without asking "how much?" How much immigration does an extra unit of social insurance crowd out? Marginal rates of transformation matter. If the answer is "almost zero," the implications for public policy are very different from what they are if the answer is "a lot." This is an empirical question on which there is some (but not a lot of convincing) evidence, and Tyler and I disagree. If I believed (as he does) that we would have a much more liberal immigration policy if we had a smaller social insurance state, I would reach the conclusions he does. This is an example of Milton Friedman's dictum that in the end almost all disagreements between economists can be phrased as disagreements over matters of fact rather than matters of value.

Footnote 2: I'm not a cosmopolite: I care about myself and my family first; my friends second; my country third; and the world fourth. There are policies that would be good for the world as a whole and yet be bad for my country, and I would have to think long and hard before advocating such policies. I don't think expanded immigration is one such (although I fear that totally open borders may be one such). But it is worth bearing in mind.

Footnote 3: When the share of commodities that can be cheaply traded across national borders is large, trade can effectively substitute for migration. To the extent that freer trade avoids (some of) the political problems generated by freer migration, we economists should concentrate our attention on the first rather than the second (and here too marginal rates of transformation matter).

But with those footnotes, then yes, Tyler is right: Increased immigration is superior to strengthening the welfare state. I just don't think it will or can happen, so I will advocate the next best thing. From a cosmopolitan world perspective, almost all of the costs of maldistribution come from income gaps between nations and very little come from within-nation inequality. Development is far more important from a world welfare perspective than social insurance within rich countries. And immigration is a powerful tool for world development."

Scroll down two posts on MR for background context, in case you haven't been following the exchange.

The bottom line: This is fun, can we deal some more? (Do not forget the Virginia governor's race, much is at stake. Perhaps Atrios will repudiate the Democrats in return for my vote.) And now Lynne Kiesling has to tell me whether I am the mutton or the lamb.