Paul Krugman hates it when conservatives mention that welfare benefits and taxation induce some people to leave the labor force. Today, he wrote of “moochers not really seeking work because they’re cradled in Paul Ryan’s hammock” – mocking the idea that some people have stopped looking for work because of government benefits.

But when the topic was how Obamacare’s subsidies and taxes will induce some people to leave the labor force, the Krugman changed his tune completely: “they gain something that is, to them, worth more…more time with their children, an earlier retirement, etc.” What a difference a day makes!

Both of Krugman’s framings pass a moral judgment on the people involved, bad or good. And both distract from the real discussion, which is how public policy changes incentives.

The Congressional Budget Office estimated last week that Obamacare will destroy 2.5 million jobs because it will tax work and add subsidies for those who earn little or nothing. Krugman’s “extremely wonkish” discussion of the CBO score cleverly obscured the issue by focusing on a very small change – the tax treatment of employer-based health care plans. A much bigger ‘notch’ is being introduced by Obamacare: Large subsidies for those who work little enough. Krugman halfway acknowledges in another post, saying Obamacare “replaces one set of distortions with a different set of distortions.”

One might think that an “Extremely Wonkish” approach to Obamacare’s distortions would be like Casey Mulligan’s work – a detailed effort to quantify the precise impact on effective marginal tax rates. Instead, Krugman gives us a passable ECON-101 exam question that has more relevance to longstanding conservative proposals to equalize tax treatment of individual health plans than it does to Obamacare. The really wonkish work, Mulligan’s, estimates that the job losses will be twice as big as CBO currently estimates.