As Paul Krugman says,

Not much point in going through Amity Shlaes’s latest: after having inadvertently revealed that she has no idea what Keynesian economics is, she’s back on the warpath against FDR, and me.

Krugman deals with her wages argument there. But she gives half her column to the unemployment statistics story again, apparently in the hope that if she makes it seem complicated and controversial, it will look as if she’s got it right. Let me spend some virtual column inches on this one more time.

Amity Shlaes lives in New York. She has a unique advantage: in many cities, access to a world-class research library, if one exists at all, is open only to members of a university community. But in New York, the Public Library is a world-class research library.

Suppose like Amity Shlaes you’re in New York. Suppose further you’re a sufficiently honest and competent researcher wishing to find out historical statistics of US unemployment in the 1930s. What would you do? You would go to the New York Public Library—for the purposes of this story it doesn’t matter whether you goes go to the main reading room on 5th Avenue and 42nd Street or the Science, Industry, and Business Library on Madison and 34th. Both are free and open, both easily accessible by a variety of major subway lines, and both have convenient hours that extend into the early evening in case you’re tied up during the working day.

If you go either to the main reading room or the SIBL you will find—and if you do not, a research librarian will be happy to show you—the current edition of Historical Statistics of the United States on open shelves—you don’t even have to page it from the stacks and wait for it.

If you open it up and look under B, “Work and Welfare”, then under Ba, “Labor”, then under “Employment and Unemployment” for a historical series on unemployment, you will find David Weir’s series from Research in Economic History, and if you read the footnotes, you will understand why Weir’s revision is regarded as superior to earlier versions. If you’re really curious, you might want to read the relevant issue of Research in Economic History to see how the debate proceeded; there’s a long forum there that includes Stanley Lebergott and takes in the work of Christina Romer (economic history is a small world). Unfortunately it looks as though NYPL doesn’t have the 1992 volume, so you might have to go somewhere else to do it.

But at this point you’re in possession of the relevant information. And if you’re not in New York or otherwise near a major research library, or maybe you don’t want to undertake this experiment yourself, you can read a summary of what you’d find here. Where I pointed out, quite politely, and in respect of which I point out, quite politely again, if you have got this far and you now choose to use a different series than the ones in HSUS, you are not doing what a sufficiently honest and competent researcher would do.

Shlaes spends a lot of time on the bona fides of Stanley Lebergott, who compiled the series that’s now considered pioneering, yet flawed. This is foolish ad hominemism; Newton was a pretty impressive fella, but Einstein improved on his physics.

And Shlaes’s method is by now suspect. It led her to praise the Bush administration’s competence in the wake of Katrina and then to acknowledge the administration’s incompetence but to blame it on the president’s wise respect for federalism. It led her to agree with Phil Gramm about our current crisis being “mental” and America being a “nation of whiners”. And it led her, as a correspondent points out by the email, to this prediction:

A house is inherently more lovable an object than cold metal or a stock certificate and so it is harder to part with. This means that the housing market cannot be as volatile as the others. In other words, while house prices may sag and sag, they will not pop.

She wouldn’t be worth responding to, you know, if she didn’t have free access to the op-ed pages of major newspapers, to the airwaves of PBS and even Comedy Central. But she does. It’s a mystery.

PS: Michael Cembalest says in his summary by email,

For the anti-FDR crowd that believes that WPA/CCC workers were not really employed: that would mean that the pre-war decline in unemployment to 10% by 1941 is even larger and more impressive, and that fiscal stimulus did an even better job. Another anti-FDR myth asserts that employment only improved via government workers; a chart decomposing civilian, government and emergency workers shows that all 3 contributed meaningfully to the employment gains of the late 1930s. I have relatives that despise FDR but can’t remember why; this phenomena seems to infuse a lot of the ideologically-based research against the fiscal policies of that era.

PPS: It also doesn’t matter that, per Shlaes, Barack Obama may have had Lebergott’s series in mind when he referred to unemployment in 1932-33. The president-elect is not, so far as I know, offering his services as an expert commenter on the subject.

PPPS: I took so long to write this, what with the patience and the reasonableness, that DeLong got in first, claiming Shlaes has ventured into “flunking the Turing Test territory.”