Semi-Daily Journal Archive

The Blogspot archive of the weblog of J. Bradford DeLong, Professor of Economics and Chair of the PEIS major at U.C. Berkeley, a Research Associate of the National Bureau of Economic Research, and former Deputy Assistant Secretary of the U.S. Treasury.

Friday, September 22, 2006

Intellectual Garbage Cleanup: Gene Epstein Edition

Sigh. There are so many more fun things I could be doing than intellectual garbage cleanup. But somebody has to do it, if only to give people doing Google searches a chance.

Today we have the unpleasant task of looking again at Gene Epstein of Barrons.

In our last episode, New York Times ombudsman Danny Okrent claimed that Paul Krugman was "blending numbers from the household survey and the establishment survey -- apples and oranges -- ... to make a more vivid political point about Bush (5/25/04)" by combining Establishment Survey estimates of payroll employment with a number "140,000 new jobs each month needed to keep up with the growing population... that... only appeared in the household survey."

Okrent's attack was false. Krugman's 140,000 per month number of new payroll jobs needed to keep up with growing population did not come from the household survey. It came from combining Census estimates of the growth rate of the adult population with the current level of payroll employment from the establishment survey. No blending. No mixing and matching. No apples and oranges. Apples only.

Okrent was simply wrong.

And he has never had the ovaries to admit it.

Gene Epstein of Barrons wrote a book, Econospinning, in which he joined Danny Okrent in virtual self-immolation. I wrote to tell him he was wrong, and why.

Now Epstein is back for a second encounter with the self-poured virtual gasoline and self-lit virtual match.


Econospinning: I quote the sentence from [Krugman's] May ‘04 column that leads up to the “math” Krugman would have us do: “And employment is chasing a moving target: It must rise by about 140,000 a month just to keep up with a growing population.” I then comment: That could only come from the Household Survey because it both explicitly and implicitly refers to measures that only the Household Survey covers. The measures implicitly referred to in the statement--which explain the link between “employment” and the “growing population”--include unemployment, labor force, and that huge negative category, not in labor force.... [Those] terms are all Household Survey measures....

DeLong reveals that... payroll employment comes from the Employment Situation Report’s Establishment, not Household, Survey. But... readers of Krugman’s column are not expected to read his mind, what he means by “employment” in what he actually wrote is unclear. While he could mean payroll employment... he could just as easily mean the somewhat different measure of employment in the Household Survey.... Krugman has just referred to... Establishment Survey figures... you would think he means payroll employment....

But... these statements are preliminary to the “math” he would have us “do.”... The confusion... [is] further compounded by the sudden use of the term “full employment,” which is also a Household Survey concept. Look it up in the glossary of Paul Samuelson’s well-regarded economics text... you find it refers to “that level of employment at which no (or minimal) involuntary unemployment exists.” Since that kind of “employment” is measured in the Household Survey, those of us who could not read Krugman’s mind could not reject that interpretation of what he meant....

[H]is figure of “about 140,000” necessarily refers to the expected growth in the labor force [and not to growth in payroll employment]. We can infer this from an example DeLong himself gives.... [Bush's] [Economic] Report [of the President] does make it clear from the context that the “employment” being referred to is payroll employment. But otherwise notice... references to the “labor force” and “unemployment rate” that come directly from the Household Survey....

The point, then, is that Krugman’s use of that term “employment” was ambiguous....

Maybe DeLong is hung up on the difference between a number literally being drawn from a survey--as the figure of 288,000 was literally drawn from the April ’04 Establishment Survey--and a figure like 140,000, a projection of growth in the labor force, which is not literally drawn from the Household Survey, but does refer to that Survey.... Nowhere does [DeLong] mention that the Current Population Survey (CPS) is conducted by the Census Bureau as a joint effort between that agency and the Bureau of Labor Statistics....

[I]f that data is about the labor force, to say that it doesn’t come from the Household Survey is rather like saying that Bill Clinton was never really “alone” with Monica [Lewinsky].

What can one say? Other, that is, than that this is really embarrassing to watch. To call this incoherent is an insult to all the incoherent people in America.


Post a Comment

Links to this post:

Create a Link

<< Home