I have complained in the past about reporters' willingness to accept corporate numbers uncritically. My favorite example is the widely reported claim that the compensation of Delphi's unionized workers averaged $65 an hour. This implied a benefits package worth more than $70,000 a year. Anyone believe that?
We have another example from the Times today. Its article on GM's buyout offer to workers reports that getting rid of 35,000 workers will save GM $8 billion a year. Hmmmm, my calculator puts that at a savings of just under $230,000 per worker.
If you think this number sounds a bit high, you would get confirmation by the end of the article. The last paragraph reports the assessment of a stock analyst that GM earnings will rise $1.25 a share for each 10,000 workers who accept the offer. With 565.6 million shares outstanding, this implies additional earnings of $707 million for each 10,000 workers, or $2.47 billion for the 35,000 workers who accepted the buyout. That comes to a more pluasible gain of $70,700 per worker.
I am not an expert on the auto industry and cannot assess how much GM will actually benefit from this buyout (the net gain will depend on whether GM can lose one-third of their workers and still produce the same number of cars, or will have to hire replacement workers), but the $8 billion figure reported in the Times article is obviously ridiculous. Someone involved in the publication of this article should have been able to recognize this fact.