I read an interesting article in this weeks Barron's about the CPI. As y'all probably know, downward revisions of the CPI are one of the most effective means of upwards redistribution of income: taxes on labor rise, because brackets and deductions rise more slowly; social security and other government benefits decrease; union contract wage increases are usually tied to the CPI in some way, etc. The writer was plainly very excited about what this all meant to rentiers, although she did try to cloak it in the usual terms ("incentives to investment", etc).
Five revisions have been implemented, and two more are on the way. Every single one (surprise!) resulted/will result in a lowered CPI. A couple of them (the least important) actually made sense. Then there was the 'substitution' one: if beef goes up by x%, the CPI will reflect less than this increase because people will react by partly switching to chicken. Hell, if chicken goes up, then retirees can switch to rice and beans, which is more healthy anyway.
The most egregious revision, however, had to do with health care. So far, hospital costs were measured in terms of cost per day. From now on, the BLS will keep track of cost _per patient_. So, the sooner insurers can kick patients out of hospitals, the 'cheaper' health care becomes, and there will be no need to give people raises. I think this revision is crucial, given that the cost of health care is projected to skyrocket again after a brief lull.
-- Enrique Diaz-Alvarez Office # (607) 255 5034 Electrical Engineering Home # (607) 758 8962 112 Phillips Hall Fax # (607) 255 4565 Cornell University mailto:enrique at ee.cornell.edu Ithaca, NY 14853 http://peta.ee.cornell.edu/~enrique