The Treasury has always been very explicit that the CPI for indexed bonds is the CPI published by the BLS, that if the BLS changes the way the CPI is made then that will change the amount of interest to be received by holders of indexed bonds, but that Congressional declarations will have no effect on indexed bonds unless the Congress legislates that the BLS shall publish such-and-such a number and say it is the CPI.
Most legislative proposals are not so willing to dictate the exact numbers that a statistical agency shall publish. Most proposals are to take, say, half a percentage point off the indexation formula, so that benefits (and tax brackets) grow at CPI - 1/2 percent per year, not at CPI percent per year, and to explain that this is being adopted because CPI - 1/2 is a better approximation to the true cost of living than is CPI. So the message that "CPI - 1/2" is for the beneficiaries rather than the bondholders is certainly true.
In all of this the BLS continues to follow its own course as it decides how to calculate its index, in which it is (a) eager to assert that it is an independent statistical agency making technocratic decisions, and (b) keenly aware that if Congress gets *really* *unhappy* with it, its budget is in danger...
I have always tended to think that CPI - 1/2 is a good idea, because my calculations are that it is on net progressive: half of the shift in the budget deficit comes from increasing taxes on the rich by lowering (by 1/2 percent per year) the brackets at which the higher marginal tax rates kick in. The spending part of it is very regressive, but on the whole it is a shift of wealth away from the rich.
Of course others disagree--and make the good point that in steady-state, after it has been implemented for a long time, those hit hardest by CPI - 1/2 are 95 year old widows with no resources other than Social Security, who lose 15% of their benefits...
Brad DeLong
-- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- "Now 'in the long run' this [way of
summarizing the quantity theory of | <delong at econ.berkeley.edu> money] is probably true.... But this| Brad De Long long run is a misleading guide to | Dept. of Economics current affairs. **In the long run**| U.C. Berkeley we are all dead. Economists set | Berkeley, CA 94720-3880 themselves too easy, too useless a | (510) 643-4027 phone task if in tempestuous seasons they | (510) 642-6615 fax can only tell us that when the storm| http://econ161.berkeley.edu/ is long past the ocean is flat again."
--J.M. Keynes