Switching to the Chained CPI Is a Good Idea

Yesterday, the AARP published a report and sent an accompanying letter to members of Congress and President Obama in opposition to switching to the chained CPI for calculating cost-of-living adjustments (COLAs) for Social Security. In the letter, the AARP argues that the chained CPI is not an accurate or appropriate COLA calculation for Social Security and suggests that its effects are regressive -- including the tax portion.  These points are simply not true.

We have written extensively on the chained CPI, including a full paper on the subject last May. As we have explained, economists of all stripes believe the chained CPI is the most accurate measure of inflation, as it is unbiased for consumer substitution between goods when relative prices change (for example, between apples and oranges). The AARP advocates going in the other direction by switching to the CPI-E (Experimental CPI for the Elderly). However, the CPI-E suffers from serious methodological flaws, including that it is only a sub-index, it has a small sample size, it doesn't account for certain elderly consumption patterns (use of catalogues, senior discounts, etc.), and it suffers from the same substitution bias as the unchained CPIs. Partly because of these concerns, CBO concluded that "it is unclear, however, whether the cost of living actually grows at a faster rate for the elderly than for younger people."

AARP also criticizes the chained CPI for being harder on older beneficiaries as the lower COLAs compound. It is true that the effect of the chained CPI grows as beneficiaries age, but this can be addressed in the context of Social Security reform -- for example, by bumping up benefits for those have been in the programs for 20 years, as the Domenici-Rivlin and Simpson-Bowles plans do.

AARP's claim that the tax impact would be regressive is also false, as we've pointed out before. The JCT study used in justifying the claim is based on percent change in tax burden rather than percent change in income as well as current law rather than current policy -- making the study virtually irrelevant for measuring the distributional impact. When one corrects those issues, as Tax Policy Center did, the distributional result is roughly even by income group.

^{Source: Tax Policy Center}

At the end of the day, switching to the chained CPI would save more $200 billion over ten years through a combination of spending reductions and revenue increases while closing one-fifth of Social Security's funding shortfall. Being able to save that amount of money from a technical change that most experts believe should be happening anyway should be a no-brainer. Though groups from the left and right may criticize it, we hope policymakers can look past this and do the right thing.