off the charts

The Chained CPI: A Response to Robert Kuttner

The President’s decision to include, in his latest “fiscal cliff” offer to House Speaker John Boehner, a proposal to use the “chained Consumer Price Index” in calculating both annual cost-of-living adjustments in Social Security and other benefits and annual inflation adjustments to various features of the tax code (such as the incomes at which tax brackets begin and end) is eliciting dismay among many progressives. At CBPP, we have long been open to such a change if, and only if, the tax savings are devoted entirely to deficit reduction (not to financing other tax cuts) and the benefit change includes protections for poor and very old beneficiaries.  The details on the Administration’s proposal are not yet available and, until they are, we’re withholding judgment on it. But, my old friend Bob Kuttner attacked us yesterday on The American Prospect’s website for our openness to the chained CPI.  Bob’s piece merits some response. First, Bob portrays the chained CPI as a massive Social Security benefit cut.  “On average,” he writes, “the cut is about 3 percent a year.” But, the average cut isn’t 3 percent per year.  It’s three-tenths of 1 percentage point per year (0.3 percent), according to the Social Security actuaries, or one-tenth what Kuttner assumed.  And according to the Congressional Budget Office, the average reduction would be a bit smaller than that — 0.25 percent per year. To be sure, the benefit loss would cumulate over time; after 20 years, it would represent a benefit cut of about 5 to 6 percent.  But, the proposal also appears to include a benefit increase after about 20 years that’s equal to 5 percent of the average Social Security benefit, returning the benefit level at that point to close what it would be under current law.  (It would begin to slowly erode again after that.) In short, this proposal would affect beneficiaries.  But the benefit reduction is much smaller than Kuttner portrays it. That leads me to my main disappointment with Bob’s piece.  The bulk of it purports to describe my thinking and motivation for being open to what he considers a horrific proposal.  Alas, Bob invents some of what he writes out of whole cloth.  He never talked to me about why I end up where I do on this issue, and he gets much of my thinking wrong. Let’s quickly dismiss a Kuttner canard.  He lays out a series of reasons why he assumes I’ve adopted this view and suggests that one could be to strengthen my access to the Obama White House.  That makes no sense.  I have been writing about the chained CPI along the same lines since 2004, when Barack Obama was still serving in the Illinois state legislature. So what is my thinking?  I share concerns about the effects of the chained CPI on beneficiaries.  But I think that some benefit cuts in Social Security are inevitable sooner or later.  The program needs some changes to make it solvent for the long term, and the chances that policymakers will restore solvency entirely through tax increases — with no benefit-reduction component — are essentially zero.  That didn’t happen in 1983, and it almost certainly won’t happen now or in the foreseeable future. Furthermore, as in 1983, benefit changes won’t be limited to high-income beneficiaries.  There are limits to how far one can cut benefits at the top without breaking Social Security’s link between payroll-tax contributions and benefits, and thereby risking undermining public support for the program.  I see these factors as basic political realities, whether we like them or not. That brings me to the key point: the chained CPI is the only Social Security benefit change that brings an increase in general tax revenues with it.  Eventually, about half of the savings from the chained CPI come from revenues, and about half from Social Security and other benefit programs.  The more that we raise in revenues, the less that policymakers will slash programs generally.  So, this — and the fact that the chained CPI is a more accurate measure of inflation (although not necessarily for the elderly) — leads me to conclude that if policymakers can build appropriate protections into the chained CPI to protect both the oldest and the poorest beneficiaries, it’s worth considering. Finally, I hope people reading Bob’s article don’t think I’m the sole liberal who’s open to the chained CPI.  The late, revered Bob Ball — the former Commissioner of Social Security who led efforts to defend the program for half a century and was a hero and beacon on social insurance issues to so many of us — put forth several Social Security plans that included the chained CPI as a way to help restore Social Security solvency.  Indeed, I first looked at the proposal after I saw that Bob Ball had included it in one of his plans.  The Center for American Progress also embraced the chained CPI several years ago and included the proposal in its Social Security solvency plan.