CPI & COLAs: Still costing us a bundle

J Brauer | © Stone Garden Economics

In preparing a lecture on the Consumer Price Index (CPI), a popular measure of consumer price inflation, I was reminded of some of the measurement and policy issues involved.

One of the issues is that the CPI is frequently used as the basis to make cost-of-living adjustments (COLAs) to a person’s base pay and/or benefits. But economists know that the way inflation is measured overstates the actual amount of inflation—there is no theoretically ideal or practically feasible way to measure it to perfection—and for this reason the Bureau of Labor Statistics, the agency that computes the CPI numbers, partially compensates for the overstatement. This leaves some of the overstatement unaccounted for, and many economists would probably be comfortable believing that it is still on the order of half a percentage point or more. That is, if the measured CPI inflation rate is, say, 2.0 percent, it may reflect an actual inflation rate of only 1.5 percent or less.

This matters greatly because a very large number of outlays in the U.S. federal budget—including social security payments, military retirement payments, and subsidies for school lunches—are automatically increased (or “indexed”) each year by the amount of CPI inflation.

For example, if your annual military retirement pay is, say, US$40,000 and measured inflation is two percent, then the retirement pay increases by two percent to a new annual pay of US$40,800. But if the actual inflation had been just 1.5 percent then the new pay should have been US$40,600, or US$200 less.

A US$200 annual overpayment seems like a very small matter indeed—until one looks up the numbers of military retirees and other beneficiaries. Thus, the annual Statistical Report on the Military Retirement System (MRS), issued in May 2013 by the Department of Defense Office of the Actuary, reports that as of 30 September 2012 (the end of fiscal year 2012), there were 1.47 million non-disability retirees from active military duty, entitled to draw retirement pay of US$42.1 billion that fiscal year. A rough estimate of the inflation-related overcompensation in this amount is about US$10 million.

But then they were an additional 96,000 military disability retirees drawing US$1.38 billion; 376,000 folks retired from the military reserves, receiving US$5.36 billion;  328,000 survivors of eligible members of the armed forces, drawing US$3.81 billion; and 58,000 folks who, after the end of the cold war, were eligible for early retirement, and costing US$938 million in FY 2012.

All these numbers add up of course: They apply not just to a single year’s worth of military retirement system payments, but for all public sector outlays, and over many years. They include for instance well over 50 million monthly nonmilitary retirement and disability checks from the Social Security Administration. In fact, Michael Boskin, an economics professor who once chaired an important congressional investigation into these matters, believes that the cumulative adverse effect has run into the trillions of dollars just for the federal budget alone. (Here is a copy of a 2004 paper laying out some of the details, and matters haven’t gotten better since.)

In addition, the use of a smaller, correct number for the CPI would “trip” some income earners into higher income tax brackets. To stick with the US$40,000 example, suppose that a CPI-adjusted bracket moves to US$40,800 (+2%) but that the adjustment should have been only to US$40,600 (+1.5%). Then a person who makes an income of US$40,700, would not be pushed into the higher bracket under the CPI-adjustment to US$40,800 and therefore cost the federal government foregone tax revenue, a higher debt load, and larger amounts of interest on debt. In contrast, with a lower tax bracket adjustment to US$40,600, the person earning US$40,700 now would pay higher income taxes.

It is easy to see that neither tax payers not recipients of federal payments like to get away from the CPI/COLA adjustment as currently practiced. As so often in politics, the fight over the “correct” number to use for the purpose of COLA adjustments is a fight over income distribution. But the current state of affairs is in fact costing us a bundle, and the right policy is to address the underlying income distribution issues, rather than to continue to fudge the numbers.

J Brauer is Professor of Economics, James M. Hull College of Business, Georgia Regents University, Augusta, Georgia, USA.

Comments are closed.