How Chained CPI Affects Social Security COLA

Using new price index to set COLA would cut benefit increases by a quarter of a percent each year.

By SHARE

As the fiscal cliff negotiations head toward actual changes in federal tax and spending laws, the White House is reportedly willing to accept some cuts in Social Security benefits. This is part of the cost to secure Republican acceptance of higher taxes on the wealthy and spending plans that would support other traditional Democratic causes, including maintaining extended unemployment benefits, which are set to expire at the end of the year.

Social security card on top of a table
Social security card on top of a table

The impact on Social Security would occur by using a different measure of consumer price changes to determine the program's annual cost of living adjustment (COLA), which is announced each fall and takes effect the following year. Extremely low rates of inflation meant there was no COLA in 2010 and 2011. The COLA was 3.6 percent in 2012 and will be 1.7 percent in 2013.

The new index, which the government has been calculating for more than a decade, is called the chained consumer price index for all urban consumers (C-CPI-U). The COLA is now determined by a version of the CPI that measures prices for urban wage earners and clerical workers (CPI-W). The CPI-W measures price changes but does not assume that people change their buying habits in response to those changes. The chained CPI, by contrast, assumes consumers change their purchase habits when prices rise—substituting cheaper cuts of meat, for example, or switching to generic store brands from more expensive branded items, or just buying less.

The Bureau of Labor Statistics (BLS) calculates that the chained CPI has risen an average of just under three-tenths of a percentage point less each year than the CPI-W during the past decade. The agency does not make projections of future changes. The nonpartisan Congressional Budget Office last year forecast that the C-CPI-U would cause the Social Security COLA to rise during the coming decade by a quarter of a percentage point less each year than it would using the CPI-W.

[Your Retirement Benefits: What to Expect in 2013.]

This change, the CBO estimated, would reduce Social Security spending by $112 billion during the 10 years ending in 2021. Since that 2011 estimate, the total has risen a bit. More importantly, its long-term cumulative effect is particularly large. In the first year using the C-CPI-U to set the Social Security COLA, payments would drop by only $1.4 billion from what they would be if the CPI-W was used. But in the tenth year, they would be nearly $22 billion less than had the CPI-W been used to set the COLA each year for the previous decade.

The White House reportedly linked its acceptance of the chained CPI with safeguards to protect lower-income and disabled beneficiaries as well as veterans. The details of those proposals have not yet been publicly disclosed.

The National Committee to Preserve Social Security and Medicare, AARP, and most other seniors' organizations decry even including Social Security in the fiscal cliff negotiations, let alone cutting benefits. The program is self-funded and has not added to government deficits, they note.

While revenues from payroll taxes are less than current Social Security benefits, the program also receives interest payments on accumulated trust fund surpluses. Those surpluses now exceed $2.7 trillion and are projected by program trustees to rise to more than $3 trillion by 2021. Under current rules, surpluses would then decline due to rising numbers of baby boomer retirees and growing annual deficits. The trust fund would be depleted in 2033, at which time annual payroll taxes would fund about 75 percent of promised benefits.

[Read: 10 Great Senior Discounts.]

While the program does need some adjustments, supporters note, they are relatively minor in comparison to other federal spending pressures, especially healthcare. Strategically, they also object to involving Social Security in broader budget negotiations and feel the program's integrity and public support are stronger if it is treated separately.

There also are strong objections to using the chained CPI to set the annual COLA. Already, supporters argue, the CPI-W understates the impact of inflation on seniors. That's because it does not reflect their relatively heavy spending on healthcare. If a different index is going to be used, they say, careful study should be given to using an experiment index, the CPI-E, that measures price changes for elderly consumers.

"There is a lot of confusion about substitution" in the BLS indexes, says John Greenlees, research chief of the BLS unit that develops its price indexes. "People think it's about shifting to something not as good when prices go up." The chained CPI, however, makes no assumptions about the specific items people might buy instead, simply that the relative amount of money they're spending on a specific item has changed. "It's not as if you're poor now and can't afford to buy what you like."

The BLS surveys prices for 211 item categories in 38 geographic areas, producing 8,018 basic price indexes. These indexes are the same for all major CPI indexes, Greenlees says, including the chained CPI. And within this first-stage indexing work, he explains, there is already some assumed product substitution taking place. In a second stage of work, differing assumptions are applied to these indexes.

[In Pictures: 10 Books Investors Should Read.]

If the price of an individual item category has gone up but overall consumer spending on that item has not gone up as much (or actually declined), the chained CPI formula concludes that consumers are not paying the full amount of the item's price change, and that their cost of living has not increased or not increased by the full amount of that price change. The CPI-W does not assume any changes in consumer behavior, so it will register a higher rate of inflation.

Further, all the indexes look at each item category and assign it a weight in overall consumer spending. Changes in an item's price thus influence overall inflation in proportion to that weight. Oil prices, for example, carry a big weight, so shifts in gasoline prices have a big influence on the overall CPI. For the CPI, these category weights are changed only once every two years. The category weights in the chained CPI, by contrast, are changed every month.

Consumer spending on all medical care, including health insurance, is a bit less than 7.1 percent of the weight of the widely used CPI, nearly 5.7 percent in the CPI-W (the index now used to set the Social Security COLA), and more than 6.9 percent in the chained CPI. For the experimental elderly price index, the CPI-E, medical spending had a weight of 11.6 percent.

Because the chained CPI calculates price inflation of the things people actually buy, Greenlees says, "the BLS is on record as saying that we think the chained CPI is a closer approximation to a cost-of-living index."