The Chained CPI Is Bad for Seniors and for Accuracy
This article is from Dollars & Sense: Real World Economics, available at http://www.dollarsandsense.org/archives/2013/0513miller.html
This article is from the May/June 2013 issue of Dollars & Sense magazine.
at a 30% discount.
The word “thuggish” comes to mind. “I’m not a number,” says the older man in a television ad funded by the seniors’ lobby AARP. ... “But I am a voter. So Washington, before you even think about cutting my Medicare and Social Security benefits, here’s a number you should remember: 50 million.”
This unyielding position, undergirded by a multimillion-dollar ad campaign, is as wrongheaded as the equivalent line-drawing of Grover Norquist and the no-new-taxes crowd. ...
[T]he brutal fact is that Social Security cannot pay all promised benefits, and a debt discussion is a useful place to make reasonable tradeoffs.
—Washington Post, “Congress should reject AARP’s self-centered appeals on Social Security,” Nov. 4, 2011.
That AARP television ad sure raised the hackles of the Washington Post editors back in 2011. The editors called AARP’s threat—to vote out any politician who supported a reduction in the cost-of-living adjustment (COLA) for Social Security benefits—”thuggish,” “self-centered,” in denial about the crisis of Social Security, and as “wrongheaded” as conservative power-broker Grover Norquist. That last one had to hurt.
Back then, the proposal to reduce the Social Security COLA by switching to the “chained” Consumer Price Index (CPI) didn’t come to pass. But now it’s back, this time as part of the 2014 Obama budget proposal and going by its technical economic name—the “superlative CPI.” Make no mistake, though. It’s the same idea now as then, and would reduce the COLAs for Social Security and veterans’ benefits, as well as the inflation adjustment for income-tax brackets.
What’s all the fuss about? The Social Security Administration currently uses the CPI–W, a measure of the price of a basket of goods and services typically purchased by urban wage-earners and clerical workers, to calculate COLAs for Social Security recipients. The “chained CPI-U,” as it is officially designated by the Bureau of Labor Statistics (BLS), grows at a slower rate than the CPI-W. Therefore, calculating the COLAs using the chained CPI will reduce future Social Security benefits by more and more each year. If that sounds to you like a roundabout way to hold down spending on Social Security, you’ve got it right.
The proposal is meant to establish Obama’s deficit-reduction bona fides and to lure Republicans and conservative Democrats into a “grand bargain” boosting tax revenues and cutting entitlements spending. For good measure, the Obama administration is selling the superlative CPI as just that—“a more accurate measure of the average change in the cost of living than the standard CPI.” And the Washington Post is once again on board, endorsing the Obama proposal for “Social Security spending restraint” as part of the “worthy end” of entitlement reform.
Using the chained CPI to reduce future Social Security spending, however, is far from the even-handed proposal the Post editors suggest. In truth, it is neither fair nor accurate. Worse yet, it would fall most heavily on some of the most vulnerable in our society—older women, veterans, and the disabled.
The CPI in Chains
To understand why, we need to look at just how the COLA for Social Security benefits is calculated. In 2013, the COLA was 1.7%, equal to the increase in the CPI-W from the third quarter of 2011 to the third quarter of 2012. For the typical Social Security retiree, this translated into about $250, boosting the average retirement benefit to just over $15,000 a year in 2013.
The CPI-W is what economists call a “fixed-weight” index. It measures the price of a fixed “basket” of 211 different items. (The basket of goods is updated every two years to keep up with changes in consumers’ buying patterns.)
According to the persistent complaints of conservative politicians and economists, however, that fixed basket results in the CPI-W overstating the rate of inflation. They argue that consumers typically purchase less of those goods whose prices are rising compared to those of other goods. Take the example provided by BLS: If the price of pork rises while the price of beef falls, consumers are likely to purchase less pork and more beef.
That’s the supposed problem the chained CPI is intended to correct. The “U” in “CPI-U,” by the way, stands for all urban consumers, a broader group than urban workers. The basket used for the chained CPI, therefore, differs from that for the CPI-W. More importantly, however, the chained CPI uses a flexible basket of goods that captures how consumers adjust their purchases in reaction to rising prices. The basket used for the final chained CPI is updated monthly. In the example of rising pork prices and declining beef prices, then, declining pork consumption means that pork prices will have less weight in the calculation of the index. By the same token, rising beef consumption means that beef will have greater weight inthe index.
The long and the short of it is that the chained CPI reports a lower rate of inflation than the fixed-basket CPI-W. The Social Security Administration estimates that using the chained CPI instead of the CPI-W would reduce annual Social Security COLAs by about 0.3 percentage points per year. If the chained CPI had been used to calculate the Social Security COLA, the average retiree would have gotten $45 less in benefits this year.
Perhaps it is these small figures that have the Post editors convinced that AARP is “wrongheaded” about the switch. The loss of benefits, however, gets larger each year, and the cumulative effect is substantial. The average 65-year-old is now expected to live about 19 additional years. According to AARP projections, a chained-CPI COLA would cost the average Social Security retiree more than $5,000 over the first 15 years of retirement and more than $9,000 over the first 20.
Nor is the chained CPI an accurate measure of the cost of living for most seniors. The typical senior spends a larger share of her income on medical care and housing than other consumers. The cost of both items has risen more quickly than other costs and it is hard to substitute for either item with other purchases. In addition, because seniors are less mobile than other consumers, it is harder for them to change their consumption patterns. This makes the chained CPI, whether or not it is valid for other individuals, inappropriate for calculating COLAs affecting seniors’ retirement incomes.
Double and Triple Whammies
The reduction of the COLA would hit hardest on some of the most vulnerable in our society.
Older women would suffer, as the National Women’s Law Center puts it, a “triple-whammy.” First, the effects of the change would increase over time, and women tend to live longer than men. (A 65-year-old woman is more than 1.5 times as likely to live into her 90s as a 65-year-old man.) By age 90, a typical single woman who retired at age 65 would have lost $15,000 of benefits from the switch to a chained-CPI COLA.
Second, women rely more heavily on their Social Security benefits than men do. Among beneficiaries 80 or older, Social Security accounts for two-thirds of women’s income, compared to three-fifths of men’s. So any reduction in benefits will cost women a larger share of their total income than it will men.
Third, older women are more economically vulnerable than older men. Among women receiving Social Security benefits, almost 10% remain in poverty, nearly twice the rate as for men. Shifting to the chained CPI would heighten the risk of poverty for these women.
Veterans would also be hit by the switch to the chained CPI. Because veterans with twenty years of service are eligible for their pensions as early as age 50, the cumulative effect of a reduced COLA would be particularly large. Disabled veterans would face a double whammy. With a chained-CPI COLA, they would collect lower Social Security benefits and lower veterans’ benefits.
The story is similar for those receiving disability benefits. The disabled typically start receiving Social Security benefits before retirement age, so their cumulative loss of future income will be much greater than retirees’. For instance, someone who began collecting disability benefits at age 30 would collect nearly 10% less in benefits at age 65 under the chained CPI, and the annual loss would get larger each year after that.
The switch to the chained CPI would also lower the inflation adjustment for income-tax brackets. While income taxes would go up across the board, more than three-quarters of the additional taxes would be paid by those with adjusted gross incomes under $200,000.Those with incomes between $30,000 and $40,000 would suffer the largest declines in their after-tax incomes, according to the Tax Policy Center.
Fair and Accurate
If accuracy were the goal of reforming the COLA, it would be far better to adopt the BLS’s new CPI-E (for “elderly”). This fixed-weight experimental index is derived specifically from seniors’ spending patterns, placing higher weights on housing and medical care than other indices, including the CPI-W. The BLS reports that between December 1982 and December 2011, the CPI-E added 0.2 percentage points to the annual inflation rate, compared to the CPI-W. (The difference between the two rates has shrunk recently as the rise in health-care costs has slowed.) Using the CPI-E would make it clear that the honest way to lower the COLA for seniors would be to rein in health costs and therefore slow the growth in their actual cost of living.
Finally, reducing benefits is neither right nor necessary to avoid the projected shortfall in Social Security payments starting in 2033. Currently, wage income above $113,700 is not subject to the payroll tax. Lifting this cap would eliminate the entire projected shortfall in one easy step. And unlike a reduction in the COLA, which would hurt the most vulnerable, lifting the cap would put the burden on some of those who benefited most from the lopsided economic growth of the last three decades.