Dean Baker and Doug Henwood both have good analysis on the cuts involved in chaining inflation. Since the rumored cuts to Social Security will hinge on this way of calculating inflation, I want to dig one level into the data to convey what it will mean and then look at some of the distributional impact.
Let's start with two groups of people. The first is urban wage earners and clerical workers, one select group of the population, who purchase a representative basket of goods and services. How much does the basket of goods they purchase increase in price over time? This cost is called CPI-W, and it is currently used for adjusting Social Security benefits. The second group is all people aged 62 and over. Since the 1980s, the government has calculated the cost of goods and services for this group as well, and it is referred to as CPI-E. What do they spend money on? Here's the relative importance of major categories of spending, provided by the BLS, for each group from December 2007:
Green is where the group spends compartively less. As we can see, the elderly spend a lot more of their (more limited) money on housing, utilities, and medical care. And as you probably know, health care costs have been rising rapidly over the past several decades. With the notable exception of college costs, the things urban wage earners spend money on haven't increased in prices as fast as what the elderly purchase. As a result, the CPI-E has increased 3.3 percent a year from 1982 to 2007, while the CPI-W has only increased 3.0 a year.
But wait, what's this chained thing that is being proposed? Picture that in response to a price increase for one good you could substitute similar items. So if the price of chicken goes up, you could eat more beef. Or if the price of a movie went up, you would rent movies more often. This substitution effect blunts some of the price increases. As such, inflation is lower when you take this into account. It's more complicated than that, but it is a start for a definition.
But we don't have a "chained" version of the CPI-E. And the items that the elderly purchase probably aren't impacted in the same competitive way. If the price of beer goes up, you can drink more wine; if the price of utilities go up, your options are limited. The areas where the elderly pay more don't have the same competitive pressures, and their geography is going to be more limited. We could get a chained version of the CPI-E if Congress told economists to make one. However it's likely not to have the cuts built in the same way.
Brad Delong, who signed a letter from over 300 economist experts and social scientists organized by EPI arguing that there's no empirical basis for the COLA change, says that "Chained-CPI" is code for "let's really impoverish some women in their 90s!" This will fall on those who live the longest and rely on Social Security the most. But can we find a way to have this impact the poor less so that it doesn't fall too hard on those with the least?
The White House is saying that there will be such a set of protections, and think tanks have proposed some, but we won't know what they'll entail until they are better reported. No matter what additional measures are proposed, it's important to understand how compressed the distribution of income is for those receiving Social Security. From the Social Security Administration, here's a chart on the importance of Social Security relative to total income by income quintile for beneficiary families over 65 years of age (Table 9.B6):
I hate using charts that have so many percents of a percent of a percent, but this data is really important. To get a sense of what this chart is telling us, let's look at a box. From this chart, in the botom 20 percent of income, or those that make $11,417 or less, 65 percent of beneficiaries families get 90 percent of their income from Social Security. So the poorest are very dependent on Social Security, and a large cut will impact them harshly.
But let's say we wave a policy wand and protect those in the bottom 20 percent. The problem is that the income here is very compressed, and that Social Security is a major source of income up the ladder. Even for those in the 60-80 percent of income bracket, 41 percent of their income comes from Social Security. The group around the middle, in the third quintile, have only around $20,000 a year to live on and get a majority of their income from Social Security.
This is not a program that just helps the destitute; it provides a broad level of income security in old age for the majority of retirees. The average elderly family receiving Social Security gets 58.2 percent of their income from the program. A quarter of families get 90 percent or more of their income from Social Security. Once you leave the top income quintile, Social Security is the major source of retirement security. It is hard to see how means-testing these across-the-board cuts will be sufficient to prevent this from having a serious impact on our most vulnerable.