In light of the recent announcement of a 5.9% cost of living adjustment (COLA) for Social Security benefits in 2022, a reporter asked what the number would have been based on the oft-advocated CPI-E, the experimental index for the elderly.
Read: Seniors get big social security hike – but here’s why they’re falling behind
My trusted colleague, Patrick hubbard, came back with 4.8%. I sighed in disappointment, reminded him that he knew the CPI-E was rising faster than the CPI-W (the index used to adjust social security benefits), and sent it back. He returned a few minutes later with the same number and one more story. Patrick is generally right. Indeed, the COLA based on the experimental CPI-E would have been 4.8% compared to the actual COLA of 5.9% (see Figure 1).
First off, let me tell you why this happened. The underlying argument for a CPI-E is that older households spend more on medical care than their younger counterparts and that the cost of medical care is rising faster than other budget items.
Read: COLA’s biggest hike in 40 years comes to Social Security in 2022 – what it means for your retirement
The weighting argument is correct. In 2007, the first year for which the CPI-E weights are publicly available, older people spent more than twice as much on health care in relation to their total expenditure as the population as a whole. A notable difference persists today (see Figure 2).
What has changed is the rate of medical inflation. Health care costs are increasing at a much slower rate than in the past. And between the third quarter of 2020 and the third quarter of 2021, they hardly increased (see figure 3). Since this low-inflation component receives twice as much weight in the CPI for seniors as in the CPI-W, the CPI-E has increased more slowly.
To me, switching to CPI-E to determine COLA has never seemed like a productive proposition. First, the difference between the two indices has diminished over time. The annual average difference from 1983 to 2002 was 0.38%. But, from 2002 to 2020, it reduced to 0.12% (see Figure 4) and, as shown, the difference in 2021 was actually negative.
More importantly, the CPI-E is not a true price index. It simply reweighted the data collected for the entire population. As a result, it suffers from several defects.
Read: The complicated calculations of Social Security, spousal benefits and when to apply
First, a relatively small number of households are used to determine expenditure patterns. Second, the prices are based on the same geographies and retail outlets used by young people. Third, the items sampled may not be the same as those purchased by the elderly. Finally, the prices used are the same as those reported for the youngest and do not take into account senior discounts.
Thus, if the decision were made to use an index for the elderly, a new index would be needed with a larger sample of older households that uses the prices of the products they buy in the places where they shop.
In short, having fun with the COLA social security index is not worth it. We have bigger fish to whip.