CPI-E More Generous, But Less Accurate, Congressional Research Service Says

The Congressional Research Service issued a report on the hypothetical use of the CPI-E, or the consumer price index for the elderly, to calculate annual inflation adjustments for Social Security benefits. In summary, the researchers found that using the measure for cost-of-living decisions would create more generous Social Security benefits, but also risk being less accurate.

The Bureau of Labor Statistics tracks several measures of inflation. The price index used to calculate the cost-of-living adjustment for Social Security currently is the CPI-W, or the consumer price index for urban wage earners and clerical workers.

The CPI-E is another measure of inflation that focuses on the spending habits of the elderly and uses a different weighting system based on how those aged 62 or older tend to spend their money, since they have different spending habits. The CPI-E is higher most years than the CPI-W because it is weighted more heavily toward health expenses, and inflation in health care has outpaced most other industries.

According to the CRS report, the CPI-W adjustment from 1985 to 2024 was 188%, but it would have been 211% had the CPI-E been used instead. The report also notes that the COLA for 2024 was 3.2%, but would have been 4% under the CPI-E. This would have increased the average monthly benefit from $1,907 to $1,922, though researchers note that, depending on other factors, the CPI-E would not always outperform its CPI-W counterpart.

Some advocate that this measure should be used to calculate the COLA for Social Security, and there is pending legislation to do just that. The Social Security Expansion Act, introduced in February 2023 by Senators Bernie Sanders, I-Vermont, and Elizabeth Warren, D-Massachusetts, would require the use of CPI-E for Social Security COLA.

Joel Eskovitz, the senior director of Social Security and savings at AARP, says that “AARP supports using a CPI-E as a more accurate measure to calculate the Social Security COLA.”

However, there are some methodological issues with its use and the CPI-E currently has no official usage, despite being tracked by BLS since 1982, according to researchers. Those include:

• The CPI-E is a measure of weighted inflation for those aged 62 and older and not those collecting Social Security per se;
• the weighting system for the CPI-E is also based on the Consumer Expenditure Survey, which does not specifically sample those 62 and older, resulting in a small sample size and higher sampling error; and
• the CPI-E also does not account for the geographic distribution of those 62 and older, which is non-random, and it does not account for senior discounts.

Eskovitz explains that “the CPI-E is only experimental, and is subject to a higher sampling error because the BLS uses a much smaller subset of data to create this index than it does for CPI-W.”

He adds: “We support a COLA that more accurately reflects the spending patterns of older Americans, and for the BLS to better develop a measurement that meets that standard.”

Eskovitz cautions that “the CPI-E is not always a more generous measurement than the existing CPI-W. In some years, it underperforms the CPI-W.” However, “the true impact of the COLA is not felt on a year-to-year basis, but instead is felt over time by the compounding effect, and there it is clear that older Americans would certainly benefit in the long run from a switch to a CPI-E because of its increased accuracy.”