The 2021 Trustees Report projected a modest 3.6 percent cost-of-living adjustment (COLA) for 2022.  As a result of pandemic-fueled inflation, the actual COLA in 2022 was 5.9 percent.  While this was the highest COLA since 1982, the National Committee believes that this estimate does not reflect the effect of inflation on today’s seniors and believes that Social Security’s COLA needs to be strengthened.

The current COLA formula has not led to an increase over 2 percent for seven out of the last eight years.  The average COLA for the past 10 years was only 1.65 percent, and there was no COLA at all for three of the past 12 years.  The 2015 COLA was zero; the 2016 COLA was 0.3 percent. Only the 2018 COLA reached 2.8 percent.  Concerns about the accuracy of the Consumer Price Index (CPI) used to calculate the COLA are reflected in a June 2020 General Accountability Office report to Congress:

“The U.S. Bureau of Labor Statistics (BLS) faces accuracy, timeliness, and relevancy challenges developing consumer price indexes (CPI) for subpopulations of blue-collar workers and older Americans.”

We urge the adoption of a consumer price index (CPI) for the elderly, or CPI-E, as a more accurate means of calculating Social Security COLAs.  An in-depth examination of the CPI-E follows.

History and Background

From the inception of the Social Security program in 1935, Congress periodically adjusted Social Security benefits so that beneficiaries’ purchasing power would keep up with inflation.  Beginning in 1972, Congress enacted a provision that requires benefits to be increased annually based on the amount the consumer price index has increased in the previous year.  At the time this provision was enacted, there was only one CPI index available for use, the CPI-W, which reflects price increases for urban wage earners and clerical workers, based on a fixed market basket of goods and services.  This index continues in use as the basis for the computation of Social Security COLAs.  Today it is derived from the Consumer Price Index for All Urban Consumers (CPI-U), which is also used to derive the CPI-E.

The CPI-W, which reflects the expenditures of wage and clerical households in urban areas, measures a population that is employed, unlike most retired Social Security beneficiaries, and displays patterns of consumer spending which are reflective of people who work.  Concerns about whether the CPI-W accurately reflects the spending patterns of those who are retired or disabled date back to the 1980s.  In 1987, as part of amendments to the Older Americans Act of 1965, Congress directed the Bureau of Labor Statistics (BLS) to develop an index focused on the elderly.   BLS then developed the Experimental CPI for Americans 62 Years of Age and Older (CPI-E), and calculated estimates of the index dating back to December 1982.

Seniors’ Spending Patterns

To determine consumers’ day-to-day living expenses and thus measure the rate of inflation consumers face in the marketplace, the government collects data on a “market basket” of goods and services.  Market basket categories include: food and beverage, housing, apparel, recreation, education, transportation, and medical care.  Research has shown that spending patterns differ between the elderly and the general population, especially in the health care category.  Seniors 65 and older spend more than twice as much on health care, and those 75 and older spend nearly three times more on health care than younger consumers.  Not only do health care expenditures steadily increase with age but health care costs have also consistently risen much faster than other market basket categories.  The current price index (CPI-W) does not take these critical differences in the elderly population into consideration.

Additionally, the June 2020 GAO report to Congress questions the reliability of the CPI-W itself because of demographic shifts away from the blue-collar occupations surveyed, such as clerical, sales, laborer and construction jobs, as well as declining response rates to surveys.

Seniors spend a significant portion of their income on out-of-pocket health care expenses not covered by Medicare. As time goes by, more and more of their Social Security benefit checks will be eaten up by rising health care costs.  According to the Medicare Trustees, 40 percent of the average senior’s Social Security check will be consumed by Medicare out-of-pocket costs by 2095, compared with 23 percent in 2021.

How the CPI-E Compares to Other Indices

The CPI-E uses the same formulas and prices as the CPI-W, but their importance is determined, or weighted, differently.  The CPI-E uses expenditure weights for households with individuals age 62 or older.  This sample size is about one-third the size used for the CPI-U, so is subject to a greater sampling error. This is one reason the CPI-E continues to be classified as “experimental”.

From 1982 to 2014, the average annual increase for the experimental CPI-E was 2.9 percent, compared to 2.7 percent for the CPI-W.  Using the CPI-E to determine the Social Security COLA would increase the expected average COLA by about 0.2 percentage points per year. The BLS acknowledges the current CPI does not “produce official estimates for the rate of inflation experienced by subgroups of the population, such as the elderly or the poor.”  An index for the elderly could make a difference for millions of Social Security beneficiaries.

According to the BLS, a move to an “official” CPI-E, would require additional research including where elderly households are located, where the elderly actually shop, and what mix of products the elderly purchase.  The June 2020 GAO report to Congress recommends BLS explore cost-efficient ways to evaluate the data currently used to produce subpopulation indexes.

The CPI-E has been under review for four decades.  It is time for the federal government to provide the resources for BLS to conduct this research, expand the sample size of individuals age 62 and older, conclude its analysis, and adopt a more accurate consumer price index for the elderly.


The National Committee believes accuracy must be the goal of changing the current consumer price index. The CPI-E represents the best alternative for correcting problems with the CPI-W for America’s seniors. The current CPI-W does not account for the unique spending patterns of the elderly and is itself flawed because of shifts in occupations and declining survey responses.  In order to implement the CPI-E, the BLS will need to explore cost-efficient ways to evaluate the data currently used to produce indexes, obtain sufficient funding to evaluate current data, conduct the research and expand the sampling needed to enable the index to move from experimental to official.  When the CPI-E becomes official it should, at last, finally represent the most accurate measure of the inflation affecting our nation’s seniors.


Government Relations and Policy, January 28, 2022