A Look at Inflation Measures and Social Security COLAs

Inflation has risen to a position of prominence among economic concerns and measures. And it affects a variety of ways prices are measured — measures that, in turn, have relevance for Social Security. Recent analyses examine how it affects those measures and the almost century-old safety net. 

Mark J. Warshawsky, a Searle Fellow and Senior Fellow at the American Enterprise Institute, in a two-part series discusses measures of prices and their relevance to Social Security in light of the high inflation rates of the last year. 

Measuring Inflation

One such measure is the CPI-U: the consumer price index as it applies to urban consumers. This measure is especially relevant, according to Warshawsky, since it represents prices affecting 88% of the U.S. population. He adds that, not only is it considered in public discussions and in actions concerning monetary and fiscal policy, but the federal government also uses the CPI-U to pay interest on Treasury Inflation Protected Securities, and to adjust portions of the federal tax code that are relevant to retirement. 

There are, however, other measures of prices that the Bureau of Labor Statistics (BLS) uses. For instance, the CPI-W — the “W” refers to urban wage and clerical workers — which Warshawsky notes is the original measure of consumer prices the BLS devised, having done so in 1913. This is a significant measure, he suggests, because:

• this cohort of employees comprises 28% of the population;
• the federal government uses it to make annual cost of living adjustments (COLAs) to Social Security and federal employee pension benefits; and 
• the CPI-U and CPI-W generally track closely. 

Warshawsky cites another measure the BLS devised, the C-CPI-U — in which the “C” means “chained” — which he writes that many experts consider to be the most accurate and current price measures used in calculating the inflation rate and that it is the best measure for use in determining COLAs. This, he says, is because the C-CPI-U holds consumer welfare constant and recognizes that consumers substitute some goods and services for others as their prices change. The federal government also uses the C-CPI-U to adjust parameters in the income tax code, he adds. 

There is yet another measure Warshawsky identifies — the R-CPI-E, for which the “E” stands for “elderly” and the “R” for research. The BLS has calculated the R-CPI-E since 1982, he says, to measure prices based on the broad purchasing patterns of those age 62 and older. This measure reflects spending on housing and medical care. There are problems, however, in relying on this measure, Warshawsky indicates. 

He notes that the R-CPI-E has reflected increases in medical care costs, but that now those increases have “cooled”; in addition, he says, inflation in housing prices mainly reflect changes in an owner’s equivalent rental, which he describes as “conceptual” and does not reflect actual transactions. And the BLS itself, Warshawsky says, has indicated that it recognizes that the R-CPI-E is a flawed way to measure inflation. 

Trends

In broad terms, says Warshawsky, the measures from January 2001 to June 2022 were largely in concert. “The mean inflation rates based on the CPI-U and CPI-W are nearly identical,” he writes, and the C-CPI-U has a high correlation with them as well; in addition, the R-CPI-E rate is a mere 0.06 percentage points higher than those rates.  

 

 

 

 

 

Despite the overall similarities, says Warshawsky, there are some differences. 

• The CPI-W measure somewhat exaggerates the highs and lows of inflation. 
• The inflation measured from the C-CPI-U is a bit below the other measures.
• There is no discernable pattern in either direction from the R-CPI-E.

Why this Matters

These measures, and their use in setting Social Security COLAs, matter in light of the important role Social Security plays in augmenting retirement income and helping to provide financial security in retirement. 

Warshawsky argues that the CPI-W “somewhat exaggerates” changes in inflation, and that such volatility makes it “undesirable” for use in setting public policy, since it “causes needless anxiety and concern about inflation measurement for benefit purposes.”

Warshawsky contends that the C-CPI-U is not only the most accurate, but also the least volatile, of the inflation measures that can be used to set COLAs for Social Security while also yielding savings that could be part of Social Security reform. 

However, he adds, if saving money is not a priority in setting COLAs, switching from the CPI-W to the CPI-U to reduce volatility and increase familiarity and transparency could be “good policy.” And Warshawsky argues against using the R-CPI-E.