The lack of an inflation adjustment in retirement benefits in programs including federal annuities and Social Security for 2016 has renewed calls to use a special index geared toward expenses of the elderly in particular, but a recent study shows that it might not make a substantial difference.

The report by the Center for Retirement Research focused on what is called the CPI-E, which is an alternative measure to the CPI-W used in adjusting retirement and certain other programs for inflation. “For a long time, critics have contended that the CPI-W understates inflation for the elderly because it does not reflect their spending patterns,” especially for health care where inflation typically runs especially high among the goods and services measured, the report notes.

The Labor Department has calculated a CPI-E since 1987, retroactive to 1982, which for example weights health care as 10.8 percent of the total expenses measured compared with 5.2 percent in the CPI-W. However, that index is not formally used to adjust any benefits but rather serves as a model.

The report said that over 1983-2002, the CPI-E outstripped the CPI-W by an average of 0.4 percentage points per year, meaning adjustments of that much more if it had been used. However, the difference has fallen sharply in more recent years, at just a 0.07 difference on average over 2002-2015. For the entire period since 1983, that would work out to an average 0.25 percentage point difference, it said.

That was largely due to a slowdown in health care costs in recent years, although that cost did bump up in 2015 and is not reflected in the report.

While some retiree advocates press for use of the CPI-E even though it might make relatively little difference, others suggest that a better course is to leave things alone. Opening a debate over which inflation measure to use would, in their view, run the risk of a switch to the “chained” CPI, which others advocate as the most accurate measure since it adjusts for changes in purchasing patterns as the costs of certain goods and services change.

Most assessments of that measure show that it would produce annual adjustments about a quarter to half a percentage point lower than the current measure.