Many investors worry that the consumer-price index is being manipulated
Is the government lying to us about how high inflation really is?
Many financial advisers report hearing this question from their clients. And the concern is understandable, since the consumer-price index affects everything from Social Security's annual cost-of-living adjustment to the interest rate earned by investors in Treasury Inflation-Protected Securities.
There can be little doubt that politicians have powerful incentives to manipulate inflation data. The pundits tell us inflation concerns were one of the biggest reasons that Democrats lost the 2024 presidential election - and that Republicans are facing poor prospects in this November's midterms.
But manipulating the consumer-price index would not be as easy as you might think, according to Alberto Cavallo, a professor of business administration at Harvard Business School who used to serve on the technical advisory committee at the Bureau of Labor Statistics, the government department that produces the CPI data. I emailed Cavallo last August, in the wake of President Donald Trump's firing of the head of the BLS, when concern about CPI manipulation was running at a fever pitch.
Cavallo's response was that "when it comes to inflation data, any significant tampering would likely be easy to detect by outside researchers." To illustrate this point, Cavallo referred to Argentina's attempt to manipulate its inflation data in 2007. "I was able to identify [the manipulation] ... even as a Ph.D. student with limited resources," he said. "Today it is much easier, thanks to companies like PriceStats, which has been tracking inflation daily in 25 countries since 2011, with near-real-time releases." And as you can see from the chart above, which is from the PriceStats website, inflation as calculated by PriceStats (now part of State Street) is very closely correlated with the as-reported CPI from the BLS.
Possible past manipulation
What this comparison shows is that the BLS is calculating CPI in the way that it says it does. But what if the methodology the BLS uses is itself biased to underreport inflation? The agency has made over two dozen changes to its methodology over the last 50 years, as listed in this table.
Some commentators have claimed that, had those changes not been made, current inflation would be as much as 6 to 8 percentage points higher than currently reported. But they are wrong. The BLS has calculated how much the CPI would have grown over the last five decades had its current methodology been used at every point along the way over those five decades. The net effect is quite modest, as you can see from the chart above, which plots the as-reported CPI with what the CPI would have been using the methodology from 50 years ago: a reduction of just 0.2 of an annualized percentage point over the last five decades. Even if we assume that the BLS made the methodological changes for nefarious reasons, the annualized difference is not big enough to have a meaningful impact in any given year, or on the outcome of any political contest.
Is the Social Security COLA being manipulated?
Retirees have a related concern about how Social Security's annual cost-of-living adjustment is calculated. They worry that the headline CPI number doesn't reflect the unique cost-of-living increases faced by those in retirement. Once again, however, I think these concerns are overblown.
Consider the above chart, which plots three different ways of calculating the CPI:
-- The headline CPI, known technically as CPI-U (the Consumer Price Index for All Urban Consumers)
-- The CPI series that Social Security currently uses to calculate its COLA, known technically as CPI-W (Consumer Price Index for Urban Wage Earners and Clerical Workers)
-- A CPI series that focuses on the particular spending habits of those in retirement, known technically as CPI-E (Consumer Price Index for the Elderly)
As you can see, the three series are very closely correlated. Cumulatively over the last four decades, the CPI-E has appreciated at only a slightly faster pace than the CPI-W - by just 0.21% annualized. To put that in context, that difference corresponds to an increase of just $4.37 in the current average monthly Social Security check.
The bottom line? Concern about possible CPI manipulation is overblown. Investors would do better to focus on how fast the CPI will rise in coming months than worrying whether the index accurately reflects what it says it does.
Mark Hulbert is a regular contributor to MarketWatch. His Hulbert Ratings tracks investment newsletters that pay a flat fee to be audited. He can be reached at mark@hulbertratings.com.
-Mark Hulbert
Comments