Every month the government releases a number called CPI – the Consumer Price Index. The media reports it as the inflation rate. Politicians cite it when defending economic policy. Financial advisors use it to calculate your “real” investment returns.
The problem is that CPI does not measure what most people think it measures. It has been systematically changed over the past 50 years in ways that consistently produce a lower number than the actual cost of living increase most Americans experience. And the government has direct financial incentives to keep that number low.
This is not a fringe theory. The methodology changes are publicly documented. The incentive structure is openly acknowledged. What is rarely discussed is what it means for your financial life – and why it matters so much more than most people realize.
What CPI Is Supposed to Measure
The Consumer Price Index was originally designed to measure the cost of maintaining a constant standard of living. The Bureau of Labor Statistics tracks the prices of a basket of goods and services that a typical household buys – food, housing, transportation, medical care, clothing, education – and measures how those prices change over time.
If that basket costs $1,000 in January and $1,030 in December, the CPI shows 3 percent inflation for the year. Simple concept. The execution is where it gets complicated – and where the government’s thumb on the scale becomes visible.
How the Methodology Has Been Changed
The CPI calculation methodology has been significantly altered multiple times since the 1980s. Each change, presented as a technical improvement, had the effect of lowering the reported inflation rate. Here are the main ones:
Substitution
The original CPI measured a fixed basket of goods. If steak got expensive, the index recorded that steak got expensive. In the 1990s, the BLS introduced substitution – the idea that if steak gets too expensive, consumers will substitute chicken. So instead of recording rising steak prices, the index partially replaces steak with chicken in the basket.
The problem: this measures the cost of a declining standard of living, not a constant one. If you can no longer afford what you used to buy and have to substitute cheaper alternatives, your cost of living has gone up – even if the substituted basket costs the same. The index masks the quality-of-life decline as price stability.
Hedonics
Hedonic adjustments account for quality improvements in products. If a new laptop costs the same as last year’s model but has twice the processing speed, the BLS may record the price as having fallen – because you are getting more for the same money.
In theory this makes sense. In practice it is applied heavily to technology and inconsistently across categories. A car today is more expensive in nominal terms than a car in 2000, but the BLS adjusts for safety features, fuel efficiency, and entertainment systems. The adjusted price often shows as flat or falling even as the sticker price climbs. Meanwhile, housing and food – which affect your actual budget far more than processing speed – receive less generous treatment.
Owners’ Equivalent Rent
This is the most significant methodological quirk in the CPI. Housing is the largest expense for most Americans, so how you measure it matters enormously. The BLS does not use actual home prices to measure housing costs. Instead it uses a concept called Owners’ Equivalent Rent – what a homeowner estimates they would receive if they rented their home.
This is a surveyed, estimated number. It is not the actual cost of buying a home. During the 2020 to 2022 period when home prices rose 40 percent in many markets, Owners’ Equivalent Rent lagged significantly behind. The CPI showed modest housing inflation while anyone trying to buy a home was experiencing catastrophic price increases. The index was not lying about rents – it was simply not measuring what most people experience as housing costs.
The Government’s Incentive to Keep the Number Low
Here is where it gets important to understand the incentive structure – because the government does not just report CPI, it uses CPI to determine payments and obligations that cost real money.
Social Security cost-of-living adjustments
Social Security benefits are adjusted annually based on CPI. If CPI shows 3 percent inflation, benefits increase by 3 percent. If actual inflation is 6 percent, the government is paying retirees half of what they need to maintain their purchasing power – and saving the difference. With tens of millions of Social Security recipients, every percentage point of understated inflation saves the government hundreds of billions of dollars over time.
Treasury Inflation-Protected Securities
TIPS are government bonds that adjust their principal with inflation – specifically with CPI. If CPI understates inflation, TIPS holders receive lower adjustments than they would under accurate inflation measurement. The government issues the bonds and controls the index the bonds are tied to. The conflict of interest is structural.
Federal employee and military pensions
Many federal pensions are also indexed to CPI. Lower reported inflation means lower cost-of-living adjustments and lower payments out of the government’s budget.
Tax bracket adjustments
Income tax brackets are adjusted for inflation based on CPI. If CPI understates inflation, brackets adjust upward more slowly than real prices rise. Workers whose wages keep up with real inflation get pushed into higher tax brackets even though their real purchasing power has not increased. This is called bracket creep and it generates additional tax revenue without Congress ever voting for a tax increase.
Add it up. Every percentage point of understated CPI reduces Social Security payments, reduces TIPS payouts, slows pension increases, and pushes workers into higher tax brackets. The financial benefit to the government of a lower CPI number is enormous – and ongoing, year after year, compounding silently.
What You Actually Experience vs What the Government Reports
Let me make this concrete. Think about what your dollar bought five years ago versus today.
Groceries. In 2020, the average American family spent roughly $400 to $500 per month on food at home. By 2024, that same basket costs $550 to $700 in most markets – an increase of 30 to 40 percent. The CPI food component showed significant increases during this period, but the overall CPI headline number was moderated by the categories with slower price growth or hedonic adjustments.
Housing. Home prices in the US rose approximately 40 percent from 2020 to 2022. Mortgage payments on a median-priced home more than doubled as prices rose and interest rates climbed from 3 percent to over 7 percent. The CPI housing component, anchored by Owners’ Equivalent Rent surveys, showed increases – but nothing close to what a buyer or renter in a hot market actually experienced.
Insurance. Auto insurance premiums rose 20 to 30 percent in many states between 2022 and 2024. Health insurance costs have increased at multiples of reported CPI for over a decade. These categories are in the CPI basket but they do not always receive the weight that their actual share of household budgets would suggest.
The economist John Williams at Shadow Government Statistics has calculated what CPI would look like using the pre-1980 and pre-1990 methodologies. His estimates consistently show inflation running 3 to 7 percentage points higher than the official BLS figure. You do not have to accept his specific numbers to recognize that the direction of the bias in the methodology changes is consistent – they all go one way, and that way is lower.
What This Means for Your Financial Life
If official inflation is understated, everything calculated from it is wrong in the same direction.
Your “real” investment return is lower than reported. If your index fund returned 8 percent and official CPI shows 4 percent inflation, your real return is quoted as 4 percent. If actual inflation is 7 percent, your real return was 1 percent. The same nominal gain. Very different actual purchasing power improvement.
Your Social Security cost-of-living adjustments are inadequate. If you are relying on Social Security to maintain purchasing power in retirement and CPI understates real inflation by even 2 percent annually, your purchasing power erodes by 2 percent per year compounded over a 20-year retirement. That is nearly 33 percent of purchasing power lost quietly, without any headlines.
Bonds are even worse than they appear. I wrote a full article on why I do not hold government bonds. The inflation math already makes them unattractive. If the inflation number used in that math is itself understated, the real return on bonds is even more negative than the official figures suggest.
And the case for hard assets gets stronger. Bitcoin’s fixed supply and gold’s scarcity are not just hedges against reported inflation. They are hedges against the full measure of monetary debasement, whether or not that debasement shows up accurately in official statistics.
This Is Not Conspiracy – It Is Documented
I want to be precise here because this topic attracts both legitimate analysis and genuine conspiracy thinking. The methodology changes to CPI are not hidden. The BLS publishes them. Economists debate them openly. The criticism of CPI does not require believing in secret cabals or coordinated deception.
What it does require is recognizing that institutions respond to incentives. The government has financial incentives to report lower inflation. The methodology changes made over decades each produced lower numbers. These are facts, not theories. Whether those changes were made in response to the incentives, as genuine technical improvements, or some combination of both is a question reasonable people disagree on.
What is not debatable is the direction of the effect. Every major methodology change since 1980 has produced a lower CPI reading than the previous methodology would have. If the changes were random technical adjustments with no systematic bias, you would expect roughly half to produce higher numbers and half to produce lower numbers. That is not what happened.
What to Do With This Information
Understanding CPI’s limitations does not require you to do anything dramatic. It does require adjusting how you think about a few things.
Do not use official inflation figures as the baseline for “good enough” investment returns. If your savings account earns 4 percent and official CPI shows 3 percent, you are not necessarily ahead. Aim for real assets and real returns, not returns that barely exceed a number that may itself be understated.
Reconsider the role of bonds in your portfolio with this in mind. The bond case already does not hold up well against honest inflation analysis. It holds up even worse if official inflation is itself understated.
Think seriously about hard assets. Bitcoin’s fixed supply does not care what the BLS reports. Gold’s scarcity does not adjust for hedonic improvements in government spreadsheets. These assets protect purchasing power against the full measure of monetary expansion, not just the officially reported portion.
And pay attention to your own experience. Your grocery bill, your rent or mortgage payment, your insurance premiums, your healthcare costs – these are real numbers. If they are rising faster than what the official inflation rate would suggest, trust your lived experience. That gap is the part of the story the CPI number is not telling.
Frequently Asked Questions
What is CPI and how is it calculated?
CPI – the Consumer Price Index – is the US government’s primary measure of inflation. The Bureau of Labor Statistics tracks the prices of a basket of goods and services including food, housing, transportation, medical care, and clothing. The percentage change in the cost of that basket from one period to the next is reported as the inflation rate. The specific methodology for what goes in the basket and how items are weighted has been changed significantly since the 1980s.
What is hedonic adjustment in CPI?
Hedonic adjustment is a methodology that accounts for quality improvements in products. If a new laptop costs the same as last year’s model but has twice the speed, the BLS may record the effective price as having fallen because you get more for your money. In theory this is reasonable. In practice it is applied heavily to technology and produces lower reported inflation by crediting quality improvements against price increases – even when consumers do not experience the quality gain as meaningful to their daily budget.
Why does the government have an incentive to understate inflation?
The government uses CPI to determine Social Security cost-of-living adjustments, TIPS bond payouts, federal pension increases, and income tax bracket adjustments. Lower reported inflation means lower Social Security payments, lower TIPS payouts, lower pension increases, and slower bracket adjustments – all of which save or generate money for the federal government. With tens of millions of beneficiaries, every percentage point of understated inflation represents hundreds of billions of dollars over time.
What is Owners’ Equivalent Rent?
Owners’ Equivalent Rent is the BLS’s method of measuring housing costs for homeowners. Rather than using actual home prices, it uses a survey asking homeowners what they estimate they would charge to rent their home. This is a lagging, estimated number that consistently understates rapid home price appreciation. During 2020 to 2022, when home prices rose 40 percent in many markets, Owners’ Equivalent Rent showed much more modest increases – masking one of the most significant cost-of-living increases most Americans faced.
How does understated inflation affect my investments?
Every investment return quoted in “real” terms uses official CPI as the inflation benchmark. If CPI understates actual inflation by 2 percent, your real returns are 2 percent lower than reported. A bond returning 4 percent with official inflation at 3 percent looks like a 1 percent real gain – but with actual inflation at 5 percent it is a 1 percent real loss. It also makes the case for hard assets like Bitcoin and gold stronger – they protect against the full measure of monetary debasement regardless of what the official statistics report.