Inflation Doesn't Have an Undo Button

I've noticed something interesting in conversations about the economy. People across income levels often talk about inflation as if it resets every year. When they hear that inflation has fallen, they assume prices should be falling, too.

It's an understandable mistake. The distinction between the rate at which prices are increasing and the prices themselves isn't always intuitive. But understanding it changes how we interpret economic headlines, evaluate our income, and make financial decisions.

Let's start with the math, then look at how inflation is measured and why the official number may not reflect what you experience at the checkout counter.

Part 1: Inflation compounds. It doesn't reset.

Imagine an item that costs $10 today.

• Year 1: Inflation is 4%. The item rises to $10.40.
• Year 2: Inflation slows to 3%. The item rises another 3%, from $10.40 to approximately $10.71.
• Year 3: Inflation slows again to 2%. The item rises to approximately $10.93.

Notice what happened. Inflation fell from 4% to 3%, then to 2%. Yet the item's price never fell.

Why? Because each year's inflation rate is applied to the current price, not the original $10.

The item's price increased by roughly 9.3% over three years, even though the annual inflation rate declined each year.

For the price to return to $10, we would need deflation, meaning a sustained decline in prices, not merely a decline in the inflation rate.

This is the first distinction worth remembering:

Disinflation means prices are rising more slowly. Deflation means prices are falling. They are not the same thing.

And this is where the conversation gets more interesting. The Consumer Price Index, or CPI, is designed to measure changes in consumer prices over time. But how does the government actually calculate it?

Part 2: How the government measures inflation in four steps

The U.S. Bureau of Labor Statistics (BLS) calculates the CPI by tracking prices for a representative sample of goods and services purchased by urban consumers. It isn't simply checking the price of a few groceries and calling it a day.

Here's the simplified process.

Step 1: Build the basket

The BLS uses Consumer Expenditure Surveys to understand what households buy and how much they spend across categories such as housing, food, transportation, medical care, and recreation.

This information helps determine the relative importance of different expenses in the CPI. The government doesn't assume that a dollar spent on airline tickets should count as much as a dollar spent on housing across the entire economy (this is important or as the Wayans would say “message"). Categories are weighted according to consumer spending patterns.

Step 2: Collect the prices

The BLS collects thousands of prices each month from selected retail outlets, service providers, and other sources across the country.

The samples cover a broad range of consumer goods and services. The goal is to compare like with like, tracking comparable products and services over time rather than simply comparing two unrelated shopping trips.

For example, the BLS has a specific methodology for airline fares that accounts for factors such as departure city, destination, advance booking, and travel day.

Step 3: Calculate the cost of the basket

Imagine a simplified basket containing rent, groceries, gasoline, and a streaming subscription.

The BLS measures how the prices of those items change and combines the changes using their respective weights. In a basic fixed-quantity illustration, the quantities remain the same so that the calculation isolates price changes rather than changes in how much a household buys.

The actual CPI methodology is more sophisticated: it uses different formulas at different stages, and the expenditure weights are periodically updated. But the underlying idea is to measure changes in the cost of comparable consumption over time.

Step 4: Calculate the index

The BLS combines the weighted price changes into an index. The percentage change in that index over a specified period becomes the reported inflation rate.

For example, if the CPI rises from 300 to 309, the increase is 3%.

That means the measured price level for the index is 3% higher than it was in the comparison period. It does not mean every product increased by 3%, nor does it mean prices are now returning to where they were before previous increases.

The CPI is an average, and averages can conceal some pretty significant differences.

Part 3: The CPI is useful, but it isn't your personal cost-of-living index

The CPI is an important economic tool. It informs policymakers, helps adjust certain benefits and contracts, and gives households a common reference point for understanding price changes.

But it is also a statistical model of consumer spending. That means its results depend on the basket, the weights, the assumptions, and the methods used to compare products over time.

Three areas deserve particular attention.

1. The weighting problem: Your budget isn't the average budget

Not every category contributes equally to the CPI.

Housing carries considerably more weight than many individual consumer categories. An increase in a heavily weighted category can have a much greater effect on the overall index than a substantial increase in a category that represents a small share of consumer spending.

Consider a hypothetical example.

Suppose airline fares rise 23%, while airline fares account for approximately 0.9% of the CPI basket. That category's direct contribution to the overall index would be roughly 0.2 percentage points, all else equal.

The fare increase is significant if you fly frequently. But it has a relatively small impact on the national CPI because air travel represents a comparatively small share of the typical consumer's total expenditures.

Now imagine that your household spends a large portion of its income on airfare. Your personal experience could be very different from the national average.

The same principle applies to rent, childcare, insurance, groceries, and medical expenses. If the categories consuming most of your income rise faster than the overall CPI, your personal cost of living can increase faster than the headline inflation rate.

The national average can be accurate and still fail to describe your financial reality.

2. The substitution effect: What if you can't switch?

The BLS uses a geometric mean formula for many categories of goods and services. In simplified terms, this method allows for the possibility that consumers will adjust what they buy when relative prices change.

For instance, if one brand of cereal becomes more expensive while a similar brand remains affordable, consumers may switch brands. The formula is designed to reflect that possibility rather than assume everyone continues buying exactly the same product in exactly the same proportions.

That is a reasonable assumption in many circumstances. But the ability to substitute isn't universal.

Imagine that the cost of your commute rises sharply. In theory, you might switch from rideshare to public transportation, or from public transportation to driving.

But what if you don't own a car? What if the nearest train station is miles away? What if your work schedule doesn't align with bus service?

The cheaper alternative may exist on paper without being a realistic option for you.

There is an important technical distinction here: the geometric mean formula is not a blanket mechanism for removing every price increase when consumers change their behavior. It is used within many individual CPI categories, while other categories use different formulas. Nor does it simply assume everyone can substitute rideshare for a bus.

The broader issue remains: a statistical allowance for substitution cannot guarantee that every household has a practical, affordable alternative.

For someone with limited transportation options, the increase in the cost of getting to work may be unavoidable, even if the index captures the category's average price change appropriately.

3. Hedonic quality adjustments: Is a more expensive product necessarily more inflationary?

This is one of the more technical, and often misunderstood, parts of the CPI.

Products change over time. A new computer might have more memory, a faster processor, or a better display than its predecessor. A replacement product may cost more, but some of that difference reflects an improvement in quality rather than a higher price for an identical product.

To account for this, the BLS uses quality-adjustment methods, including hedonic regression, in certain categories. These methods estimate the value associated with product characteristics so that the index can distinguish changes in quality from changes in the price of comparable quality.

Consider a simplified example.

A television costs $500 one year. A comparable new model costs $600 the next year, but it includes a feature that the statistical model values at $100.

If the entire $100 difference is attributed to improved quality, the quality-adjusted price is unchanged. The measured price increase for comparable quality could therefore be 0%, even though the consumer must pay $100 more to buy the new model.

That is the basic logic behind quality adjustment.

Now consider the consumer's perspective. Perhaps the new feature is something you don't need. Perhaps you would happily buy the older model, but it is no longer available. Or perhaps a streaming service raises its monthly fee and advertises additional features you never requested.

Your bill still rises. You still need to find the money.

There is a nuance worth emphasizing: hedonic adjustments are not automatically applied to every price increase, and a provider's marketing claims do not, by themselves, determine the BLS adjustment. The statistical methods are intended to estimate the value of actual changes in product characteristics.

Still, the underlying question is fair: How much of an upgrade's estimated value reflects something consumers genuinely value, and how much reflects a feature they never wanted in the first place?

Quality adjustment is an attempt to make prices comparable over time. It isn't a direct measurement of how affordable a product feels to every consumer.

Part 4: What should we take away from all of this?

None of these limitations means the CPI is useless or that the government is necessarily manipulating inflation figures. Although there can be a serious case to be made to contradict this. Measuring millions of different consumer transactions across a changing economy is an inherently difficult task.

But it is equally misguided to treat one national inflation number as a complete account of every household's financial situation.

Three things can be true at the same time:

1. Inflation can decline while the overall price level remains elevated.
2. The CPI can accurately measure average price changes while individual households experience very different increases.
3. A statistical adjustment can be reasonable in economic terms while still failing to capture how a particular expense affects someone's budget.

The practical lesson is to pay attention to both the macroeconomic data and your own numbers. If I took a stance: pay more attention to your numbers, the macro is noise.

Track what you actually spend on housing, food, insurance, transportation, healthcare, and other recurring expenses. Compare those costs over time. Look at whether your income is keeping pace with your own spending increases, rather than assuming that a raise matching the headline inflation rate necessarily leaves you in the same financial position.

And when you hear that inflation has fallen, ask the follow-up question: Has the rate of price increases slowed, or have prices actually declined?

Those are two very different economic outcomes.

For a useful starting point, try the official BLS CPI Inflation Calculator. Enter an amount and compare different years to see how the purchasing power of the dollar has changed over time.

The goal isn't to distrust every economic statistic. It's to understand what a statistic measures, what it doesn't measure, and how to use it intelligently.

Financial literacy starts with asking better questions. Inflation is a good place to start.

Sources and further reading

1. U.S. Bureau of Labor Statistics, CPI Handbook: Calculation. Explains the formulas used to calculate basic and aggregate consumer price indexes, including expenditure weights and geometric mean formulas.
2. https://www.bls.gov/opub/hom/cpi/calculation.htm
3. U.S. Bureau of Labor Statistics, Consumer Expenditure Surveys. Explains the household expenditure data used to inform CPI weights.
4. https://www.bls.gov/cex/
5. U.S. Bureau of Labor Statistics, Frequently Asked Questions about Hedonic Quality Adjustment in the CPI. Explains why and how the BLS adjusts for product quality changes.
6. https://www.bls.gov/cpi/quality-adjustment/questions-and-answers.htm
7. U.S. Bureau of Labor Statistics, Measuring Price Change in the CPI: Airline Fares. Details how airline fares are sampled and measured, including the category's relative importance in the index.
8. https://www.bls.gov/cpi/factsheets/airline-fares.htm
9. U.S. Bureau of Labor Statistics, CPI Inflation Calculator. A tool for comparing the purchasing power of the dollar across years.
10. https://www.bls.gov/data/inflation_calculator.htm