Measurement Failure

Claim

CPI is a useful inflation measure, but it is not a complete measure of the cost of living.

Signal

The Bureau of Labor Statistics says CPI is sometimes termed a “conditional cost-of-living index” because it does not attempt to quantify all factors that affect the cost of living, including some governmental, environmental, health, education, safety, water quality, and crime factors.

Why It Matters

A household can feel pressure that is real but only partly visible in CPI. The number may be accurate inside its method while still failing to capture the whole lived experience.

Caveat

This does not mean CPI is fraudulent. It means CPI has a defined scope. The Drift Library should use this card to clarify scope, not to imply bad faith.

Source Drawer

Claim

CPI measures the shelter service of owner-occupied housing, not the purchase price of the house as an asset.

Signal

The Bureau of Labor Statistics states that owned housing units themselves are not priced in the CPI Housing Survey. The CPI treats owned housing units as capital or investment goods distinct from the shelter service they provide.

Why It Matters

A household trying to buy a home is answering a different question from an index measuring the shelter service consumed by current occupants. The method can be sound inside CPI and still be the wrong measure of access to ownership.

Caveat

Rental equivalence is a real methodology, not a trick. This card should be used to explain the gap between consumption measurement and lived housing access.

Source Drawer

Claim

BLS adjusts CPI item replacements when quality differences could otherwise appear as price change, and the published method varies by item category.

Signal

When a sampled item leaves the market, BLS selects a replacement. If the replacement is less comparable, a quality difference can enter the observed price change. BLS publishes category-specific methods that include hedonic adjustments, cost-based adjustments, and no adjustment for many listed categories.

Why It Matters

Quality drift is hard to measure. If a product gets smaller, worse, less durable, more restricted, or more bundled with fees, the visible price may not tell the whole story.

Caveat

The method page establishes that BLS uses different procedures across item categories. It does not show that a particular adjustment is wrong, or that every lived change in durability, service, or restriction is missed. A category-specific claim needs the category-specific method.

Source Drawer

Claim

BLS researchers found that from 2006 to 2023, average annual R-CPI-I inflation was 0.28 percentage point higher for the lowest equivalized-income quintile than for the highest.

Signal

The research series reweights CPI-U elementary indexes using spending patterns for five income groups. Equivalized income divides before-tax household income by the square root of household size so that the ranking accounts for household size.

Why It Matters

One national average can be accurate for its target population while missing a persistent difference between groups that buy different mixes of goods and services.

Caveat

R-CPI-I is a research index, not the headline CPI. The result is an average over 2006–2023, not a claim about every year or household. BLS notes that the series reuses CPI-U areas, stores, rental units, items, and prices, may have more measurement error than official indexes, and can be revised.

Source Drawer

Claim

BEA publishes distributional statistics that divide U.S. personal income across household groups instead of treating the national aggregate as a complete household account.

Signal

BEA’s Income Data Analysis Tool can build tables and charts by income concept, year, and income quantile. The June 2026 release also added an income-stratified PCE price index showing inflation rates by income decile.

Why It Matters

The national total answers how much personal income exists. Distributional tables ask where it went. Those are related questions, but a clean aggregate cannot answer both by itself.

Caveat

Income quantiles are still groups, not individual household histories. BEA labeled its 2025 distribution figures experimental nowcasts and its 2024 figures provisional in the June 2026 release. The estimates combine several source types and can be revised.

Source Drawer

Claim

NIST’s state-by-state inventory lists unit-pricing provisions in some jurisdictions and none in others, so retail unit-pricing law is not uniform across the United States.

Signal

The NIST table links unit-pricing laws for some states while showing no listed provision for others. The legal baseline therefore depends on jurisdiction, and the details depend on the covered product, store, label, and exemption.

Why It Matters

When a unit price is missing or presented on a different basis, comparing package sizes takes more work. A shelf price can stay still while the price per ounce, count, or use changes.

Caveat

The NIST inventory is a reference table, not legal advice or a measure of what every store displays. Retailers may provide unit prices voluntarily, and local rules, product coverage, label quality, and enforcement can differ within a state.

Source Drawer

Claim

Using BLS CPI-U index values, the price level was 94.6% higher in May 2026 than the 2000 annual average.

Signal

The all-items CPI-U was 172.2 for the 2000 annual average and 335.123 in May 2026. Dividing 335.123 by 172.2 gives a 94.6% increase. The inverse calculation means one 2000 dollar had about 51 cents of CPI-measured purchasing power in May 2026.

Why It Matters

Nominal dollars keep the same name while the amount they can buy changes. That can make a wage, savings balance, or long-term target look stable even when its broad purchasing power has moved.

Caveat

This compares an annual average with one monthly observation. CPI-U is a national average for urban consumers, not a household-specific inflation rate, and the 51-cent figure is an index conversion rather than a claim about every purchase.

Source Drawer

Claim

BLS describes CPI as a conditional, not complete, cost-of-living measure and excludes factors such as changes in income taxes and social or environmental conditions from its defined scope.

Signal

CPI tracks price change for a representative basket of consumer goods and services. BLS includes sales and excise taxes tied to those purchases, but excludes taxes not tied to a purchase and investment items such as stocks, bonds, real estate, and life insurance.

Why It Matters

A measure can be rigorous and still be narrower than the question people ask it to answer. CPI is built to track consumer price change, not to total every pressure on a household’s standard of living.

Caveat

An exclusion is not evidence of manipulation. CPI remains an official price statistic with a defined purpose, and it includes many consumption costs and government user fees that looser critiques sometimes assume are missing.

Source Drawer

Claim

Revised BLS data show average hourly earnings rose 0.2% in May 2026 while real average hourly earnings fell 0.2% after CPI-U adjustment.

Signal

The revised table shows CPI-U up 0.5% over the month. From May 2025 to May 2026, real average hourly earnings fell 0.8%; real average weekly earnings fell 0.3% over the month.

Why It Matters

A larger number on a paycheck does not by itself establish greater purchasing power. Nominal pay and the price adjustment have to be read together.

Caveat

This is an average across jobs on private nonfarm payrolls, not a household income measure. It does not describe every worker, occupation, region, tax situation, benefit package, or work schedule.

Correction

The archived May release called its values preliminary. BLS later revised the monthly nominal change from 0.3% to 0.2%, the monthly real change from -0.1% to -0.2%, the annual real-hourly change from -0.7% to -0.8%, and the monthly real-weekly change from -0.2% to -0.3%. The exhibit ID is unchanged so the correction remains attached to the original record. On October 1, 2026, the citation was pinned to the August 12 release, whose Table A-1 retains these May observations; the rolling current-release link no longer shows May.

Source Drawer

Claim

BEA’s September 30, 2026 annual update put the May 2026 U.S. personal saving rate at 4.4% of disposable personal income.

Signal

The revised May rate is 4.4%. BEA defines it as personal saving divided by disposable personal income. It describes the share left after personal outlays, not the percentage of households that saved.

Why It Matters

The measure asks what remained nationally after income met outlays. It can show movement in aggregate margin, but not which households did the saving or which had nothing left.

Caveat

This is a dated estimate for May, subject to further revision. It cannot show distribution by income, age, region, household type, or balance sheet.

Correction

Updated October 1, 2026: BEA’s June 25 release originally estimated the May saving rate at 3.0%. The September 30 annual update revised it to 4.4%. The original release remains linked below as the record of the earlier estimate.

Source Drawer