Economic release reporting problems and fixes

What to take away

• An index level, percentage change, and percentage-point change are different quantities.
• Comparisons fail when periods, populations, or adjustments differ.
• Advance estimates should carry their revision status.
• A national average is not a personal bill.
• Correlation in one release does not prove cause.

Economic stories become misleading through small-looking shortcuts. A missing word such as annualized, real, average, or preliminary can change the entire claim. The vocabulary is defined in the economic releases and household effects guide.

Problem 1: lower inflation becomes lower prices

Symptom: A slowing rate is described as prices falling.

Fix: Report the rate and index direction separately. State whether prices rose more slowly, stayed flat, or fell over the named period. The distinction is worked through in translating inflation to household context.

Problem 2: percentage points become percent

Symptom: A rate moving from 4 percent to 5 percent is called a 1 percent increase.

Fix: It rose 1 percentage point, which is a 25 percent increase relative to the starting rate. Choose the expression that answers the story's question.

Problem 3: incompatible periods share a chart

Symptom: A monthly rate, quarterly annualized rate, and year-over-year rate appear as if directly comparable.

Fix: Convert only when the method permits and disclose the calculation. Prefer separate panels or a table with explicit periods.

Problem 4: revisions vanish

Symptom: The new estimate dominates while a large revision to the prior period appears at the bottom.

Fix: Report revisions that change the recent pattern in the opening. Preserve the old vintage only when showing how estimates evolved. Health desks solve this with status labels; compare health evidence and alert types.

Problem 5: nominal sales become volume

Symptom: Higher dollar sales are called more goods purchased without a price adjustment.

Fix: State the release's adjustment. The Census Bureau's current Advance Monthly Sales report labels its estimates as adjusted for seasonal, holiday, and trading-day effects but not for price changes. The specific figures on that page will change; the article relies only on the stated measurement distinction.

Read the release's category definition as well. Retail-establishment data group sales by the primary business of the establishment, which may differ from the product a reader has in mind. A store category can sell many kinds of goods. A product can be sold through several store categories. The article should not convert one classification into the other without supporting data.

Check current and prior vintages before describing momentum. An advance estimate may be superseded by a larger monthly sample, and a benchmark can revise a longer history. A chart that silently mixes vintages can create a false change at the join.

Problem 6: average earnings become everyone's raise

Symptom: An average increase is applied to all workers or households.

Fix: Name the employee population, hourly or weekly measure, composition effects, and price deflator. The BLS real earnings technical note explains that its constant-dollar earnings series combine Current Employment Statistics earnings with CPI deflators and use different CPI series for specified employee groups. That is a defined aggregate measure, not a payroll record for each worker.

Problem 7: the release proves the cause

Symptom: One category movement is attributed to policy, weather, or sentiment without causal evidence.

Fix: Separate description from explanation. Use additional data, timing, mechanism, and research before making a causal claim. Policy attribution has its own trap list; see government decision reporting problems.

Problem 8: the forecast becomes the benchmark of truth

Symptom: A result is called good or bad only because it differed from a market consensus.

Fix: Report the forecast as one comparison. Explain the measure's level, trend, revision, uncertainty, and real-world meaning.

Problem 9: charts preserve no vintage

Symptom: A chart updates automatically, while the text describes an older release.

Fix: Display retrieval date, vintage, series, units, and adjustment. Archive the data used for the published conclusion.

Problem 10: household examples look official

Symptom: A hypothetical budget appears beside government data without clear separation.

Fix: Label the example, publish inputs and arithmetic, and state that it does not estimate a typical household.

Common questions

Is beating a forecast the main story?

Not necessarily. Forecasts matter to markets, but readers also need the measure, trend, revisions, and household or business context.

Can seasonally adjusted and unadjusted data appear together?

Yes, with a clear reason and labels. Do not subtract or compare them as if they were one series.

Does a revised estimate mean the agency made a mistake?

Not by itself. Many programs publish early estimates and incorporate more complete data later.

When can a release support causal language?

When its design or supporting evidence identifies cause, not merely simultaneous movement.