Real GDP components expressed in chained dollars generally do not add up to the published total outside the reference year. The aggregate and its components track changes in quantities using changing price weights, then receive their own dollar scaling. Adding those separately scaled series can produce a different number. This property is called nonadditivity. It does not mean that output has disappeared.

This educational explanation uses the US Bureau of Economic Analysis method. The example is hypothetical, covers two complete years and assumes an economy with only two final goods. It contains no imports, intermediate inputs or inventories. Familiarity with percentages and square roots is helpful.

Why the dollar label can mislead

Current dollar GDP values production at the prices of the period. Real GDP aims to separate changes in quantities from changes in prices. BEA uses a Fisher quantity index, which combines comparisons using prices from both adjacent periods. Successive comparisons are linked together. The BEA methods handbook, chapter 4 explains the calculation.

A chained dollar series scales its quantity index to a dollar amount in a reference year. The BEA definition of chained dollar estimates warns that these values should not be used to calculate component shares or contributions to aggregate growth. A shared dollar label does not give every series a common fixed set of prices.

A two good example you can recalculate

Year 0 is the reference year. Prices are hypothetical dollars per unit. Quantities are units produced during each year. Goods A and B are distinct products of unchanged quality.

Hypothetical annual prices and quantities
Good Year 0 price Year 0 quantity Year 1 price Year 1 quantity
A $1 100 $1 200
B $1 100 $2 100

Nominal GDP is $200 in Year 0 and $400 in Year 1. The second total is $200 for A plus $200 for B. Its 100% increase combines a quantity change for A and a price change for B.

To isolate quantities, first value both years at Year 0 prices. The ratio is 300 divided by 200, or 1.5. This is the Laspeyres quantity comparison. Then value both years at Year 1 prices. The ratio is 400 divided by 300, or about 1.333333. This is the Paasche quantity comparison.

The Fisher quantity factor is the square root of their product. Using the exact fractions, the square root of (300 divided by 200) times (400 divided by 300) equals the square root of 2, or approximately 1.414214. Aggregate real growth is therefore about 41.42% from Year 0 to Year 1. This is an annual comparison, not an annualized quarterly rate.

Scale the aggregate factor by reference year GDP of $200. Year 1 real GDP is approximately $282.84 in Year 0 chained dollars. Separately, A has doubled in quantity, so its chained value rises from $100 to $200. B has unchanged quantity, so its chained value stays at $100. The component values sum to $300, which exceeds the aggregate by approximately $17.16.

That gap follows from the calculation. It is not an omitted industry, a statistical estimate of hidden activity or a rounding error. In the reference year, both components total $200 and match GDP. In Year 1, the relative price of B has changed and the separate component scaling no longer preserves addition.

Choose the measure that answers your question

For a spending share, use matching current dollar values. A accounts for $200 divided by $400, or 50%, of nominal GDP in Year 1. Dividing its chained value of $200 by chained GDP of $282.84 gives about 70.71%. That ratio is not its nominal spending share.

For growth in one series, a quantity index and its corresponding chained dollar series give the same percentage change, as BEA explains in its guidance on using real measures. For the contribution of a component to aggregate growth, use BEA's published contributions tables. Those contributions are expressed in percentage points and use the aggregation method, rather than simple differences between chained dollar levels.

The toy economy does not reproduce every GDP compilation rule, particularly components that can be negative. It demonstrates why forcing an additive total from chained levels can mislead. Check the units, reference year, period and release vintage before reconciling a real table. For context, read GDP, growth and productivity and how to read economic releases.

Sources and editorial note

The linked BEA guidance and the December 2023 methods chapter were checked on 14 November 2026. All example values are original hypothetical inputs, not current US statistics. Review is due by 14 May 2027, or sooner if the methodology changes.

This educational text was prepared with a generative language model, checked against the linked sources and recalculated. No independent human review is claimed. The original main cartoon was created with a generative image tool for FOLITIC. It illustrates a measurement puzzle and does not depict an actual economy or a precise formula. This article provides general education, not a personal investment recommendation.