Data Note
The euro area's growth rate reflects the relative size of its economies and the construction of its volume accounts. A hypothetical aggregation separates union-wide output growth from the experience of an equally weighted member.
The growth rate of the euro area is a statement about the change in the union's aggregate output. An equally weighted average of national growth rates is a statement about the average member. They need not coincide because an economy representing a large share of production does not receive the same economic weight as a much smaller member.
Eurostat's quarterly national-accounts metadata explains the presentation of current-price and volume measures and warns that chain-linked member-country levels do not generally sum to the chain-linked aggregate. Its account of European aggregation sets out the role of country levels and compatible prices. Neither method amounts to averaging the growth headlines with one equal vote for each country.
Consider a hypothetical monetary union containing only A and B. Express both periods at one compatible set of prices so that the values are additive. A initially produces 900 units of output and B produces 100. In the next period, their output rises by 1 percent and 5 percent, respectively.
| Economy | Initial output | Next output | Output growth |
|---|---|---|---|
| A | 900 | 909 | 1% |
| B | 100 | 105 | 5% |
| Union | 1,000 | 1,014 | 1.4% |
The unweighted mean is 3 percent. Aggregate growth is 1.4 percent because A has nine times B's initial output. A contributes 0.9 percentage point to the aggregate increase and B contributes 0.5 percentage point. The smaller economy grows faster but contributes less additional output.
A survey of how widely growth is distributed might reasonably report the median member or the proportion of economies expanding. Those measures describe dispersion or breadth. Replacing the aggregate rate with one of them would change the research question rather than correct the official measure.
The example uses additive values deliberately. Summing separately published chain-linked country levels and treating the result as a new official real aggregate can introduce another error. A proper reconstruction needs the agency's aggregation inputs, common price basis and linking method. An initially plausible country-weighted approximation should be labeled as such when it does not reproduce the official construction.
Geographic coverage needs equal care. Comparing the output of a larger membership group this year with that of a smaller group last year mixes production growth with a change in the statistical boundary. A constant-composition series addresses a different question from a series reflecting membership as it changes. The series label and metadata should establish which one is being used before a growth rate is calculated.
The same discipline applies to vintage. If one national estimate has been revised and the regional aggregate still reflects an earlier release, a researcher may be combining observations that were never compiled together. A reconciliation should record the extraction date and available estimates rather than treat every temporary difference as a failure of aggregation.
Union-wide output is relevant to the overall economic environment, but a firm's geographic sales exposure can differ sharply from GDP weights. A supplier concentrated in B may experience demand more closely related to B's economy than the union aggregate in the example. The appropriate business-exposure index would use the firm's relevant market weights and would need its own interpretation.
Three measures can therefore be reported without contradiction: aggregate output growth, the distribution of country growth, and a firm's weighted market exposure. Their weights should be explicit. Clear labeling preserves the distinction between the scale of the monetary union's expansion and the experience of its individual economies.