Your linked site states: > The harmonised indices shall be annually chain-linked Laspeyres-type indices. And further down: > Elementary aggregate indices for

Your linked site states: > The harmonised indices shall be annually chain-linked Laspeyres-type indices. And further down: > Elementary aggregate indices for the HICPs are computed as the ratio of the geometric average of prices or the ratio of the arithmetic average of prices. The arithmetic mean of price relatives may only be applied in exceptional cases and where it can be shown to be comparable. I read this as: - individual indices for countries are calculated using the Laspeyres [1] formula, which is effectively an arithmetic mean of weighted prices of different products, not a geometric mean - the individual indices of countries are then "aggregated" by calculating the geometric mean That'd mean that outlier countries would not move the average much so there you'd have underestimation of inflation to a certain degree. But it wouldn't mean that individual products/components of the consumption basket that significantly increased in price would be undererstimated through the geometric mean. Right? Just calculating a geometric mean of countries' indices to get an index for EU wouldn't make much sense, though .. they'd have to weigh the indices with population, GDP or total consumption. So the actual calculation to get the HICP is surely more complicated than that. But if they really use Laspeyres then the underestimation due to geometric mean shouldn't be an issue here. [1] https://en.wikipedia.org/wiki/Price_index#Paasche_and_Laspeyres_price_indices https://en.wikipedia.org/wiki/Price_index#Paasche_and_Laspey...