chainweighted

Chainweighted, also written as chain-weighted, is a term used in statistics and economics to describe a class of index construction methods that update the weights assigned to components as a sequence evolves over time. The technique aims to reduce substitution bias by allowing the relative importance of components to shift as consumer preferences change. In practice, chain-weighted indices are built by linking short-term indices whose weights are refreshed from the most recent expenditure shares, producing a single continuous measure.

Methodology: At each period t, weights reflect current expenditure shares for the components. A short-term index is computed using those weights (a Laspeyres-like calculation). The index for period t is then chained to the index for period t-1, so that the overall series remains continuous as weights change. Over time, this produces a chain-weighted index such as a chained consumer price index (C-CPI).

Applications: The best-known use is in chained CPI measures used by statistical agencies to track inflation and to calculate cost-of-living adjustments. Chain weighting is also used in other price indices, cost-of-living analyses, and in some cases, in financial indices that must reflect changing composition of baskets or portfolios.

Advantages and limitations: Chainweighted indices better capture substitution effects and evolving consumer patterns than fixed-weight indices. They are more responsive but require more frequent data on expenditure shares and more complex data processing. Revisions to weights can cause added volatility and complicate interpretation.

See also: Fisher index, Laspeyres index, Paasche index, chained index, cost-of-living adjustments, inflation measurement, consumer price index.