Process / pipelineEconomicsIndex numbers & aggregationPipeline

Törnqvist Index

Also known as: Tornqvist Index, Tornqvist-Theil Index, Translog Index, Tornqvist Price Index

OriginatorLeo Törnqvist; superlative theory by W. Erwin DiewertYear1936Sources2Related methods8

The Törnqvist index is a superlative index number used to aggregate many individual prices or quantities into a single measure of overall price change or quantity change between two periods. It is a weighted geometric mean of the individual price (or quantity) relatives, where each item's weight is the average of its value shares in the two periods. Because it is 'exact' for the flexible translog aggregator function, it is the standard tool for constructing productivity indices and is widely used in national accounts, productivity statistics, and price measurement.

Key highlights

  • Superlative: exact for the flexible translog aggregator, so it approximates a true cost-of-living or production index to second order.
  • Uses two-period average value shares, capturing substitution between goods and avoiding the substitution bias of fixed-basket indices.
  • The geometric (log) form treats price increases and decreases symmetrically and integrates cleanly with growth-accounting log-difference identities.
  • Forms the theoretical basis for productivity measurement: the ratio of Törnqvist output to input indices is an exact Malmquist productivity index.

Intuition

This section is available to Pro members. Upgrade to Pro

How it works

This section is available to Pro members. Upgrade to Pro

When to use it

Use the Törnqvist index when you need to aggregate heterogeneous prices or quantities into a single index and you have value (price times quantity) information for two periods so that value shares can be computed. It is the preferred index for productivity measurement and is widely used for chained price and quantity indices in national accounts, because it is superlative — exact for a flexible second-order aggregator — and therefore approximates a true cost-of-living or production index closely. It is most appropriate for chaining adjacent periods, where price and quantity changes are moderate. It is less suitable when value-share data are unavailable, when comparing distant periods directly (chain drift can arise with volatile, bouncing prices), or when a strict additive-decomposition or transitivity property is required, in which case the Fisher index or a multilateral method may be preferred.

Strengths & limitations

Strengths
  • Superlative: exact for the flexible translog aggregator, so it approximates a true cost-of-living or production index to second order.
  • Uses two-period average value shares, capturing substitution between goods and avoiding the substitution bias of fixed-basket indices.
  • The geometric (log) form treats price increases and decreases symmetrically and integrates cleanly with growth-accounting log-difference identities.
  • Forms the theoretical basis for productivity measurement: the ratio of Törnqvist output to input indices is an exact Malmquist productivity index.
Limitations
  • Requires both prices and quantities (value shares) in each period; it cannot be computed from price relatives alone.
  • Does not satisfy the factor-reversal test exactly, so the product of the Törnqvist price and quantity indices need not equal the value ratio.
  • Not transitive, so chained comparisons over many periods can suffer chain drift when prices bounce.
  • Slightly harder to communicate and decompose additively than the simpler Laspeyres or Paasche fixed-basket indices.

Common pitfalls

This section is available to Pro members. Upgrade to Pro

Applications

This section is available to Pro members. Upgrade to Pro

Frequently asked

What makes the Törnqvist index a 'superlative' index?

An index is called superlative, in Diewert's 1976 terminology, when it is exact for a flexible functional form — one that can provide a second-order approximation to an arbitrary aggregator function. The Törnqvist index is exact for the translog function, meaning that if the underlying cost or production function is translog, the Törnqvist index recovers the true index exactly. Because the translog flexibly approximates any smooth technology to second order, superlative indices like Törnqvist closely track the true cost-of-living or productivity index without assuming a restrictive functional form, which is why they are preferred over fixed-basket alternatives.

How does the Törnqvist index differ from the Fisher ideal index?

Both are superlative indices that use information from both periods and closely approximate a true economic index, and they typically give very similar numbers in practice. The Törnqvist index is a geometric mean of price relatives weighted by average value shares, exact for the translog aggregator; the Fisher index is the geometric mean of the Laspeyres and Paasche indices, exact for a quadratic aggregator. The key practical difference is that Fisher satisfies the factor-reversal and time-reversal tests exactly — its price and quantity indices multiply to the value ratio — whereas Törnqvist does not satisfy factor reversal exactly. Törnqvist, by contrast, decomposes additively in logs, which is convenient for growth accounting.

Why are two-period average value shares used as weights?

Using the average of the base-period and current-period value shares is precisely what makes the index superlative and lets it capture substitution behaviour. A Laspeyres index fixes weights at the base period and overstates inflation because it ignores that consumers shift away from goods that become relatively expensive; a Paasche index fixes weights at the current period and understates it. Averaging the two shares splits the difference symmetrically and corresponds to the exact aggregation of a flexible translog function, so the resulting index neither systematically over- nor under-states the true change.

Sources

  1. 1.
    Diewert, W. E. (1976). Exact and superlative index numbers. Journal of Econometrics, 4(2), 115–145.
  2. 2.
    Caves, D. W., Christensen, L. R., & Diewert, W. E. (1982). The economic theory of index numbers and the measurement of input, output, and productivity. Econometrica, 50(6), 1393–1414.

You have read it. What now?

Cite this page

ScholarGate. (2026, June 22). Törnqvist Index. ScholarGate. https://scholargate.app/economics/tornqvist-index