Trading Gains and Productivity: A Tornqvist Approach

This article looks at alternative Tornqvist measures of a country's trading-gain and terms-of-trade effects, as they have been proposed in the literature starting with the seminal work of Diewert and Morrison (1986), and their link to standard measures of productivity. It strongly argues in favour of using the price of domestic final demand as a deflator when computing real Gross Domestic Income (GDI), and, by the same token, the trading gains and labour productivity measures. It shows that the trading gains then generally consist of two parts, a pure terms-of-trade component and an additional relative-price component, the latter of which can be interpreted as a real-exchange-rate effect. National and international statistical agencies, with the notable exceptions of Statistics Canada and the U.S. Bureau of Economic Analysis, tend to report incomplete trading-gain statistics in that they omit the second component. Consequently the real GDI estimates they publish must be viewed as flawed. Taking trading-gains into account has no direct effect on the measurement of total factor productivity, but it does affect the measures of average and marginal labour productivity when related to real GDI and its deflator. Numerical estimates for Switzerland are reported as an illustration.