Trading gains: terms‐of‐trade and real‐exchange‐rate effects

• Published date: 01 December 2023
• Author: Ulrich Kohli
• Date: 01 December 2023
• DOI: http://doi.org/10.1111/roiw.12615

bs_bs_banner

Review of Income and Wealth

Series 69, Number 4, December 2023

DOI: 10.1111/roiw.12615

TRADING GAINS: TERMS-OF-TRADE AND

REAL-EXCHANGE-RATE EFFECTS

BY ULRICH KOHLI∗

University of Geneva

This paper considers alternativemeasures of a country’s trading gains, i.e.,the extra income that it earns

(or loses) as the result of changes in the relativeprices relevant for international trade, and which makes

up the difference between real gross domestic product (GDP) and real gross domestic income (GDI).

Looking at both the Laspeyres and the Törnqvist aggregation, we show that the trading gains really

consist of two components, a terms-of-trade effect and a real-exchange-rate effect. Nearly all national

statistical agencies, receiving no rm guidance from international organizations in this matter, merely

consider the rst effect, which suggests thatthe so-called trading-gain estimates they publish are incom-

plete and misnamed. Even more seriously,it implies that the corresponding measures of real GDI they

derive are conceptually awed.A straightforward way to circumvent these difcultiesis to use the gross

domestic nal expenditure price index as deatorwhen computing real GDI and the trading gains. Some

numerical estimates for Australia are provided as an illustration. The paperalso identies the underly-

ing linear and Translog realGDI functions for which the Laspeyres and Törnqvist terms-of-trade and

real-exchange-rate effectsare exact.

JEL Codes: C43, E13, F11

Keywords:real exchange rate, real GDI, terms of trade, trading gains

1. INTRODUCTION

Australia has enjoyed substantial improvements in its terms of trade over the

past several decades.As shown by international trade theory,an improvement in the

terms of trade, other things equal, is income and welfare enhancing. There is some-

times a suspicion, however, in Australia and elsewhere that ofcial statistics, most

prominently real grossdomestic product (GDP), do an inadequate job at capturing

this increase in real income.1

Australia is not an isolated case. Canada, on the wave of a massive increase

in commodity prices in recent years, has been facing a similar situation, and so is

This is a much-expanded and revised version of a paper originally preparedfor the SSHRC Con-

ference on Price and Productivity Measurement, Vancouver, B.C.,July 2004. I wish to thank W. Erwin

Diewert forhis comments on an earlier draft, and I am also very grateful to the editor and an anonymous

referee for valuablecomments and suggestions.

*Correspondence to: Ulrich Kohli, Emeritus Professor of Economics, Geneva School of Eco-

nomics and Management, University of Geneva, Geneva, Switzerland (ulrich.kohli@hotmail.com;

ulrich.kohli@unige.ch).

1See The Australian,March 8, 2005, for instance. This question has been the subject of a major study

focusing on Australiaby Diewert and Lawrence (2006); their ndings is that over the period 1959–2003

the improvementin the ter ms of trade has contributedclose to 5 percent to real income growth, but this

was before the recent commodities boom.

© 2022 International Association forResearch in Income and Wealth.

975

Review of Income and Wealth, Series 69, Number 4, December 2023

Norway. Not all examples arenatural resources related. Switzerland, over the course

of the past several decades, has experienced a signicant improvement in its terms

of trade leading some observers to argue that the growth of real value added has

been underestimated by real GDP by close to half a percentage point annually, and

that the growth pessimism thathas developed in that country over time is therefore

not fully justied.2

As argued by Diewertand Morrison (1986) in their seminal article, an improve-

ment in the terms of trade has an impact on productivity similar to a technologi-

cal progress. The country essentially gets more for less. Yet, unlike a technological

progress, a terms-of-trade improvement is treated by the national accounts as a

price phenomenon, rather than as a real development. A drop in import prices, for

instance, will not only tend to increase nominal GDP (rightly so), but it will also

raise the GDP price deator (although no price has gone up), leaving real GDP

little changed, even though real income must unambiguously have increased.3

Much of international economic theory treats trade as taking place in n-

ished products,and thus occurring after production. This view is rather misleading.

Indeed, most trade is in raw materials and intermediate products, and even most

so-called nished goods that are traded are not ready to meet nal demand. Just

about all imported “nished” products must still transit through the domestic pro-

duction sector where they are subject to a number of transformations (handling,

transportation, insurance, wholesaling, repackaging, storing, retailing, and so on).

In this process,they are combined with domestic factor services, so that a signicant

proportion of the nal price tag is typicallyaccounted for by domestic value added.4

Similarly,exports are not ready to meet nal demand, as they must still owthrough

the foreign production sector and go through a number of changes. They are thus

conceptually different from goods and services intended for domestic absorption.

In that sense, nearly all traded goods are intermediate goods or middle products.5

By the same logic, all goods intended for domestic use can be viewed as nontraded

goods. In truth, international trade is an intimate part of production. Trade is just

another way by which some goods can be transformed into other goods. It should

make little difference to economists whether products are transformed into others

through a physical process, through a chemical reaction, or through trade. If the

terms of this transformation become more favorable, it should not matter whether

this is the result of a technological advance (whichmight come from abroad as well)

or whether it is due to better exchange conditions.

Among the statistical agencies, the U.S. Bureau of Economic Analysis (BEA)

stands out for long having published series of command-basis real GDP, generally

2See Kohli (2004a,2005), and The Economist, February14, 2004.

3In fact, as shown by Kohli(1983,2004a), an improvement in the terms of trade will tend to reduce

real GDP if it is measured by a Laspeyres quantity index.

4Of course, it is always possible to nd exceptions, e.g., e-commerce, which enables consumers to

import goods directly,but economic models are not meant to be perfect descriptions of reality. Instead,

to be useful, they must remain simple and focus on the essential elements of what is to be analyzed.

The alternative, conventional approach to international trade, byassuming that all trade is in nished

products and thus occurs after production, is a much moredrastic departure from reality.

5See Burgess (1974) and Kohli(1978,1991); the term “middle products” has been coined by Sanyal

and Jones (1982).

© 2022 International Association forResearch in Income and Wealth.

976