Executive summary
This paper presents a new measure of aggregate trade restrictions (MATR) using data from the International Monetary Fund’s Annual Report on Exchange Arrangements and Exchange Restrictions. MATR is strongly correlated with existing measures of trade restrictiveness but is more comprehensive in terms of country and time coverage. It is available for an unbalanced sample of up to 157 countries during 1949-2019. We use MATR to re-examine how trade restrictiveness varies with the business cycle, and how the macroeconomy looks in the aftermath of changes in trade restrictiveness. For the sample as a whole, MATR is typically a-cyclical but this average finding is heterogeneous across income groups: aggregate trade restrictions are a-cyclical in advanced economies but are counter-cyclical in emerging market and developing economies, especially in response to increases in unemployment. As to macroeconomic effects, increases in MATR are robustly associated with declines in GDP and in labour productivity (as well as being adverse for a range of other macroeconomic indicators).
Introduction
Trade policies are an important instrument in economic policy toolkits, and accordingly have received considerable scrutiny in the empirical economics literature which uses available measures to gauge their economic impact. Yet, it remains difficult to measure quantitatively the extent of trade restrictiveness across a large set of countries over a long period of time. While there is a plethora of trade policy indicators, most of them – except for tariff data – are available only with limited time and country coverage (see Estefania-Flores et al, 2022, for a discussion). To address this limitation, we present a new way to quantify policy towards international trade at the aggregate level. Our measure of aggregate trade restrictions (hereafter ‘MATR’) is based on data from the IMF’s Annual Report on Exchange Arrangements and Exchange Restrictions (hereafter ‘AREAER’). The measure is constructed through an in-depth process that combines information in the AREAER online database (available from 1999 onwards) with narrative accounts of how restrictive official government policy is towards the cross-border flow of goods and services, obtainable in printed versions of the AREAER country-year specific reports (from 1949 onwards). We show that our indicator is strongly correlated with existing measures of trade restrictiveness but more comprehensive in terms of country and time coverage: it is available for an unbalanced sample of up to 157 countries over the period from 1949 to 2019 (and of course is updatable as more data become available). The aggregate level of the data makes it particularly useful to assess the macro-economic dimension of restrictions, including the co-movement of restrictiveness with the business cycle. There is a longstanding literature examining how specific trade policy measures (tariffs, quotas and temporary trade barriers) respond to fluctuations in economic activity. While this literature provides convincing evidence that trade policy tended to be counter-cyclical – that is, rising during periods of economic downturn – before the Second World War1, the evidence using post-war data is less clear cut. For example, using a large panel of data, Rose (2013) showed that trade protectionism does not systematically increase during economic downturns. In contrast, Knetter and Prusa (2003) found that real exchange appreciations increase anti-dumping filings in Australia, Canada, the EU and the USA between 1980 and 1998. Bown and Crowley (2013) estimated the impact of macroeconomic fluctuations on import protection policies for five industrialised economies – the United States, European Union, Australia, Canada and South Korea. They found evidence of strongly countercyclical trade policy in the two decades leading up to the Great Recession, as countries resorted to new temporary trade barriers (TTBs) in response to increases in unemployment rates and real exchange rate appreciations. Similarly, Furceri et al (2023) used high-frequency TTB sectoral data covering 1220 sectors in 25 countries during 1989-2019, and found that retaliation through trade barriers increases in periods of high unemployment. High-frequency and granular data may make it easier to identify cyclical responses of trade policy. Our new measure of aggregate trade restrictiveness affords us the opportunity to re-examine the connection between trade policy and the business cycle. We present results on how MATR varies with the business cycle, and whether cyclicality varies over time and across countries. Our results suggest that, on average, MATR appears to be largely a-cyclical in the sample as a whole, but this result is heterogeneous across countries: MATR tends to be a-cyclical in advanced economies (AEs) but counter-cyclical in emerging market and developing economies (EMDEs), especially in response to unemployment.
MATR data
The MATR is built on data from the IMF’s Annual Report on Exchange Arrangements and Exchange Restrictions (AREAER). The measure combines information in the AREAER online dataset (available from 1999) with the narrative accounts of policies across countries related to the international flow of goods and services. The latter was compiled from the printed version of the IMF’s AREAER countryspecific reports from 1949 onwards. The in-depth details on the measure, including underlying method for compiling narrative accounts, is described in Estefania Flores et al (2022). MATR is based on the IMF’s AREAER binary variables on: (i) exchange measures; (ii) arrangements for payments and receipts; (iii) imports and imports payments; (iv) exports and exports proceeds; and (v) payment and proceeds from individual transfers and current transfers. Each of these categories include sub-categories2. The simplest version of the MATR is the unweighted sum of 22 variables (Table 1). The underlying components of MATR (the ‘fundamentals’) give granular measures of different facets of policy by using information on tariffs, non-tariff barriers, and restrictions on requiring, obtaining, and using foreign exchange for current transactions. The MATR, has several desirable properties: (i) it is based on sensible, plausible policy inputs from a transparent, accessible, reliable source; (ii) each of the underlying fundamentals is quantitative, based on clear criteria, and the fundamentals include a host of non-tariff barriers as well as tariffs; (iii) normalisation issues are avoided since the measure is an aggregate of binary components. The MATR is available for a large, unbalanced panel of most economies from 1949 through 2019, and it is regularly updated.3,4 The coverage increases from about 30 economies in 1949 to more than 100 countries in 1973, and over 150 countries by 2000, as shown in Figure 1. The MATR is an intrinsically aggregate measure rather than a weighted average of disaggregated microdata (in contrast to the aggregate tariff); it does not have sectoral variation, ie it is inherently macroeconomic or aggregate in nature. Moreover, it codes the existence of restrictions, not their intensity or efficacy. That said, and as shown in Table 2, MATR is strongly correlated with existing measures that capture the intensity of trade restrictions.
Acess the full publication here
About the Authors
Julia Estefania-Flores is a Research Analyst at the International Monetary Fund
Davide Furceri is Deputy Division Chief of the Development Macroeconomic Division at the International Monetary Fund
Swarnali A. Hannan is a Senior Economist at the International Monetary Fund
Jonathan D. Ostry is a non-resident fellow at Bruegel, as well as Professor of the Practice of Economics at Georgetown University in Washington, DC; he is also a Research Fellow at the Center for Economic Policy Research in London and serves on the advisory board of the World Economic Forum’s Global Risk Report in Geneva.
Andrew K. Rose is Professor Emeritus of Economic Analysis and Policy a the Haas School of Business.