The first systematic attempt at monitoring agricultural policies was undertaken by FAO thirty years ago. The genesis of the efforts to monitor agricultural policies in the 1970s was the necessity to understand and quantify the impact of agricultural protection in developed countries on world markets and hence on agriculture in developing countries. It had been recognized since the 1950s that developed country price support policies generated surpluses and restricted import markets to the detriment of world trade in farm products. GATT reports had called for the development of quantitative indicators to facilitate international negotiations, but these had not been forthcoming (GATT, 1958 and 1962). Some quantitative modelling had been attempted, but this suffered from the lack of information about the nature of the policies themselves and about their quantitative dimensions (see for instance the model by UNCTAD).
All the early studies on agricultural protection in developed countries had a “trade distortion” focus, though some had as an objective the improvement of domestic efficiency as well as the reduction of externalities. This implied an emphasis on policies specific to particular commodities, on price as opposed to non-price policies, and on short- and medium-term policies rather than long-term structural and technology policies. All took into account policies that directly impacted on prices, but all ignored factor markets. Most of the studies also ignored exchange rate misalignments, though it was understood that exchange rate changes were significant in agricultural markets. All were “partial” in nature, ignoring the impact of non-agricultural policies on the agricultural sector.
Four attempts to measure agricultural protection in developed countries will be discussed below. These include the FAO International Agricultural Adjustment (IAA) work, initiated in 1973 and updated and expanded through 1985, which introduced Producer and Consumer Subsidy Equivalents (PSEs and CSEs); the OECD PSEs, published regularly from 1986 until the present; the USDA PSEs, calculated for a different group of countries and published from 1985 to 1992; and the WTO AMS notifications, made by every member that entered domestic support commitments in their schedules (mostly developed countries) from 1995 to the present date.
FAO International Agricultural Adjustment
The first systematic attempt to monitor developed country farm policies was undertaken in the context of a study on international agricultural adjustment, the way in which the agricultural economy of the world adjusted as countries interacted with each other through trade and aid and influenced each other through their domestic policies (FAO, 1973, 1975). The project was motivated by the realization that these policies were having a profound impact on world markets for agricultural goods. The 1973 paper lays out the rationale for monitoring such policies and the links with the issues of trade liberalization in the GATT. Two main indicators were developed: the Producer Subsidy Equivalent (PSE), and the Consumer Tax Equivalent (CTE).9 Several measures were developed as potentially useful for trade talks, including the tariff equivalent (TE, calculated as a weighted average of PSE and CSE) and the Foreign Exchange Displacement (FE) of the policies, designed to indicate the magnitude of the external impact of a country's agricultural policies.10
The estimates developed in this paper, as an illustration of the method, were for five countries (Canada, France, Germany, the United States and the United Kingdom) and five commodities (wheat, barley, maize, milk and sugar). The time period for the data was 1968-70, and policy detail was from published sources. The report was presented to the FAO Conference in 1973 as a part of the plans for a regular monitoring of IAA.
The 1975 paper revised the method and terminology somewhat, and developed the link with trade negotiations. The consumer indicator was re-labelled the Consumer Subsidy Equivalent (CSE) and its sign changed. PSEs and CSEs were calculated for six countries (adding Australia and Japan and considering the EEC as one unit) and six commodities (including rice), over the period 1968-74. The longer time period gave an opportunity to judge the usefulness of the indicators at a time of variable world prices. The international impact of these policies was monitored by means of the Trade Volume Effect (TVE), calculated, as in 1973, from information on elasticities of supply and demand and the PSEs and CSEs.
Regular reports on the process of international agricultural adjustment were given to governments at the FAO Conference in 1977, 1979, 1981, 1983 and 1985. In each of these reports the PSE and CSE was presented as evidence of the evolution of their agricultural policies in the context of the IAA Guidelines.11 As the interest in the IAA guidelines waned, the quantitative monitoring proved less urgent for FAO: by the Seventh IAA Progress Report in 1991 the OECD PSE estimates (see below) had replaced those from the FAO in the monitoring of the developed country policy guideline.
OECD trade mandate
The Organisation for Economic Cooperation and Development (OECD) has a long-standing interest in agricultural policies, but reports on the topic tended to be qualitative and descriptive. In 1982 the OECD Ministerial Council committed itself to the task of tackling the problems of agricultural policies and their effects on trade (Legg, 2003). Ministers approved a Mandate that required the Secretariat to analyze “approaches and methods for a balanced and gradual reduction in protection for agriculture and the fuller integration of agriculture within the open multilateral trading system”. In doing so they were to conduct an examination of national policies that have a significant impact on international trade. The 1987 report (National Policies and International Trade) was the result of this endeavour (OECD, 1987). Modelled on the FAO PSEs, but with important differences, the OECD began to monitor domestic agricultural policies of its members and calculate subsidy equivalents for the major commodities. The focus of the OECD work was on the need to reform domestic policies in part to avoid trade impacts but also to improve domestic policy targeting.
The OECD methodology, as it has developed over time, now comprises a three-part classification of policy measures. The PSE itself is made up of Market Price Support (MPS) and Budgetary Payments; the General Services Support Estimate (GSSE), including research and development, infrastructure, marketing and promotion payments and inspection services; and any transfers from taxpayers to consumers (see Table 2). The total of these three items is known as the Total Support Estimate (TSE). This is equivalent to the sum of transfers to consumers and producers less any budget revenues, such as from import tariffs.
The OECD PSEs represent the most ambitious attempt to date in monitoring agricultural policies. The PSE data set is both the longest series of quantitative policy indicators (in any sector) and the most inclusive in terms of countries and commodities.12 It is also unique in that government officials and statisticians tackle part in the oversight of the estimates, giving the resulting estimates an implicit degree of approval. As a result it has widespread acceptance among governments and the media, and is the most frequently used measure for modelling policies and constructing models with policy variables.
USDA PSE and CSEs
The United States government showed an early interest in the monitoring of domestic farm policies in the OECD countries, but decided to calculate its own PSEs (USDA, 1987; 1988).13 The work was undertaken by the Economic Research Service of the United Stats Department of Agriculture (USDA), and continued for a number of years. Although the United States was included in the ERS-calculated subsidy equivalents, they were primarily devised to indicate extent of protection in other countries, and hence be useful in support of United States objectives for trade liberalization in the Uruguay Round.14 The countries included in the study were those of significance as markets or potential export competitors. As a result, several developing countries were included in the USDA estimates, in contrast to the OECD PSEs that were being developed at the same time. The results indicated the extent of subsidy and tax implied by the policies included in the study, a similar range of policies to that chosen by the OECD (see Table 3).
WTO AMS Notifications
The World Trade Organization also monitors agricultural policies of member countries. The Uruguay Round Agreement on Agriculture included the provision for monitoring compliance with commitments to reduce domestic support that was deemed to be trade distorting. The commitments were expressed as an Aggregate Measure of Support (AMS), known popularly as the amber box (WTO document WT/AG/NG/S/1, 2000). Though based loosely on the OECD PSEs, the concept of the amber box is somewhat different from the PSE, as it is based largely on budgetary payments and the implied transfer of administered domestic prices relative to a fixed reference price. Thus it includes some support that would otherwise be captured in the price effects of border measures (in effect double counting this support, as it is also monitored in market access and export subsidy commitments) but it excludes many elements of subsidy that are included in the PSE. Thus, total domestic support as measured by the WTO includes transfers in the green and blue boxes, the special and differential treatment category for developing countries, a de minimis exemption and the Aggregate Measure of Support (AMS), in turn comprised of market price support and budgetary payments. The relation between these measures is illustrated in Figure 1.
Measurement of protection levels in developing countries lagged behind that in developed countries, though protection indicators were measured for individual developing countries in the 1970s and 1980s. 15 However, primarily through several studies organized by the World Bank, a small number of cross-country support and protection estimates emerged in the 1990s. Other institutions followed up on this work, particularly the International Food Policy Research Institute (IFPRI) and FAO. This section reviews experience with a number of these studies.16
The World Bank has been the leading proponent of the monitoring of developing country policy in the agricultural sector. The earliest and most comprehensive study in this area was a large cross-country analysis of the political economy of agricultural pricing policy (Krueger, Schiff and Valdés, 1988, and Krueger, Schiff and Valdés, 1992). The study covered 18 developing countries during the period 1960-1985. It reports the NPR for the major import-competing and export activities distinguishing between “direct” and “indirect” effects of sectoral and economy-wide policies. NPR estimates follow a direct price comparison approach. The direct price comparison between border and farm prices adjusted for transport costs to or from producer and consumer locations, storage costs, quality differences, and other elements in the marketing margins, at the prevailing nominal exchange rate. The “indirect” nominal protection rate was measured as the proportional difference between the domestic agricultural price in relation to a price index of the nonagricultural sector, on the one hand, and the value of the relative price when measured at the equilibrium exchange rate and in the absence of industrial trade interventions, on the other hand. There are three major elements on the calculation of the indirect effects. First the depreciation of the real exchange rate required for the elimination of the non-sustainable part of the current account deficit; second, the depreciation of the real exchange rate due to the removal of trade interventions; and third, the increase in the price of agricultural tradable products relative to non-agricultural tradables due to a removal of industrial policy interventions. The first two are changes in the price of tradable relative to non-tradables; the third is a change of prices within the tradable category. On average, the net effect of direct and indirect interventions has been an enormous income transfer out of agriculture - averaging 64 percent of agricultural gross domestic product a year during the period 1996-84. Indirect effects were the main source of this implicit and explicit taxation of agriculture.
More recently the World Bank undertook a surveillance of agricultural price and trade policies in Latin America during major policy reforms (World Bank, 1996). This study covers eight countries during the period 1984-1994. The study examined how agricultural protection had changed since the initiation of economic reforms in these countries, what happened to real farm prices, and what was the current status (at the time) regarding tariffs and quantitative restrictions in the various countries. Using a direct price comparison approach, the study reported NPRs, ERPs, PSEs, and ERAs for the major agricultural import-competing and export activities. This analysis did not adjust for a possible misalignment of the exchange rate. The study presented a “decomposition analysis” to examine the relative effect of fluctuations in the real exchange rate, border prices and domestic trade policy on the evolution of domestic real farm prices. An analysis of the evolution of producer prices shows that between 1986 and 1995, in seven out of eight countries, all major agricultural producer prices declined in real terms. For most countries, the decline for both importables and exportables was larger during the reform period than in the previous years. The main factor explaining the decline in real domestic farm prices was the exchange rate appreciation observed during the early 1990s, a phenomenon that was amplified by tariff reductions and, in some cases, by a fall in border prices. This fall in real farm prices led to increased political pressure by farmers for protection of import-competing sub-sectors.
A further study by the World Bank monitored agricultural support policies in transition economies (World Bank, 2000). This study covers six transition economies during the period 1994-97. Using a direct price comparison approach, the study presents various estimates of agricultural support policies, including trade and price policy interventions and government non-price related subsidies on production incentives and on net farm income. The study examined to what extent the economic environment prevailing in 1994-97 provided an appropriate and sound basis for adjustment towards a more internationally competitive agricultural sector. Based on a common methodology for all countries, the study reports estimates of NPRs, ERPs, and ERAs for the major agricultural import-competing and export activities. The report presents a synthesis of the various indicators for all the countries included, which is followed by individual country agricultural policy notes describing the salient features of agricultural policies at the time. This study does not adjust for a possible misalignment of the exchange rate, though it does present a “decomposition analysis” to examine the relative effect of fluctuations in the real exchange rate, border prices and domestic trade policy on the evolution of domestic real farm prices.
EU North Africa
In addition to the World Bank studies mentioned above, other institutions with an interest in developing country policy have begun to be active in the area of monitoring. A cross-country comparative study on agricultural policies in North Africa has been undertaken on the initiative of the University of Montpellier, in France, with funding from the European Union. This study, entitled Le soutien aux produits agricoles et aux filières agro-alimentaires reports estimates of agricultural support measures for Egypt, Morocco, Turkey, and Tunisia for the period between 1994 and 1999. The study is directed by Petit and Allaya at Montpellier, and the project commissioned the studies to local economists in these countries. The study estimates NPRs and ERPs by direct price comparison at the prevailing nominal exchange rate. The terms of reference for the project follow the approach used in the World Bank studies on Eastern Europe and Latin America described above. It is a partial equilibrium study, with no adjustment for the “indirect effects” that were included in Krueger/Schiff/Valdés. On product coverage, for example, the study on Morocco covers wheat (soft and hard wheat, and rain-fed and irrigated as separate activities), sugar (sugar beet), milk, all import-competing activities, and on exportables covers oranges and tomatoes. The treatment of water pricing for irrigation is of particular interest in these countries and it is discussed in these studies. There was a considerable delay in the submission of the final reports, due primarily to “mobilisation des équipes nationales”.
FAO/ROA and IFPRI South Asia Studies
In addition to those studies mentioned above, two other projects should be mentioned. The on-going FAO study on the Roles of Agriculture (ROA) includes a component on agricultural support measures. This study includes a total of 11 countries in Asia, sub-Saharan Africa, North Africa and Latin America. This analysis presents estimates of nominal and effective rates of protection for the late 1990s, for the major agricultural import-competing and export activities (approximately six activities in total). Preliminary results of this study were presented at an FAO workshop in October 2003.
Work at IFPRI is underway on protection in South Asian agriculture, directed by Ashok Gulati. A study on agricultural protection in India has been published (Gulati and Narayanan, 2003) building on earlier work by Gulati and Purcell for the World Bank. The World Bank study followed the Krueger/Valdés/Schiff approach. The latter book emphasises the offsetting impacts on Indian agriculture of taxes on inputs and subsidies on outputs, leading to net negative protection for several crops. As this work is not yet cross-country in scope it will not be reviewed further here, As this work is not yet cross-country in scope it will not be reviewed further here, but IFPRI's planned studies will be complementary to any FAO work.
9 These measures were expressed in both per unit and percentage terms. In addition, the actual values of the transfers were calculated as the Producer Subsidy Value (PSV) and the Consumer Burden (CB), along with the Exchequer Cost (EC) and the effective protection. These were grouped as "domestic performance measures".
10 Two further indicators were also presented in that report, one the ratio of the TE to the PSE represented the extent to which the transfers captured in the PSE are trade distorting, and the other the ratio of the FE to the PSV reflecting the external impact per unit of transfer.
11 The IAA reports were grouped around eleven guidelines endorsed by governments at the 1975 Conference. The FAO Director-General was required to report progress toward meeting the objectives of the programmes Guideline 3 stated that: "National policies of developing countries should provide appropriate incentives for farmers to expand production and to promote the adaptation of structures within farming both to permit optimum use of available and suitable technology, and to promote social equity and fuller integration of the rural population into the national economy; national policies of developed countries should aim at the most rational use of their resources, having regard to the special needs and interests of developing countries and taking into account the need to ensure world food security." The policies of developing countries were not systematically monitored, but the PSEs gave an indication of the trends in the levels of support afforded to developed country producers.
12 In 1999 the OECD changed the label from Producer (Consumer) Subsidy Equivalent to Producer (Consumer) Support Estimate, thus retaining the acronym. Several member states of the OECD had been unhappy with the implication that all measured programmes were "subsidies" and preferred the more neutral term "support". Of course the phrase "support equivalent" never did imply that the transfers were themselves in the nature of subsidies.
13 Among the reasons for an alternative set of estimates was the fact that the United States had an interest in protection in non-OECD countries; the convenience of being able to choose commodity and policy sets for particular purposes; and the fact that there was a long lag before the OECD countries authorized the release of the PSE calculations.
14 The United States proposed the use of the OECD PSEs as a negotiating mechanism in the Uruguay Round. This was broadly agreeable to the Cairns Group but initially rejected by the EU. However, the EU itself proposed a "support measurement unit" based on fixed reference prices. The eventual Aggregate Measure of Support was variant on these measures, including administered prices and subsidies but not the impact of border measures.
15 To our knowledge the first comparative studies measuring protection in several developing countries were the study commissioned by the OECD: Little, Scitovsky and Scott "Industry and Trade in Some Developing Countries", Oxford University Press, 1970, and by B. Balassa, "The Structure of Protection in Developing Countries" J. Hopkins University Press, 1971" Although focused on the industrial sector, both studies discuss the potential bias against agriculture created by the protection of industry. Both studies use the Effective Protection framework.
16 There are several quantitative models that make use of protection estimates for developing countries. Some are general equilibrium and some are partial equilibrium. These include the GTAP model (T. Hertel et. al.) and those employed by the ERS/USDA, the OECD model, the IFPRI model (Rosegrant), the IMF model, Capri, the Australian ABARE model, and the FAO-UNCTAD model. These models rely on WTO-provided estimates of the actual tariffs (often through the TRAINS database) for developing countries, as PSEs are not available except for the very few which are OECD members (Mexico, Turkey, Korea, Poland).