Commodity Price Volatility and Inclusive Growth in Low-Income Countries
Chapter

Chapter 17. Trade Distortions and Food Price Surges

Author(s):
Rabah Arezki, Catherine Pattillo, Marc Quintyn, and Min Zhu
Published Date:
October 2012
Share
  • ShareShare
Show Summary Details
Author(s)
Will Martin and Kym Anderson 

Introduction

Prices of grains and other storable commodities are characterized by long periods in the doldrums punctuated by short but intense price spikes (Deaton and Laroque, 1992). Those spikes are of concern, not least because they can have large impacts on poverty in developing countries (Ivanic and Martin, 2008). Accounts of the food price spikes of 1973–74, 2006–08, and 2010–11 include discussion of a wide range of contributing factors such as exogenous shocks to supply or demand, below-trend stock levels, speculative behavior, and trade policy responses to the shock. Johnson (1975) emphasized policy responses in his analysis of the 1973–74 price spike, as have most of the available assessments of the 2006–08 shock (Robles, Torero, and von Braun, 2009; Baffes and Haniotis, 2010; Bouët and Laborde, 2010; Hochman and others, 2010; Timmer, 2010). Several suggest that export restrictions (and possibly also import subsidies) played an important role in these price spikes, just as intensified export subsidies and triggered import restrictions played a significant role in 1986–88 when international food prices slumped. However, we are unaware of any attempts to quantify the aggregate contribution across countries of trade policy responses to international price surges.

In this chapter, we address this issue directly. Following Freund and Özden (2008), we assume that national trade policy responds to the risk of losses for significant groups by insulating the domestic market to some extent from international price fluctuations for staple foods. This is consistent with the behavior of many governments, and it provides an economic rationale for the econometric estimation of price transmission elasticities. We use a standard conceptual framework to derive a simple equation that provides at least a rough way to estimate the contribution of market-insulating policy behavior to international price spikes for homogenous farm products. We subsequently examine evidence from two major upward price spikes (1973–74 and 2006–08) for the key commodities of wheat and rice. We then discuss the policy implications in the final section of the chapter.

International Price Volatility and National Policy Responses

Consider a weather- (or financial market–)induced exogenous shock to the global market for a food staple that causes a surge in its international price. Suppose that, in response, exporting countries impose or raise an export tax or tighten export restrictions (or lower any export subsidy), and importing countries reduce their tariff or other import restrictions (or introduce or raise an import subsidy) to reduce the rise in their domestic price. If both sets of countries try to reduce the impact of the shock on domestic prices to the same extent, their attempts will be collectively futile. This is very easy to show graphically in the case in which first the exporters and then the importers seek to block completely the effect of an increase in the price of food resulting from an initial shock.

For an individual small exporting country, the effect of the increase in its (explicit or implicit) export tax is to reduce the domestic price relative to the newly raised world price. The same effect occurs in a small importing country that reduces its (explicit or implicit) import tariff. If a sufficient number of exporting countries intervene in this way, their export restrictions cause the world price of the good to rise further, thereby reducing the impact of each country’s initial action on its domestic price. This situation is depicted in Figure 17.1, where the excess supply curve of the exporting-country group is ES and the excess demand curve of the importing-country group is ED following the exogenous shock but prior to any changes to trade restrictions. If an export tax is then applied, the world price needed to obtain any given level of exports is higher, because part of the export price is paid to the exporting government. This is reflected in the ES curve moving up to ES′,1 the effects of which are to raise the world price from Pw to P′w and lower the domestic price from Pw to Pd.

Figure 17.1Key Impacts of an Export Restriction

Source: Authors’ calculations.

In the situation depicted in Figure 17.1, the exporting-country group gains from the improvement in their export price. However, production incentives are reduced and consumers have an incentive to increase their demand, even though export prices are higher. The global social cost associated with these incentives is given by the triangle abc. That can be subdivided into a loss to private agents in the importing group of area bcPωPω, a loss to private agents in the exporting group of area baPdPω, and a gain to the government or export quota holders in the exporting-country group of area acPwPω. Whether the exporting countries as a group enjoy a net gain from restricting exports depends on whether the upper rectangle (the terms of trade gain) is larger than the lower triangle (the social cost). Because the social costs rise with the square of the export tax equivalent (PwPd), while the terms of trade gain is likely to rise roughly linearly with the rate, the benefits to the exporter group will become negative if the export tax rate becomes sufficiently large. By contrast, importing countries unambiguously lose from the export restrictions as they transfer income to the exporter and reduce net consumption.2

In Figure 17.2, exporters attempt to completely offset the impact of the initial increase in the price of the good by shifting the ES curve to ES′. Importers seek to achieve the same insulation by reducing tariffs (or paying import subsidies) so as to shift the ED curve to ED′. As is evident in Figure 17.2, the combined effect of these policy changes is to leave domestic prices in both importers and exporters at the postshock level Pw and to raise the international price from Pω to Pω. Despite the attempts of both the importing- and exporting-country groups to fully offset the original increase in price to Pw, domestic prices and quantities are unchanged at their postshock level (P0 in Figure 17.2). The only effect of these policies is to compound the terms of trade shift against the importing-country group and in favor of the exporting group, generating a transfer from the former to the latter of (PωʺP0). Q in Figure 17.2 (in addition to that caused by the initial exogenous shock). This is in sharp contrast with a move from autarchy toward free trade, which is able to substantially reduce price risk through diversification of market outlets because the correlations between commodity output shocks across countries are very limited (Johnson, 1975).

Figure 17.2Impacts of Equal Export Barrier Increases and Import Barrier Reductions

Source: Authors’ calculations.

When Countries Partially Insulate Against Global Price Changes

The empirical evidence based on a large sample of developing countries over the period 1995–2007 (Anderson, 2009) using a price comparison methodology (Anderson and others, 2008) indicates that agricultural distortions differ substantially across commodities and countries, with rates for individual commodities changing over long periods and year-to-year changes being negatively correlated with movements in real prices. This evidence suggests that average rates of protection at any time differ substantially across commodities in response to political economy pressures and that the average rate of protection for commodities may trend upward or downward depending upon the evolution of these pressures. In addition, it appears that policymakers attempt to smooth domestic prices of some key commodities relative to international prices. This preference for policies that insulate domestic prices from short-term changes around a desired level that differs from world prices in a way that rewards politically influential interest groups has been represented typically using relatively ad hoc combinations of average protection rates and price insulation coefficients. It seems desirable to be able to specify a welfare function that motivates such preferences.

An objective function that can represent this type of preference builds on Jean, Laborde, and Martin (2010a) and is closely related to Freund and Özden (2008). Suppose that policymakers in a single small country seek to minimize the following money-metric political economy welfare loss function:

where higher values of W indicate greater costs of deviating from the policymakers’ preferred equilibrium in which domestic prices are aligned to the strength of different interest groups and do not deviate from their desired long-term levels; p is the domestic price vector; α is a matrix representing the political economy costs of deviations from the vector of desired domestic prices under the intervention regime, p;h is a vector of weights that represents the preference for higher, or lower, average domestic prices for individual commodities; e is the expenditure function; g is the GDP function representing the value of output in the country; z = e – g is the net expenditure function; and zp is the derivative of this function and, hence, by duality, the country’s vector of net imports; and p* is a vector of world prices. The (pp*) term is a vector of border price interventions such as trade taxes and subsidies or quantitative trade restrictions.

Note that the last three terms of equation (17.1) are a standard Anderson and Neary balance of trade function. If policymakers seek to minimize this function alone, the optimal tariff, pp*, will be zero and the balance of trade function can be used to measure the cost of deviations from zero tariffs. As shown by Jean, Laborde, and Martin (2010b), inclusion of the hp term makes the political economy function consistent with nonzero interventions.

The h function captures in a reduced form a wide range of political economy incentives for intervention such as the relative ability of particular sectors to lobby for assistance (see Hillman, 1982; Anderson and Hayami, 1986; Lindert, 1991; and Anderson, 2010). This formulation captures the essence of the policy preference for sector-specific profits in Freund and Özden (2008) in a more general, but less specific, context. It allows factors such as countervailing lobbying by downstream users and the differential impact of protection on returns to factors emphasized by Anderson (1995) to be taken into account. As shown by Jean, Laborde, and Martin (2010a, 2010b), the values of h can potentially be inferred from information on relative levels of protection3 across sectors and the price responsiveness of traded quantities.

The first term in equation (17.1) represents the cost of deviations from average domestic prices. These arise from a range of factors, including the inherent risk aversion of particular groups, and credit market imperfections that make it difficult to smooth consumption in response to income or expenditure shocks. The diagonal elements of this matrix would generally be expected to be positive—because deviations from average domestic prices (which include the effects of the chosen tariff) raise costs to some groups and hence induce some political pain. The off-diagonal elements might be positive or negative depending upon whether changes in other prices alleviate or exacerbate the political pain.

Equation (17.1) is very similar to the social welfare function in Freund and Özden’s (2008) equation. One major difference is that equation (17.1) is based on deviations from the expected domestic price, rather than a reference price. A second is that equation (17.1) punishes deviations in either direction from the expected price. A third is that the cost of deviations enters quadratically rather than linearly. These differences in formulation reflect the particular situation of agricultural commodities, especially in developing countries. The expected price can be seen as a rational expectations counterpart of the reference price in Freund and Özden (2008). With agricultural commodities in poor countries, deviations in either direction from the expected price involve social costs because staple foods make up a large share of the incomes of poor consumers with limited access to credit as well as being important for the income of some poor producers.

In contrast to the puzzling situation in the manufacturing trade, where only antitrade interventions are frequently observed (Rodrik, 1995), we also observe export subsidies and import subsidies in different states of agricultural markets as well as antitrade measures such as export and import restrictions. It also seems likely that the costs of being away from the expected domestic price level are increasing in the size of the deviation. One piece of evidence for this hypothesis is the dramatic increase in the incidence of export restrictions and import subsidies when prices increase sharply (World Bank, 2008).

If we differentiate equation (17.1) with respect to prices, we obtain

This yields

The expected value of (pp*) is therefore given by

Rearranging equation (17.3) and substituting for p from equation (17.4), we obtain

which can be rearranged to get

and finally,

Equation (17.8) is difficult to interpret in its general form, and in reality, it seems unlikely that policymakers or analysts would have good information on either the full matrix of slopes of the import demand function, zpp, or penalties, α, for deviations from the average domestic price. In applied work, the typical response to this problem (see Feenstra, 1995) is to focus only on the diagonal elements of the relevant matrices. If we do this with equation (17.8), we are left with a relationship between the price distortion rate and deviations from the average world price of a particular commodity:

Because αi is positive and zii is negative, the coefficient on (p¯ip¯i*) in equation (17.9) lies between zero and one. Its (highly plausible) implication is that the higher the world price relative to its trend value, the lower will be the rate of distortion. This coefficient is, in fact, one minus the coefficient of price insulation used by Tyers and Anderson (1992). Equation (17.9) generates the potentially testable hypothesis that policymakers minimizing an objective function such as equation (17.1) will adjust their rates of agricultural price distortion to partially offset deviations of world prices from their trend value. This provides a rationale for the popular approach of characterizing policies using simple, apparently ad hoc approaches such as the elasticity of price transmission.

The extent to which countries can reduce the instability they face by transferring it to other countries will depend on which countries seek to insulate and what reaction other countries make to these insulating policies. Tyers and Anderson (1992) made an ambitious attempt to assess the extent to which the policies adopted by major participants in the world market for grains and other agricultural staples affect the volatility of world market prices and of domestic market prices. They concluded that the coefficient of variation of world prices for food would fall from 34 to 10 if all countries agreed to eliminate their price-insulating policies. In most of the 16 developing economies they considered, the coefficient of variation for domestic prices would fall substantially if all countries refrained from using the type of price-insulating prices they have used in the past. In a number of these cases, such as Bangladesh, South Africa, and Thailand, the reductions in domestic price instability were estimated to be dramatic, with the coefficient of variation in Bangladesh, for instance, falling from 26 to 8. In the few cases in which the coefficient of variation of domestic prices was estimated to rise, the increases were much smaller.

If policymakers in importing countries were concerned primarily about the impact on their terms of trade of the imposition of export restrictions, they might respond by raising tariffs on their imports. However, during episodes of international food price spikes, the response has typically been the opposite: Tariffs on food imports are reduced in an attempt to avoid adverse impacts on domestic consumers. This response reduces the cost imposed on the importing-country group by its own protectionist barriers. However, it will compound the increase in world prices resulting from the initial price shock and the policy response by exporters. It will also add to the exporter group’s terms of trade benefits resulting from the initial upward price shock and from its own imposition of export restrictions.

Global Impacts of Price Insulation

Insulation generates a classic collective-action problem akin to when a crowd stands up in a stadium to get a better view: No one gets a better view by standing, but any that remain seated get a worse view. This collective action is, unfortunately, not just completely ineffective—it generates an international public “bad” by amplifying the volatility in the world price of the product, and hence the volatility of the income transfers associated with terms of trade changes.

To assess the implications of price insulation for a homogenous product’s international price, p*, we begin with the global market equilibrium condition:

where Si is the supply in region i, pi is the region’s producer price, vi is a random production shift variable for that region, Di is demand in region i (assumed to be not subject to shocks from year to year), and Pi is the consumer price in region i. We assume that pi = (1 + tp).p*, where tp is the distortion rate between the producer price and international price, and that Pi = (1 + tc).p*, where tc is the distortion rate between the consumer price and international price. With a focus on border measures, we can use a single variable for the power of the trade tax equivalent, T = (1 + t), where t = tp = tc.

Totally differentiating equation (17.10), rearranging it, and expressing the results in percentage change form yields the following expression for the impact of a set of changes in trade distortions on the international price:

where p^* is the proportional change in the international price, υ^i is an exogenous stochastic shock to output such as might result from better or worse weather than average, ηi is the elasticity of demand, γi is the elasticity of supply, Gi is the share at international prices of country i in global demand, and Hi is the share of country i in global production. That is, the impact on the international price of a change in trade distortions in country i depends on the importance of that country in global supply and demand, as well as the responsiveness of its production and consumption to price changes in the country, as represented by γi and ηi. With large proportional changes in trade policies and other shocks, the effects are no longer purely additive as in equation (17.11), and we need to take into account the interaction between these two proportional changes.

A notable implication of equation (17.11) is that a uniform policy response by all countries (T^ is the same for all i) will make the elasticities of supply and demand irrelevant to the impact on international prices—if all countries alter their distortions by a uniform amount, the international price will change by an exactly offsetting amount, leaving domestic prices unchanged.

If we assume that output cannot respond in the short term and that inventory levels are low enough that stock adjustments have limited effect, then γi = 0. If we further assume that the national elasticities of final demand for the product (ηi) are the same across countries, then equation (17.11) suggests that we can estimate the contribution to international price changes resulting from changes in national trade policies as simply the negative of the consumption-weighted global average of the T^is.

Incidentally, if we consider the case in which protection varies endogenously in response to changes in the international price, trade distortions are no longer an exogenous source of shocks, and international prices will change only in response to exogenous shocks such as weather-induced shocks to output. In this case, the counterpart to equation (17.11) is

where θi is the elasticity of transmission from the international price to the consumer price in country i and φi is the elasticity of transmission from the international price to the domestic producer price. Where we focus only on trade measures, such that these elasticities of price transmission are the same, it follows that the impact of price insulation on the international price is larger the smaller are those price transmission elasticities. If the short-term elasticity of price transmission is, for instance, 0.5 in all countries (a finding in line with that of Anderson and others [2010b], for key commodities such as rice and wheat since 1985 and consistent with the results in Tyers and Anderson [1992] for earlier periods), the impact of any exogenous shock on the international price will be twice as large as it would be with full price transmission.

In this situation, the variance of the international price will be four times as large as it would be in the absence of price insulation. If all countries used the price transmission elasticity of 0.15 implied by the 85 percent compensating duty under the proposed Special Safeguard Mechanism (Hertel, Martin, and Leister, 2010), then the impact of any shock on the international price would be magnified by a factor of 6.7 and the variance by a factor of 44.

The Uruguay Round agreement of the World Trade Organization attempted to address this problem by banning variable import levies and other directly insulating policies and by counting protection provided by measures involving administered prices under both the market access and domestic support measures. However, the Uruguay Round bindings on import tariffs and subsidies are at levels well above historically applied rates in most cases, providing room for countries to raise applied rates without infringing their WTO commitments. Furthermore, no effective disciplines yet apply in the WTO to variations in export restrictions. With that in mind, we turn now to seeing how much of a contribution insulating behavior of national governments had on international prices of rice and wheat in price spike periods before and after the Uruguay Round, that is, around 1974 and 2008.

The Cases of Rice and Wheat

The two food commodities that have received the most attention because of price surges are the key staples of wheat and rice. The length of their international price spikes around 1974 were broadly similar to those around 2008, but the height of the spike—particularly for rice—was greater in 1974. The recent price rises were more gradual except in the final months, so we consider an extra year in the lead-up to the 2008 spike.

Estimates of the Tis are available for all key rice and wheat countries in the form of nominal assistance coefficients (NACs) from three sources. Anderson and Valenzuela (2008) provide them through 2004 for developing countries and through 2007 for high-income countries (summarized in Anderson [2009]). They are similarly available for high-income countries for 2008 in OECD (2010). For developing countries, Anderson and Nelgen (2010b) provide estimates based on FAO and World Bank data on producer and border prices, respectively, for 2005–08. The most recent developing country estimates are less reliable than the NAC estimates in Anderson and Valenzuela (2008) for several reasons. One is that the coverage is not as extensive because domestic prices are not available for some countries. Another is that the FAO’s producer prices and World Bank international prices are not always as reliable as previously used domestic and border prices from national statistical agencies. The FAO (2010) producer prices in U.S. current dollars were converted into an index set at 100 for 2004, and the 2004 U.S. dollar prices in Anderson and Valenzuela (2008) were updated using the changes in these indices through 2008. Likewise, the Thailand 5 percent broken rice and Canadian wheat prices from the World Bank were converted to indices set at 100 for 2004, and the 2004 border prices in Anderson and Valenzuela (2008) were updated using changes in those indices through 2008.

These NAC estimates are reported in Table 17.1 for the two upward price spike periods. For each of the regions shown, as well as for the world as a whole, the patterns are strikingly similar; that is, the NAC fell as the international price rose. The proportional changes in NACs in the first half of each spike differ across products and country groups, however. As shown in Figure 17.3, the proportional change was very similar for high-income and developing countries in the 1970s spike, albeit only half as large for wheat as for rice. In the more recent spike, the proportional change for high-income countries was somewhat smaller in the case of rice and very much smaller in the case of wheat than that for developing countries.

Table 17.1Weighted Average Ti s for Rice and Wheat,1 1972–76 and 2005–08
197219731974197519762005200620072008
Rice
World1.300.930.540.900.981.331.241.150.72
High-income countries3.062.291.261.712.353.352.662.282.10
Developing countries1.030.730.450.820.871.251.191.100.66
Asia1.030.710.420.820.861.241.171.080.66
Africa0.970.600.380.660.860.991.091.290.70
Latin America and the Caribbean1.001.080.890.900.981.511.461.390.86
Wheat
World1.150.810.810.950.941.191.141.020.86
High-income countries1.110.830.800.900.921.201.171.041.03
Developing countries1.220.770.811.020.961.181.131.000.75
Asia1.330.820.881.020.941.211.151.010.70
Africa1.030.750.630.820.921.151.030.931.08
Latin America and the Caribbean0.950.570.671.111.071.021.020.970.84
Source: Anderson and Nelgen (2010b).

Weights are consumption shares for the sample countries.

Source: Anderson and Nelgen (2010b).

Weights are consumption shares for the sample countries.

Figure 17.3Percentage Changes in Tis for Rice and Wheat, High-Income and Developing Countries, 1972–74, 1984–86, and 2005–08

Source: Anderson and Nelgen (2010b).

Note: HICs: high-income countries; DCs: developing countries.

Assuming that output was able to respond only to a limited degree in the first half of each spike and that the national elasticities of demand (including stock demand) are similar across countries for each product, we set the γi s to zero and use the equation

to estimate the contribution to international price changes of price-insulating behavior resulting from national price-insulating policy behavior is the (negative of the) consumption-weighted global average change in the national Ti s. For rice, the cumulative decline shown in the world row of Table 17.1 was 46 percent between 2005 and 2008, which is in the same order of magnitude as the decline between 1972 and 1974 of 58 percent. For wheat, the globally weighted T^ was –28 percent over the 2005–08 period, compared with –30 percent in 1972–74.

According to World Bank price data, the world price of rice increased by 127 percent between 2005 and 2008, and the price of wheat increased by 114 percent. By taking the interactions between the proportional changes in trade policy and other factors into account, we can estimate the magnitude of the nontrade shocks. Comparing these with the estimated trade shocks suggests that in 2005–08, more than 45 percent of the explained change in the international price of rice was due to the changes in border restrictions that countries used in an attempt to insulate themselves from the initial increases in price. For wheat, the corresponding estimate was 29 percent. In 2008 alone, the change in protection on rice explains almost half of the 90 percent increase in rice prices observed for that year.

One important and encouraging difference between the 2008 price surge and the earlier one around 1974 is an apparent sharp reduction in the extent of price insulation in high-income countries. For rice, their NAC declined 45 percent between 1973 and 1974, whereas it fell only 8 percent between 2007 and 2008. In the case of wheat, the comparable numbers were 28 percent and 12 percent.

Although desirable, the reduction in insulating behavior by these countries has a very limited beneficial impact in the world market for rice because members not classifying themselves as developing countries in agriculture account for only 3 percent of world rice consumption. For wheat, for which these countries account for 27 percent of world consumption, the benefit is likely somewhat greater. However, it is clear from these trade shares that the key trade policy influence on the stability of world markets is what happens in developing countries.

Within the group of developing countries, there are also very substantial differences in the extent of price insulation. As shown in Figure 17.4, it appears that domestic price rises for wheat in the 2006–08 period were restrained much more in Asia than in other world regions. This suggests that, in contrast with the case considered in Figures 17.1 and 17.2, price-insulating policies may not have been completely ineffective in stabilizing prices in all regions. Rather, their effect may have been to reduce the volatility of domestic prices in some regions while increasing this volatility in others.

Figure 17.4Indices of Real International and Producer Prices of Rice and Wheat, Developing Countries’ Unweighted Average, 2006–10

(2005 = 100)

Source: FAOSTAT (FAO Statistics) producer prices (www.fao.org) and international reference prices from the World Bank’s Prospects Group (www.econ.worldbank.org).

Are rice and wheat representative of other farm products in terms of insulating behavior by governments? There is no global database for all farm products for the most recent spike period, but there is a database for the upward spike of 1974–76 and the slump of 1984–88. Anderson and Nelgen (2010a) decomposed the nominal rate of assistance (NRA) estimates for the overall agricultural sector of all 75 countries in the Anderson and Valenzuela (2008) database into the various border and domestic measures for developing and high-income countries. The annual estimates summarized for the upward-spike period of 1972–76 and the downward-spike period of 1984–88 are reported in Table 17.2.

Table 17.2Contributions to Total Agricultural NRA1 from Different Policy Instruments by Region, 1972–76 and 1984–88(Percent)
1972197319741975197619841985198619871988
a. Developing countries
Border measures
Import tax equivalent22228677898
Export subsidies4001111111
Export tax equivalent−26−18−24−22−9−20−10−14−19−22
Import subsidy equivalent−6−5−5−2−1−1−1−1−1−2
All border measures–22–21–28–16–4–14–3–6–11–15
Total NRA (including domestic measures)3–14–29–17–2–15–2–5–9–13
b. High-income countries
Border measures
Import tax equivalent25181521303334504942
Export subsidies4212224775
Export tax equivalent0−10000−1000
Import subsidy equivalent–1–3–3–1–100000
All border measures27171322313537575646
Total NRA (including domestic measures)29181324324652706959
Source: Anderson and Nelgen (2010b).Note: NRA: nominal rate of assistance.

All entries were generated by dividing the producer subsidy equivalent of all (including domestic price, non-product-specific, and “decoupled”) measures by the total agricultural sector’s gross production valued at undistorted prices.

Source: Anderson and Nelgen (2010b).Note: NRA: nominal rate of assistance.

All entries were generated by dividing the producer subsidy equivalent of all (including domestic price, non-product-specific, and “decoupled”) measures by the total agricultural sector’s gross production valued at undistorted prices.

In both of these periods, export restrictions were the dominant instrument for developing countries; they became more and then less important in the upward-spike period of 1972–76, and conversely in the downward-spike period of 1984–88. In high-income countries, there were virtually no taxes or other restrictions on exports, but export subsidies followed the same path as import tariffs over those spike periods; that is, U-shaped during the upward spike, inverted U-shaped in the downward spike. Together these estimates suggest that the experiences with rice and wheat were not inconsistent with the pattern for farm products in general, especially when bearing in mind that the NRA estimates in Table 17.2 include numerous nontradable products whose NRAs tend to remain close to zero and hence dampen year-to-year fluctuations in the aggregate estimates.

Conclusions and Policy Implications

Trade policy changes—and particularly export restrictions—are frequently discussed as contributing factors to food price surges. This chapter examines the role of trade barriers in contributing to surges. It first highlights the collective-action problem associated with the use of these measures as stabilization policies, noting that the use of these measures by all countries would be ineffective in stabilizing domestic prices, while magnifying international price instability associated with exogenous shocks to food markets. We develop a simple approach to assessing the contribution of price-insulating trade policy actions to international price changes for individual agricultural commodities and use this approach to estimate the extent to which changes in trade policy measures have contributed to price surges for the key staple foods of rice and wheat.

Our analysis shows that changes in trade policies contributed very substantially to the increases in world prices of these staple crops in both the 1973–74 and 2006–08 price surges. In the 2006–08 surge, insulating policies affecting the market for rice explain 45 percent of the increase in the international rice price, whereas almost 30 percent of the observed change in the international price of wheat during 2006–08 can be explained by the changes in border protection rates.

The evidence in Figure 17.3 suggests that at least high-income countries altered their NACs less in the most recent price spike period than in the two previous ones. That is not inconsistent with the fact that the Uruguay Round Agreement on Agriculture, which came into force with the creation of the WTO in 1995, involved commitments to bind tariffs and subsidies. The finding that developing countries are still very active users of variable border measures and especially export restrictions is also not inconsistent, given that developing country bindings are well above applied rates and that the WTO has no effective restrictions on agricultural export measures. However, more comprehensive empirical analysis over a broader range of products is needed before it would be possible to say how much of these changes can be attributed to the presence or absence of WTO disciplines.

Because bindings on import tariffs and subsidies, even for many high-income countries, were made at levels well above historically applied rates, plenty of “wiggle room” for countries to raise applied rates without infringing their commitments to other WTO members remains. Furthermore, with no effective disciplines yet being applied to export restrictions, the WTO membership has yet to address the other half of this “beggar-thy-neighbor” problem. If a special safeguard mechanism were to be introduced as part of a Doha Development Agenda agreement, the problem would be become even worse (Hertel, Martin, and Leister, 2010). An obvious solution is to seek a collective agreement to limit the extent of price-insulating policy use. Perhaps the most recent experience with price spikes in 2006–08, and again in 2010–11, will make WTO members more willing to address this issue.

References

    AndersonJ. E. and J. P.Neary2005Measuring the Restrictiveness of International Trade Policy (Cambridge, Massachusetts: MIT Press).

    • Search Google Scholar
    • Export Citation

    AndersonK.1995Lobbying Incentives and the Pattern of Protection in Rich and Poor Countries,Economic Development and Cultural Change Vol. 43 No. 2 pp. 40123.

    • Search Google Scholar
    • Export Citation

    AndersonK.ed. 2009Distortions to Agricultural Incentives: A Global Perspective 1955–2007 (Washington: Palgrave Macmillan and World Bank).

    • Search Google Scholar
    • Export Citation

    AndersonK.ed. 2010The Political Economy of Agricultural Price Distortions (Cambridge and New York: Cambridge University Press).

    AndersonK.J. L.CroserD.Sandri and E.Valenzuela2010Agricultural Distortion Patterns since the 1950s: What Needs Explaining,in The Political Economy of Agricultural Price Distortionsed. byK.Anderson (Cambridge and New York: Cambridge University Press) pp. 2578.

    • Search Google Scholar
    • Export Citation

    AndersonK. and Y.Hayami1986The Political Economy of Agricultural Protection: East Asia in International Perspective (London: Allen and Unwin).

    • Search Google Scholar
    • Export Citation

    AndersonK.M.KurzweilW.MartinD.Sandri and E.Valenzuela2008Measuring Distortions to Agricultural Incentives, Revisited,World Trade Review Vol. 7 No. 4 pp. 675704.

    • Search Google Scholar
    • Export Citation

    AndersonK. and S.Nelgen2010aHow Do Governments Respond to Food Price Spikes? Lessons from the Past,Journal of International Commerce Economics and Policy Vol. 1 No. 2 pp. 26585.

    • Search Google Scholar
    • Export Citation

    AndersonK.2010bTrade Barrier Volatility and Agricultural Price Stabilization,Policy Research Working Paper No. 5511 (Washington: World Bank).

    • Search Google Scholar
    • Export Citation

    AndersonK. and E.Valenzuela2008“Estimates of Distortions to Agricultural Incentives 1955 to 2007” (Washington: World Bank).

    BaffesJ. and T.Haniotis2010Placing the 2006/08 Commodity Price Boom into Perspective,Policy Research Working Paper No. 5371 (Washington: World Bank).

    • Search Google Scholar
    • Export Citation

    BouëtA. and D.Laborde2010Economics of Export Taxation in a Context of Food Crisis: A Theoretical and CGE Approach Contribution,Discussion Paper No. 00994 (Washington: International Food Policy Research Institute).

    • Search Google Scholar
    • Export Citation

    DeatonA. and G.Laroque1992On the Behavior of Commodity Prices,Review of Economic Studies Vol. 59 No. 198 pp. 123.

    FAO (Food and Agriculture Organization)2010FAOSTAT database. Available via the Internet: www.fao.org.

    FeenstraR.1995Estimating the Effects of Trade Policy,in Handbook of International Economics Vol. 3ed. byG.Grossman and K.Rogoff (Amsterdam: Elsevier).

    • Search Google Scholar
    • Export Citation

    FreundC. and C.Özden2008Trade Policy and Loss Aversion,American Economic Review Vol. 98 No. 4 pp. 167591.

    HertelT.W.Martin and A.Leister2010Potential Implications of a Special Safeguard Mechanism in the World Trade Organization: The Case of Wheat,World Bank Economic Review Vol. 24 No. 2 pp. 33059.

    • Search Google Scholar
    • Export Citation

    HillmanA.1982Declining Industries and Political Support Protectionist Motives,American Economic Review Vol. 72 pp. 118087.

    • Search Google Scholar
    • Export Citation

    HochmanG.D.RajagopalyG.Timilsina and D.Zilberman2010Quantifying the Causes of the Global Food Commodity Price Crisis,Policy Research Working Paper (forthcoming; Washington: World Bank).

    • Search Google Scholar
    • Export Citation

    IvanicM. and W.Martin2008Implications of Higher Global Food Prices for Poverty in Low-Income Countries,Agricultural Economics Vol. 39 pp. 40516.

    • Search Google Scholar
    • Export Citation

    JeanS.D.Laborde and W.Martin2010aFormulas and Flexibility in Trade Negotiations: Sensitive Agricultural Products in the WTO’s Doha Agenda,World Bank Economic Review Vol. 24 No. 3 pp. 50019.

    • Search Google Scholar
    • Export Citation

    JeanS.2010bPolitical Costs of Policy Reform,paper presented at the European Trade Study Group Annual ConferenceLausanneSeptember.

    • Search Google Scholar
    • Export Citation

    JohnsonD. G.1975World Agriculture, Commodity Policy, and Price Variability,American Journal of Agricultural Economics Vol. 57 No. 5 pp. 82328.

    • Search Google Scholar
    • Export Citation

    LindertP.1991“Historical Patterns in Agricultural Policy” in Agriculture and the Stateed. byC. P.Timmer (Ithaca, New York: Cornell University Press).

    • Search Google Scholar
    • Export Citation

    MartinW.1997“Measuring Welfare Changes with Distortions” in Applied Methods for Trade Policy Analysis: A Handbooked. byJ.Francois and K.Reinert (Cambridge: Cambridge University Press).

    • Search Google Scholar
    • Export Citation

    OECD (Organisation for Economic Co-operation and Development)2010Producer and Consumer Support Estimates database 1986–2009. Available via the Internet: www.oecd.org.

    • Search Google Scholar
    • Export Citation

    RoblesM.M.Torero and J.von Braun2009“When Speculation Matters” Issue Brief No. 57 (Washington: International Food Policy Research Institute).

    • Search Google Scholar
    • Export Citation

    RodrikD.1995Political Economy of Trade Policy,in Handbook of International Economics Vol. 3ed. byG.Grossman and K.Rogoff (Amsterdam: North Holland) pp. 145795.

    • Search Google Scholar
    • Export Citation

    TimmerP.2010Reflections on Food Crises Past,Food Policy Vol. 35 pp. 111.

    TyersR. and K.Anderson1992Disarray in World Food Markets: A Quantitative Assessment (Cambridge and New York: Cambridge University Press).

    • Search Google Scholar
    • Export Citation

    World Bank2008“Rising Food Prices: Policy Options and World Bank Response” paper prepared as background to the meetings of the Development CommitteeApril. Available via the Internet: siteresources.worldbank.org/NEWS/Resources/risingfoodprices_backgroundnote_apr08.pdf.

    • Search Google Scholar
    • Export Citation
The authors are grateful to Signe Nelgen for excellent research assistance. The research on which this chapter is based benefited from support by the Multi-Donor Trust Fund for Trade. The views expressed are the authors’ alone and not necessarily those of the World Bank. This chapter builds on World Bank Policy Research Working Paper No. 5645.Will Martin is Research Manager of Agricultural and Rural Development of the Development Research Group at the World Bank; Kym Anderson is George Gollin Professor of Economics at the University of Adelaide, Australia.
1If quantitative export restrictions were imposed instead, the rights to export would become valuable, with the holders of the export rights receiving the benefits that would accrue to the government if an export tax had been used.
2See Martin (1997) for approaches to measuring this welfare impact.
3In the current context, h depends on the average rate of protection.

    Other Resources Citing This Publication