University of Amsterdam
MSc Thesis Economics
Did food exporting nations’ policies affect global food prices?
Author: Supervisor: Fan Zhang Drs. Naomi J. Leefmans Studentnumber: 10390294 Second reader: Dr. Dirk Veestraeten
Abstract
In this thesis a 2SLS model has been applied, to empirically investigate whether or not food-exporting countries’ policy distortions to rice, wheat and maize producer incentives during recent years have causal effects on the increasing global rice, wheat and maize prices. The testing period was from 1990 to 2011, and a panel data method was used in order to get a sufficient sample size. Policy distortions to food producer incentives are claimed by recent relevant literature to be associated with the global food crisis, and the thesis has investigated the significance of the relationship. The main findings indicated that both rice and maize exporting countries’ policy distortions to rice and maize producer incentives significantly contributed to a higher global rice and maize prices, while the policy distortions to wheat producer incentives in wheat exporting countries were not significantly responsible for the increasing global wheat prices.
Acronyms
2SLS Two-stage Least Squares ADF Augmented Dickey-Fuller ARDL Autoregressive Distributed Lag
BIC Schwartz’s Bayesian Information Criterion CPI Consumer price index
DAI Distortions to Agriculture Incentives EU European Union
FAO Food and Agriculture Organization IMF International Monetary Policy NRA Nominal Rate of Assistance
OECD Organization for Economic Cooperation and development PAE Poor agrarian economy
RIE Rich industrial economy UR Uruguay Round
VAR Vector Autoregressive
VECM Vector Error Correction Model WBG World Bank Group
Table of Contents
Chapter 1 ... 5
Introduction ... 5
Chapter 2 ... 8
Literature Review... 8
2.1. Theories from existing literature ... 9
2.2. Empirical research of existing literature... 13
Chapter 3 ... 17
Data and Methodology ... 17
3.1 Data Description ... 18
3.1.1. Nominal Rate of Assistance (NRA) ... 18
3.1.2. Other Variables ... 20
3.2. Sample Description ... 21
3.3. Methodology ... 25
Chapter 4 ... 27
Empirical analysis ... 27
4.1. Effects of policy distortions to rice producer incentives ... 28
4.2. Effects of policy distortions to wheat producer incentives ... 30
4.3. Effect of policy distortions to maize producer incentives ... 31
4.4. A focus on rice-exporting countries ... 32
4.5. Conclusions of the empirical analysis and further research ... 34
Chapter 5 ... 35
Conclusion ... 35
Reference ... 37
Chapter 1
Introduction
Global food prices experienced an upward spike on record in late 2007 to early 2008, and stayed at a high level afterwards. High food prices have strong negative effects on poor households who spent most of their income on food staples, since they barely have mechanisms to cope with the soaring food prices at their disposal. These households were hardest hit by the food crisis, and they reacted by reducing the quantity and quality of food consumed. In addition, they cut back money spend on basic needs and investments in human capital such as education or health. Short-term threats caused by high food prices could thus have long-run effects on poverty and development.
High food prices not only made the poor households worse-off, but also had an overall negative impact on the economy. Decline in consumption due to weak purchasing power led to a less active market. The World Bank Group had estimated that around 44 million people were put into poverty and food insecurity due to the 2008 food crisis. Such a large amount of unsatisfied population can fuel political unrest, and turmoil related to high food prices have already happened in many developing countries including Egypt, Mexico, Morocco, Uzbekistan and so on1. What’s more, soaring food prices imposes pressure on the overall price level, especially in developing countries, where food has a large share of the consumer price index (CPI). IMF (2008) claimed that high food prices accounted for a 26.3
1
‘High Food Prices: Impact and Recommendations for Actions’ Paper prepared by FAO, IFAD and WFP for the meeting of the Chief Executives Board for Coordination on 28-29 April 2008, Bern,
percentage increase of the headline inflation all over the world, and 44.6 percentage increase of the headline inflation in poor African countries.
The relatively low global food prices for the passing two decades came to an end in late 2007, and fighting with increasing food prices to ensure food security became more important and emergent for all the countries no matter whether they are poor, rich, importers, or exporters. To deal with soaring global food prices, governments need to first figure out the determinants of the high food prices. There are some factors claimed by recent relevant literate to be responsible for the increasing global food prices. The first one is the extreme weather. The world food supply is suffering more calamity and higher volatility due to more frequent droughts, floods and other types of extreme weather (Anderson et al., 2009). Russia’s wildfires, Pakistan’s flooding, unseasonable heat waves in Australia and United States etc., These serious weather conditions occurred in many major food-exporting countries, which resulted in a worst corn crop supply in late 2007 and contributed to high global food prices on record. The second factor is the increasing demand for biofuel productions. According to FAO’s Statement of food and agriculture 20132, 37 percent of grain production was used to make the ethanol production in the United States and 80 percent of vegetable oil production went into biodiesel in the European countries. The largest part of the new demand for agricultural production is from biofuels, which are considered to drive up price levels of wheat and maize (Chau, J.P. 2008; Chen,S., and M. Ravallion. 2007). The third factor that is responsible for the increasing food prices is the rising oil prices. In the past decade, world food prices and oil prices moved up and down more or less in the same direction (Appendix 1). Modern agriculture uses farm machines, which are fueled by oil, to enhance productivity; and transporting agricultural products to ultimate consumers is also an oil-consuming process. Therefore an increasing oil price puts pressures on the costs of food productions and thus, food prices can be driven up. There is a consideration that the volatility and fluctuations of crude oil prices may lead to a further increase in global food prices (Bloomberg, 2011). The fourth factor is policy distortions. Besides these factors mentioned above, many papers (Anderson et al., 2009, Anderson and Valenzuela, 2010; Anderson and Nelgen, 2012; Borchert, Gootiiz and Mattoo, 2012; Bouet and
Laborde, 2010;. Chapoto, Anton and Jayne, 2009) indicated that policy distortions to food producer incentives in food exporting countries also had effects on the global food prices. When world food prices were increased by some exogenous shocks such as extreme weather or increasing oil prices, individual countries would respond to the initial increase in global food prices by imposing policy distortions within domestic markets. More specifically, food-exporting countries may increase export taxes in order to reduce food exportation and thus to ensure a sufficient domestic food supply and low domestic food prices. Food importing countries on the other hand may for instance reduce import tariffs so as to lower the price of imported foods and provide domestic customers a reasonable food prices. According to FAO’s (2008)3, in late 2007 there were 68 countries that imposed distortions to food producer incentives including import subsidies, export controls and tariff reductions etc., to stabilize domestic food prices when global food prices were soaring. Policy distortions to food producer incentives are countries’ reactions to the initial increase in global food prices that was caused by exogenous shocks. However, these policy distortions are claimed by recent relevant literature to be responsible for driving up global food prices to an even higher level. Director General of the WTO, Pascal Lamy mentioned in 2011 “…export restrictions play a direct role in aggravating food crises… Each individual country had unilateral incentives to impose policy distortions in the domestic market but as a whole the world was not better off, and the global food prices were driven up.”
The thesis focuses on three main food products, rice wheat and maize, and aims at answering the question whether or not these food-exporting countries’ policy distortions to rice, wheat and maize producer incentives have had causal effects on the increasing global rice, wheat and maize prices. The proxy variable for policy distortions to producer incentives of different food products, which will be used in this thesis, is the ‘Nominal Rate of Assistance’ (NRA) that was constructed by Anderson et al., in 2007. A 2SLS model will be applied to investigate the effect of NRARice,t, NRAWheat, t, NRAMaize,t on the global rice, wheat and maize prices.
3
Agricultural Policies in OECD Countries: Monitoring and Evaluation 2007. Paris: OECD. OECD and FAO (Food and Agriculture Organization). 2008. OECD-FAO Agricultural Outlook 2008– 2017. Paris:
Existing literature does mention the correlation between policy distortions to food producer incentives and the rising global food prices during recent years, but no causality tests have been done of the effect for particular food products such as rice, wheat or maize. In addition, previous relevant literature (Anderson & Neary, 2005; Anderson, Kurzweil, Martin, & Sandri, 2008; Anderson & Valenzuela, 2008; Mensbrugghe, 2009) used NRA data only up to 2009. This thesis tries to make a difference by using an updated NRA database, which was extended to 2011, and to investigate the causal effect of policy distortions on global rice, wheat and maize prices.
The remainder of this thesis is structured as follows: Chapter 2 is the literature review part, which firstly summarizes the theories used to analyze the correlation between policy distortions to food producer incentives and global food prices; and provides the empirical findings about the correlation from relevant literature. Chapter 3 introduces the data and methodology that will be used in the empirical analysis. Chapter 4 presents the regression results and discusses the implications, and Chapter 5 draws a conclusion by highlighting the implications of the obtained results.
Chapter 2
Literature Review
This chapter firstly, in section 2.1, summarizes the theories that were used in previous literature to analyze the impact of policy distortions to food producer incentives on global food prices. Secondly, in section 2.2, the chapter presents several empirical studies, which confirmed that policy distortions to food producer incentives were associated with the soaring global food prices.
2.1. Theories from existing literature
Will Martin and Kym Anderson (Oct 2010) suggested that in the past few years, many countries sought to insulate their domestic markets from international food price shocks by imposing policy distortions to producer incentives such as trade distortions (export taxes/import tariffs). However, these distortions didn’t effectively prevent domestic food prices from rising, but turned out to result in an even higher global food prices. Signe Nelgen and Kym Anderson (2012) hold a similar idea, and they discussed the following three situations to analyze the correlation between policy distortions to food producer incentives and the global food prices.
In the first situation, Will Martin and Kym Anderson (Oct 2010) assumed that when exogenous shocks such as extreme weather initially drove up the global food prices, exporting countries were the only country groups that responded through policy distortions (in this case, export taxes). By imposing these policy distortions, exporting countries aimed to reduce food exportation in order to provide the domestic market a sufficient food supply and to keep domestic food prices at a low level. Figure 1a below indicates that in the beginning, the equilibrium world food price was P. If exporting countries impose export taxes, food producers will face higher exporting costs and thus, they export less. The available global food supply will decrease, so the supply curve will shift to the left (from S to S’), which leads to an even higher world food price at P’’. The policy distortions (export taxes in this case) could lower food-exporting countries’ domestic food prices from P to P’. The social costs of this policy distortions is area abc, which can be divided into a loss acPP” to food importing countries, a loss abPP’ to food exporting countries, and tax revenues bcP’P” to exporting counties. Therefore, the importing countries are definitely worse off, but the result for exporting countries is ambiguous, whether they get better off depends on their terms of trade gains and social costs. Anderson mentioned that4, “ the social costs rise with the square of the export tax, while the terms of trade gain is likely to rise linearly with the rate, the benefits to the exporter group will become negative if the export tax rate becomes sufficiently large”. In conclusion, under the first situation,
4 Anderson, K. and S. Nelgen (2012a), ‘ Agricultural Trade Distortions During the Global Financial
exporting countries’ policy distortions to food producer incentives could lower domestic food prices but contribute to higher global food prices.
Figure 1a Food exporting countries react to the rising food prices (World market food supply and demand)
Source: Signe Nelgen and Kym Anderson (2012)
In the second situation, authors assumed that when an exogenous stuck stroke the global food market and enhanced food prices, the importing countries were the only country group that imposed policy distortions. In order to insulate the domestic market from the rising global food prices, importing countries reduced import tariffs to lower the price of imported foods and thus, to provide domestic customers with low food prices. These strategies could increase domestic demand for imported food products, so the demand curve D in figure 1b will shift to D’. The global food prices is pushed up from p to p’, while the domestic food price of importing countries is lowered from p to p”. Although consumers in importing countries can enjoy a relatively lower domestic price, these nations suffer from a higher terms of trade loss, in addition, tariffs can’t be reduced forever, if tariffs went to zero, importing countries may need to depend on import subsidies to attract food imports. In conclusion, under this situation, policy distortions lowered the importing countries’ domestic food prices but contributed to a higher global food prices.
Figure 1b Food importing countries react to the rising food prices (World market food supply and demand)
Source: Signe Nelgen and Kym Anderson (2012)
The third situation is most realistic; authors assumed that both exporting and importing countries responded to the exogenous shocks that pushed up global food prices. Figure 1c shows that distortions from exporting countries makes the S curve shift to S’, and importing countries immediately react by lowering import tariffs, which makes the demand curve move to the right. The new world price is P”’, it is much higher than the initial price P, and is also higher than P” and P’, which were discussed in the first and second situation. When both exporting and importing countries try to insulate domestic markets from the rising global food prices by imposing policy distortions to producer incentives simultaneously, the impacts of these distortions on the domestic price from two groups offset each other. In other words, policy distortions didn’t lower domestic food prices in both country groups, the domestic price after distortions stays at P, which is the same as the price without any policy distortions. In conclusion, when many countries impose policy distortions to producer incentives at the same time, these strategies can’t help individual countries to lower their domestic food prices, but can lead to higher global food prices.
Figure 1c both exporting and importing countries react to the rising food prices (World market food supply and demand)
Source: Signe Nelgen and Kym Anderson (2012)
The three situations discussed above all indicated that policy distortions to producer incentives could result in higher global food prices.
During the food crisis years, exporting countries from advanced economies imposed policy distortions less aggressively compared with exporting countries from developing regions and thus, one could expect the impacts of developed countries’ policy distortions on global food prices were less significant than the effects of developing countries’ policy distortions. Anderson, Martin, van der Mensbrugghe (2009) discussed some reasons why different exporting countries chose different policy distortions to producer incentives. Their research showed that compared with rich industrial economies, poor agrarian economies were less likely to protect domestic food producers but rather acted on behalf of urban households (consumers) when global food prices were increasing. Governments from less developed regions are likely to impose policy distortions in order to reduce food exportation and thus, to provide domestic market a sufficient food supply and to keep domestic food prices from rising. The paper suggests that in a poor agrarian economy (PAE), farm products represent a large part of urban household expenditures and therefore the real wages of urban employees are strongly affected by food prices. Compared with farmers who are numerous but usually small groups and poorly organized, urban households have more power in influencing the government’s policy decisions, therefore governments
in PAEs are likely to impose policy distortions in the interest of urban households rather than farmers. However, in a rich industrial economy (RIE), much more workers are hired in industrial sectors other than farm jobs; intervening in the price of agricultural products relative to industrial products has a smaller effect on the price of mobile workers in a RIE than that in a PAE, therefore RIE governments face less opposition to protect domestic farmers. In addition, farmers in more developed countries are better organized and have bigger influence in governments decision-making process compared with their counterparts in developing economies. Therefore, when global food prices are soaring, compared with developing nations, exporting countries from advanced economies are less likely to impose policy distortions to producer incentives that aims at reducing exportation and lowering domestic food prices.
The above theories from existing literature highlighted that firstly, exporting countries’ policy distortions to food producer incentives could result in a higher global food price. Secondly, compared with exporting countries from developing regions, exporting countries from advanced economies imposed policy distortions to producer incentives less aggressively when global food prices were soaring, In chapter 4 of the thesis, an empirical test confirms that the impacts of developed countries’ policy distortions on global food prices were less significant than the effects of developing countries’ policy distortions.
2.2. Empirical research of existing literature
Anderson et al. (2007) introduced a proxy variable ‘Nominal Rate of Assistance (NRA)’ to measure the policy distortions to producer incentives. The nominal Rate of Assistance estimates the percentage by which the gross returns to food producers have been raised above or lowered below the level without government policy distortions; in other words it measures the extent to which the domestic food producer price is higher (if NRA is positive) or lower (if NRA is negative) than the border price of a like product without any policy distortions.
The NRA includes the following components: (1) NRAborder: policy distortions to
producer incentives at the border, which refers to trade policy distortions such as export taxes or import tariffs; (2) NRAdomestic: policy distortions to producer incentives
in the domestic market, such as production taxes or subsidies; (3) NRAexchange:
distortions to producer incentives that are caused by exchange rate system when food products are traded internationally; (4) NRAinput: distortions from the intermediate
inputs, which means that food producer incentives will be affected when intermediate inputs are taxed or subsidized; (5) NRAindirect: indirect policy distortions to producer
incentives that come from other sectors than agriculture. By adding up all the components, policy distortions to producer incentives are quantified, and Anderson et al. (2007) constructed a NAR database5, which was updated in 2011 and provided NRA data of 75 food products for 82 countries available from 1955 to 2011. The thesis will use NRARice, NRAWheat, NRAMaize as proxies to measure policy distortions
to rice, wheat, maize producer incentives and thus, test the causal effects of these distortions on soaring global rice, wheat, maize prices separately. Chapter 3 will present a more detailed discussion about what NRA is and how to calculate it.
Anderson (2007) claimed that when countries imposed policy distortions to producer incentives in order to keep domestic food prices at a lower level than global food prices, the NRA would start to decline and even became negative if these distortions were very aggressive and domestic prices were indeed reduced to a lower than global price level. Declining NRA thus means increasing policy distortions that aims at lowering domestic prices, since these policy distortions are claimed to be positively associated with increasing global food prices, thus NRA is expected to be negatively related to soaring world food prices (figure2).
5 Anderson et al., (2011) ‘Estimates of Distortions to Agricultural Incentives, 1955-2011’, world bank
working paper.
http://econ.worldbank.org/WBSITE/EXTERNAL/EXTDEC/EXTRESEARCH/0,,contentMDK:219600 58~pagePK:64214825~piPK:64214943~theSitePK:469382,00.html
Increasein global food prices
Decline in NRA Increasein policy distortions to producer incentives that aims at lowering domestic
food prices
Means Relate to
Figure 2. Expected correlations between NRA, policy distortions to producer incentives, and global food prices
Kym Anderson (2009)compared NRA data with global food prices during the period 1950-2009 and indeed found a negative correlation between NRA and global food prices. Empirical findings indicated that when countries imposed policy distortions to producer incentives in order to keep domestic prices from increasing global food prices, NRA started to decrease, and NAR in many food-exporting countries became negative during food crisis period 2007-2008. Anderson also compared NRA of developing countries and that of developed countries, and he claimed that advanced economies were less likely to insulate the domestic food market from rising global food prices by imposing policy distortions to producer incentives; while less developed countries reacted more aggressively to the soaring global food prices and imposed more policy distortions to producer incentives compared with their developed counterparts. Chapter 4 of the thesis will show empirical findings that wheat exporting countries’ policy distortions to producer incentives didn't have significant effects on the increasing global wheat prices, while rice exporting countries’ policy distortions were significantly responsible for the soaring global rice prices. Major wheat exporting countries are from the developed world, while major rice exporting countries are less developed economies. Because developed countries were less likely to impose policy distortions to producer incentives to fight increasing global food prices compared with their developing counterparts and thus, policy distortions from food exporting developed countries didn’t significantly affect global food prices while policy distortions from less developed regions did.
Anderson, Ivanic and Martin (2011) used the NRA to proxy the policy distortions to producer incentives and investigated the relationship between policy distortions and the soaring global food prices. Authors found the similar results as many studies mentioned that when global food prices were initially enhanced by exogenous shocks such as extreme weather, policy distortions to producer incentives would lead to an even higher global food prices. Individual countries have unilateral concerns to impose policy distortions but the expected benefits of these distortions were partially offset by the higher global food prices. Authors also claimed that when individual countries imposed trade restrictions to fight against the higher global food prices, those countries that distorted producer incentives more aggressively could enjoy a relatively lower domestic food price at the cost of other countries’ higher domestic food price. Developing countries as a whole imposed policy distortions more
aggressively than advanced economies and thus, the higher international prices are ‘exported’ to developed countries. However, this transfer of higher food prices also happened among developing countries. China and India for instance, as two big developing economies, distort their domestic food markets more frequently than the rest of the developing world, therefore domestic food prices in these two countries stayed in a relatively lower level compared with that in other countries. Although these policy distortions to producer incentives helped to keep some countries’ domestic food prices at a relatively lower level, these distortions as a whole could lead to the increasing global food prices.
Ruta,Rocha and Giordani (2011) focused on trade policy distortions especially export restrictions, and they tested the effects of export distortions on global food prices. Authors hold the belief that when an exogenous shock such as extreme weather or soaring oil price drove up global food prices, if many individual countries react by imposing export restrictions, the world food price can be enhanced to an even higher level. The higher prices will attract more other countries to respond similarly, which in turn exacerbates the world food prices and induces further export restrictions. They set up a 2SLS model to test the effect of export distortions on global food prices. In Pkt =α0 +α1In ERkt+α2 In Rainkt +α3 In Rainvarkt +α4 In Energykt + δk + λt + εkt,
In ERkt = β0 +β1 In Pkt +β2 In Electionskt + δk + λt + ϑkt,
Where P stands for global food prices and ER is short for export restrictions, rainfall, variance in rainfall and energy price were used as control variables in the first stage. Authors used Elections as an instrument for ER in the second stage to reduce the threat of reverse causality, since global food prices and export restrictions affect each other. They indicated that from the political economy point of view, people own specific factors are more easily organized as an interest group, and are more likely to be politically formed together to lobby the government. For instance, they stated in their paper that Lobby groups that representing producers’ interests will generally oppose (support) the implementation of export restrictions (subsidies). Governments weigh more about organized interests, and their trade policy distortions are more likely to be affected by different lobby groups in the proximity of elections in order to gain votes. In conclusion, authors believe that during election period, export policies are influenced by different lobbying groups, but elections don't directly have impacts
on global food prices and thus, elections is a reasonable instrumental variable for export policies.
In their paper, authors found empirical evidence that individual countries’ exporting restrictions significantly contributed to the increasing global food prices during 2006 to 2008. The thesis will apply a similar 2SLS to test whether or not the food exporting countries policy distortions to food producer incentives caused the increasing global food prices. NRA would be the proxy to measure policy distortions and Elections would be the instrument variable for NRA to reduce the threat of reverse causality. Chapter 3 of the thesis will have a detailed discussion about these variables and methodology.
This Chapter firstly summarized theories why policy distortions to producer incentives especially distortions imposed by exporting countries could be positively associated with the increasing global food prices; and why developed countries imposed much less policy distortions to fight the soaring global food prices compared with their developing counterparts. Secondly this chapter introduced some methodologies to proxy policy distortions to producer incentives and to test the effect of trade policy distortions on global food prices. In addition this part summarized some empirical findings of existing literature to show that food exporting countries’ policy distortions to producer incentives were indeed positively related to the increasing global food prices, and that developed countries imposed less distortions to insulate the domestic market from the soaring global food prices compared with developing countries.
Chapter 3
Data and Methodology
This chapter will present the empirical analysis and test whether or not food-exporting countries’ policy distortions to rice, wheat and maize producer incentives caused the increasing global rice, wheat and maize prices. Existing literature mentioned the
correlation between policy distortions to food producer incentives and rising global food prices, but there were no causality tests of the effect for particular food products such as rise, wheat or maize. In addition, previous relevant literature used data updated to 2009, while the thesis extended the research period to 2012 and thus, complements the existing literature.
This chapter will firstly, in section 3.1, present all the variables that are used in the empirical analysis, and a detailed introduction of the Nominal Rate of Assistance (NRA) will be included. The second part of this chapter, section 3.2, will present the sample, and several tests such as normality test will be done to make sure the sample is appropriate to do the empirical analysis. Third part of this chapter, section 3.3, will introduce the methodology and set up the model.
3.1 Data Description
3.1.1. Nominal Rate of Assistance (NRA)
In order to test the effects of exporting counties’ policy distortions to producer incentives on global food prices in a regression, one needs to first find a proxy to quantify these policy distortions. In some earlier articles, Bhagwati (1971) described that these policy distortions created a gap between the marginal social costs paid by buyers and marginal social returns received by producers. Harberger (1971) pioneered to estimate these social economic costs caused by distortions through welfare methods, and he noted that the estimation of a policy distortion is the gap between the price received by producers and the price paid by customers in the case of no divergences6. Anderson et al (2007) introduced a variable ‘Nominal Rate of Assistance’ (NRA) as a proxy to estimate the policy distortions to food producer incentives. The Nominal Rate of Assistance estimates the percentage by which the gross returns to food producers have been raised above or lowered below the level that without government policy distortions; in other words it measures the extent to which the domestic food producer price is higher (if NRA is positive) or lower (if
6 Anderson, K., M. Kurzweil, W.Martin, D. Sandri, and E. Valenzuela. 2008a. “Methodology for
Measuring Distortions to Agriculture Incentives.” Agricultural Distortions Working Paper 02, World Bank, Washington, DC.
NRA is negative) than the border price of a like product without any policy distortions. When countries imposed policy distortions to food producer incentives in order to keep domestic food prices at a lower level than the global food prices, NRA would decline and become negative if these distortions were aggressive and reduced domestic prices to a lower than global food price level.
When calculating NRA, the authors took into account the following main components:
1) Distortions at the national border. These distortions mainly refer to trade distortions such as import tariffs or export restrictions, which can be calculated as NRA covered by border price support (NRAborder).
2) Distortions within the domestic market. These distortions refer to policy distortions to producer incentives imposed by individual countries within their domestic markets, such as production taxes or production subsidies. These kinds of distortions can be calculated as NRA covered by domestic price support (NRAdomestic). Adding up
NRAborder and NRAdomestic, we get the total nominal rate of assistance to output
(NRAoutput).
3) Distortions from the exchange rate system. These distortions occur when exchange rate policies affect the prices of tradable food products. Anderson (2007) noted that in the case of multiple exchange rates system, the distortions to domestic food producer incentives caused by foreign currency depend on the particular exchange rate applied to that goods each year, and these distortions can be calculated as NRA covered by exchange rate distortions NRAExchange.
4) Distortions to inputs in food production. These distortions occur when farm production includes intermediate inputs, which are taxed or subsidized. Incentives of producers are affected by these policy distortions, therefore when computing NRA these factors should be taken into considerations. These distortions can be calculated as NRA covered by distortions to input NRAinput.
5) Indirect policy distortions to producer incentives. Base on Lerner (1936) Symmetry Theorem7, Anderson (2009) claimed that indirect distortions via non-agricultural
sector also needed to be taken into considerations when calculating total NRA. For example, if government invested more in non-agricultural sectors, one could expect the value of mobile resources that would otherwise have been employed in agricultural sector are bided up, there would be fewer resources in agricultural sector compared with the situation under free market case. So NRAindirect is calculated to
cover these indirect policy distortions to food producer incentives.
By adding up all these 5 components, policy distortions to producer incentives are quantified as NRA. Anderson et al (2009) built a ‘Distortions to Agricultural Incentives’ (DAI) database8 for World Bank, which was updated in 2011 and provided NRA data for more than 75 food products in 82 countries during the period 1950-2011. The thesis uses the total NRA as a proxy of policy distortions to producer incentives.
3.1.2. Other Variables
Except the independent variable NRA, the following variables are also included in the analysis. The global food price is the dependent variable that is taken from the World Bank database. Control variables are chosen base on the model developed by Ruta, Rocha and Giordani (2011) and thus, the energy price level as well as rainfall level are controlled for. Many literature (Eberstadt, 2007; Pardey, Alston, and Jones, 2008; Schnepf, 2008) claimed that population growth has effects on global food prices since it could enhance the demand for food, therefore, population growth is included in the regression to avoid omitted variable bias. Since policy distortions to producer incentives (proxied by NRA) and global food prices could affect each other, there exists the threat of reverse causality. To get unbiased analysis results, an instrumental variable ‘Electionsit’ is applied to instrument NRA. Data of variable ‘Electionsit’ is
found in individual countries’ department of statistics. Electionsit equals to 1 when the
government is in the proximity of election, otherwise it equals to zero. More detailed discussion about the model will be presented in section 3.3 of this chapter.
8 Anderson et al., (2011) ‘Estimates of Distortions to Agricultural Incentives, 1955-2011’, world bank
working paper.
http://econ.worldbank.org/WBSITE/EXTERNAL/EXTDEC/EXTRESEARCH/EXTPROGRAMS/EX TTRADERESEARCH/0,,contentMDK:21012395~pagePK:64168182~piPK:64168060~theSitePK:544 849,00.html
3.2. Sample Description
This chapter includes a description of the sample that is going to be investigated in the empirical analysis. Figure 3 shows that in the past four decades, the global food prices level stayed relatively stable. This thesis chose a sample from 1990 to 2012, because these two decades was a period that global food prices started to rise and experienced a significant increase (2007-2008 and afterwards); and also a period that individual countries imposed policy distortions to fight increasing food prices (Martin, Will, and Anderson, 2012; Van der Mensbrugghe, Valenzuela and Anderson, 2010.) Figure 4 shows the changes in NRAs, one can see the governments increased policy distortions gradually and imposed aggressive distortions around late 2007. NRAs were shown negatively associated with the food prices and thus, policy distortions to producer incentives could be expected to be one of the factors that increased the global food prices during these years.
Figure 3 Global food prices during 1970 to 2014 ( in dollars)
Source: FAO database
Figure 4 the trend of NRAs for rice, wheat and maize
Source: World Bank DAI database 0,0 50,0 100,0 150,0 200,0 250,0 19 70 19 75 19 80 19 85 19 90 19 95 20 00 20 05 20 10
Global food prices
-2,5 -2 -1,5 -1 -0,5 0 0,5 1 1,5 2 19 90 19 92 19 94 19 96 19 98 20 00 20 02 20 04 20 06 20 08 20 10 20 12 NRA rice NRA wheat NRA maize
There are three sub-samples. The first one includes the top 10 rice-exporting countries: Thailand, Vietnam, Pakistan, India, Cambodia, Uruguay, China, Egypt, Argentina, and Brazil. According to the World Bank database, rice exportation from these countries accounted for more than 95.7% of the total global rice exportation therefore these countries can represent rice-exporting countries as a whole and thus, the sample can be seen as a proper sample to test the causal effect of rice exporting countries’ policy distortions on global rice prices.
The second sub-sample includes top 10 wheat-exporting regions: United States, UK, France, Germany, Canada, Australia, Russian Federation, Ukraine, Turkey and Kazakhstan. World Bank database shows that wheat exportation from these regions accounted for more than 97.2% of the global wheat exportation. These countries can represent wheat-exporting countries as a whole, and thereby the sample can be used to test whether or not wheat-exporting countries policy distortions to producer incentives contributed to an increasing global wheat prices.
The third sub-sample includes 10 maize-exporting countries: United States, Brazil, Argentina, Ukraine, India, France, Russian Federation, South Africa, Paraguay and Serbia. Maize exportation from these countries accounted for more than 94.9% of the global maize exportation. Thus the sample is seen as a valid sample to do the empirical test for the relationship between policy distortions to maize producer incentives and global maize prices.
As discussed in chapter 2, several studies stated that when individual countries imposed policy distortions to producer incentives in order to fight with the increasing global food prices, the NRA started to decline, and the NRA was claimed to be negatively associated with global food prices. Figure 5a, 5b and 5c present the correlation between NRARice, NRAWheat, NRAMaize and global rice, wheat and maize prices for the period 1990 to 2012.
Figure 5a correlation between Rice NRA and global Rice Prices (dollars)
Source: computed base on updated DAI database, FAO annual rice price
Figure 5b correlation between wheat NRA and global Wheat prices (dollars)
Source: Computed based on updated DAI database, FAO wheat price data
Figure 5c correlation between wheat NRA and global maize prices (dollars)
Source: Computed based on updated DAI database, FAO wheat price data y = -479,09x + 545,52 R² = 0,7456 0 200 400 600 800 -0,5 0 0,5 1
Effect of Rice NRA on Rice Price y = -244.37x + 228.58 R² = 0.54366 0 100 200 300 400 -0,2 0 0,2 0,4
effect of wheat NRA on wheat price y = -254.727x + 118.18 R² = 0.69535 0 100 200 300 -2 -1 0 1
Correlation between NRA wheat and global Maize
All figures indicated that the NRAs were strong negatively correlated to World food prices. More specifically, one unit decrease in NRARice is associated with 4.799
increase in global rice price; one unit decrease in NRAWheat is related to 2.43 increase
in global wheat price; one unit decrease in NRAMaize indicates 2.54 increase in global
maize prices. A decline in NRA means that there was an increase in countries’ policy distortions to producer incentives that aimed at keeping domestic food prices at a lower level. Since decreasing NRA is associated with increasing global food prices, one can expect that increase in policy distortions to food producer incentives is associated with increase in global food prices (figure 6), and the next chapter will test whether or not this relation is significant.
Figure 6: Expected relations among NRA, policy distortions, and global food prices
Source: Anderson et al (2009)
Another finding from these above figures was that the correlation coefficient of NRArice was the largest and that of NRAwheat was the smallest. One explanation could
be that main rice exporting countries are from the developing world such as Vietnam, Thailand, India; while main wheat exporting countries are advanced economies, such as US, EU countries, Australia, and Canada. As Anderson (2009) mentioned, since 1976 developed world used less policy distortions to disturb the domestic food producer incentives compared to their developing country counterparts, therefore the smaller magnitude of the correlation coefficient of NRAwheat, was in line with the fact
that advanced economies imposed policy distortions less aggressively. 9Δy= -1*(-479%)=4.79 M ea ns Decrease in NRA Increase in Policy distortions Increase in global food prices
To ensure that all the variables in the sample can be included in the analysis, the following tests were performed to check the data set before running the regressions.
Tests Results
Test for Outliers* No outliers were detected in all the variables.
Test for Multicollinearity** The highest bivariate correlation is about 0.45 between NRA and rainfall level. Multicollinearity occurs when the bivariate correlation is close to 1, so one can expect that there is no serious Multicollinearity problem in the data set. Test for Normality *** All the included (scale) variables are
normally distributed.
Note: * Tukey’s Range Test (Appendix 2). The upper bound = 3rd quarter percentile +(2.2*(3rd quarter percentile-1st quarter percentile), the lower bound = 1st quarter percentile –(2.2*(3rd quarter percentile-1st quarter percentile). Outliers are those values smaller than the lower bound or higher than the upper bound.
**Appendix 3 is the correlation matrix, which shows the bivariate correlation for all variables.
***The following tests are conducted: One-Sample Chi-Square Test for nominal variable, and One-Sample Kolmogorov-Smirnov Test for scale variables. The null hypothesis is that the distribution of particular variable is normally distributed. Appendix 4 shows that we can’t reject the null hypothesis.
In conclusion of these tests, all the dependent variable, independent variables and control variables in the sample data (1990-2012) can be used in the empirical analysis.
3.3. Methodology
2SLS models are used in the thesis to test the causal effects of policy distortions to rice, wheat, and maize producer incentives on global rice, wheat, maize prices. Because policy distortions and global food price affect each other, an instrumental variable approach can help to reduce the threat of simultaneous causality. The sample period of the analysis is from 1990 to 2012. Since it’s not a long time series, the thesis adopted panel data to get a much bigger sample size. The advantages of using panel data are first, it offers researchers a large number of observations for a limited time period across different sections; second, it enhances the degrees of freedom and also alleviates the collinearity among dependent variables.
In Pt = α0 + α1NRAit+ α2In Rainit +α3In EnergyPit + α4 PopulationGit +λi + φt + θit (1)
NRAit = β0+β1 Electionsit + λi + φt + πit (2)
Where in equation (1), In Pt corresponds to the global food price in time t. NRAit is the
nominal rate of assistance in country i during time t, which is a proxy for policy distortions to producer incentives. In Rainit stands for the rainfall level, In EnergyPit
measures the energy price level and PopulationGit estimates the population growth in
country i during time t. In equation (2), Electionsit is the dummy variable, it equals to
one if country i is in during the time of election. λi is the state fixed effect for a set of
food exporting countries, φt is the time fixed effects and θkt measures the error term.
The model is based on a similar 2SLS model that was introduced by Ruta,Rocha and Giordani (2011). As discussed in chapter 2, these authors used a 2SLS model to test the causal effect of trade policy distortions on global food prices. In their first stage regression, rainfall level and energy price level were chosen to be the control variables. In the second stage regression, trade policy distortion was instrumented by ‘election’. This paper adds a third control variable in the fist stage regression, namely the population growth. It is necessary to be controlled since it could also have a direct effect on global food prices. (Bellemare and Marc, 2011; Crafts and Toniolo, 2008; Chau.our, 2008.) One can expect that the coefficient associate with NRAit will have a
negative sign. As discussed in chapter 2, when individual governments impose policy distortions to producer incentives, in order to keep domestic food prices from rising under the pressure of soaring global food prices, the NRA started to decrease. The decline in NRA means the increase in policy distortions, and the higher policy distortions the higher global food prices. Therefore a decline in NRAit is related with
an increase in In Pt and thus, one can expect a negative sign of the coefficient in front
of NRAit.
In the second equation (2), the variable ‘Electionsit’ is used as an instrumental
variable to instrument the nominal rates of assistance. An IV should satisfy two conditions, namely instrument relevance and instrument exogeneity.
Instrument relevance means that Corr (Z, X) ≠ 0. Giordani, Rocha and Ruta (2012) used ‘Elections’ as an IV for trade policy distortions, and because trade distortion is a part of policy distortions to producer incentives (NRAborder is a component of NRA),
one can expect ‘Elections’ to be related to NRA too. In addition, Anderson (2009) mentioned that during recent years the NRAborder was the main contributor of NRA,
policy distortions to producer incentives were primarily caused by trade distortions, and thus the correlation between the IV and the instrumented variable is not zero. Figure 7 indicates the correlation that two variables are significantly related.
Figure 7 Bivariate Correlation
Instrument exogeneity requires Corr (Z,u) = 0. Rocha and Ruta (2012) mentioned that election was not directly related to global food prices. The correlation test (Appendix3) indicates that Corr (Election, Price) is close to zero (0.05), and the magnitude of 0.05 is not significant. Therefore one can expect that the instrumental exogeneity be met. In conclusion, the variable ‘Electionsit’ satisfies the two IV
conditions and thus, it can be a valid instrument of NRA.
Chapter 4
Empirical analysis
This chapter will present three regressions to investigate whether or not the food exporting countries’ policy distortions to rice, wheat, maize producer incentives have causal effects on increasing the global rice, wheat and maize prices during the period 1990 to 2012.
Table 1: regression results Model1 (Rice-exporting countries) In PRice,t Model2 (Wheat-exporting countries) In PWheat, t Model3 (Maize-exporting countries) In PMaize,t NRARice,t -103.6606*
Energy Price in rice-exporting countries
191.186*
Rainfall level in rice-exporting countries
-0.363169
Population growth in rice-exporting countries
-23.32673
NRAWheat,t -66.51157
Energy Price in wheat-exporting countries
181.5222*
Rainfall level in wheat-exporting countries
-0.517343
Population growth in wheat-exporting countries
-16.2344
NRAMaize,t -75.1037*
Energy Price in maize-exporting countries
162.0693*
Rainfall level in maize-exporting countries
0.0361217
Population growth in maize-exporting countries
11.062774
Note: 1.* Means significant at a 0.05 level 2.Price in dollars
4.1. Effects of policy distortions to rice producer incentives
The regression results are listed in the table 1. One can first focus on the relationship between policy distortions to rice producer incentives and the global rice prices. The coefficient of NRARice is around -103 and the impact is significant at a 0.05 level. The
coefficient can be explained in the way that one unit decrease in NRARice (Δx= -1),
will increase global rice price by $1.0310. A decrease in NRARice means that
individual countries are imposing policy distortions to rice producer incentives in order to keep domestic rice price at a relatively low level under the pressure of 10Δy= -1*(-103%)=1.03 Dependent variable Independent variables
soaring global rice prices. Because a unit decrease in NRARice leads to $1.03 increases
in InPRice, one can make a conclusion that exporting countries’ policy distortions to
rice producer incentives have casual effects on increasing global rice prices.
One may also want to explain the effect of a one unit increase in NRA. As discussed in chapter 2, when there was an increase in NRA, it means that governments imposed policy distortions to encourage domestic market and increase domestic food prices to a higher level, under the pressure of decreasing global food prices and depressed domestic food producers (Anderson et al., 2007). However, during the thesis-studying period (1990 to 2011) the global food prices was increasing, and NRARice & NRAMaize
were found have a decreasing trend; NRAWheat was relatively stable but also had a
downward spike during 2007-2008. Therefore the thesis only considers the effect of declining NRA on the increasing global rice, wheat and maize prices. In other words, the thesis only discusses the effect of policy distortions that aimed at keeping domestic food prices at a lower level under the pressure of soring global food prices. In this first model, the rainfall level has a negative effect on the global rice prices but it is not significant. The reason could be that rainfall hasn’t changed a lot in these selected rice exporting countries during the passing two decades and thus, it can’t explain the huge increase in global rice prices. Another variable, energy price level has a significant positive effect on the world rice prices. Many studies mentioned that modern global food system is fuel-dependent since farm machinery needs oil products to fuel and the transportation of farm products also depends on fossil liquid or non-fossil liquid like biofuels that are mainly made from crops. Therefore a change in the energy price has a significant effect on global food prices in general. Growth rate of population, as the third control variable in the regression, didn’t have a significant effect on the global rice prices. Although some commentators11 have argued that global population growth is responsible for the increasing food prices, this model didn't find the proof. The reason could be that the global population growth rates have declined dramatically since 1980, and grain availability has continued to outpace
11
World in grip of food crisis. IANS, Thaindian News. 7 April 2008, ^ Burgonio, TJ. Runaway population growth factor in rice crisis—solon. Philippine Daily Inquirer, 30 March 2008.
population according to United States Census Bureau & United Nation’s report in 200812.
4.2. Effects of policy distortions to wheat producer incentives
Table 1 also presents the regression results of model 2, which focuses on the relationship between exporting countries’ policy distortions to wheat producer incentives and the global wheat prices. The coefficient of NRAwheat is about -67.
Although the coefficient has a negative sign as expected, it is not significant at a 0.05 level considering a low p-value. The regression result indicates that exporting countries’ policy distortions to wheat producer incentives didn’t significantly contribute to the increasing global wheat prices in the period 1990 to 2012. Some following reasons could be taken into consideration to explain why policy distortions to rice producer incentives had significant effect on increasing global rice prices, while the effect of policy distortions to wheat producer incentives on global wheat prices was not significant.
Firstly, figure 8 gives an overview of NRAWheat of wheat exporting countries and
NRARice of rice exporting countries during the period 1990 to 2012. One can see that
NRAWheat was far more stable than NRARice, and there was no significant decrease in
NRAWheat during the passing two decades except a little fluctuation between late 2007
to early 2009. On the contrary, NRARice had a very obvious declining trend and there
was a huge decrease in value during the food crisis period. Significant decrease in NRA means that countries are imposing aggressive policy distortions to producer incentives to fight soaring global food prices, however this was not the case for wheat exporting countries, but only applied for rice exporting countries. Most of those world leading wheat exporting countries are from the developed world, such as the United States, EU-27, Canada, Australia, while the main world rice exporting countries are from the developing world, such as India, Thailand, Vietnam, Pakistan, Cambodia, Brazil. As discussed in Chapter 2, many literature mentioned that after 1970s, developed countries were less likely to impose aggressive policy distortions to fight soaring global food prices, while developing countries continuously used these
12 World Population Information. United States Census Bureau. Data updated 27 Marrh 2008.
Population annual growth rate, 229 countries 1955–2050 (UN Population Division's quinquennial estimates and projections). United Nations. Last updated on 17 July 2007.
strategies to insulate domestic food market from the increasing world food prices. This explains why policy distortions to producer incentives imposed by wheat exporting countries, most of which are developed economies, didn't have a significant effect on global wheat prices.
Figure 8. Overview of Rice NRA and Wheat NRA
Source: worldbank DAI database
Secondly, wheat producers from advanced economies such as EU-27 countries, United States, Australia, Canada, are better organized and have more lobby power compared with their counterparts in developing countries. When global food prices were initially driven up by some exogenous shocks such as extreme weather or energy prices, individual countries would react by imposing policy distortions to producer incentives in order to keep domestic food prices at a low level. Food producers in developed countries could play a more active role in resisting these distortions, because they have relatively stronger bargaining power and thus, countries’ policy distortions that fight with soaring global food prices at the cost of producers were not aggressive during the passing two decades. NRAWheat in these advanced economies
were quite stable, it didn't experience significant changes and so, the impact of NRAWheat on InPWheat was not significant.
4.3. Effect of policy distortions to maize producer incentives
The third regression results listed in table 1 are about model 3, which focuses on the relationship between policy distortions to maize producer incentives and global maize prices. The coefficient of NRAmaize is around -75, it is significant at a 0.05 level. This
-2,5 -2 -1,5 -1 -0,5 0 0,5 1 1,5 2 19 90 19 92 19 94 19 96 19 98 20 00 20 02 20 04 20 06 20 08 20 10 20 12 Wheat NRA Rice NRA
implies that one unit decrease in NRAmaize will contribute to $0.75 increases in
InPmaize. However, the effect of NRAmaize on InPmaize is smaller than the effect of
NRARice on InPRice, and the R square in model three is much lower than the one in
model one, it can only explain 46% of the variability in the global maize price. Not like top rice exporting countries, of which majorities are developing countries, world top maize exporting group including both developing and developed countries, and these two country types have similar shares of maize exportation. Since developed countries were less likely to impose policy distortions to producer incentives while developing countries distorted domestic food market more aggressively (Anderson et, al., 2009), therefore, the overall effect of policy distortions to maize producer incentives on global maize prices was smaller than the effect of policy distortions to rice producer incentives on world rice prices.
4.4. A focus on rice-exporting countries
As seen in the previous sections, policy distortions to producer incentives from rice-exporting developing countries had significant effects on increasing the global rice prices; and the rice prices increased much more than other grains, tripling within the food crisis period. This section will focus on rice-exporting countries and investigate what kind of policy distortions did these countries impose and the reasons behind these policy decisions.
Numerous studies (Demeke 2008; Pandey 2008; Warr 2008; Ivanic and martin 2008; Aksoy and Isik-Kiknelik 2008; World Bank 2008b; Timmer 2008; Child and Kiawu 2009; Abbott and de Battisti 2009; Headey 2010;) mentioned that main rice exporting countries were more likely to impose aggressive policy distortions to fight the soaring global rice prices and to keep domestic prices at a lower level. In 2007 India raised the floor price of rice export to prevent non-basmati rice from flowing out the country (Childs 2008). Almost at the same time, China imposed an export tax on rice, and in early 2008 Vietnam government announced that domestic rice producers couldn’t get any new permissions to sign rice-exporting contracts with foreign buyers. In March 2008, Egypt and Cambodia banned the export of rice completely, and a month later, Pakistan enhanced the lowest price for allowing export, while Brazil started to temporarily prohibit rice exports.
Many articles (Shigetomi, Kubo and Tsukada 2011; Abbott et al., 2009; Aksoy et al., 2008; Benson et al., 2008; Childs et al., 2009; Maros and Martin 2008) indicated that India and Vietnam, these two major rice-exporting countries that provided more than 60% of the global rice exportation, imposed the most aggressive policy distortions during the food crisis period.
The India central government executed a series of rice export restrictions since October 2007, and levied a straightforward ban on the export of rice in April 2008, and these policy distortions to producer incentives remained in force until 2010. Schigteomi et al., (2011) identified India’s policy distortions as ‘a mixed policy approach both for farmers and consumers’. It means that on the one hand, government restricted rice exportation by levying export taxes in order to offer domestic consumer a sufficient rice supply; on the other hand government subsidized domestic food producers to encourage them to produce rice. According to Chen and Ravallion(2009), 80 percent of the poor in India were engaged in agriculture, thus Indian government imposed policy distortions in the interests of both poor producers and poor consumers.
The Vietnam government stopped signing rice-exporting contracts with foreign buyers in 2008. Direct after Vietnam imposed these distortions, the Thai rice benchmark surged to the record highest level above $1,000 per ton, and global rice prices became much higher. Vietnam government adjusted its policy distortions after the crisis period, but several restrictions on rice export are still in force until now. Childs et al., (2010) discussed that Poverty and malnutrition were serious problems in Vietnam. Given the poverty line at US $2 per day per capita, there were ½ of the populations in Vietnam that suffered from the poverty. People’s tolerance to rising food prices was low, and thereby Vietnam reacted to the soaring global food prices more aggressively than some other countries.
In summary rice-exporting developing countries’ policy distortions to producer incentives were, to some degree, inevitable. Considering these developing countries’ internal economic and political needs, these policy distortions were understandable even though they resulted in undesirable results— an even higher global rice prices. An implication from the result could be that: rice-exporting countries need to reform domestic food market and enhance the tolerance to fluctuations in global food prices.
They are required to reduce policy distortions and to help build a more stable global food market.
4.5. Conclusions of the empirical analysis and further research
The previous sections present the following findings: firstly, developing counties were more likely to impose policy distortions to producer incentives during the passing two decades. Secondly, rice exporting countries’ policy distortions to rice producer incentives had significant effects on driving up global rice prices. Thirdly, wheat exporting countries’ policy distortions to wheat producer incentives didn’t significantly contribute to the increasing global wheat prices. Finally, maize exporting countries’ policy distortions to maize producer incentives were found significantly responsible for the soaring global maize prices. However, the effect on prices was smaller compare with the effect of policy distortions to rice producer incentives on rice prices.
This thesis used the latest worldbank DAI database13, but the data about nominal rates of assistance (NRA) was only updated to 2011 and thus, the thesis chosen a test period end in 2011. One could expect a slightly different regression results through using updated NRAs once they are available, but the main findings from this thesis such as the significant negative effect of NRARice on the global rice prices should not
change, because during the passing two or three years the overall situation in the global food market hasn’t changed a lot.
The thesis focused on the relationships between exporting countries’ policy distortions to rice, wheat, maize producer incentives and global rice, wheat, maize prices; however, a further research could test much more agricultural products, then one can have a more broad understanding about to which extent could policy distortions in exporting countries affect global prices of different farm products.
13
World Bank DAI Database
http://econ.worldbank.org/WBSITE/EXTERNAL/EXTDEC/EXTRESEARCH/0,,contentMDK:219600 58~pagePK:64214825~piPK:64214943~theSitePK:469382,00.html
Chapter 5
Conclusion
The global food prices increased dramatically in early 2007 to 2008. After peaking in the second quarter of 2008, the world food prices declined for a short period and peaking again in 2011 at a relatively higher level than the level achieved in 2008. More specifically, FAO statistics indicated that during the food crisis years, the global rice prices increased by 217%, the global wheat prices rose by 136% and the global maize prices was driven up by 125%. The food crisis caused political and economical instability in both developing and developed nations, and individual countries started to take actions to fight the soaring food prices. Many recent articles (Anderson et al., 2010; Broadberry,S., and D.A. Irwin, 2007; Wallis and Weingast, 2009; Otsuka and Tamano, 2006; Ravallion and Chen, 2007; Thompson, Dewbre and Martini, 2007; Valenzuela, Van der Mensbrugghe and Anderson, 2009) indicated that in order to provide a sufficient food supply in domestic market and to keep domestic food prices from rising under the pressure of increasing global food prices, countries reacted by imposing policy distortions to producer incentives such as increasing export taxes, subsiding food producers, lowering import tariffs etc.. These policy distortions are believed to be positively associated with an even higher global food prices, but there were no existed literature that tested the causality for particular food products such as rise, wheat or maize. In addition, previous relevant literature that did quantitative researches used data updated to 2009, while the thesis extended the research period to 2012 and thus, complements the existing literature.
The thesis tested whether or not food exporting countries’ policy distortions to rice, wheat, and maize producer incentives had significant causal effects on increasing the global rice, wheat and maize prices. The findings are following: (1) Rice exporting countries’ policy distortions to rice producer incentives had significant effects on driving up global rice prices. (2) Wheat exporting countries’ policy distortions to wheat producer incentives didn’t significantly contribute to the increasing global wheat prices. (3) Maize exporting countries’ policy distortions to maize producer incentives were found significantly responsible for the soaring global maize prices. However, the effect on prices was smaller compared with the effect of policy
distortions to rice producer incentives on rice prices. Existing literature (Signe Nelgen and Kym Anderson, 2010 Anderson, Martin and van der Mensbrugghe, 2011; Ruta,Rocha and Giordani, 2011; Carter et al., 2006; Furtan 2006; Orden, Paarlberg, and Roe, 1999) and the empirical evidence from this thesis indicated that: (1) There were no significant aggressive policy distortions to producer incentives that were imposed by developed nations; (2) compared with developed nations, developing countries were more likely to use policy distortions to producer incentives in order to insulate domestic food market from the soaring global food prices. Because top rice exporting countries are from the developing world while major wheat exporting countries are advanced economies, it’s reasonable to get the empirical result that rice exporting countries’ policy distortions to producer incentives had significant effect on global rice prices, while the effect of wheat exporting countries’ policy distortions didn't significantly contribute to the increasing global wheat prices.
Policy distortions to producer incentives, especially those distortions in developing countries had significant effects on increasing the global food prices. These distortions were individual countries’ reactions to soaring global food prices, but in return they drove up global food prices to an even higher level. In order to build a more stable world food market and to avoid the unexpected global food crisis, international organizations should intervene (Anderson 2012), and multilateral cooperation is needed to make sure main food exporting countries especially these major food-exporting developing countries will not impose aggressive policy distortions to producer incentives due to unilateral interests. In addition, less developed food exporting countries need to take actions to reform domestic agriculture (Shigteomi 2011; Calpe 2005; Dawe, David 2002), and become less vulnerable to future surges in global food prices. If these countries have higher tolerance to the soaring world food prices, they would not react by imposing very aggressive policy distortions, and thus the effect of these distortions on global food prices could be limited.
Reference
Agricultural Policies in OECD Countries: Monitoring and Evaluation 2007. Paris: OECD. OECD and FAO (Food and Agriculture Organization). 2008. OECD-FAO Agricultural Outlook 2008– 2017. Paris: OECD.
Aksoy, M.A. and B.M. Hoekman (2011). "Food Prices and Poverty: Introduction and Overview". In Aksoy and Hoekman eds. Food Prices and Poverty . World Bank.
Anderson, J.E., and J.P. Neary. 2005. Measuring the Restrictiveness of International Trade Policy. Cambridge, MA: MIT Press.
Anderson, K., M. Kurzweil, W.Martin, D. Sandri, and E. Valenzuela. 2008a. “Methodology for Measuring Distortions to Agriculture Incentives.” Agricultural Distortions Working Paper 02, World Bank, Washington, DC.
Anderson, K., R. Lattimore, P. J. Lloyd, and D. Maclaren. 2007. “Distortions to Agriculture Incentives in Australia since World War II.” The Economic Record 83 (263): 461-82
Anderson, K., and E. Valenzuela. 2007. “Do Global Trade Distortions still Harm Developing Country Farmers?” Reivew of World Economics 143 (1):108-39
Anderson, K., and E. Valenzuela. 2008. Global Estimates of Distortions to Agricultural Incentives, 1995 to 2007. Core database at World Bank.
Anderson, Kym, Johanna Croser, Damiano Sandri, and Ernesto Valenzuela. 2010. “Agricultural Distortions Patterns Since the 1950: What Needs Explaining?” In the Political Economy of Agricultural Price Distortions, ed. Kym Anderson, 25-80. New York: Cambridge University Press.
Anderson, K. and S. Nelgen (2012a), ‘ Agricultural Trade Distortions During the Global Financial Crisis’, Oxford Review of Economic Policy 28 (1)
