The impact of coffee market reforms on producer prices : a case study of Honduras

(1)

Faculty of Economics & Business

Master of Science in

Economics: International Economics & Globalization

Master Thesis

The Impact of Coffee Market Reforms on

Producer Prices: A Case Study of Honduras

Author: Supervisor:

Monique Rood Dr. Boe Thio

Student ID: 10182594 Second reader:

(2)

2

Statement of Originality

This document is written by Monique Rood who declares to take full responsibility for the content of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

3

Abstract

This thesis investigates the impact of the policy reforms in the Honduran coffee market during the late 1980s and the early 1990s. ECM estimates together with cointegration estimates have been used to illustrate the impact of the policy changes on the received producer prices as a share of the world price, and the nature and speed of the price transmission from the world to domestic market. The data consist of annual coffee prices during the time span 1980-2009. The empirical results provide evidence that the share of producer price in the world price has increased significantly since the introduction of the reforms. Moreover, the results suggest a greater integration between the world and domestic prices. However, the reforms made the pass-trough of negative world price changes easier than the pass-through of positive world price changes. These findings imply that implementing liberalization in the Honduran coffee market does not necessarily ensure a better relationship between the world price and the prices received by the producers. In the absence of the right conditions for efficient private markets to evolve and the appropriate infrastructure, some impact of market reforms may be limited or to a disadvantage of the producers.

(4)

4

1. Introduction

Coffee is the third most consumed beverage globally, after water and tea, so it comes as no surprise that coffee is one of the largest traded commodities worldwide. Colombia and Brazil have always been the top two coffee producing countries for most of the twentieth century (Karp, 1993). However, this situation has recently changed due to the significantly increased coffee production in Vietnam. Vietnam has replaced Colombia as the second largest coffee producing country followed by Indonesia and Ethiopia. Despite the smaller individual contribution, Honduras ranks sixth globally in coffee supply and coffee export by volume. According to the Foreign Agricultural Services Honduras is responsible for 7% of world coffee supply (USDA, 2016).

Coffee has been an important driver for economic activity and a major export commodity in developing countries. In addition, coffee is an important cash crop, a big source of employment, foreign exchange and revenue. However, coffee producers in developing countries like Honduras have historically received a small share of the international price of green (unroasted) coffee prior to the liberalization of the coffee sector in the late 80s early 90s. According to Worako, Van Schalkwijk, Alemu and Ayele (2008) the share received by the producers ranged between 30% and 45% of the export price of green coffee. One of the reasons often mentioned in the literature for this small share is heavy government intervention in the coffee sector in the form of fixed producer prices, regulations and market inefficiencies. The two main reasons for the government interventions are first of all revenue collection, since agriculture is usually the largest export crop in a number of developing countries; it offers an easy accessible tax base. The second reason mentioned in the literature is reduction of the price volatility by governments in order to lower the risk for producers that are dependent on export prices. Moreover, governments appointed marketing boards to execute these governmental policies. This gave the marketing boards monopsony power in the coffee sector, by executing policies like the regulation of domestic marketing and the control of purchasing and exporting of commodities. The fixed producer prices, governmental regulations and the monopsony power of the marketing boards drove a wedge between the price received by growers and the world price.

(6)

6

Next to the individual domestic government interventions, the international coffee market was regulated by the International Coffee Agreement (ICA). The ICA is an agreement between main coffee exporting and importing countries that was established to increase the price at which member countries could trade coffee through an export quota system. After the collapse of the ICA in 1989, many developing countries have undertaken economic reforms and liberalized domestic coffee markets under Structural Adjustment Programs (SAPs) by the International Monetary Fund (IMF). The economic reforms together with the SAPs were aimed at improving producer prices and increasing trade efficiencies, by implementing liberalization in the export crop markets and allowing private agents to operate as traders instead of the governmental owned marketing boards (White & Leavy, 2001). In addition, the reforms were expected to increase market efficiencies from policy imperfections at the production and marketing levels (Worako et.al, 2008) since the removal of distortionary prices and trade barriers should bring the domestic prices closer to international prices (Baffes & Gardner, 2003). According to Kilima (2006) the success of such reforms depends on the strength of the transmission of price signals between international and domestic producer prices.

Honduras, as one of the largest coffee producing countries, has also undertaken economic reforms in a reaction to the collapse of the ICA in 1989. Coffee is one of the most important industries in Honduras, both in terms of processing and growing. As of 2003, coffee accounts 3,5 percent of Gross Domestic Product (GDP) and 47% of Honduran export (Thomas, 2003). According to the Central Bank of Honduras, coffee is the number one agricultural export product and provides a large contribution to the foreign exchange reserves (USDA, 2016). Today the coffee industry provides employment to more than a million people, representing one out of eight Honduran workers. The (coffee) reforms were a large part of the overall economic reforms in Honduras aiming at more efficient markets, lower marketing margins and higher producer prices.

The aim of this paper is to analyze the impact of the policy reforms on the coffee sector in Honduras during the late 1980s and early 1990s. The main objective of this study is whether producers of coffee beans received a higher share of the world price

(7)

7

for coffee after reforms in Honduras. A second objective is to assess the impacts of the policy reforms on the nature and speed of transmission of world prices to domestic prices. It is expected that the reforms (1) will bring the domestic price levels closer to the world price and (2) will improve the speed and nature of the price transmission of world price changes to the domestic prices. Honduras is chosen for this case study since the literature lacks information on the impact of the reforms in Honduras, while Honduras is a very important player in the coffee sector. Honduras has undertaken policy reforms adequately long enough that results should be measurable in the coffee price behavior. This empirical study applies a cointegration approach, employing an autoregressive distributed lag model and an error-correction model to answer the above questions. Section II presents a detailed overview of the economic market reforms in the coffee sector and the economic reforms in Honduras. Section III presents the literature review followed by the empirical model and data discussed in section IV. Section V provides the results of the empirical results, VI offers a detailed discussion and section VII presents the conclusion.

(8)

8

2. The Coffee Market and Economic Reforms in Honduras

2.1. Organization of the International Coffee Industry

Prior to the economic reforms, the coffee producing countries were categorized by heavy government intervention in agricultural and financial markets. Domestic governments did not only intervene in commodity markets, but accepted the intervening practices as part of the policy framework. Development theory coupled with practical political considerations encouraged the continued interventions. Governments pursued tax-based policies on agricultural commodities in order to promote industrial development (Akiyama, 2001). Next to these domestic government regulations, the global coffee sector was also regulated by the ICA as proposed by the International Coffee Organization1 (ICO). Members of the ICO are coffee producing and coffee-importing countries and are responsible for over 97 percent of world production of coffee (Akiyama, 2001). The first ICA agreement was signed in 1962. The ICA established a target price for coffee and export quotas that were assigned to each member country. The assigned quotas were loosened when the coffee price rose above the set price, and tightened when the price fell below the set price.

Many coffee-producing governments regulated the coffee supply and production by state-owned institutions; this led to heavy spending on and protection of the agricultural sectors. These state-owned marketing boards were part of the coffee production chain and were responsible for the non-faming activities like the coffee quality, the promotion of coffee production, the reallocation of resources needed for exports regulated by the ICA, marketing and export activities and the coffee price formation. Figure 1 shows the coffee production chain, the production chain knew four parties involved, the farmers, the global traders, the roasters and the bodies that tied these parties together; the marketing boards. The global traders sourced coffee from different origins and sold the coffee to the roasters. The roasters branded the coffee and sold the coffee on to retailers. The farmers and marketing boards were a focus point in the SAPs (Xavier, 2011).

1_{The International Coffee Organization (ICO) is the main intergovernmental organization for coffee} exporting and importing countries. The organization tries to organize the coffee sector through international cooperation. The ICO supervised the International Coffee Agreement (ICA). The latest agreement between the member countries was signed in 2007.

(9)

9

Figure 1: Actors in the Honduran coffee marketing chain before economic reforms

Source: Xavier, 2011

The stabilizing exports quotas introduced by the ICA together with domestic regulated markets in coffee producing countries made the international coffee market relatively stable. The relative success of the ICA knew several reasons: (1) the participation of the consumer countries with the quota system, (2) the existence of the marketing boards in coffee producing countries, where governments were in control of decision-making, (3) the acceptance by Brazil for a shrinking coffee market share that resulted from the ICA, and (4) a common strategy of import substitution in producing countries (Daviron, 1996). Trading rules for the coffee sector were clear, changes were politically negotiated and the generated profits were equally divided between consumer and producer countries. However, in 1989 the ICA quota system collapsed due to free-riding schemes and disagreements on quotas concerning the system of coffee exports to non-member countries and the distributed quota of Arabica and Robusta coffee, the two main coffee types.

2.2. Causes of Liberalization

From 1986 onwards the general price level of coffee had already been declining drastically. Governments were forced to grant financial help or to assist in looking for donor countries or international organization for financial help (Akiyama, 2001). In 1986 the IMF established Structural Adjustment Facilities (FAS). The FAS offered concessional financing to the world’s poorest countries (IMF, 1995). The IMF wanted to respond to the balance of payments difficulties of these countries. Countries had to adopt the IMF structural adjustment programs (SAPs) before they were qualified for

Coffee Farmers Marketing Boards Government /

ICA / Export quota system

Coffee Traders

Local Market International Market

(10)

10

these loans. The implementation of the SAP is country specific, so the implementation process and time differed per country. These SAPs were aimed at liberalizing the domestic and export coffee market with the ‘right price’ as a stimulating mean for increasing productivity and growth (Worako et.al, 2008). In other words, the SAPs tried to bring the producer prices closer to the world prices of coffee, without interventions of marketing boards and policy imperfections since the deregulation of the marketing system provided new opportunities for the private sector to engage in all tiers of the marketing chain (Worako et.al, 2008).

However, the fall of the ICA in 1989 together with the already declining coffee prices led to fiscal and financial problems in coffee producing countries. Policies on the production and quality of coffee were left to individual countries and the producing countries increased exports. As a consequence, excessive volumes of coffee entered international markets, coffee prices became more volatile and the overall quality of coffee began to decline (Ponte, 2002). Many coffee-producing countries were in the process of deregulating, privatizing and liberalizing the coffee production, which led to uncertainty among coffee growers.

2.3. Liberalization in Honduras

Before the collapse of the ICA most Honduran households grew coffee for domestic use or for local markets. The coffee production was mainly in hands of smallholders (Williams, 1994). Honduras started to develop itself as a coffee exporting country in the 1950s with help from the United States government who provided technical and financial assistance. It was not until the 1970s that the Government of Honduras (GOH) started to promote the coffee industry actively. New laws were introduced. The Honduran Coffee Institute (IHCAFE) was formed by the GOH to implement governmental policies and to encourage coffee production. In addition, the national development bank started to provide loans to medium-sized growers and the nation held trophy competitions for best quality coffee to promote high quality coffee standard. One well-known competition is the Cup of Excellence. This initiative, taken by IHCAFE, is a competition between small-scale farmers that rewards coffee producers with a special trophy and a higher price for the high-quality produced coffee. Next to the Cup of Excellence, IHCAFE actively promoted for the production

(11)

11

of differentiated coffees, like specialty coffees. These specialty coffees would generate direct benefits to smallholders in the form of a higher price per bag for the coffee beans (Tucker, 2013). These policies and initiatives were the starting point of a continuing expansion in the coffee production and quality in Honduras.

In Honduras the coffee sector also started to liberalize in response to the collapse of the ICA. Foreign trade was liberalized; new programs replaced the import substitution policies and government interventions focused on tariff reduction and agricultural trade liberalization by removing price controls and guarantees. Further, the reforms strengthened the individual property right to land and cooperative lands were privatized. In 1992 the Law for Modernization and Development of the Agricultural Sector (LMDSA) was accepted. The LMDSA replaced the 1975 Agrarian Reform Law, replacing some key statues including the commitment to eliminate pieces of land with five hectares or less. The LMSDA also promoted the titling of land to individuals and implemented the Land Titling Project were approximately 150.000 titles were granted between 1992 and 2000. BANADESA, the agricultural development bank of Honduras, became the main source of credit for small farmers. The GOH restructured BANADESA to stimulate profitable bank lending by liberalizing agricultural interest rates. This made it possible for small and medium sized growers to borrow credit to stimulate coffee production and increased the access to capital across economic classes. This in turn strengthened land access, agricultural organization and productivity. However, formal credit remained still in hands of the high-wealth households and land rentals were still a very small percentage of the total land area, which made the smallholders operations constrained. But farmers that did receive technical assistance and credit proved to increase their productivity significantly (Boucher et.al, 2005).

Honduras started to adopt the SAPs in 1991 with the aim of achieving sustainable economic growth. The IMF support was a medium-term economic program for the period 1992-1995 (IMF, 1995). The program made progress in the structural areas, however due to the presidential elections in 1993 the financial imbalances widened. A new program was developed for the period 1994-1997 to restore the macroeconomic stability and intensifying structural reforms, with focus on

(12)

12

the social sector in Honduras.

2.4. The Honduran coffee market

The coffee price is one of the most volatile prices among the agricultural commodities (Worako et.al, 2008). Figure 2 depicts the Honduran producer price and world price of coffee as measured by the ICO. It can be seen from Figure 2 that the general price level of coffee had already been declining drastically from 1986 onwards. The prices showed improvement in the initial post-reform period (1992-1997) but dropped drastically in the late 97s due to the increased world production. Moreover, the impact of the market reforms had been more effective in early stages of the liberalization process. Reasons mentioned in the literature are the institutional adjustments during those early stages and another reason brought forward by Ponte (2002) is that the market reforms tend to increase price volatility because of the absence of stabilizations mechanisms that were active in the coffee sector before the fall of the ICA. Between 2001 and 2002 the coffee price approached its lowest level, this period is known as the ‘coffee crisis’. According to Ponte (2002) the coffee oversupply was at the root of the coffee crisis after the collapse of the ICA.

Figure 3 shows the domestic price as a percentage of the world price. The prices paid to the producers were on average 57% of the world price before the reforms, and increased after the reforms to an average of 63%. This is a relatively small increase; explanations pointed out by the literature are the neglecting attitude of farmers that did not keep track of world market changes and collusive behavior by traders (Krivonos, 2004). That is, even if many traders are operating in the market, the positive effect of the liberalization on the producer prices may be undercut if the coffee traders engage in collusive behavior. In addition, the coffee crisis had serious impact on the smallholder producers (Daviron & Ponte, 2005). Farmers took measures to diversify their livelihoods by reducing their dependence on the coffee production, at least in the short run. The diversification increased their resistance to fluctuations in the coffee market and other shocks like weather events and national, economic and political shocks (Tucker, 2013). Regardless of the decline in the average price level and the diversification during the reforms, the volume of exports

(13)

13

in Honduras had increased steadily as can be seen in Figure 4. This might have been due to the reform policy measures such as currency devaluations, the elimination of the export quota and taxes on coffee exports, improved infrastructure and other marketing infrastructure (Worako et.al, 2008).

Figure 2:

Source: International Coffee Organization

Figure 3:

Source: Own estimations

0 50 100 150 200 250 300 350 400 450 500 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 US c ents pe r lb

Price series: Honduras domestic price and the word price

Producer Price World Price 0% 10% 20% 30% 40% 50% 60% 70% 80% 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 Domes&c price as a percentage of the Interna&onal price Domes2c price as a percentage of the Interna2onal price

(14)

14

Figure 4:

The impact of the coffee crisis led to interventions by the GOH. During the coffee crisis the GOH decided to provide different terms on loans to support the coffee producers. Most of the loans granted had payment periods of twenty years. In addition, the GOH created the Law of Financial Reactivation of the Coffee Producing Sector in 2003, in order to prevent coffee producers from abandoning coffee production. The Law together with the changed loan terms helped the coffee producers fund savings to sustain their production. Also the National Coffee Fund (NCF) was created to support the coffee sector and was responsible for the maintenance and construction of roads in coffee producing areas that helped the transport of coffee from the mountains to the markets.

0 1,000 2,000 3,000 4,000 5,000 6,000 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 In th ou sa nd 6 0kg b ag s Coﬀee Exports in Honduras Honduras Linear (Honduras)

(15)

15

Furthermore, the GOH privatized IHCAFE in 2000 to create an organization that could respond quickly to the needs of the coffee industry. IHCAFE’s main activities are to oversee coffee production, increasing productivity, regulation quality and promotion and the finance practices of the Honduran coffee sector. Further, the Honduras coffee expansion took advantage of the frost and drought in Brazil in 2004. As a consequence, the Brazilian harvest was low that year which led to a price increase for green coffee. This encouraged greater investments to new growers in Honduras who began planting with national help. The greater investments led to more productivity in terms of more revenue per hectare of land as can be seen in Figure 5 (Eakin, Tucker, & Castellanos, 2006).

Figure 5: Productivity in Honduras’ coffee sector

Source: FAOSTAT and World Bank staff estimates

Deregulation of the marketing system after the reforms has opened up opportunities for the private sector to participate in the marketing chain for coffee in Honduras. Traders could now negotiate contracts with parties overseas and payments were easier transmitted to coffee growers (Krivonos, 2004). So the marketing chain changed as a consequence of the reforms. A large number of competitive buyers and sellers entered the market and became active in the chain. According to IHCAFE (2005) there were in 2005 about 70.000 coffee growers in Honduras, about 3000 intermediaries nationwide, and about 44 exporting firms in Honduras. The coffee post-reform marketing chain in Honduras is as follows: First in the chain are all the small-scale

(16)

16

coffee farmers, who have very little influence on the coffee price. The second link in the marketing chain is the intermediary, who is responsible for the trade of coffee. Some intermediaries have price setting power, however this depends on the size of the intermediary company.The exporting firms are the third link in the chain. These firms trade directly with the intermediaries, but usually have representatives working in the rural areas. The exporting firms deal with the distribution of the coffee and make coffee trading process more efficient. The final link in the national chain is the roaster, who is responsible for the domestic market and rarely export. The roasting companies have their own distribution channels and marketing strategies (Fromm & Dubón, 2006). Figure 6 shows an overview of the marketing chain.

Figure 6: Actors in the Honduran coffee marketing chain after economic reforms

Source: Fromm & Dubón, 2006

The liberalization space and scope varied between the developing countries. The start of the liberalization process in Honduras is a little speculative since the reforms were implemented gradually over time in the early 90s together with the implementation of the SAPs as discussed above. However, for this research the year of reform is taken as 1992. The reason for 1992 is as follows: in 1992 the SAPs were fully implemented by the IMF and at the same time foreign trade was liberalized; new programs replaced the import substitution policies, and government interventions focused on tariff reduction and agricultural trade liberalization by removing price controls and guarantees. Coﬀee Farmers Intermediaries Exporters / Private agents Roaster Coﬀee Traders Local Market Interna.onal Market

(17)

17

2.5. Concluding remarks

Before the liberalization, the governmental marketing board played a prominent large role in the coffee producing countries as discusses above. However, the Honduran coffee marketing chain has transformed as a consequence of the economic reforms. One comparing the marketing chain before and after the reforms (Figure 1 and Figure 6) can see a state-owned marketing chain controlled by the GOH and the ICA before the reforms and marketing chain operated by private agents and traders after the reforms. Changes in the international dynamics of the coffee market started the liberalization of commodity markets and institutions. First, producers were allowed to sell small amounts of coffee on the domestic market. Then, the governmental involvement in the coffee production and export ended and growers were allowed to sell their coffee beans to private agents. According to the CFC 2000 Report “the key benefits of the liberalization is that increased competition throughout the marketing chain leads to a reduction in marketing costs and growers receiving a higher proportion of the export unit value” (CFC, 2000). This thesis will investigate whether the share of international prices received by producers have indeed increased as a consequence of the liberalized coffee sector and marketing chain. In addition, the price transmission between world and domestic prices is analyzed.

(18)

18

3. Literature Review

As an introduction to this empirical research this section explores the responsiveness of domestic commodity prices to the international commodity markets. The coffee market reforms included the removal of price controls, market liberalization and privatization of state-owned institutions. The different measures implemented during the reforms have been widely documented in the literature. In theory the economic reforms should operate in a beneficial way, however imperfect markets may have a distortive influence.

Lukanima and Swaray (2014) examine the impact of commodity market reforms on producer price volatility with the use of empirical research. The authors try to link volatility with differences in market structures. In addition, the study provides a detailed discussion on policy implications for the East African coffee market. The authors employ a standard generalized autoregressive conditional heteroscedastic (GARCH) time-series prices using producer prices. The estimates show an increase in the producer price volatility during the reforms, however the authors show more impact during the early stages of the reforms. Fafchamps and Vargas (2008) examine the price transmission from international prices to prices received by coffee growers, traders and exporters in Ghana. The study uses price information collected in three surveys covering all the levels of the Ugandan value chain. The authors compare prices that producers actually received to those reported by trades and exporters. Uganda is used as a test case because the marketing and export of coffee there was fully liberalized. The authors find that a rise in the world price is transmitted to all domestic prices except to the price paid to the producers. The price transmitted to the producers is smaller than changes in the international price, especially for world price increases. One reason mentioned in the study for the price asymmetry might be the ignorance of producers concerning increases in wholesale prices.

Mundlak & Larson (1992) estimate a direct relationship between the domestic price and world market price using a logarithm formula that measures the elasticity of

(19)

19

transmission Food and Agricultural commodities and the exchange rate. The empirical research is based on 58 countries and finds evidence of almost prefect price transmission for the pool of commodities. However, independent OLS regressions for wheat, coffee and cocoa show lower estimates of price transmission than the estimates for the pooled commodity regression. This result indicates a high-level of distortion for these tree commodity markets.

With the use of a vector autoregression (VAR) model Shepherd (2004) examines the impact of the fall of the ICA’s export quota system for the coffee market chain. The results show evidence that the elimination of the export quota system did not result in an improved price transmission from producers to the world market. One reason mentioned for the failure in impact is the market power exercised by the large and established coffee processors. However, Gemer and Struthers (2007) find evidence of a significant increase in the coffee price volatility after the elimination of the ICA in Ethiopia using a GARCH model.

A different approach estimating a relationship between two price time-series is the Error Correction Model (ECM). This model includes a dynamic component that captures the effect of the adjustment of the dependent variable when it deviates from the long-term equilibrium level. Hallam and Zanoli (1993) explore the relevance of the ECM in their paper for studying agricultural supply responses and argue that the ECM provides a useful theoretical framework for studying agricultural supply responses after detailed theoretical analyses of the prevailing frameworks. Cramon-Raubadel (1998), Tambi (1999) and Abdulau and Rieder (1995) all lend support to the use of ECM, especially when testing for asymmetric price transmission in agricultural markets. Asymmetric price transmission controls for the difference in impact of a negative and positive price change in another connected market. In addition, price series of agricultural commodities are often non-stationary, so estimating a price impact of a change in one price on another should be based on appropriate methods that allow for non-stationary variables like the ECM (Meyer, 2004).

As a side note it should be stated that the presence of price intervention does not imply imperfect price transmission. Full transmission can occur with a constant

(20)

20

export tax, however other policies such as export quotas may disrupt the price transmission. Positive price transmission is usually found when time-series data are used since administered prices tend to change over time in response to world market prices. The ECM allows the evaluation of price transmission under different policies (Krivonos, 2004).

Many researchers have devoted their time to the examination of price transmission with the error-correction model. The results of Baffes and Gardner (2003) differ substantially from the results reported by Mundlak & Larson (1992) presented above. Baffes & Gardner (2003) use an error correction model estimator to examine the responsiveness of domestic prices to fluctuations in the world market. The authors estimate the price transmission for ten different commodities on a country-by-country basis for eight countries using data between 1970 and 1990. In addition, the authors go a step further in their research, measuring the effect of policy reforms and testing whether the reforms have improved the price transmission in the commodity markets. This approach introduces the use of structural breaks, corresponding to the years of the market reforms. The results showed for most countries limited effect of the reforms on price transmission. Out of 31 countries only 6 countries showed a closer relationship with the world market following the reforms. An explanation given by Conforti (2004) is that price transmission can also occur with the presence of government interventions. If this is the case, then policy reforms do not need to have a large effect on the degree of price transmission.

Focusing entirely on the coffee sector, Mofya-Mukuka and Abdulai (2013), Xavier (2011), Worako et.al (2008) and Krivonos (2004) take a closer look at the impact of the reforms on the price transmission between the domestic market and world market using an ECM.

Mofya-Mukuka and Abdulai (2013) follow the approach by Baffes and Gardner (2003) with the use of structural breaks to take account of policy shifts in Zambia and Tanzania. The authors use a two-regime momentum threshold cointegration model (MTAR) for this research to focus on the limitations of the linear two-stage cointegration model proposed by Engle and Granger (1987) and find significant evidence of an increase in the price received by the producer as a share of

(21)

21

the world price after economic reforms in Zambia. The producer share in the world price increased from 61.5% before the reforms to an almost complete integrated domestic market after the reforms were the producer price share is 96.7%. However, the authors find evidence of a reduction in the share of Tanzania’s prices. One explanation for the different results in Zambia and Tanzania might be the difference in reforms that both countries underwent. The liberalized coffee market in Zambia led to a positive impact on the producer prices, while the reforms in Tanzania increased the government interventions, which in turn led to a negative impact on the producer price. In addition, the authors find evidence of price asymmetry in both countries from the world market to domestic market.

Similar results are reported by Worako et.al (2008) with the use of the ECM specifications and data ranging from 1992 to 2006. The authors include dummy variables in the model to test for policy shifts and asymmetric price transmission in the Ethiopian coffee market. The estimates show significant positive results for the share of producer price in the world price for all the coffee types since the introduction of the reforms. The estimate of the producer price in the world price increased significantly from 45% in the pre-reform period to 65% in the post-reform period. In addition, the results show that the Ethiopian coffee price adjusts easier to world price changes in the post-reform period than it did prior to the reforms. The price transmission for negative world price changes increased from 17% before the reforms to 26% after the reforms. However, the pass-through for positive world price changes decreased from 32% before the reform to nearly zero price transmission after the reforms. The authors argue that the success of reforms could be constrained by many structural shortcomings in the domestic market like actions of traders, government policies on export taxes and exchange rates and marketing infrastructure in the form of transportation and transaction costs. Krivonos (2004) follows a similar approach, employing an Autoregressive Distributed Lag and error correction model using the two sets of dummy variables to capture the role of asymmetric impact shocks using data from 1994-2004. The producer price as a share of the world price increased significantly in 13 out of 14 countries. For example, in India the producer price increased from 56% to 86% as a result of the reforms. Furthermore, the estimates show an improved speed at which domestic prices react to changes in the

(22)

22

world price for most countries, for example increasing from zero adjustment speed to 0.2 in Ethiopia. In addition, Krivonos (2004) finds evidence that the short-run price transmission from the world market to domestic market has improved as a consequence of the restructuring of marketing-boards, however, the pass-through of negative world price changes is easier than the transmission of the positive world price changes, causing negative effects for the producers in the short run. However, in some countries the reforms were not far-reaching enough or the reforms were implemented more gradually over time, resulting in less significant results.

Xavier (2011) uses a slightly different approach. Instead of using the two-stage Engle and Granger (1987) test for cointegration, he uses an alternative one-two-stage model suggested by Gómez, Lee and Koerner (2010). Gómez, Lee and Koerner (2010) explain that when dealing with smaller samples the one-stage model is more appropriate in which the error correction term is employed directly in the estimating equation. The findings indicate that reforms induced a stronger relationship between domestic and international prices for Colombia, but not for Ghana and Ivory Coast. One possible explanation may be the institutional arrangements in Ghana and Ivory Coast. Also in Colombia the prices moved closer towards a long-term equilibrium, while in Ghana and Ivory Coast the price series exhibit divergences.

The assumption taken in this study is that exporting countries act as price takers on the world market. This last aspect is found significant in the study Karp & Perloff (1993), who argue that the two largest coffee exporting countries (Brazil and Colombia) act as price-takers instead of collusive behavior.

To sum up, the impact of the policy reforms have shown mixed results in the presented literature. Overall the studies show a significant positive impact of the reforms on the prices paid to producers as a share of the world price. Furthermore, the studies show mixed results on the nature and speed of the price transmission from the world market to the domestic markets, some studies show significant improvement in the price-transmission, while others show negative results. However, most studies find a significant difference in the price-transmission as a result of the reforms. In addition, significant results are found for price asymmetry that made the pass-through of world market decreases easier than the transmission of positive world price

(23)

23

changes. However, the relationship between the world market and domestic market is sensitive to the completeness and impact of the reforms and differ substantially across commodity and country. This paper follows the dynamic approach taken by Baffes and Gardner (2003), partly from Xavier, (2009). In addition, this study incorporates asymmetric price shocks by the use of dummies like Krivonos (2004) and Warako et.al (2008) to investigate the policy shocks on the Honduran coffee market. The main question addressed is whether the reforms have resulted in a closer relationship between the world market price for coffee and the internal prices paid to growers.

(24)

24

4. Empirical Model

This section starts with the introduction of the methodology, describing the model used and the interpretation of the variables. Subsequently, the section will give an overview of the data and the descriptive statistics.

To test for cointegration between domestic and world prices, an Error Correction Model (ECM) can be used. There are several strategies to estimate the error-correction model. The study of Engle and Granger (1987) suggest the use of a two-step approach that is followed here. In this study the error-correction model is specified to measure the price adjustments of the prices received by the producer as a share of the world price as a result of the economic reforms, a policy shock.

4.1 The Model

Following Krivonos (2004), the specifications for the error-correction model start with an Autoregressive Distributed Lag (ADL) model with lagged values of the price received by the Honduras coffee producers as a share of the world price, from now on the domestic price, and the world price of coffee as the independent variable. Consider the ADL (1,1) model with two non-stationary time series 𝑃_!!_{, 𝑃}

!! and two

single lags 𝑃_!!!! _and _𝑃

!!!! :

𝑃_!! _{= 𝛼 + 𝛽}

!𝑃!! + 𝛽!𝑃!!!! + 𝛽!𝑃!!!! + 𝜇! (1)

Where 𝑃!! is the domestic coffee price at time t, 𝑃!! is the world coffee price at time t,

and 𝜇_! is the random error term. 𝛽_!, 𝛽_! & 𝛽_! need to be estimated. Equation (1) captures the effect of a unit change in the world price that will affect the domestic price by 𝛽_! plus the error term. The error term takes accounts for seasonal and policy variables that do not have a direct influence on the domestic and world coffee prices. Formula (1) can be rearranged to an error specification model with one lagged value

(25)

25

as in the study of Krivonos (2004) (see appendix A for derivation), resulting in equation (2):

𝑃!! − 𝑃!!!! = 𝛼 + 𝛿(𝑃!!− 𝑃!!!! ) + 𝜃(𝑃!!!! − 𝛾𝑃!!!! ) + 𝜇! (2)

Where 𝛿 = 𝛽_!, 𝜃 = − (1 − 𝛽_!) 𝑎𝑛𝑑 𝛾 =!!!!!

!!!!

In this model 𝜃(𝑃_!!!! _{− 𝛾𝑃}

!!!! ) is the error correction term that measures the

deviations of the domestic price in the previous period from the long-run equilibrium. In other words, it is the cointegration relationship between 𝑃_!!_{& 𝑃}

!!. Equation (2)

captures the extent of price transmission from the world market to the domestic market. 𝛿 measures the short-run responsiveness of the domestic price to a one-time shock in the world price and 𝜃 measures the speed of adjustment of the domestic price towards its long-run equilibrium.

To capture the impact of the reforms on the domestic price and to test for asymmetric price transmission, two different sets of dummies are used in this research. The first set contains policy dummies and mirror the market liberalization:

𝐷_!!"# = 1 𝑖𝑓 𝑎 𝑝𝑒𝑟𝑖𝑜𝑑 𝑝𝑟𝑖𝑜𝑟 𝑡𝑜 𝑚𝑎𝑟𝑘𝑒𝑡 𝑟𝑒𝑓𝑜𝑟𝑚

𝐷_!!"# = 0 𝑖𝑓 𝑎 𝑝𝑒𝑟𝑖𝑜𝑑 𝑎𝑓𝑡𝑒𝑟 𝑚𝑎𝑟𝑘𝑒𝑡 𝑟𝑒𝑓𝑜𝑟𝑚

The second set of dummies indicates whether a world price increase or decrease has occurred:

𝐷_!∆_{= 1 𝑖𝑓 ∆𝑃} !< 0

𝐷!∆= 0 𝑖𝑓 ∆𝑃!≥ 0

Where ∆𝑃_𝒕 = 𝑃_𝒕− 𝑃_𝒕!𝟏.

As mentioned previously, in this research the assumption is made that the reforms started in 1992 in Honduras. In 1992 the SAPs were fully implemented by the IMF and at the same time foreign trade was liberalized; new programs replaced the import

(26)

26

substitution policies, and government interventions focused on tariff reduction and agricultural trade liberalization by removing price controls and guarantees. This should have, theoretically, brought the producer prices closer to the world market prices because private agents took over the role of governmental-owned marketing boards.

4.1.1. Test for Stationarity

To determine whether the data used is appropriate for the ECM model, the first step is to test whether the time-series are stationary and cointegrated. The analysis starts by testing whether both variables have a unit root. That is to test if the first differences of both variables are stationary. Stationarity refers to a probability distribution of the time-series to be constant over time and to be integrated of order zero. In addition, if two individual time-series are cointegrated, meaning the existence of a linear relationship between two non-stationary series that is stationary, then the first differential of the series is stationary and can be used in the estimation of the ECM (Jamora, Wurriehausen, Mengel, Greb, Cramon-Taubadel, 2016). If the variables are integrated of a different order, then there is no long-term relationship between the variables. The stationary properties of the price-time series are tested using the Augmented Dickey-Fuller2 test (ADF). If the null hypothesis can be rejected at the 10% Dickey-Fuller significance level, the price-time series can be considered stationary.

If the variables are stationary and a unit root is found, the Engle-Granger cointegration test is done. The cointegration relationship is the basis for the error correction model, and can be tested using a basic long-run model without a structural break using equation (3):

2_{The ADF test for a unit root without a trend involves estimating the following equation for a} time series variable 𝑃_𝒕: ∆𝑃_!= 𝜇 + 𝜆𝑃_!!!+ 𝜂_!∆𝑃_!!!+ 𝜀_! ! !!! where k is the number of lags of the first differences used. The null hypothesis is 𝜆 = 0, which is tested against 𝜆 < 0. The failure to reject the null hypothesis indicates that the time-series are nonstationary.

(27)

27 𝑃_!! _{= 𝛾𝑃} !! + 𝜏! (3)

Where 𝛾 needs to be estimated and can be interpreted as the producer price as a share of the world price for coffee since the constant is restricted to zero, and 𝜏! is the

random error term.

Three OLS regressions are performed and estimated: separate regressions for the period before and after the reforms using equation (3). The third OLS regression is a pooled regression with a structural break, where a dummy variable is used to take account of the shock that can lead to a jump of the variable of interest:

𝑃_!! _{= 𝛾}

!𝑃!!∗ 𝐷!!"#+ 𝛾!𝑃!! ∗ (1 − 𝐷!!"#) + 𝜏! (4)

Model (4) allows estimating different slope coefficients of the world price before and after the reforms. 𝑃_!!_{is equal to the domestic price,}_𝑃

!!∗ 𝐷!!"#is the world price in

the period before the reforms (<1992), 𝑃_!! _{∗ (1 − 𝐷}

!!"#) is the world price in the

period after the reforms (>1992) and 𝜈_! is the error term. 𝛾_! 𝑎𝑛𝑑 𝛾_! measure the share of the producer price in the world price in the period before and after the reforms. To evaluate the long-run relationship between the world and domestic price, the residual of the three OLS regressions is tested using the ADF test for a unit root.

4.1.2. The Final Model

Following Krivonos (2004) for the final model, given that the ADF test and the cointegration relationship support the validity of the variables, the equation tested is:

∆𝑃_!! _{= 𝛿}

!∆𝑃!!∗ 𝐷!!"#∗ 𝐷∆+ 𝛿!∆𝑃!!∗ 𝐷!!"#∗ (1 − 𝐷∆) + 𝛿!∆𝑃!!∗ (1 − 𝐷!!"#) ∗ 𝐷!∆+

𝛿!∆𝑃!!∗ (1 − 𝐷!!"#) ∗ (1 − 𝐷!∆) + 𝜃!𝜏!!!∗ 𝐷!!"#+ 𝜃!𝜏!!!∗ (1 − 𝐷!!"#) + 𝜀! (5)

Where ∆𝑃_!!_{is the domestic price at time t, ∆𝑃}

(28)

28

in the period before the reforms at time t, ∆𝑃_!! _{∗ 𝐷}

!!"# ∗ (1 − 𝐷∆) is a world price

increase before the reforms at time t, ∆𝑃_!!_{∗ (1 − 𝐷}

!!"#) ∗ 𝐷!∆ is a world price

decrease after the reforms at time t, ∆𝑃_!! _{∗ (1 − 𝐷}

!!"#) ∗ (1 − 𝐷!∆) is a world price

increase after the reforms at time t. In addition, 𝜏_!!!∗ 𝐷_!!"# is an interaction term between the predicted residual from equation (4) before the reforms at time t-1, 𝜏_!!!∗ (1 − 𝐷_!!"#) is an interaction term between the predicted residual after the reforms and 𝜀! is the error term.

Equation (5) predicts the absolute price transmission of a change in the domestic price as a result of a one-time shock to the world price. 𝛿! & 𝛿! describe the

short-run responsiveness of a world price decrease, and 𝛿_! & 𝛿_! describe the short-run responsiveness of a world price increase. The short-run responsiveness can be tested using the ADF test on the 𝛿s. The 𝜃s, the ‘error correction’ terms, are the parameters measuring the pre- (𝜃!) and post-reform (𝜃!) speed of adjustment of the domestic

price to the long-run equilibrium. This implies that the larger 𝜃s, the faster the domestic price will return to the suggested long-run equilibrium relationship with the world price. Cointegration can be tested using a t-test on the 𝜃s.

All prices used in the empirical analyses are in current US dollars. The implicit assumption used in the model is the fact that there exists a perfect price transmission of the exchange rate to domestic prices. The reforms are expected to strengthen the relationship between the world price and the domestic price as a consequence of the reforms. That is, the producer price as a share of the world price is expected to increase significantly as a result of reforms. Furthermore, both the short-run price transmission and the speed of adjustment of the domestic price to world price are expected to improve significantly as a result of the reforms. This is tested by the models describe above.

4.2. Data Description

(29)

29

in Honduras are used, expressed in US cents per lb. The time-period for this research reaches from 1980 to 2009. The time-period chosen for this research is wider than in the studies discussed in chapter 3. The reason for the wider time frame is as follows: since the 2001-2002 coffee crises had a significant impact on coffee prices as discussed in chapter 2, it may have an influence on the estimates of interest in this research and may give the wrong interpretation of the impact of the policy reforms in the coffee sector in Honduras. That is, choosing a wider time-frame will give the possibility to examine the long-run impact of the market reforms in the period after the coffee crisis when the coffee prices started to rise again.

The data used in this research is collected by the International Coffee organization. The prices paid to growers are the average prices paid to the grower at farmgate level as reported to the ICO by IHCAFE and establish the average of all grades purchased from the farmers. The ICO converts the Honduran currencies to US cents with the monthly average exchange rate published by the IMF. Although there is not an official definition for the world price of coffee, most studies use the commodity prices as published by the ICO. The ICO distinguishes four groups of coffee: Colombian milds, Brazilian milds, other milds and Robustas. Honduras’s main coffee export is Arabica coffee and belongs in the group other milds. The average price of other milds will be used as the world market price, also collected by the ICO. Many studies use the Indicator Prices published by the ICO. The indicator price is the price calculated based on the market share of exports of each group of coffee weights. However, the price of Robusta coffee is on average substantial lower than the price of Arabica coffee as can be seen in Figure 7. Since Honduras only producers the high quality Arabica coffee the price of Other Milds is taken as world price instead of the Group Indicator Price.

(30)

30

Figure 7: Price Series – Robusta and Arabica International Prices

Table 1 provides an overview of he descriptive statistics of the estimating sample. The variable World Price describes the world price of the Arabica Coffee and the variable Domestic Price describes the price received by producers in Honduras at farmgate level.

The mean of the domestic price in Honduras is 71 cents per pound and the mean of the world price is 118 cents per pound. The descriptive statistics are also given for the estimating sample before (Table 2) and after the reforms (Table 3). The mean and standard deviations of the world and domestic price are lower in the period after the reforms as can be seen from the data. One explanation for this might be the 2001-2002 coffee crisis, since the coffee prices reached a historical low in that period. After the coffee crisis, the coffee prices increased to a higher level than before the coffee crisis, which can also be seen in Figure 2. The average price received by the producers as share of the world price has on average increased from 57% prior to the reform to 63% after the reforms, measured by dividing the average domestic price by the average world price before and after the reform year 1992.

0 50 100 150 200 250 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 US c ents pe r lb Robustas Other Milds

(31)

31

Table 1: Descriptive Statistics of the Estimating Sample, 1980-2009

Table 2: Descriptive Statistics of the Estimating Sample Before Reforms, 1980-1991

Table 3: Descriptive Statistics of the Estimating Sample After Reforms, 1993-2009

Descrip(ve Sta(s(cs

Mean Standard Devia(on Minimum Maximum

World Price 118.38 35.92 61.52 194.69

Domes(c Price 71.11 22.99 34.32 121.71

Descrip(ve Sta(s(cs

World Price 131.07 29.82 84.98 194.69

Domes(c Price 74.94 19.95 50.35 121.71

Descrip(ve Sta(s(cs

World Price 112.61 37.23 61.52 189.06

(32)

32

5. Results

It is clear from Figure 2 that the Honduran coffee price has consistently been lower than the world price for Arabica coffee from 1980 onwards. As stated previously, the individual time series’ show that prices moved closer together right after the policy reforms until the coffee crisis in 2001, especially for price decreases, however after the coffee crisis the prices moved apart again. Before the reforms, the prices paid to the producers were on average 57% of the international coffee price, and after the reforms this share has increased to an average of 63% ((𝑃! _𝑃_!) ∗ 100%). According to these numbers, the reforms did improve the relationship between the world and domestic price partially, however a deeper analysis is required for a complete picture. It should be noted that even though Honduras has implemented reforms, remaining non-tariff barriers to trade that are not explicitly stated might underminethe effects. Therefore, the effect of liberalization should be investigated by empirical research.

5.1. Result of Stationarity Test

An ADF test is used to test whether the price variables are non-stationary, that is whether they contain a unit root. The estimated are reported in Table 4. The first price differences allow rejecting the unit root hypothesis at the 5% level for the world price and at the 1% level for the domestic price. So the first differences of the price time-series follow a unit root process, and can be used in the ECM, which is consistent with the overall findings in the literature.

(33)

33

Table 4: Stationarity of the domestic and the world prices

5.2. Cointegration Between Prices

Subsequently, the test for long-run cointegration between the world price and domestic price is estimated using equation (4) using the approach adopted by Engle and Granger (1987). In this method the domestic price is first regressed on the world price and a constant for the period before the reform. Subsequently, another regression is done for the period after the reforms. Thereafter, to examine the long-run relationship between the series, the residuals from the linear combinations are tested using the ADF test. The null hypothesis tested is that the world and domestic coffee prices are cointegrated and follow a unit root. If a unit root is found, then the residuals are not stationary and the null hypothesis of cointegration can be rejected. However, if no unit root is found, there is evidence that the variables are cointegrated. Stata has a standard code for the Engle-Granger test where the two-step approach is done at once, the results are shown in Table 5 below:

Augmented Dickey-Fuller Test (without trend) ADF sta;s;cs Price level First differen;al ICO Average World Price Other milds Arabica -2.086 -3.883*** Producer Prices Honduras -2.469 -4.151*** * Null of the unit root rejected at 10% significance ** Null of the unit root rejected at 5% significance * ** Null of the unit root rejected at 1% significance Yearly lag length is used

(34)

34

Table 5: Cointegration between the domestic and world prices

Table 5 shows no clear evidence for cointegration in the period before the reforms. An explanation may be that before the liberalization the degree of government interventions was high in the Honduran coffee market, so the expectation that the domestic price was driven by government policies rather than the world price is valid according to the estimates. In the period after the reforms clear evidence for cointegration is found between the world and domestic price. So it can be concluded that right after the reforms the domestic price is driven by changes in the world price rather than policy changes by the GOH. In the pooled model with structural breaks the null hypothesis of no cointegration is rejected at the 1% significance level. This relationship implies a long-run relationship between the two time series in the period before and after the reforms, so prices tend to convergence to a long-run equilibrium.

5.3. Prices Received by Producers

Next, an ECM was estimated for the Honduran coffee market for which cointegration with the world market was found. Table 6 shows the estimates of the target share of the prices received by producer as a share of the world price before and after the reforms (γ_! & γ_!) together with the ECM estimates. In Honduras the reforms significantly increased the target share of the world price received by the producers.

Augmented Dickey-Fuller Test (without trend)

ADF sta;s;cs

Reform year Before A?er With structural break

Honduras arabica 1992 -2.029 -3.681** -3.736***

* Null of the unit root rejected at 10% significance ** Null of the unit root rejected at 5% significance *** Null of the unit root rejected at 1% significance Yearly lag length is used

(35)

35

The target share increased from 0.57 before the reform to 0.63 after the reform. From the results of these estimates it can be concluded that the reforms did create a significant better integration between the world market and the domestic market.

Table 6: ECM with asymmetric price transmission and the target share of world prices

5.4. Results of Price Asymmetry

The estimates for the short-run price transmission are also listed in Table 6 and indicate that in all cases the short-term transmission has changed significantly. The short-run transmission increased from 0.66 before the reforms (1980 – 1991) to 0.73 after the reforms (1993 – 2009) for a world price decrease. This implies that negative world price changes are transmitted to the domestic market by 66% before the reform and by 73% after the reform. Meaning, that negative world price changes are easier transmitted to the domestic market now than before the reforms. Subsequently, the short-run price transmission decreased from 0.95 before the reforms to 0.65 after the reforms for a world price increase. So the short-run price transmission for a positive world price change decreased significantly after the reforms.

Concluded, the price transmission for world price changes shows significant results. The liberalization made the pass-through of negative world price changes easier than the transmission of the positive world price changes. This is an

Honduras

Type of coﬀee Arabica

Reform year 1992

Short-term transmission Price decrease Before reforms δ1 0.659***

ADer reforms δ3 0.726*** Price increase Before reforms δ2 0.945*** ADer reforms δ4 0.651*** Speed of adjustment Before reforms θ1 0.291 ADer θ2 0.093 Target share of the world price Before reforms γ1 0.570*** ADer γ2 0.625*** •  significant at 10% ** significant at 5% *** significant at 1%

(36)

36

unbeneficial outcome for the producers, since they bear the full burden of world price decreases but do not profit from world price increases in the period after the reforms. In addition, the estimates of the short-run price transmission show that the reforms had an asymmetric impact on the negative and positive price transmission. It can be concluded from the estimates that there exists price asymmetry in the coffee market in Honduras, since the domestic price reacts differently to a world price increase as to a world price decrease.

5.5. Adjustment Speed of Prices

The speed of adjustment is shown in Table 6 indicated as θ_! & θ_!. The speed of adjustment for the domestic price changed from 0.29 prior to the reform to 0.09 after the reforms. However, the speed of adjustment of the domestic price to the long-run equilibrium doe not show significant results in this research, so no real conclusion can be drawn from the estimates presented in this empirical research. This result is however not consistent with the literature. In most studies the speed of adjustment significantly changes as a result of the reforms. One reason might be that the reforms were far less restrictive in Honduras than in other countries. In order for the reforms to have an impact on the adjustment speed of prices, the scope of reforms should have been more persistent and far reaching.

(37)

37

6. Discussion

The previous section describes the main empirical findings for Honduras; this section will summarize the main results and compare them with the findings in the literature. In addition, the effect of the policy reforms is formulated and the answer to the research question is given. In addition, the implications of the research are discussed.

The cointegration analysis between the world and domestic price reported no cointegration relationship in the period before the reforms, but showed significant results in the period after the reforms. This could be explained by the fact that domestic prices were driven by government policies rather than world prices in the period before the reforms, but the liberalization changed this by opening up the domestic markets to international trade where domestic prices are driven by changes in the world price rather than policy changes by the GOH. Krivonos (2004) found similar evidence for a long-run cointegration relationship in 9 out of 14 countries in her study in the period after the reform.

The received producer price as a share of the world price has increased significantly in Honduras, from 57% prior to the reforms to 63% after the reforms. Moreover, it supports the expectation made in chapter 1, that the reforms would improve the integration between the world and domestic market. The study by Mofya-Mukuka et.al (2013), Worako et.al (2008) and Krivonos (2004), mentioned in chapter 3, all find evidence for the significant increase in the received producer price as a share of the world price as a result of the reforms. The significant estimates for Ethiopia reported by the study of Worako et.al (2008) showed a producer share varying from 45% in the pre-reform period to 65% in the post-reform period. Similar results are shown for India (Krivonos, 2004) and in Zambia (Mufya-Mukuka et.al, 2013). In India the producer share increased from 56% to 85% as a result of the reforms and in Zambia the share increased significantly from 61.5% to an almost

(38)

38

complete integrated domestic market were the producer price share is 96.7%. However, as mentioned previously, compared to the competing coffee producing countries presented, the significant increased producer share as a share of the world price remains low in Honduras. One reason for the small variation in the adjusted producer price as a share of the world price in the different countries might be the persistence and magnitude of the reforms per country.

Furthermore, the estimates of the ECM model showed mixed results. The liberalization made the pass-through of negative world price changes easier than the transmission of the positive world price changes from 66% to 73%, however the price transmission for positive world price changes decreased significantly in Honduras from 95% to 60%. Subsequently, the adjustment speed does not show any significant results for Honduras. Moreover, the results show less significant and less improved results as predicted in chapter 1. However, these results are consistent with the study of Worako et.al (2008) for the Ethiopian coffee market where the world price increase was transmitted on average 32% before the reform to nearly zero price transmission after the reforms. Subsequently, the pass-through for a world price decrease changed from 17% before the reforms to 26% after the reforms. In contrast, the study of Krivonos (2004) shows either unchanged or positive results for the short-run price transmission for world price increases in most countries studied. For example in Kenya, Tanzania, India and Uganda the price transmission increased significantly from close to zero before the reforms to on average of 40% after the reforms. In addition, the study of Krivonos (2004) shows significant increases in the speed of adjustment in most countries. Remaining non-tariff barriers to trade that are not explicitly stated may cause the limit impact of the reforms on the price transmission. These might undermine the effects of the reforms.

How do the findings presented in chapter 5 relate to the research question and to the policy reforms implemented in the Honduran coffee market: 1) have producers of coffee beans received a higher share of the world price for coffee after reforms in Honduras and 2) have the policy reforms strengthened the world coffee market and the domestic coffee market in terms of the nature and speed of transmission of world

The impact of coffee market reforms on producer prices : a case study of Honduras

Master of Science in

Economics: International Economics & Globalization

Master Thesis

The Impact of Coffee Market Reforms on

Producer Prices: A Case Study of Honduras

Statement of Originality

Abstract

Table of contents

1. Introduction

2. The Coffee Market and Economic Reforms in Honduras

3. Literature Review

4. Empirical Model

5. Results

6. Discussion