Studies on Market Integration and Price Transmission between Global and Domestic Markets Domestic Markets

results in price increases than shocks that cause price decreases. The Ethiopian grain markets were reported to be characterized by substantial market inefficiencies.

Kim et al. (2016) investigated whether the US rice markets were spatially integrated and whether these markets were integrated across varieties. They used vector error correction model, ADF and PP unit root tests, IRFs, and forecast error variance decomposition to achieve the objective of their research. The average monthly f.o.b. prices US$/cwt⁵ from major milling centers located in each specific region were used. The price series included: Arkansas long, Arkansas medium, Louisiana long, Louisiana medium, Texas long, California medium and California short. The analysis covered a period from January 1980 to July 2015. Arkansas-Missouri region was identified as the leading price in the variety of long and medium grain markets. Arkansas-Missouri medium grain was observed to play an additional important role in the long grain markets. California short grain market appeared to move somewhat independently (weekly exogenous) in the short-run, but its price movement is affected by Arkansas-Missouri medium grain in the longer term.

1.3.2. Studies on Market Integration and Price Transmission between Global and

and consumer prices (Baquedano et al., 2011). However, the mere transmission of price signals cannot be interpreted as success of trade policy reforms. That is, the transmission of global prices is not the only objective rather the degree to which domestic producers and consumers are better-off may be more important. Moreover, the magnitude of global to domestic price transmission is important on many grounds: (1) it is reflected in local prices, which in turn affect the welfare of producers and consumers by depreciating or appreciating their disposable income; (2) it determines the adjustment of producers and consumers; (3) it signals changes in the prices of other substitutes and complementary commodities; (4) it determines the policy response of governments; (5) it serves as a measure of market integration and success of market reforms; and (6) it signifies the stability of global market and vulnerability of countries to global shocks (Sharma, 2003; Dawe, 2009; Minot, 2011; Ghoshray, 2011; Abidoye and Labuschagne, 2014).

Food prices experienced a dramatic increase between 2007/08 and 2008/09, which in turn affected domestic prices in different countries. This was triggered by a complex set of factors such as (1) higher oil prices and increase in demand for biofuel, which led to increase in the prices of agricultural inputs such as fertilizers; (2) income growth in India and China;

(3) weather related supply shocks; (4) depreciation of U.S. dollar; (5) faster income growth accompanied by slow yield growth; and (6) under investment in agriculture in the earlier decades (Dawe, 2009; Minot, 2011). The so-called food price crisis undermined the economic assumption that free trade stabilizes domestic markets in the importing countries (Braha et al., 2015; Baquedano and Liefert, 2014). Some of the commonly reported effects on domestic markets of the global food price shock are: (1) reducing the disposable income or purchasing power of poor households (especially the net food buyer households in urban and rural areas), compelling them to spend more on cheaper and less nutritious foods by withdrawing income from other uses (Ghoshray, 2011; Minot, 2011; Baquedano and Liefert, 2014); (2) creating

balance of payment problems for the net food importing countries as a result of increased cost of imports; (3) igniting social unrest in the form of demonstrations and riots in some countries;

(4) imposing export restrictions and bans by some of the exporting countries; (5) depreciating trust of the importing countries in the stability and reliability of global food market as a source of food imports; and (6) resuming the interest of countries in food self-sufficiency, trade impediments and strategic grain reserves (Minot, 2011).

It is assumed that an increase in the prices of agricultural products is good for producers.

But, the simultaneous increase in the prices of agricultural inputs and outputs undermines the potential benefits to farmers of higher output prices. For instance, the global prices of urea more than doubled, i.e., increased by 107% between the fourth quarter of 2003 and 2007, when global prices of wheat, rice and maize were also experiencing increasing trend. This along with incomplete price transmission reduced the incentive of rice farmers in Asia, which in turn affected their supply response (Dawe, 2009). In addition to the supply elasticity and the extent of price increase (Dawe, 2009), the supply response may also be determined by type of farming, i.e., subsistence or commercial, the increase in domestic prices relative to global prices, the level of farmers’ market participation and the degree to which agricultural markets are integrated among themselves as well as with their respective global markets.

As mentioned in the outset, studies on price transmission and market integration are heterogeneous in many aspects, which make direct comparison of their results across countries implausible. Thus, some of the recent studies on price transmission from global to domestic markets are individually presented below.

Rapsomanikis et al. (2003) besides describing the concepts, estimation procedure and methods of measuring market integration and price transmission, also examined cointegration and price transmission among coffee and wheat markets using monthly prices from January 1990 to December 2011. Johansen’s cointegration test and vector error correction models

(symmetric and asymmetric) were employed. The findings showed that the coffee markets in Ethiopia and Uganda were cointegrated with their respective global markets whereas Rwandan coffee markets were not integrated to their world market. Also, the Egyptian wheat markets were cointegrated with the global wheat market, which was said to be motivated by trade liberalization and exchange regime reforms. The floor price policy of Egyptian government did not inhibit cointegration between domestic and global wheat markets. Similarly, the government auction system did not affect integration of Ethiopian coffee market to the global market. Contrarily, the export taxes and government procurement policies were found to prevent cointegration between Rwandan and global coffee markets. This indicates that government interventions have mixed effects on the global to domestic price transmission.

Conforti (2004) investigated price transmission from global to domestic markets as well as across supply chains for several agricultural commodities in 16 selected countries. Monthly and annual price series were analyzed using vector error correction and autoregressive distributed lag models to achieve the research objective. The study identified three regular features of price transmission: (1) the extent of price transmission was generally lower in the African countries whereas it was relatively complete in the Asian countries. But, the picture was somewhat more mixed for Latin America; (2) the magnitude of vertical price transmission between producer, wholesale and retail levels within the countries appeared to be generally higher than the extent of global to domestic price transmission; and (3) the high and fast price transmission was comparatively more frequent for cereals, followed by oilseeds whereas it was mostly slower for livestock products. Moreover, instances of price transmission were observed even for the well-known deeply regulated commodities such as wheat in Egypt, cereals in India and Pakistan and rice in Indonesia. This suggests that either government interventions cannot prevent domestic prices from following changes in the global prices or policy makers manage domestic prices in relation to trends in global market.

Imai et al. (2008) examined global to domestic price transmission for certain agricultural commodities, i.e., wheat, maize, rice, fruits and vegetables, in China and India. Annual price data from 1966 to 2005 were analyzed using error correction models to achieve the research objective. The results showed that adjustment of domestic prices towards their equilibrium with global prices, in the short- to medium-run (0 – 3 years), was generally faster for China than India. Besides, the magnitude of adjustment was observed to be relatively larger for wheat, maize and rice prices as compared to that for fruits and vegetables in both China and India.

Majority of the domestic prices were found to have a long-run relationship with their global counterparts but the extent of price transmission was generally incomplete, which was attributed to government interventions and failure to exploit spatial arbitrage opportunities.

Baquedano et al. (2011) conducted a study with a purpose to examine market integration and price transmission among the domestic and border prices of export cash crops (cotton and coffee) and import food crops (rice) in Mali and Nicaragua. The generalized error correction model was used in the data analysis. They found that Nicaraguan agriculture was more integrated with global market than that of Mali. Besides, the producer prices of cotton in Mali adjusted relatively slower to border prices as compared to that of Nicaragua. The elasticity of price transmission from border to domestic prices was reported to be larger for the exportable and importable of Nicaragua than Mali. This difference was attributed to Mali’s landlocked situation, poor road system and state control over the cotton industry.

Ghoshray (2011) researched global to domestic price transmission among selected food (rice, wheat, maize, tea, coffee, sugar, beef, palm oil, coconut oil and soybean oil) and non-food (rubber) commodities for certain Asian countries (China, India, Indonesia, Laos, Philippines and Thailand). Engle and Granger (1987), and Enders and Siklos (2001) cointegration tests and vector error correction model with threshold adjustment were employed in the data analysis. The study found that 9 of the 13 Asian markets studied were cointegrated

with their respective global market such that they adjusted asymmetrically towards the long-run equilibrium. Indian rice, Indonesian sugar, and Chinese sugar and soybean oil markets were not cointegrated to their corresponding global markets. The elasticity of price transmission was reported to range from 0.11 to 0.89 with a median value of 0.51% or 51%. This is close to the price transmission elasticity reported by Minot (2011) for sub–Saharan Africa. However, both indicate incomplete price transmission in the Asian and African markets. Although India and Philippines somehow stabilized domestic prices, changes in the global prices were found to be transmitted into their domestic markets. The study reported asymmetric price transmission for tea in India, rice in China and Philippines, beef in Thailand, and soybean oil in China such that decreases in the global prices were incompletely or slowly transmitted to domestic markets than increases. Government interventions may or may not impede market integration, but the equilibrium adjustment was observed to be relatively slow in markets that are characterized by government interventions. The study suggests Asian countries to either move towards self-sufficiency by increasing productivity or restrict imports.

Minot (2011) studied the relationships between global and domestic prices of food grains (maize, wheat, rice and sorghum) in 12 sub-Saharan African countries. The study analyzed a total of 83 monthly price series with periods ranging from March 1994 to December 2008. In addition to simple trend analysis, Johansen’s cointegration tests and linear vector error correction models were used in the data analysis. It was reported that 13 of the 62 price series were cointegrated with the global market. The largest share of cointegrated price series were observed for Malawi, Mozambique and Ethiopia whereas none of the foodgrains prices in Zambia, Uganda and Kenya were integrated to the global market. Moreover, rice markets were found to be better connected to the world market as compared to maize markets. This was attributed to the reliance of most sub-Saharan African countries on rice imports and their closeness to self-sufficiency in maize. The study estimated a median elasticity of price

transmission of 0.54. This implies that 54% of a change in the global prices of food grains may be transmitted to domestic markets in sub-Saharan Africa. The study suggested self-sufficiency policies and negotiation about reducing export restrictions through the World Trade Organization (WTO).

Greb et al. (2012) examined price transmission between global and domestic markets of cereals in developing countries and compared their estimations with those of previous similar studies (30 papers). The study used monthly price data for wheat, rice and maize covering all the countries included in the GIEWS database of the Food and Agricultural Organization of the United Nations (FAO). The results showed that 79% of markets pairs in the literature sample were cointegrated whereas their own estimates showed 43% of the markets pairs to be cointegrated with the global market. Although this difference was attributed to previous studies authors’ bias in reporting only significant results, this study also suffers from limitations such as using a single specification of error correction model for all markets pairs, which may question the validity of the estimates. However, the average elasticity of price transmission was similar between the literature sample and authors’ estimates, i.e., 0.74 and 0.76 respectively. This indicates that about 75% of a unit change in the global prices was transmitted to domestic markets in developing countries. The speed of adjustment towards equilibrium is reported to be slow across countries and commodities with an average value of -0.11 in the literature sample and -0.09 from the authors’ estimates. It is mentioned that Thai 5% broken rice and US No.2 Hard Red Winter Wheat were commonly considered as global reference prices for rice and wheat in 55% and 24% of the cases, respectively.

Baltzer (2013) reviewed and summarized the evidence on transmission of global wheat, rice and maize prices to domestic markets in fourteen developing countries during the so-called global food price crisis of 2007–2008. It draws on several case studies one for each of the selected countries, viz., Bangladesh, Brazil, China, Egypt, Ethiopia, Kenya, India, Malawi,

Mozambique, Nigeria, Senegal, South Africa, Vietnam and Zambia. The study reported a great deal of variations in the pattern of global to domestic price transmission with almost no price pass-through in China and India, very close relationship between global and domestic prices in Brazil and South Africa, and substantial price spikes in Ethiopia and Nigeria. Price stabilization policies, public policy failure, incomplete market integration and coinciding domestic shocks were considered the possible determinants of these variations.

Abbott et al. (2014) investigated the relationship between border and domestic prices in Vietnam with a purpose to determine the influence of inflation on domestic prices. The study was based on aggregate annual prices from 1999 to 2008 for different sectors of the economy.

A number of logarithmic and first difference models, some of which have error correction term, were employed in the analysis. The findings revealed that 15% of agricultural export sectors, 18% of agricultural import sectors, 40% of manufacturing export sectors and 18% of manufacturing import sectors were integrated with the global market. Conversely, 38% of agricultural export sectors, 33% of agricultural import sectors, 20% of manufacturing export sectors, 35% of manufacturing import sectors and 100% of energy import sectors were considered to be segmented from the global market. Moreover, intermediate, non-tradeable and protected goods were less-likely to be integrated with the global market. The results indicated the presence of imperfect price transmission in majority of the sectors, suggesting imperfect transmission of prices from global to domestic markets. Lastly, domestic prices were explained better by inflation than the global prices.

Abidoye and Labuschagne (2014) examined the relationship between global and domestic maize prices in South Africa. The study used monthly maize prices from January 2000 to December 2010 and employed a three-regime threshold vector error correction model to achieve the study objective. It is reported that the South African and global maize prices were cointegrated in Regime-1 and Regime-3 but they were not cointegrated in Regime-2. The

elasticity of price transmission from global to local maize prices was 0.98 in Regime-1 and 0.97 in Regime-3, which denotes almost perfect price transmission. That is, 98% and 97% of changes in global maize prices were eventually transmitted to South African maize markets in Regime-1 and Regime-3, respectively. While the elasticity of price transmission was similar in the two regimes, the speed of adjustment toward equilibrium was showed to be faster in Regime-3 (0.46) than Regime-1 (0.36).

Baquedano and Liefert (2014) studied the long-run relationship between urban consumer markets in developing countries and global markets for selected agricultural commodities.

Monthly prices of wheat, rice, maize and sorghum for 61 countries during 2000s were analyzed using the single equation error correction models. The findings showed that wheat, rice, maize and sorghum prices in the developing countries under study had a long-run relationship with both world prices and real exchange rates. Besides, the extent of global to domestic price transmission was low, i.e., ranging from 0.16 for sorghum to 0.32 for wheat, for the commodities studied, which suggest only partial transmission of global price changes to the domestic markets. Similarly, the adjustment of domestic prices to their long-run equilibrium with global prices was slow with an average value of 0.15. The time required to restore equilibrium was minimum for maize and rice but it was the lowest for wheat. It was reported that cointegration was more common for the commodities that were heavily traded such as wheat and rice as compared to thinly traded products of maize and sorghum.

Braha et al. (2015) examined spatial price transmission from global to domestic markets of certain agricultural and livestock commodities, viz., wheat, maize, barley, beef meat, and chicken meat, in Kosovo. Monthly price data were used for a period from January 2005 to December 2012. Johansen’s and threshold cointegration tests, and vector error correction models with TAR and M-TAR adjustments were employed in the data analysis. The study reported the existence of long-run equilibrium relationship between global and domestic prices.

Asymmetric adjustment was observed in wheat, beef, barley and chicken markets. The price transmission was one way from global to Kosovan markets and the country was regarded as vulnerable to shocks in global food prices.

Irazou (2015) investigated the relationship between the global and Togolese rice prices during January 1991 to December 2013. Linear and non-linear cointegration tests and error correction models were used in the data analysis. The study found that Togolese rice prices had a long-run equilibrium relationship with the global rice prices. Almost 80% of changes in global rice prices were transmitted to local rice markets in the long-run. In addition, Togolese rice prices were reported to adjust asymmetrically with respect to the global prices such that the adjustment was faster to negative than positive deviations from the long-run equilibrium.

That is, increases in global rice prices were transmitted faster than decreases. The asymmetric price adjustment was attributed to the presence of market power, high transport cost and government interventions.

Selliah et al. (2015) studied the effect of global food price spikes on domestic inflation process in Sri Lanka using consumer price indices (all items, food and non-food) wholesale price index and global food price index for a period from January 2003 to December 2013.

Johansen’s cointegration test, vector error correction models and impulse response functions were, inter alia, used in the data analysis. The results revealed that the pairs of domestic and global price indices were cointegrated. Global food prices were likely to be transmitted to domestic prices with the elasticity of price transmission being larger for food than non-food prices. In addition, transmission of global food prices was reported to induce domestic inflation in the country. The study invited short- and long-run measures for ensuring food security and price stabilization in Sri Lanka.

1.3.3. Studies on Market Integration and Price Transmission among Global Markets