1

Final Draft, September 16, 2003

ASSESSMENT OF IMPACTS OF TRADE LIBERALIZATION: THE COLOMBIAN’S RICE SECTOR

FEDERACION NACIONAL DE ARROCEROS DE COLOMBIA THE UNITED NATIONS ENVIRONMENT PROGRAMME

2

SUMMARY Colombian rice sector was the exception among the short cycle crops that reduced in 29% of cultivated area during 1990’s, as consequence of the implementation of the new economic development model named “economic openness”. In Colombia, this model consisted of reduction of tariffs and protection for a whole range of goods, mainly food. During this same period, food imports increased about 388%; surprisingly, products such as rice and maize increased in a high rate, and they are highly produced and consumed in Colombia. Colombian rice sector reactivated since 1996 due to the government stimulus to accomplish stability in prices such as storage subsidies and protection from CAN (Andean Community of Nations) and third parties (price fringes, administered trade, etc.). In spite of that, analyzing price behavior in real terms for the period 1991-2001 shows a decrease of 38%. A similar behavior was presented by the production costs because of foreign exchange rate trend and rationalization of inputs, achieving a 28% decrease. Seed variety Fedearroz-50 contributed to an increase in average productivity in more than a ton per hectare, reducing real costs per hectare in 34%. These latter results of price and cost decreases in magnitudes over 30%, could make think of excellence of Colombian rice competitiveness. By 2003, Colombia not only presents advances in competitiveness but also is further from international quotations for rice price. There are two main reasons: the first is that international prices decreased in 50% during the second half of 1990’s. The second is that subsidies, the main international quotation component, increased in 280%. It resulted interesting to observe the behavior of national rice prices versus the international quotations, proving that there is a strong relationship between both markets despite the protection mechanisms implanted for rice in Colombia. In the first section of this document, a connection between Colombian and international rice market was searched. In the former, there is a substitutability of rice for other imported goods such as wheat and maize that pulled prices down as imported good prices were lowering down. Second topic was an evaluation of the economic impact of economic liberalization over Colombian rice sector, measuring impact of international subsidies over several variables related to rice production at Colombia. Reduction of cultivated area for the short cycle crops in Colombia was caused by low prices for substitute goods in consumption and land use, leaving as only alternative rice cultivation. Growth of this crop and the lack of alternatives for cultivators, created a situation of mono-cultivation with serious implications for environment. Rice cultivation used to be rotated with sorghum, soy, maize and cotton, and the reduction of profit levels for production induced producers to stay with rice for every year long to maximize income. Increase of rice productivity through a sort of technological alternatives established new challenges to solve the troubles coming from mono-cultivation, characterized by its genetic, physiologic, and morphologic uniformity. First, loss of nitrogen has increased sowing density, encouraging the development of production-adverse factors such as

3

weeds, with consequences over the cost level in terms of active ingredient applied to cultivation, and a deterioration of environment. On the other hand, this study identified the environmental impact of every physical element affected and analyzed the advances achieved to use environment in an appropriate way. Integrated Pest Management is suggested as the most effective weapon to reduce the use for agrochemicals. Another important aspect involved in the production chain has been the “Convenio de concertación para una producción más limpia”, and an agreement with the agro-industrial sector to solve the problem of rice hull management. The Union of Producers and the Environment Ministry designed the environmental guide for the rice sub-sector, focused to guarantee production sustainability. Other national level agreements started to be implemented inside of rice production chain for water use, among other factors, and there is a whole list of the proposals to ensure the appropriate management of cultivation. Inside of the cost study, at the next section, it was observed that the only activity that increased in absolute value was use for agrochemicals. To understand this aspect, there was developed an econometric model based in cost functions to analyze changes in productivity for agrochemicals, finding that productivity for rice in general decreased during 1990’s. The increase of costs was via quantities and not via prices of agrochemicals per hectare. Another interesting result was that Fedearroz-50 rice variety reduced used quantities of chemicals and fertilizers. Finally, some substitution possibilities for chemicals were found such as labor. Execution of this document had an special meaning for rice growers union, because at its very beginning there was the definition of the path that international trade would take due to the creation of ALCA, ATPEA, and the possibility of an agreement with United States of America and also the Mercosur. This research worked as a way to show the consequences of 10 years of trade liberalization and gave all the analysis tools to anticipate the consolidation of trade agreements for the next years.

4

ACKNOWLEGMENTS

Several institutions and individuals cooperated to become to the end of this Project. The Federación Nacional de Arroceros de Colombia (Fedearroz) sponsored and implementes the proposal to the United Nations Environmental Program (UNEP) to evaluate the impact of economic liberalization over rice. Technical staff from Agriculture, Environment, and External Trade ministries contributed criticizing first drafts of this document. Finally, to Universidad del Rosario for allowing one of its Master in Economics students finish his thesis with us. Major acknowledgment must be made to Technical Committee composed by: Jaime Jiménez, Jairo Hómez, Diego Herrera, Edwin García and José Levis under direction of Néstor Gutiérrez. They are responsible for each one the topics on the final document. Other people that helped out in the interested Comitee were: Humberto Vásquez, Alberto Navarro, Henry Ramirez and Ciro Quintana from Usocoello, Yesid Castro consultant, Samuel Zambrano and Oscar Ballesteros from Departamento Nacional de Planeación, Jaime Pérez, José Ignacio Torres and Alejandro Peláez from DANE, Ivan Sombrerero and Sandra Avellaneda from Induarroz, Jorge Enrique Cubillos from Moliarroz, Alejandro Vélez and María Elvira Núñez from SAC, Francisco Acevedo, and Irma Baquero from CORPOICA, Jorge Torres, Alberto Mejía, Juan José Sierra, Gonzalo Sarmiento rice producers, Luis Eduardo Quintero and Ricardo Sánchez from Ministry of Agriculture, Nohora Helena Cruz from Bolsa Nacional Agropecuaria, Ruth Suarez and Fabio Buitrago NGO consultants, Milton Borrero, Lorena Prieto, Hernando Herrera, Rosa Lucia Rojas and Carlos Guzmán from FEDEARROZ, Fernando Barberi y Diana Martínez from Universidad del Rosario, Jorge Enrique Muñoz from Universidad de los Llanos, Luis Sanint from FLAR. From UNEP initiative was promoted by Hussein Abaza and from Federación Nacional de Arroceros by Rafael Henández L., CEO. Logistic support came form UNEP’s Sophie Forster, Rahila Mughal and Desiree León, and Lilí Sarmiento from FEDEARROZ. Project was financed by UNEP and Federación Nacional de Arroceros de Colombia with resources coming from Cuota de Fomento Arrocero.

5

PROLOGUE

Last decade of Twentieth Century was characterized by a movement for trade liberalization to encourage economies that had competitive advantages producing tradable goods characterized last decade of Twentieth Century. Low Developed Countries (LDCs) adopted this neo-liberal model making it easier to trade goods, reducing tariffs and unifying aids and commercial restrictions in subsidies. Illusion of these new paradigms was to increase share in agricultural trade, mainly. First years of twentieth-first Century have registered an economic recession process in almost all of the LDCs, with high unemployment rates, income concentration, and lower growth rates. Food imports grew spectacularly in those countries while in developed countries exports and employment in agricultural activities raised. Main interest from international institutions monitoring welfare and development indicators is to evaluate sustainability of the development model induced by trade liberalization between countries. A way to understand its impact in LDCs economies is to study agricultural goods. For this reason, rice cultivation was picked up because it is a semiannual crop that is basic as food for low-income population and one of the employment generators in rural zones. Behavior of rice sector was evaluated during the last 13 years to measure direct and indirect impact of trade liberalization. It was found that performance of the sector was linked to the international markets in spite of being protected. Consequences of sector exposition to new trade policies generated negative impacts in social and economic terms, and in environment. As a complement, a mathematical model was calculated to evaluate input and output productivity in the sector. Conclusion of document is that in spite of real cost and producer price reduction, rice sector is every year more vulnerable to international markets because of distortions coming from aids and subsidies of price-forming countries that are the major rice exporters.

6

I. INTRODUCTION, COLOMBIAN RICE SECTOR: 1990-2002

Adoption of economic openness happened by the very beginning of 1990’s just as in several Low Developed Countries from Asia and South America. Main component of such model consisted of trade liberalization with a reduction on tariffs. One of the main observed consequences of this process on agricultural sector during the last twelve years are reduction of harvested areas, mainly of cereals, and an increase on imports on those products. By 1991, Colombia imported 816.000 tons of agricultural products; by 2001 the amount reached 3’986.000 tons. Main increases of that value were maize that went from 8.000 tons to 1’579.000, and products such as beans, sorghum, soy, wheat, and rice.

Behavior of rice during that period is very important because it is main food source of low-income households and because it is an employment supplier at rural zones. Of short-cycle crops, maize occupies first place in terms of harvested area with 575.000 Has, rice is in second with 448.000 Has, and thereafter the rest of crops (cassava, potatoes, beans, sorghum, cotton, ñame, wheat, tobacco, sesame, barley, and peanuts) with less than 200.000 Has. During the last 12 years, harvested rice area presents a whole cycle (Graph 1.1), starting 1990 (400.000 Has) with a downward trend until 1996 (270.000 Has), and since then an upward trend reaching 490.000 in 1999. During last 3 years, harvested area has fluctuated because of weather conditions.

GRAPH 1.1RICE CULTIVATED AREA IN COLOMBIA, 1990 - 2003.

250,000

300,000

350,000

400,000

450,000

500,000

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

YEARS

Hectárs

Source: Arroz en Colombia, 1980-2003 FEDEARROZ.

7

Harvested area decreased 29% for semiannual crops from 1990 to 2001, as Graph 1.2 shows. Making a comparison, only rice and vegetables, that occupies a small area increased for that period. Mechanized rice harvested area increased 20% in that time frame. One can deduct from those facts that increase in rice was given for an increase of small farming by a reduction of rotation between rice and other substitutive transitory crops.

During the last 3 years, rice production has been over 2 million of dry paddy rice with a cycle during the year, as Table 1.1 shows. Commonly, Colombia has a deficit in rice production during first semester of the year, and a superavit during the second, which requires storing rice from the latter to the former period. In average, Colombia has a 5% deficit to be covered with imports. Crop productivity has had changes since 1966 when “Green Revolution” techniques showed up. Production went from 2 to 5 tons of green paddy rice per hectare. During the next two and a half decades, technological efforts centered in keeping those yields, that increased positively in one hectare since 1997 with a new seed variety: Fedearroz 50. Currently, average yield in rice is 6.5 tons per hectare of green paddy or 5.2 of dry paddy. It is important to point out that Colombia, a tropical country, produces two harvests a year, for a total production of 10.4 tons of dry paddy per hectare a year.

0

20

40

60

80

100

120

140

Cotton

Barley

Soy B

eans

Sorghu

mW

heat

Benn

etCorn

Kidney

Beans

Cassa

va

Pottato

s

Vegeta

bles

Mecha

nized

Rice

1990 2000Source: SAC, CEGA,FEDEARROZ.

%

GRAPH 1.2 SEMIANNUAL CROPS (Comparison). 1990 vs 2000, 1990 = 100

8

TABLE 1.1 AREA, PRODUCTION AND YIELDS: MECHANIZED RICE

(IRRIGATED+UPLAND RICE), COLOMBIA, 1990 – 2002.

Hectares Tons of Dry

Paddy Tons of Dry

Paddy per Ha

1990 371,965 1,633,455 4.39

1991 332,594 1,475,112 4.44

1992 316,180 1,441,342 4.56

1993 274,545 1,277,387 4.65

1994 297,587 1,400,103 4.70

1995 296,717 1,386,082 4.67

1996 256,450 1,213,583 4.73

1997 262,934 1,177,625 4.48

1998 306,332 1,505,589 4.91

1999 468,031 2,330,085 4.98

2000 447,553 2,204,950 4.93

2001 448,999 2,147,282 4.78

2002 * 407,909 2,016,025 4.94 * Estimated Source:FEDEARROZ

Dynamics of Colombian rice sector is easy to evaluate because there were two national censuses in 1988 and 19991. Number of rice producers increased during the period between censuses from 17.517 to 28.128 (60.5%). Number of rice harvested farms also increased from 19.779 to 33.435 (67.3%). Colombian agriculture frontier has not changed significantly during the last 30 years because irrigation districts have not been modified during that period. It is easy to conclude that the size of rice production units decreased to absorb the number of producers. Majority of Colombian rice producers have cultivations of less than 3 Has, only 1.7% of cultivations have areas over 100 Has. Small producers that obtain their income and employ their labor force in this activity produce Colombian rice. Optimal land and infrastructure have a “sui generis” process during the last years for warm weather crops such as rice, sorghum, and others because of generation of money from drug traffic. Land has become in a sort of saving institution with low risk, increasing its price. For this reason, majority of rice producers (54.1%) rent land instead of owning land. From this fact appear a lot of consequences: first an excessive valuation of that scarce resource. Second, main part of rice producers rent land because they don’t have enough

1 Fedearroz, 1990-2000, I y II Censo Nacional Arrocero, Bogotá.

9

capital to buy it. Third, management of land and water from a land renter is different from the management an owner would do. Rice generates 65.000 direct employments among agro-industry and technical assistance. But main impacts coming from rice is based on economic activity influence along the zones it is produced; nearby cities depend almost entirely on this activity. The whole commerce from input trade, bank activity, mechanic workshops, markets, urban and rural transportation, health services, oil trading, etc, depend directly on rice production. Multiplier factor coming from this activity is important for income generation for tropical Colombian zones. For this reason, when it is said that imports of cheaper goods to Colombia benefit consumers, “rice community” is not taken into account and a loss of purchasing power in that community would imply a loss for a big community. On the other hand, rice consumption has increased during the last years. Between 1980’s and 1990’s, rice consumption showed a decreasing from 40 kilos per capita in 1980 to 21.8 kilos per capita in 19972. From 2000 National Statistics Department (DANE) estimates in its Household Survey, used to calculate consumer price index, rice consumption. They found that consumption for year 2000 was 40 kilos per capita and 41.5 for 2001. They also found consumption levels of about 63 kilos per capita at Colombian Atlantic Cost, and lower levels for central zone, that contains main part of population. Higher consumption levels were found at Cartagena and Montería at Colombian North Zone. Consumption depends on income level, and it has been going down. In spite of that, rice consumption has increased3, a reason to think about a general impoverish of Colombian people, who substitute expensive goods for nutritive goods with lower relative price. That is, a moving towards “inferior” goods for low-income households. For Colombian case, when relative prices of goods such as meat, milk products and vegetables increase, consumption of rice increases. Comparing Colombia, Peru, and Brazil with some other Latin American countries, rice consumption is about 42 kilos per capita while average is lower for other countries. For example, Venezuela reaches 15 kilos per capita while weighted average for Latin America is 27 kilos per capita of white rice. Colombian rice sector has important institutions from behind. There is a Union that represents producers from 60 years ago. This institution is financed by a 0.5% tax over product sells. Fedearroz researches and spills technology, but also union trades producers. Five years ago, Colombian Agriculture Ministry created National Rice Council consisting of related unions, Agriculture, Trade, and Economics ministries. Its main function is to join main aspects related to rice such as need for imports, minimum prices, and trading conditions for rice. Some topics concerning research, consumption, and product trading are discussed. Biological research has played an important role in Colombian rice crop. By 1961, paddy rice yields were about only 2 tons per hectare. Between 1966 and 1973, yields increased by 5 tons. During the latter period, “Green Revolution” occurred leaded by Colombian Agricultural Institute (ICA), International Center of Tropical Agriculture (CIAT) and Fedearroz. These institutions shared basic, management, and technological diffusion research until 1980’s. Since then responsibility is exclusive for Fedearroz.

2 Fedearroz, 2001, Arroz en Colombia 1980-2001. Bogotá, Anexo 19, pg.160. 3 DANE, Encuesta Nacional de Hogares.

10

During last five years, rice yields in Colombia showed an important increase of more than a ton per year because the new seed variety. In a comparison among other Latin American countries, Colombia has a similar technical performance. Yields for irrigated rice, in terms of dry paddy rice, is about 5.2 tons per hectare. Brazil, Uruguay, and Argentina that produce only one crop per year only improve this performance. One of Colombia’s advantages with respect to other American producers is the production of two crops per year with an important sort of varieties. This fact facilitates regional specialization and reduces risks of massive production of one only variety. By 2002, producers had 16 excellent culinary quality and acceptation varieties. PRICES Colombia adopted a competition-based price system. There are basically four price levels: a raw materials level that is paid directly to producers in terms of green paddy (25% of humidity and 5% of impure matters). This price depends on grain quality as a percentage of humidity, and over all of location from industrial zones. It could be said that differences between same quality rice are given by transportation costs. Second price level corresponds to whole sales processed white rice. This rice is traded in 75 kilo bulks and it is used as raw material for final product distribution in packs of less than 5 kilos for sell. There is a market for packed rice in the country in a sort of varieties: 1 kilo, 5 kilos, 12.5 kilos. It is usual to find positioned brands in the market. Quality of this rice depends on percentage of fractionated grains. At consumer level, there is another price level, the highest. Final consumer buys rice at the markets, and local shops for consumption that ranges into one or two weeks. At low-income levels, it is used to buy rice daily. Absolute price level for prices at producer level was about U$177 for a ton of green paddy or U$209 for a ton of dry paddy. White rice for whole sale purposes had its price about U$342 for whole sales and U$464 for consumers, per ton. Relationship between rice price for consumers and green paddy is about 2. Relationship between rice price for consumers and raw materials is about 2.6. Rice market has at least three agents: producers, industrial producers, and traders. The number of agents determines competition at each level. At the producer level there are many agents, approaching a competition behavior. At the second level there is a reduction of agents, and currently there are about 130 participants. The number has been lowering during the last ten years. The lower number of participants is located at the third level, there are two or three agricultural industries and three big supermarkets. This last group belongs to the people that imports price at lower prices than national rice. Prices at producer and white rice level have seasonal along the year. During first semester, raw materials have a price increase and during second semester a decrease. This phenomena is caused by a surplus in August, September, and October. Graph 1.3

11

analyses price variability for the last twelve years, showing a reduction during the period 1997-2001 because of imports and a subsidy to incentive rice storage.

GRAPH 1.3 SEASONALITY OF PRODUCER REAL PRICES, COLOMBIA, AVERAGE 1989-1996, 1997-2002

95.00

97.50

100.00

102.50

ENE FEB MAR ABR MAY JUN JUL AGO SEP OCT NOV DIC

1989-1996 1997-2002

Percentage

Source: Arroz en Colombia 1980 - 2002, FEDEARROZ.

Note: Indice calculated with the average movil related method

There are some other factors that affect price behavior for Colombian rice, besides seasonality. Expectations for imports affects price level mainly during harvest season. Minimum price level approved at National rice Council, substitute goods prices (such as potatoes, spaghettis, and bread with a 95% imported component), incentives to storage, administered trade agreements, and available income behavior are the main variables that affect price levels in Colombia. In monetary terms, for the 1989-2002 period price has been increasing from $ 84.134 to $496.469. In real terms there is an opposite trend, a reduction of about 38% during the same period, as Table 1.2 shows. This fact means that rice has been lowering Colombian price index.

12

TABLE 1.2 REAL AND CURRENT PRICES FOR GREEN PADDY RICE

FIRST SEMESTER 1989-2002 CURRENT REAL $ / Ton (1978 = 100) / Ton

89/A 84,148

9,290

90/A 92,141 7,959 91/A 111,594 7,336 92/A 133,524 6,889 93/A 139,236 5,824 94/A 200,047 6,785 95/A 221,326 6,192 96/A 269,151 6,270 97/A 339,378 6,636 98/A 430,676 7,042 99/A 409,364 5,946 00/A 413,254 5,485 01/A 515,222 6,331 02/A 496,469 5,741

Source: FEDEARROZ.

Colombian rice prices in dollars maintain a margin with relation to main exportation markets around the world. There is no causality relationship as Graph 1.4 shows.

- 38 %

13

GRAPH 1.4. REAL PRICES IN DOLARS, WHITE RICE, COLOMBIA, UNITED STATES, TAILAND AND

VIETNAM. 1996 - 2002.

100

200

300

400

500

600

700

800

ENE 96 ENE 97 ENE 98 ENE 99 ENE 00 ENE 01 ENE 02

US$ / Ton

Source: FEDEARROZ, Creed Rice Market Report.

COLOMBIA

UNITED STATES

TAILAND

VIETNAM

COSTS Production costs depend on technological level, production system, input prices, and production zones. For Colombian case, there are two production systems: Irrigated Rice and traditional upland rice. The former is based on water availability through productive cycle, it is highly mechanized, uses modern inputs, and has good yields. In some activities, capital and labor are combined. Traditional upland rice uses rainfall and depends on rain regime. Distribution of rain through the year varies depending on the geographical location. Technological level is the same as irrigated rice. Risk levels are higher on this system because of randomness of raining seasons. Main problem of this system consists of activity concentration, mainly at harvesting and transportation of the crop, causing huge storages in small time frames. Production system does not determine the size of productive units nor property. There are producers of every size and kinds in both systems. 67% of mechanized rice belongs to irrigated rice, and 33% to traditional upland rice. There is a third production system named manual traditional upland rice, produced at the jungle zones of north Colombia and in Chocó. Rice is sown manually by the shore of rivers in small properties for self-consumption. Yields are about 1 ton per hectare. Occupied area by this production system is 6% that corresponds to 1.5% of total production. However, almost half of Colombian rice producers sow this kind of rice. Importance from this system is that it is the base for feeding communities isolated from markets. In absolute terms, irrigated rice cost is U$1.000 per hectare, as Table 1.3 shows. Around 57.2% from that amount correspond to land rent, sowing, fertilization, and weed control. In the majority of rice sowing countries, land rent does not belong to costs because it is taken

14

as a part of financial benefits from land return. In Colombia, more than a half of rice producers (irrigated and traditional upland rice) are land renters, this value is the first cost component to be paid in cash. For the case of sowing process, seed value is the most important item, with a density of about 230 kilos of seed per hectare. Fertilization occupies second place as a percentage: 300 kilos of urea, 50 kilos of DAP, 100 kilos of KCL, distributed into four applications during 4 months of sowing process. Costs coming from weed control have been increasing, and it corresponds to 13.1 % of total cost. In average three weed control applications are made. Need for constant rice sowing and bad land management have increased use of agrochemicals. Integrated Pest Management (IPM) is a positive sum practice for Colombian case. This has turned into a common practice that consists of an evaluation of infesting levels and risks to take a decision to manage pests. There is a series of indicators about damage levels to apply or not to apply insecticides. Use of insecticides has reduced, turning into only 2.2% of total per hectare cost.

15

TABLE 1.3 ESTIMATED NATIONAL PRODUCTION COSTS PER HECTARE,

PERCENTAGES , IRRIGATED RICE, FIRST SEMESTER 2002. NATIONAL SHARE AVERAGE 02/A Pesos % 1.- TECHNICAL ASSISTANCE 26,042 0.9 2.- LAND RENT 476,549 16.2 3.- PREPARATION 150,431 5.1 4.- SOW 363,410 12.3 5.- IRRIGATION 183,410 6.2 6.- FERTILIZATION 459,645 15.6 7.- WEED CONTROL 385,989 13.1 8.- PEST CONTROL 68,057 2.3 9.- SICKNESS CONTROL 180,678 6.1 10.- RODENTS AND DESPALILLE 34,634 1.2 11.- RECOLLECTION 215,105 7.3 12.- TRANSPORTATION 100,639 3.4 SUB-TOTAL 2,644,589 89.7 13.- MANAGEMENT 79,338 2.7 14.- INTERESTS 177,120 6.0 15.- DIRECT TAXES 46,248 1.6 TOTAL COST / HECTARE 2,947,295 100.0 ton/ha GREEN PADDY 6.16 COST / TON. GREEN PADDY 478,612 Ton/ha DRY PADDY 5.23 COSTO / TON. DRY PADDY 563,073 Source: FEDEARROZ.

Land preparation is relatively cheap at Colombia because depreciation of machinery has been taken into account years ago. Two kinds of rakes are used two or three times a season. With relation to costs between production systems, water, fertility, and rent make the difference. Zones with irrigation districts have high land rent costs, even 6 times higher than traditional upland rice. Yields per hectare have increased during last years because of Colombian Peso depreciation. Producers have adjusted to these changes and cost has reduced 34%, Table 1.4 shows. This change should have benefited sector’s competitivity levels with respect to international community, however, it is not enough to reduce the gap because of drastic reduction of international prices.

16

TABLE 1. 4 PRODUCCION COSTS PER TON, CURRENT

AND REAL PRICES, COLOMBIA, FIRST SEMESTER , 1989-2002

CURRENT REAL

$ / Ton (1978 = 100) / Ton

89/A 89,990 9,396

90/A 107,543 8,713

91/A 135,697 8,387

92/A 147,564 7,148

93/A 168,892 6,729

94/A 199,316 6,454

95/A 247,012 6,597

96/A 288,076 6,387

97/A 361,531 6,777

98/A 415,524 6,500

99/A 445,407 6,358

00/A 473,603 6,173

01/A 514,710 6,213

02/A 563,073 6,511

Source: FEDEARROZ INTERNATIONAL TRADE Economic openness model in Colombia, reduced tariffs in exchange for a reduction of production subsidies from developed countries. This brought an increase in food imports of about 388%, and a reduction of sown area for semiannual crops of about 25.9% during 1991-2000 period. Rice sector became a net importer since 1991. In average, Colombia imported during 1990’s 236.000 tons of dry paddy rice to cover deficit caused by excess demand, Table 1.5.

- 34 %

17

TABLE 1.5: IMPORTS OF RICE IN COLOMBIA, 1990 – 2002.

IMPORTS Origen Dry Paddy - Ton

1990 (39,457)

1991 (119,255)

1992 116,553

1993 52.702

1994 356,892

1995 170,876

1996 284,668

1997 252,100

1998 * 449.200 USA 69%, VEN 16%, ECU 15%

1999 ** 270.000 ECUADOR 98%

2000 *** 178.644 VEN 71%, ECU 21%, USA 2%

2001 a/ 321.847 ECU 48%, VEN 20%, USA 27%

2002 b/ 127.662 ECUADOR 98%

Source: FEDEARROZ * 330.000 Tons From registered trade + 119.200 Tons ilegally entered from Venezuela and Ecuador ** 76.500 Tons from Ecuador and 193.500 from contraband *** 70.000 Tons of white rice and 5.000 Tons of dry paddy / contraband per month /a. 147.445 Tons from Ecuador, 62.000 Tons from contraband 112.402 Tons from other countries.(INDUARROZ) Comité Nacional de Estadísticas. /b 123.262 tons from administered trade (CAN) and 4.400 2001 B safeguard

The highest import levels happened in 1998 with 499.000 tons. Suppliers of imported grain have changed through time depending on the supply conditions, imposed by Rice National Council. This institution considered imports of dry paddy rice to benefit industry.

18

Import excluded traders and any entity that made business with foreign countries. Main provider was United States that increased production because most of Latin-American countries followed up Colombia. Ecuador and Venezuela have exported rice to Colombia, by the way. As Ecuador has increased areas, and Venezuela had a depreciated Bolívar, Colombia turned into main importer for these countries. From 1999, those countries exported to Colombia because of Andean trade agreements with a zero tariff politic. There is some trade with United States because of prices, payment facilities, and industry productivity. Trade with Ecuador keeps restrictions because of quality and low production availability volumes in short time periods. National industry mixes this rice with native to sell it as a product. United States and Thailand that have reduced their prices since January of 1998, in more than 50%, guiding international prices. This meant a loss in competitivity with a pass-through effect that has reduced Colombian prices too. As a net result, during last three years production costs and product prices have reduced sharply (about 30%), Colombian rice has not improved its competitive levels because of reduction in international prices.

19

II. TRADE POLICY DEVELOPMENTS COLOMBIA AND THE MULTILATERAL AGREEMENTS Account: Despite Colombia participated in the GATT creation conversations, it applied for a provisional adhesion to the Trade Agreement at the end of 1968. That application was formalized until February of 1974, and the Council of Representative Countries accepted it in July of 1975. In October of 1978 Colombia presented its definitive application to be part of the GATT and the contracting parties approved it by November, 1979. By April of 1980 the ultimate protocol of adhesion was signed after the internal administrative formal steps were provided. The Colombian Congress approved it through Law 49 of 1981 and the ratification was presented in Geneva September that year, turning Colombia Contracting Party number 86 in October the 3rd, 1981. Colombia negotiated 36 tariff positions before entering the GATT, 33 of those belonged to the industrial, petrochemical and metal-mechanical sectors. In the agricultural sector, tariffs for peas, lentils (15%) and apples (20%) were consolidated, they are still effective. THE AGRICULTURE AGREEMENT –URUGUAY ROUND– Advances in the trade liberalization of agricultural products achieved in negotiation rounds of the GATT were significantly modest and the international trade of commodities was flooded by distortion mechanisms such as export subsidies, restrictive administrative mechanisms, restrictive tariffs and trade impediments. That situation enforced the necessity of a new round of negotiations where the agricultural trade issue would receive special attention. So, by 1986 a round of negotiations began which main objective was to establish discipline in the international trade for agricultural products. The negotiations were carried out for 8 years and they were summarized in the Act of Marrakech in April 15th 1994. Until the late 1980´s, Colombia developed and used an import substitution policy, which was modified radically through the Market Openness model at the early 90´s. It implied a deep reformation of the economy, eliminating the price controls, abolishing the sustentation policies in production costs, reestablishing the imports without quantitative restrictions or government monopolies, and using rates and market prices without subsidies. Then, the agricultural sector got lots of relevant changes mainly in the internal production composition and a significant growth of the import volumes. Simultaneously, with the new model lots of new weather situations took place and they made the international prices fall; additionally, the process of Colombian currency appreciation implied a reduction in the share of agriculture in the GDP. An important group of products of the Colombian agriculture was affected by those economic measures: its imports grew in a significant way while sowed areas decreased simultaneously. Sowed areas decreased in almost 400.000 hectares, 10% of the total agricultural products sowed area. After 4 years of the same results -and since the beginning of 1994-, the agriculture revitalization became one of the policy main objectives. Not only because of productive reasons but also because of social nature reasons and political stability. From a

20

commercial perspective, this has meant that a restricted number of products, including rice, were affected by such protection mechanisms such as tariffs, stabilization systems, imports administration and even safeguards. THE COMMITMENTS OF THE AGREEMENT At the Uruguay Round, Colombia acquired tariff commitments for all of its tariff sub parties, reflecting its commitments with the tariff pricing suited in the negotiations. The country also acquired reduction payments commitments for policies that were identified as export subsidies y also committed to reduce aid measures that belonged to the Amber Box. Due to its developing country condition, the implementation period, the grants and subsidies reduction commitments were established in agreement with the Special and Differentiated Treatment for the developing countries as follows: Ø A 10-year implementation period starting from 1995 until the end of 2004. Ø The Global Aid Measure, which includes supports of general character and those of

products in particular, a 13% decrease of the consolidated amount along a 10-year period. For the rice, the consolidated value included: - Sustentation prices, which were valued in US$ in the average value of $55.7 millions (1986 – 1988). - A production credit subsidy for US $23.5 millions approximately. - Other payments for specific products valued in US $11.3 millions. - So the Global Aid Measure total for rice was valued in US $90.5 millions. To summarize, Help Global Measure for the period of reference 1986-88 was US $397.7 million dollars, which 22,7% corresponded to the rice, reflecting the importance that this product has had for the Colombian agricultural economy.

Ø In these circumstances, the de minimis internal grant that Colombia could apply -

which would not be tied to reduction- would reach $39,7 million dollars. This percentage has not been covered in the recent years. .

Ø The export subsidies which refer to the payments given to the exporter performance

including those mentioned in the article 9 of the Agreement make reference to issues like direct payments, non commercial inventories sales with minor prices than the domestic ones, production taxes to subsidize exports and internal cost reduction measures, discriminating against the markets and subsidies incorporated in the raw materials.

Colombia’s reduction commitment, as a developing country, is about 14% regarding the volume and about 24% of payments budget within a 10-year period. The rice’s payment level, for the base period, was about US $118,3 million for 18.911 tons of production. For the year 2002, and if the chronogram would have been fulfilled, the payment value could have been US $96,2 millions for 16,852 tons volume. Nevertheless, Colombia revised its notification finding that what it considered as exports subventions corresponded to an indirect taxes refund, so a Corrigendum was made and it does not apply to exports subventions.

21

Additionally, -as it will be analyzed- Colombia does not export rice and moreover imports the good depending on its internal demand necessities.

Ø In market access, the Agreement pretended access conditions to be more

transparent, predictable and competitive.

The tariff reduction, which Colombia committed to set, was about 24% average with a minimum, for each tariff group, of 10%, within a 10-year implementation period. So the tariff basic type for the rice was 210% and the right’s consolidated type 189%. These levels correspond to the results of the estimated tariffs equivalent to the base period and defined for the negotiation. On the other hand, the initial minimum rice stock established to access the market, sub party 1006, was about 39.598 tons with an applicable tariff type of 80% and a final stock of 75.188 tons, with the same tariff. This stock -like we will see later-, has been successfully completed and moreover, it has been overcome during several periods and the applied tariff has been inferior.

Ø Special safeguard contingent: the agricultural special safeguard established at the

Agreement, which allows additional tariffs imposition in case of an unexpected increase of the imports or in case of an import prices decrease, has not been used by the country on any product despite that the rice and other 56 product’s tariffs up to 4 digits could be aided by this measure.

Now there is list of a number of main products, which can be covered by the Colombian special safeguard contingent. Up to 10 digits, they totalize up to 271.

o Rice o Cotton o Meat and derived o Poultry and derived o Milk and derived o Wheat o Sorghum o Flours o Peas o Beans o Chickpeas o Starch o Peanut o Oleaginous seeds o Animal fats o Vegetable oils o Sugar and derived o Tropical juices o Bran and Cakes o Alcohols o Malta

22

THE ABSORPTION AGREEMENTS AND THE INVESTMENT MEASURES AGREEMENT RELATED WITH TRADE Another element of Agricultural Policy related with the market access is the crop absorption. Colombia requests importers to demonstrate having acquired national production previously or commit to acquire it to authorize the import of commodities. This policy was notified to the WTO between the frameworks of the dispositions of the Agreement on Measures of Investment related with Trade. In conformity with article 5 of this Agreement, the developing countries are entitled to the application of this measure for five years after having gone WTO agreement has gone into effect, that is to say until January the first of 2000. Article 5.3 offers the possibility to continue the transition period having into account financial, economic and development needs of the country that requests it. Colombia requested this right for a seven years extension to the Commodities Trade Council, to be able to continue with the absorption policy. Other 9 countries followed Colombian requests, but the only country that qualified was Colombia. The Colombian application was supplemented with a request for an exemption under Article 9 of WTO Agreement to implement absorption policy to beans, which started to be applied by 2001. Finally, Colombian request was accepted with some modifications, for a 4 year maximum, under the condition of a discharging program. Discharging chronogram presented by Colombia answered to products sensibilities, leaving products like the rice, the wheat, corn, Soya, beans and poultry for the end. Colombia also committed, not to apply the policy, to any other product that was not striped in the dismounts chronogram. Table 2.1. TABLE 2.1 COLOMBIAN DISMOUNTS CHRONOGRAM FOR SOME PRODUCTS OF THE AGRICULTURAL SECTOR FOR THE WTO Dismount at December 31 2001 (19 Sub-parties)

Dismount at December 31 2002 (13 Sub-parties)

Dismount at December 31 2003 (40 Sub-parties)

- Ducks and Gooses - Suets and Fatty - Dogs and Cats Food - Stearic Acid - Oleic Acid

- Oil of Soya - Palm Oil - Fatty and Oils - Sunflower Oil - Cotton Oil - Oil of Coconut - Oil of Turnip - Corn Oil - Oil of Benne - Oleaginous Seeds

- Roosters and Hens - Turkeys - Beans - Wheat - Barley - Corn - Rice - Sorghum - Wheat flour and corn - Malt - Gluten of Wheat - Starch - Prep. For animals

Source: Ministry of Agriculture

23

On the base of notification of the obligatory crop absorption to access to the imports from TRIMS Agreement and as verification mechanism, Colombia has been using the approvals mechanism or non-automatic licenses, for a weak agriculture products group, applying a concept of agriculture-productive chain. This mechanism is mainly applied for cereals and oleaginous and it means the previous purchase of the local production, according to conditions to suit among producers and industrials by the government guarantee. For the specific rice case, it is defined a seasonal contingent, concerted in the National Rice Council which the farmers, industrial, merchants and the government representatives attend. The government issues a regulation in which dates and promise of purchase of the future production are established. Under these circumstances, the volume of imports, including that coming from the countries of the Andean Community and from members of the WTO, has overcome in several years the access levels, even in 1999 when because of the import excesses of the previous year caused an accumulation of inventories and a bigger production to the prospective, allowed a low volume in terms of white rice to the imports. THE AGRICULTURE AGREEMENT AND THE COEXISTENCE OF SUB-REGIONAL AGREEMENTS The Agriculture Agreement allows the existence of economic integration agreements among nations where they apply preferential conditions between them. The Andean Community –CAN- it is not still accepted completely as Custom Union, although trade between member countries is liberated in a high proportion. Antecedents. The Andean Community was born at 26 of May of 1969, when a South American group of countries in the Andean area, including Chile, subscribed the Cartagena Agreement -Andean Pact- with the purpose of establishing a Custom Union for ten years.

The Andean integration process crossed through different stages along three decades. From a basically closed conception of integration toward inside, in accordance with the import substitution pattern, it had a reorientation towards an open regionalism model. Nowadays the Andean Community of Nations CAN is a sub-regional organization constituted by Bolivia, Colombia, Ecuador, Peru and Venezuela, and whose main objectives summary in: - To promote the balanced and harmonic development of their member countries under

equity conditions. - To accelerate the growth by means of the integration and the economic and social

cooperation. - To impel the participation in the regional integration process, with an objective to the

gradual conformation of a Latin American market common. - To keep and improve the level of their inhabitants life. Trade among Andean Community Countries is liberated in the majority of the sub parties, with the exception of those contemplated in the Decision 414 that regulates the

24

progressive linking of Peru to the Custom Union, Peru has the dispensation of incorporating its products progressively to the integration process, with term until the year 2004. THE ANDEAN SYSTEM OF PRICE STABILIZATION Colombia -like it was mentioned previously- advanced since the end of 1980’s from a unilateral economic opening process that implied a radical change in the handling of its international trade. For some sectors of the economy and especially the agricultural sector the special circumstances of the world market conditioned the use of imports cost stabilization mechanisms. Fluctuations in the international quotations of agricultural commodities took to a imports cost stabilization mechanism translated into the application of a fluctuating tariff in function of the international market prices. In the event of an important drop of the international prices, in connection with the historical average of those prices, the basic ad-valorem tariff that is relatively low is increased by a variable component, with maximum limit of the consolidated tariff by the WTO. In the event of price increases over the average, tariffs decrease to zero. By means of the application of this mechanism, provisional and surprising fluctuations of international prices that affect the domestic market making it unstable, affecting revenues or producers decisions about the time of cultivations, and in other events harming consumers, forcing them to acquire expensive products in the international markets. Colombia used the System of Fringes of Prices for the import of the commodities starting 1991. Since 1994, with the expedition of the External Tariff and the beginning, of the traffic of the Andean countries from a free trade zone towards an Custom Union, the Andean countries suited the combined use of a System of Fringes of Prices that Colombia, Venezuela and Ecuador applied with the derivative differences of the separate negotiations that these countries made in the WTO. Table 2.2 TABLE 2.2 ACCESS COMMITMENTS TO THE ANDEAN COMMUNITY MARKETS OF NATIONS IN THE WTO, FOR RICE. Bolivia Colombia Ecuador Perú Venezuela

Safeguard No Yes No No Yes Tariff In 2004 (%) 40 189 67.5 68 122

Current access 0 13.681 0 0 30.197 Minimum access 0 75.118 0 0 30.197

Source: Ministry of Agriculture, Secretary of the Andean Community.

25

OBJECTIVES OF THE ANDEAN PRICE FRINGES The Andean agricultural price fringes system has as main objective to stabilize the import cost of a special group of agricultural products which are characterized by price instability or distortions in the international markets. Member countries apply additional rights to the Common External Tariff when reference price for those products get under some determined levels. In the same way, member countries apply discounts to the Common External Tariff to reduce import cost when price goes over a determined level. This system covers a group of sensible products for the internal production of Andean countries such as rice, palm oil, soy oil, soy, sugar, poultry pieces, wheat, pig meat, milk, and maize. CONSOLIDATED TARIFFS AND APPLIED TARIFFS FOR THE COUNTRIES OF THE ANDEAN COMMUNITY Tariffs consolidated by the countries of the Andean Community differ one to each other because of previous commitments or because of internal policies such as it is the case of Peru that consolidated most of its agricultural products to 30%. This situation is explained partly by the composition of the agricultural production of the Andean countries and the relative importance of this in the total production of the economy. This way while Colombia, Ecuador and Bolivia are, in different grades, net exporters of agricultural products, Peru and Venezuela make part of the list of net importers of food. APPLIED CUSTOM TARIFFS BY THE CAN The resulting custom tariff of the System of Fringes has varied, from tariffs of 33% in 1995, when the application of the Andean System of Fringes began, to 83% in 2001, Table 2.3. As it will be analyzed, the resulting custom tariff of the fringes of prices is related with the international prices and these last are closely linked with the subsidies of the main producing countries.

26

TABLE 2.3 TARIFFS, TOTAL OF THE ANDEAN SYSTEM OF PRICE STABILIZATION,

THE CASE OF RICE, 1991-2001. Source: Ministry of Agriculture and Secretary of the Andean Community.

THE ANDEAN SAFEGUARDS Sub regional trade is regulated by the normative of the Andean Community that allows the non discriminatory application of safeguards. By means of this mechanism, countries can limit the imports, to the scope of the necessary ones to cover the transitory shortages of supply and to level their prices in the internal market. Also, they can be applied when interferences are presented in the internal market because of low prices to those that imports can enter a country. APPLICATION OF REGULATORY MEASURES By means of Decree 2439 of 1994, the National Government held to the instrument of the grant of good approvals the imports of some agricultural products. The administration of those good approvals corresponds to the Ministry of Agriculture and Rural Development, together with these products; the rice is coming from any origin country except for the countries members of the Cartagena Agreement. Later on, with Resolution 004 of May 25, 1995, the Superior Council of External Trade (CSCE) suspended the imports of rice coming from Venezuela up to December 31, 1995. With the Ordinance 2353 of December 29, 1995, were settled down good approvals to the imports of rice coming from Venezuela and of Chile up to June 30, 1996.

Year Anual Average (%)

Mínim (%) Máxim (%)

1991 41 35 50

1992 22 20 30

1993 37 24 55

1994 34 16 56

1995 33 18 53

1996 16 9 20

1997 21 10 34

1998 21 20 28

1999 46 22 66

2000 70 51 90

2001 83 78 87

27

From then on the Superior Council of Trade Exterior has maintained the non-automatic licenses for the imports of rice coming from the countries of the Andean Community and the contingents ones. The past February 4 of 2002, the recent safeguard to the imports of rice coming from the Andean Community was imposed, for a contingent of 123.000 tons in terms of rice dry paddy; this measure had validity up to November 30 of 2002 and at the moment the legal steps of the extension of the Ordinance that establishes the safeguard are taken, for one year more and for a contingent of 150,000 tons of rice in shell. THE IMPORTS ORIGIN The handling that has given Colombia to meet the requirement of non automatic licenses of the external trade of rice and linking the import to the purchase of the national crop and the application of the system of fringes of prices for third countries, have strengthened the competitive capacity of the rice produced by the member countries of the Andean Community that enjoy tariff preference. By this way the Ecuadorian production mainly has become the source of supply and provision of the Colombian market. Except for some years, in those that the production of the Andean countries has not been enough to satisfy those consumptions of the local market, mainly because of the difficulties related with the climate and “El Niño” phenomenon, it has been necessary to go demand supply of third countries.

THE NEW AGRICULTURE ROUNDS AND OTHER COMMERCIAL AGREEMENTS OF COLOMBIA. The negotiations on the agriculture began at the beginning of 2000, in concordance with the article 20 of the Agreement on Agriculture of the WTO. By November of 2001, before the Ministerial Conference of Doha, 121 governments presented a great number of negotiation proposals. DOHA DECLARATION The Doha declaration on which the countries in development had on big hopes did not mean an instruction to achieve big advances in the agricultural trade liberation. The Declaration already reconfirms the long-term objective suited in the current Agreement on the WTO: to establish a system of fair trade directed to the market by means of a program of fundamental reformation. The program embraces reinforced norms and specific commitments in terms of official aid and protection to the agriculture. The objective is to correct and to prevent the restrictions and distortions in the world agricultural markets. Without prejudicing the results, governments of the member countries commit to celebrate global negotiations guided to achieve the following: - Market Access: substantial improvements - Export Subsidies: reductions in all the ways of grants, with the aim to get their

progressive removal - Internal help: substantial reductions of the causing help and contributions of distortion

of the trade

28

In the Declaration it is agreed that the special and differenced treatment for the developing countries is integral part of all the elements of the negotiations, as much of the new commitments of the countries as well of any norms and pertinent new or revised disciplines. It establishes that the result will be effective in the practice and to allow to the developing countries to assist its necessities, particularly regarding alimentary security and rural development. The ministers also take note of the non-commercial concerns (such as the protection of the environment, the alimentary security, the rural development, etc.) already reflected in the negotiation proposals presented. They confirm that in the negotiations they will keep in mind those concerns, according to what is foreseen in the Agreement on Agriculture. The stage of proposals in modalities is close to conclude and it is expected that starting March of 2003 the properly period of negotiations begins. In the framework of the Doha´s Program for the Development, negotiations have now up to March 31, 2003 to establish those “modalities” or goals (including the quantitative goals) necessary for the attainment of the objectives enunciated in the Ministerial Declaration of Doha: “substantial improvements in the access to the markets, reductions in all the ways of grants to the export, towards their progressive removal, and substantial reductions of the internal aid which are giving rise to distortion of the trade.” This stage will determine the form that will have the final result of the negotiations. COLOMBIA AND THE CAIRNS GROUP The objectives of Colombia for the new negotiation of the Agreement on the Agriculture are framed under two environments. The first associated to the positions of the Cairns Group and that embraces the common interests to its 17 members. The second has to do with the peculiar interests of the country and that are related, by a part, with the necessity of maintaining certain protection instruments to the national agricultural production in answer to the existent distortions in the world markets; and to other elements associated to the complexity of the civil conflict and the substitution of illicit cultivations. CAN - MERCOSUR NEGOTIATION The expectation of this negotiation for the rice was important because of the presence of the big producers from the south cone, that showed competitive levels, with macroeconomic situations that determined the possibility of later devaluations. Also, it was eminent the application of Uruguay and Paraguay of receiving a special and differentiated treatment to access Andean markets of agricultural products. The development of the negotiations was expected to be concluded by the end of the year but has taken a long time and the validity of the current preferences has continued, deferring the negotiation for next year. In this negotiation has not been defined a treatment for the sensitive products, rice is one of those. ALCA. The definition of custom exemptions for the Area of Free Trade of the Americas called ALCA in Spanish, is programmed for the February the 15th 2003, and the improvement applications for the middle of 2003.

29

Rice is one of the most important products for the countries of the Americas, including the United States. Access negotiation in this product will be surely framed simultaneously by the definition of custom exemptions chronograms conditioned to the reduction in subsidies to the export and the internal help that some developing countries have denominated “conditionality.” The balance in the negotiation that up to now has been guided towards the definition of the conditions in those bigger access will be granted, will become the central topic of the discussion and about it will not have complete clarity until 2004. CONCLUSION The unilateral opening that Colombia carried out starting from the final years of the 1980’s meant the wide opening of the market of the Colombian rice. Colombia has turned to a major dependency of third parties since 1992, when the country was a net exporter. Since 1995 when the Agreement On Agriculture of the WTO started, an increase of the external dependence of the provision of rice has not been noted for consumption so as to conclude bad effects over internal production. Since then, Colombia has applied the mechanisms foreseen in the Agreement, except for the Agricultural Special Safeguard, for the handling of its external trade. The applied tariff has been traditionally lower than the one consolidated in the WTO, especially because a high percentage of its imports is made inside the Andean Agreement without tariffs. At the present time, the high international subsidies for rice that put the product as one of the goods whose prices are more affected by the subsidies to the exports and the internal aids mean lower prices and high fringe tariffs. In consequence, the competitive situation of rice produced in the Andean countries has strengthened itself in front the one of third countries. The new negotiations of the Agreement on Agriculture that scarcely begin the incoming year facing again the interest of the United States, that reinforced their mechanisms of aid, but that proclaims a liberation of the market, and the European Union that wants to be able to reduce its aid mechanisms without losing self-sufficiency, given the high costs that for the common policies means the agricultural issues and especially now when new countries of minor development are linking to the Community. The new negotiations for the AFTA with the MERCOSUR members are still in stages of treatment of definition to the sensitive products, of definition of lists of custom exemptions, or by a manner there are not signs about what will be the treatment that will be given to rice although from the internal agricultural policy point of view it will receive special treatment.

30

III. SOCIAL AND ECONOMIC IMPACT OF TRADE LIBERALIZATION POLICIES

Perception that Colombian rice producers have about social and economic impact of trade liberalization is originated at the protection that grain exporters give to producers and commercial sector, therefore it is necessary to study the impact of this variable on the sector. As different reports have shown4, internal aids to producers, export subsidies and other support measures have impacted international quotations of rice. Mainly it is shown in the downward trend of its international prices and its volatility. Rice is considered by many as a strategic product for alimentary security, source of foreign exchange for governments, historically producer countries have intervened in production, processing and commercialization of rice through several policies to sustain price or income, tariffs, rates, or export subsidies. Some studies show that subsidies and commercial barriers contribute to price volatility and have adverse effects over rice production level in other countries. Furthermore, subsidies reduce international rice prices and that weight falls over those countries with more limited government support, as Colombia. This way, in countries where producers enjoy support from government, sowed land is maintained artificially high, while in other countries with less support, sowed land in reducing. This section pretends to measure impact caused by international subsidies over some of the variables related to rice production in Colombia. Second topic is related to the empirical evidence that results from comparing internal rice prices of Colombia with those of the international markets. It seems that Colombia is a price taker country in spite of the protection measures coming from the national government as a policy to compensate national production because of international price distortions. Reduction of national rice prices of one third, and cost reductions during the last 12 years would indicate that national producers are exposed to international markets in spite of the direct tariff protections and some other indirect protections. Income that should have been generated with protections did not happened for some reason that linked national price to international price of rice. Section consists of five main topics. First describes international rice market, signaling traded volumes and main actors in this market. Second, it is shown an analysis of subsidies to the OECD (Organization for Economic Cooperation and Development)5 producers, showing which countries subsidize grain, magnitudes, and particular characteristics. Third, it is presented an econometric exercise to measure the effect of

4 Valdez, Alberto y Seitz, Joachim, “Distortions in world food markets in the wake of GATT: Evidence and policy Implications” In World development, Volt 23, No.6., 1995, Pp. 913-926. “The Costs of protectionism to Developing Countries: an analysis for selected agricultural products”, World Bank Staff Working Paper No. 769, 1986. “Agricultural Protection in OECD Countries: its costs to less developed countries”, International Food Policy Research Institute, Research Report No 21, 1980. 5 Component countries: European Union-15, Australia, Canada, Japan, Korea, Mexico, Switzerland, Turkey y United States.

31

subsidies on international grain quotations. Over these results, some measures were evaluated of the impact on prices of the elimination of subsidies. Fourth topic analyses mechanisms that connect national rice market to international one. Fifth topic measures the impact of subsidies to OECD countries on some Colombian rice sector variables. Last section presents conclusions. MAIN PRODUCTION AGENTS AND GLOBAL RICE TRADE To measure the impact that subsidies could have had over the behavior of rice prices, it is necessary to analyze production and international trade behavior. Data presented here were calculated based on information from FAO. During 2001 338.8 Million of tons of white rice were produced but only 6% of that amount was traded at the international markets. In other words, rice production attends internal demands mainly, and surplus is traded at the international markets. World rice production is highly concentrated. Twelve countries compile 90% of it: China, India, Indonesia, Bangladesh, Vietnam, Thailand, Myanmar, Philippines, Japan, Brazil, United States, Korea. China and India produce 31% and 22,5%, respectively. With the exception of United States, Thailand, and Vietnam, production is focused to attend internal demand. United States exported 49% of its production, Thailand 37%, and Vietnam 16%. Another big producer is European Union-15 with a total share of 0,4% of world production, directing 95,5% to international markets. United States and European Union cases deserve an special attention, their per capita consumptions were about 11,3 kg and 5,5 kg in 2000 are under the world average of 64 kg., reinforcing the theory that they focus their production to international markets. These countries have a constant trend to consume wheat and its derivates. In fact, United States registers a per capita consumption of wheat of about 124 kg and European Union 249 Kg, over world average of 96 kg. WORLD SUBSIDIES TO RICE PRODUCERS IN OECD COUNTRIES Support that developed countries offer to their producers in agricultural sector is very big, by 2001, OECD countries received US$230.744 Million in direct support to producers and US$310.9596 Million in total support. Rice is one of the supported products, receiving more subsidies in relation to its international price (amount of produced ton in relation to the international price). In fact, while the international reference price for rice7 in 2000 was US$248, subsidy per ton at OECD countries was US$1050, more than tour times international quotation. On the other hand, rice registered a Nominal Protection Coefficient (NPC) of 5,288, meaning that price to the producer is 5,28 times higher than the international price.

6 Incluyes General Services not included in the PSE. 7 Refers to CIF Bangkok rice. 8 El NPC is the Nominal Protection Coefficient. If NPC is 1, international price and proce to producer are equal. No protection exists.

32

Another way to understand this amount is that OECD countries grant to rice producers a nominal protection of 528%. (Table 3.1). TABLE 3.1 NOMINAL PROTECTION COEFFICIENT FOR EUROPE, UNITED STATES

AND OECD COUNTRIES, PER PRODUCT, 2000.

Arroz 0.89 1.43 5.28Azucar Refinada1 2.39 1.50 2.11Carne de Cerdo 1.19 1.00 1.18Carne de Pollo y Otras Aves 1.59 1.00 1.20Carne de Res y ternera 1.89 1.00 1.29Cebada 0.94 1.00 1.34Leche 1.70 1.90 1.77Maíz2 1.08 1.14 1.18Oilseed3 1.01 1.21 1.22Sorgo 1.00 1.34Soya 1.00 1.00 1.34Trigo 1.00 1.15 1.15

PromedioOECDProducto Union Europea

EstadosUnidos

Source: OECD, Database 2001. Product that receives more subsidy per produced ton measured by the Producer Subsidy Equivalent (PSE)9, is rice, followed by meats, barley, and milk. Graph 3.1 shows shares of aid over products during 2001.

9 PSE = Producer Subsidy Equivalent.

33

GRAPH 3.1 SUBSIDIES PER PRODUCED TON AT OECD COUNTRIES, 2001.

Source: OECD, Database 2001. For 2001, aids for rice producers at OECD countries was US$29.335 Million, representing 10% of total subsidies for agricultural production at those countries and 32% of the World total production of the grain. These amounts show the magnitude of the subsidy and its impact over prices.10 The equivalent subsidy (PSE) received by each ton of rice produced at OECD countries was US$1.165 per ton during 2001, 4,5 times its international price of US$214 for the same year, taking as a base the Bangkok CIF price for rice. If it is supposed that the other countries around the world do not subsidize producers, PSE would be of US$64 per white rice ton, 30% of the international price. This shows the magnitude of market intervention and implication that these subsidies have in the international price formation. In spite of the last considerations, we must have into account that not all of the production, nor the exports are affected directly by subsidies granted by OECD countries. For that reason, subsidies at granting countries are analyzed. Major amount of subsidies is granted by Japan and Korea, non-exporter countries, though some other countries as United States, European Union-15, Mexico and Australia grant juicy subsidies. We must consider that OECD countries have a 5,7% share of white rice world production but only 23,3% of world exports. This indicates that in spite the magnitude of their subsidies, the other producers and exporters have an important impact over international prices.

10 Value of production was calculated as the result of world production in 2001 (380 million of metric tons) time average price US$214. The value was divided by OECD subsidies equivalent to US$ 24.340 millions.

US$ / Ton

1,050 930

333 293214

140 136 6765

60 55 410

200

400

600

800

1,000

Ric

e

1,200

Bar

ley

Pig

Mea

t

Chi

cken

Mea

t

Milk

Wh

ite

Su

gar

Oil

See

ds

Soy

Bea

ns

Sor

ghum

o

Cor

n

Whe

at

Cal

f Mea

t

34

It is important to note that the magnitude of aids, the share in world market, varies among OECD countries that grant subsidies to rice producers. This redirects analysis to countries that grant subsidies according to production volume and share in world market due to that the combination of these variables can determine the impact that a country may have over international markets, mainly prices. Japan is the most subsidies-granting country for rice. During 2001 subsidies reached US$15.935 Million, 65,5% of the total amount granted by OECD countries. Despite production levels are high, (7,7 million of tons, 2% of total world production, ninth place among world rice producers), exported volumes are insignificant. Imports are important, (9% of total consumption, 656.000 metric tons). On average, for 2001, implicit subsidy per ton was US$ 2.166, 10 times international price. Korea is the second country granting subsidies to the product. By 2001 US$7.244 Million were granted, representing 30% of total OECD grants. Korea is an important producer, 4,6 million of tons placing the country at the 13th place among the producers of this cereal with 1,2% of world production. Korea imports 4% of its consumption. Implicit subsidy per produced ton reached US$1.546, 7,2 times international price. Jointly, Japan and Korea concentrate 95,5% o subsidies granted to rice producers at OECD countries, 3,1% of production an 4% of world imports. Given these relations, impact of subsidies over world process should not be so big as the relationship established among subsidy per ton and international price. Notwithstanding, impact over its internal production must be high. If subsidies disappear, imports should cover demand turning those countries into net importers. If this is the effect of abolition of subsidies, world demand for rice would increase in the magnitude of the reduction of production in those countries, affecting international prices. United States is the third place in terms of subsidies to rice producers. During 2001 the government conceded US$784 million occupying the first place in the world ranking of producers with 6 million of ton. PSE reached US$128, representing 60% of international price for a ton of rice. This country has an special characteristic, production is focused to international markets. Almost 49% of its production is sold at the international markets, it is the fourth exporter and has a share of 11,7% of total exports. European Union-15 is the fourth producer granting subsidies to rice producers. In 2001 EU-15 conceded US$322 million in subsidies over a 1,6 million and exports of 1,4 million of tons. Their exports represent 92,5% of production, placing the group as the seventh world exporter (6,1% of total). Implicit subsidy per produced ton was about US$199. In spite of that, UE-15 is a rice net importer: During 2000 imported 2 million metric tons, overcoming its production and export levels. Australia, subsidies granted to rice reached in 2001 only US$7 million over a production of 805.000 tons and exports for 622.000 tons (55% of production). As a consequence, implicit subsidies are relatively low, US$8 per produced ton. These last countries grant 5% of subsidies y OECD countries to rice producers, represent 2% of world production and 21% of exports. Given these relationships, impact of their subsidies should not be high, in spite their share in exportations.

35

Evolution of subsidies per ton grants to rice producer at OECD countries show an increasing trend until 1995, and from 1996 tend to be reduced to levels of 1980’s. There is also an inverse relationship between behavior of Producer Subsidy Equivalent (PSE) per ton produced at OECD countries and international price of rice (Bangkok reference). This relation is clearer in the case of countries as United States, European Union, and Japan. For the case of United States, there is a strong relationship provided.11 This obeys to that country’s subsidy policy to maintain income to producers, so subsidy increases if price decreases and vice versa. Not establishing causality among these two variables, there are several studies that confirm that a subsidy makes costs cheaper and so internal prices turn to be over world prices. This price generates a higher production level over the market, resulting in exportable surpluses that press down international quotations. High subsidies granted to rice in absolute and relative terms, have had implications over the behavior of international prices. In particular two effects: Downward trend in quotations, and higher volatility. IMPACT OF SUBSIDIES OVER INTERNATIONAL RICE PRICE It is commonly affirmed that subsidies granted to agricultural producers have a negative impact over commodities world prices, and so they affect their world production. Since 1980’s, several studies about the effects of agricultural policies at industrialized countries point out that interventionist policies induce international prices to be under the market levels. Valdez summarizes the results of some of these studies until 1995, comparing international prices with a base year, with prices with no government intervention on agriculture. He concludes that agricultural policies at developed countries depress international prices in a significant way. For the case of rice, the effect over international prices by application of those policies varies between -2% (Valdez and Zietz) and -43% (UNCTAD-WIDER). Differences come from the base year for analysis and the liberalization level assumed at each document.12 As a consequence, production of countries that cannot finance producers through internal aids or subsidies to exports and keep low protection levels at borders or bad administrative controls get sowed areas, incomes, employment, and other variables affected. On the contrary, developed countries that regulate market tend to over-produce, generating exportable surpluses depressing international prices. Estimations of subsidies granted to rice producers impact at developed countries over international prices cannot be realized immediately given the existence of important substitutions in the cereal consumption, mainly rice and wheat. In that sense, it is necessary to have into account the impact of subsidies granted to wheat producers around

11 Correlation coefficient was –61%. 12 Valdez, Alberto y Seitz, Joachim, “Distortions in world food markets in the wake of GATT: Evidence and policy Implications” In World development, Volt 23, No.6, 1995, Pp. 913-926.

36

the world and its impact on rice prices. For the case of Latin America there are several studies that prove that substitutability.13 It is important to have into account that for estimation of subsidy impact on prices it is mandatory to realize its existence. The core question is: Which is the impact level of international subsidies granted to rice over its international prices? To establish this relationship, a model was estimated relating international price of white rice with subsidy per ton at OECD countries to rice producers, white rice production and the international price of wheat.14 Introduction of wheat effect is justified through the substitution effect among these two cereals. Furthermore, as shown later, there is an equilibrium relationship in the long run among the prices of these two products. Estimation was calculated using panel data for 1986-2000. This econometric method should be understood as a combination among time series with cross section data. Panel contains information for six OECD countries (United States, Mexico, Australia, Japan, Korea, and the European Union) that grant subsidies to rice producers. A lagged dependent variable was included as explicative variable, it is supposed that today’s price level would determine prices tomorrow, obligating the use for a dynamic panel. In general terms, the following equation was estimated:

tititiiti XPP ,,1, , υβαη +++= −

where P corresponds to the international price of rice, X is a vector containing other explicative variables, η is a constant capturing differences among countries, and υ is an error. Data from different variables were standardized, so its mean was rested and divided over its standard deviation with the objective of eliminating heteroskedasticity problems on the error due to the differences among the variables from different countries. For interpretation effects, it is important to have into account that the resulting estimations measure impact of the surnamed variables over the international price of rice, in terms of the standard deviation of each one. So a transformation of parameters is required to interpret results in terms of changes.

13 Maria Helena Ramírez Tolosa, ¿Porqué no se deben reducir los aranceles para la importación de trigo en Colombia?, Comité de Seguimiento a la Competitividad de Arroz de Colombia, Informe Final, Abril 2002. Martínez Covaleda, Héctor, “El efecto sustitución de consumo entre trigo y arroz en Colombia: una medición econométrica”, Revista Induarroz, No.3, Bogotá, Agosto 1999. Abreau y E. Ablan, 25 años de cambios alimentarios en Venezuela (1974-1994), Fundación Polar, Venezuela. GRADE, Evaluación del sistema de sobretasas a la importación de trigo y harina de trigo en Perú, Mimeo, Lima, 1993. Derek Byerlee, Ventaja comparativa y política agrícola para la producción de trigo en Ecuador, CIMMYT, México, 1976. 14 As international Rice prices and wheat the CIF Bangkok and CIF del Trigo Hard Red Winter No 2 prices have been used. Production data from FAO.

37

Results are presented at Table 3.2. As expected, international price of rice corresponds positively to its lagged value and to the price of wheat, and contrary, it has an inverse relationship to the total quantity of rice that is produced around the world and to the subsidy. TABLE 3.2 SUMMARIES OF THE RESULTS FROM THE MODEL TO EXPLAIN PRICE FORMATION AT INTERNATIONAL RICE MARKET, 1986-2000.

Dependent Variable Constant Price

Internacional Rice (-1) Wheat

internacional Price

World Rice

Production PSE/Tml

For the rice

Rice Internacional Price 0.0321164 0.281531 0.619201 -0.257307 -0.168245 P-value 0.000 0.000 0.000 0.000 0.001 Confinase coefficient of error at 5%. R 2 =0.7091898 P-Value Wald (joint): 0.000 P-Value Wald (dummy): 0.000

Source: Subsidies OECD, Production FAO, Prices Rice Bangkok – FEDEARROZ, Wheat’s prices Hard Red Winter No.2, Ministry of Agriculture. As observed, in terms of Standard deviations, behavior of international wheat price is the variable that explains the most movements of international rice price. Positive sign of this parameter shows that prices of rice and wheat have the same direction, so if rice price get reduced, wheat price reduces too but in a smaller magnitude. Subsidies granted to rice producers explain in a minor way international price of the product. This way, if subsidy granted to rice producers increases one standard deviation, international rice price reduces 0.168 standard deviations. This relationships in dollars suggest that if the PSE increases in US$233 per ton, international price of rice reduces in US$9,3 per ton. If we assume a total abolition of subsidies at the OECD countries, that for 2000 was US$1.050 per metric ton, and we keep other variables constant, international rice price should increase in US$42, a 17% increase with respect to the price observed for 2000 (US$246). See Table 3.3

38

TABLE 3.3 CHANGE OF INTERNATIONAL RICE PRICES IF A TOTAL OR PARTIAL

(50%) REDUCTION ON SUBSIDIES AT OECD IS ACHEIVED, 2000.

Start Prices Final Price Change % Start Price Final Price Change % European Union 246 246.1 0.04% 246 246.0 0.02% Australia 246 246.0 0.01% 246 246.0 0.00% Japan 246 275.1 11.82% 246 260.5 5.91% Korea 246 257.6 4.71% 246 251.8 2.36% México 246 246.1 0.02% 246 246.0 0.01% United States 246 247.1 0.45% 246 246.5 0.22% Total OECD 246 287.9 17.04% 246 267.0 8.52%

Total Low of PSE/Tm l 50% Low of PSE/Tm

PSE/Tm is the subsidy percentage in rice, per Ton. Source: PSE/Tm Database OECD 2001. If we assume a direct transfer to the national price, an increase of the internal price of this magnitude (17%) would imply an increase on the Colombian rice production of about 18,4% (426.782 tons) and 88.913 hectares.15 In a pessimistic stage, a reduction of 50% on the subsidy at OECD countries, US$525 per ton, international price would increase US$21, a 8,5% increase of the price registered in 2000. Taking only the case for United States, a total abolition of subsidies would increase international price US$1 (0,45%). By the way, a reduction of 50% in subsidies per ton at that country would increase international price only 0,22% in terms of 2000 prices. On the contrary, total or partial abolition of subsidies granted by Japan and Korea to rice would have major incidence on international quotations. In fact, abolition of subsidies at those countries would imply an increase in prices of 11,8% and 4,7%, respectively. As estimated coefficients show, magnitude of the effect of total or partial abolition of subsidies is not as important as awaited. This happens because of the low share of OECD countries on world markets. Due to strong substitution relationship among rice and wheat, and that wheat price affects rice quotation, the same exercise was calculated for wheat as Table 3.4 shows. As awaited, international price of wheat responds positively to its lagged value, and inversely to changes on production levels and to the subsidy.

15 Here we assume a paddy rice production for 2001 of 2.313.811 tons, a yield of 4.8 tons per hectare and a price elasticity of 1.085, estimated by Observatorio Agrocadenas from Agriculture Ministry and Desarrollo Rural de Colombia.

39

TABLE 3.4 SUMMARY OF THE MODEL EXPLAINING RELATIONSHIP AMONG

INTERNATIONAL RICE PRICE AND INTERNATIONAL WHEAT PRICE, 1986-2000. Source: Producción FAO, Precio Trigo Hard Red Winter No.2 Ministerio de Agricultura, PSE OECD Database 2001. Just as in the last exercise, series were standardized, so the results must be interpreted in terms of standard deviations from the originals. In this sense a reduction of one standard deviation on the PSE per ton would increase in 0,17 standard deviations international quotation for that product. As observed in Table 3.5, total elimination of subsidies granted to producers at OECD countries would increase international quotations about 12,2%, this means that the ton of wheat would change from US$134.9 to US$151.4. If the reduction of subsidy per ton was of 50% international wheat price would increase in 6,1%. Highest effects would be presented at European Union and United States.

TABLE 3.5 CHANGE IN WHEAT INTERNATIONAL PRICES WITH A PARTIAL OR TOTAL ABOLITION OF PRODUCER SUBSIDIES TO OECD COUNTRIES, 2000.

Start Price Final Price Change % Start Pricel Final Price Change % European Union 134.9 144.6 7.2% 134.9 139.7 3.6% Australia 134.9 136.7 1.4% 134.9 135.8 0.7% Japan 134.9 137.5 1.9% 134.9 136.2 1.0% México 134.9 136.8 1.4% 134.9 135.8 0.7% United States 134.9 142.1 5.4% 134.9 138.5 2.7% Total OECD 134.9 151.4 12.2% 134.9 143.1 6.1%

50% Low of PSE/Tm Total Reduction of PSE/Tm

Source: PSE/Tm Database OECD 2001 Applying these results on the rice equation, it shows an additional effect an increase of 6% on the price of rice as a result of a higher quotation of wheat at the international markets. As a consequence, elimination of rice subsidies has a major implication on the price of rice (17%) than the increase on the price of wheat as result of the abolition on the subsidy to the latter. This means, that a total elimination of subsidies at OECD to rice and wheat

Dependent Variable Constant Wheat (-1)

Internacional Price

Wheat World

Production

PSE/Tm

Wheat

Wheat Internacional Price -0.0915525 0.683801 -0.501542 -0.178331 P-value 0.000 0.000 0.000 0.078* *Significantl 8% R 2 =0.5380313 P-Value Wald (joint): 0.000 P-Value Wald (dummy): 0.000

40

would cause an increase of about 17% to 23% on the international quotations for those grains. To summarize, we can affirm:

1.) Subsidies granted to rice by OECD countries affect international prices inversely. The higher the subsidy, the lower the international prices.

2.) Magnitude of the impact of subsidies on white rice prices is not as important as awaited. An abolition of 100% of subsidies at the OECD countries increases price in 17%. This is explained because OECD countries only represent 5% of world production and 21% of total exports.

3.) In spite United States ad European Union are important rice producers and exporters abolition of subsidies for these countries would have a minor effect than OECD or Japan or Korea. This shows that those countries do not shape prices, subsidy is just one of the variables, maybe one of the least important, to form prices.

4.) Abolition of subsidies over rice has a bigger effect over rice price than increase of quotation of wheat as effect of abolition over the subsidy on this product. So if price for wheat increases 12,2% or 6,1% as effect on the abolition of subsidies to wheat, keeping everything else constant, price of rice would increase 6% and 3%, respectively.

5.) If we assume an increase of 17% and 23% in the national price as a result of a combined effect from subsidy abolition to rice and wheat at OECD countries, paddy rice production would increase in about 426.782 and 577.411 tons respectively, representing an increase in production of 18,4% and 25%. With the current yield levels (4,8 tons of dry paddy per hectare) this would be 88.913 and 120.294 new hectares, respectively.

RELATIONSHIP BETWEEN INTERNATIONAL AND COLOMBIAN PRICES Among the community of economists specialized on commerce, it is discussed the convenience or inconvenience of protections to certain economic sectors based mainly in the income that are generated inside of their productive chains. Colombian rice sector has had some protections during openness as described previously. Notwithstanding, during this research prices have decreased 38% in real terms during the last 12 years, something very similar has happened to costs. These variables that affect income generation for producers have generated a similar behavior to that characteristic of prices. Producers question why in spite of protection internal prices decreased so much. Furthermore, if both national and international prices have the same trend, what happened if protections should have isolated prices from international movements?. Several hypothesis appeared to explain this phenomenon. One was to analyze the possible long run relationships among international prices of wheat and rice, and among the import cost of wheat and Colombian rice price.16 This

16 International rice price corresponds to white rice 100% B. CIF Bangkok, with 10% of broken grains. International rice price corresponds to CIF Hard Red Winter No 2.National price corresponds to best quality rice. All prices in dollars per ton and were deflected by United States Price Consumer Index.

41

analysis pretends to discover the transmission mechanism of international prices to national prices. LONG RUN RELATIONSHIPS AMONG NATIONAL AND INTERNATIONAL PRICES OF RICE AND WHEAT First approach to the topic shows graphs with the main long run trends for the particular case of rice. The first interesting relationship compares internal Colombian prices with internal United States prices. CIF prices show a strong relationship in terms of volatility, mainly during the last decade. Graph 3.2 shows different scales for the series to show coincidences in a better way. At bird’s view, there is a relationship among the series.

GRAPH 3.2 CIF PRICE OF US RICE AND NATIONAL PRICE, 1970-2000 (In dollars)

Source: Agriculture Ministry and FEDEARROZ. Second relationship that can shed some light about rice price is with wheat price. CIF prices at international markets for rice and wheat for the 1970-2002 period show on Graph 3.3 a strong relationship in volatility and trend. During the 1970’s, volatility of rice price was high but lowered after 1980’s. Last years shows that gap among rice and wheat prices is lowering.

US$ / Ton USA

US$ / Ton Nal

42

GRAPH 3.3 INTERNATIONAL PRICE OF RICE AND WHEAT, 1970-2002.

Source: Agriculture Ministry and FEDEARROZ. Alter having observed the strong relationship among international prices of rice an wheat, we compare prices inside national markets. Graph 3.4 shows symmetry among paddy rice in dollars and wheat price, in dollars. There is a lag in the symmetry starting 1976 and a perfect similitude in the trend of both series.

US$ /Ton

43

GRAH 3.4 NATIONAL PADDY RICE PRICE AND INTERNATIONAL WHEAT PRICE, 1970 - 2002.

Source: Agriculture Ministry and FEDEARROZ. From this graphic analyze it can be inferred that there is a close relationship among international rice and wheat prices with Colombian national paddy rice. Notwithstanding, during 1990’s these relationships among CIF prices of rice and wheat and Colombian national price turn symmetric. To verify any actual relationship among national and international rice prices, a co-integration analysis was used, with null hypothesis of the existence of long run relationships among the variables. This exercise was made to test the relationship among national price of rice in dollars and the international wheat price, and among the international prices of rice and wheat to verify any possible substitution relationships. A co-integration relationship among the variables mentioned above shows that the perturbations or shocks do not only have a temporal effect over the variables, but a permanent effect. Time series exercises were made for the observed quotations for rice (CIF Bangkok) and wheat (Hard Red Winter #2 from United States) during the period starting June 1991 to October 2002, and the tariffs from the price fringes for each of these products. Data were deflected with united States CPI, so comparisons are in constant terms. Analysis results using Dickey-Fuller test showed that CIF prices of rice and wheat, import costs of wheat and the national price for rice have unit root, so they are co-integrated. To reinforce this hypothesis a KPSS test was calculated confirming results from the first test.

US$ / Ton

44

Once the existence of a unit root is defined for each variable, a co-integration test was performed, to determine if movements among variables tend to converge in the long run. Results of the exercise confirm existence of co-integration among importation cost of wheat and national rice price, and international CIF prices of rice and wheat. This means that the price of Colombian rice and the import cost of wheat maintain a long run equilibrium relationship, the same occurs with the international prices of rice and wheat. Any movement over any variable will make the other move in the same direction, maintaining the equilibrium relationship. RELATIONSHIP AMONG PRICES OF RICE AND WHEAT Off to now, a strong relationship among Colombian rice and international rice prices has been tested, and also among Colombian rice prices and imported wheat prices. What follows is to explain why these relationships occur given the isolation process of Colombian rice production for the last 12 years. A possible explanation and, maybe the most effective one, is given through the substitutive products of rice. Several studies have emphasized over the existence of a strong substitution among human consumption of rice and wheat derivatives.17 There are empirical evidences that prove this association. A review of the per capita consumption for rice and wheat for 120 countries with data from FAO18 shows that on average between 1991 and 2000, 50% of the selected countries are

17Maria Helena Ramírez Tolosa, ¿Porqué no se deben reducir los aranceles para la importación de trigo en Colombia?, Comité de Seguimiento a la Competitividad de Arroz de Colombia, Informe Final, Abril 2002. Martínez Covaleda, Héctor, “El efecto sustitución de consumo entre trigo y arroz en Colombia: una medición econométrica”, Revista Induarroz, No.3, Bogotá, Agosto 1999. Abreau y E. Ablan, 25 años de cambios alimentarios en Venezuela (1974-1994), Fundación Polar, Venezuela. GRADE, Evaluación del sistema de sobretasas a la importación de trigo y harina de trigo en Perú, Mimeo, Lima, 1993. Derek Byerlee, Ventaja comparativa y política agrícola para la producción de trigo en Ecuador, CIMMYT, México, 1976. 18 Per capita consumption for all countries reported by FAO was calculated for 1990 and 2000. Data in tons were taken for imports, exports, production, and apparent consumption. This amounts were divided by the annual population reported by FAO. Rice production in terms of white rice.

Critical Values

CriticalValue

L- Max 90 Traza 90

Nothing 19.86 22.95 24.80 12.39Máxim once 4.95 10.56 4.95 10.56Nothing 11.25 10.60 14.40 13.31Máxim once 2.15 2.71 2.15 2.71

Cimt - Arnal

Acif - Tcif

H0: Númbers of Cointegration

Vectors

Verosimilitud

Ratio L-Max

Traza

Cointegration Relationship - Test de Johansen

45

below the average. Excluding this percentage, 65% of the countries evidence a trend with high negative correlation among the consumption of rice and wheat: High consumption of wheat represents low consumption levels of rice. Table 3.6 On the other hand, a 33% of the countries on the sample register high levels of rice consumption and low levels of wheat consumption. TABLE 3.6 COUNTRIES AND SHARE OF RICE AND WHEAT CONSUMPTION, 1961-

1970, 1990-2000. 1961-1970

Consumption pattern versus World average Country Total Percentage

% excluding poor countries Under world average

1. Less rice , more wheat 131 62% 2. Less rice and wheat 35 17% 44% 3. More rice less wheat 0 0% 0% 4. Less rice more wheat 44 21% 56% Total 210 100% 79 Source: FAO. Authors

1990-2000 Consumption pattern versus

World average Country Total Percentagel % excluding poor countries Under world average

1. Less rice, less wheat 104 50% 2. More rice, less wheat 35 17% 33% More rice, more wheat 2 1% 2% 4. Less rice, more wheat 69 33% 65% Total 210 100% 106

Source: FAO. This data shows that countries with high per capita wheat consumption with respect to the world average are accompanied by low per capita consumption of rice, suggesting the inverse relationship among these goods. By the same way, last charts suggest that some changes have occurred on the consumption patterns for these grains. While for the 1961-1970 period only 44 countries registered more wheat than rice consumption (over the world average), during 1991-2000 the amount increased to 69, while those sustaining more per capita consumption of rice than wheat kept constant at 35. Countries that increase wheat consumption substitute rice for wheat. Several empirical studies for Colombia have shown the important substitution effect of wheat derivatives over white rice consumption. The most recent study was made for the “Comité de Seguimiento a la Competitividad del Arroz” in 2000 (Ramírez, 2002). Before presenting the results coming form that study, it is necessary to say that wheat prices of Colombian wheat follow very close international prices, having into account that imports are regulates by SAFP for wheat19.

19 Correlation coefficient among CIF price and import cost is high, 75%. Notwithstanding, national wheat represents only 4% of apparent consumption.

46

As observed graphically, relationship among imported wheat price and rice price in Colombia is relatively high (48% with no lags). Notwithstanding, as proved later, price transmission does not occur simultaneously, it converges in the long run via consumption substitution. Here are the results of the study mentioned above. The first fact to prove was to establish a relationship among internal white rice prices and its substitutes, this is wheat derivatives and the maize flour and the potatoes, measured by the CPI index starting 1991 and finishing 2002. By method, the hypothesis of long run equilibrium is tested through a co-integration analysis. If prices to consumer of white rice, spaghettis, bread, potatoes, maize flour, keep the equilibrium relationship, a linear combination of them would present a trend to converge to an equilibrium value in the long run, if the effects on the economy are only temporal. Cointegration analysis was made under the methodology of Johansen. In first place, the Dickey and Fuller test evidenced that the price series had unit root, they were non stationary. This means that the prices individually do not converge to a value in the long run and shocks over them would have permanent effects, making their long run movements unpredictable. Furthermore, this series had a deterministic term. Notwithstanding, despite individual prices are not stable in the long run and alterations over them may last for a long time, it can happen that their combination inside a market tends to a long run equilibrium eliminating shock effects and generating a joint relationship that make relative prices move in the same direction inside of the market. To test this hypothesis a Vector Auto Regressive-VAR model was estimated using Akaike, Schwarz, and Maximum Likelihood (ML) criteria and serial correlation tests to determine the appropriate lag number. Test suggested two and three lags. This exercise used three lags since there is a guarantee for no serial correlation problems. Finally, Johansen test was performed to determine if price indexes keep a joint relationship in the long run. It was found that price indexes keep a long run relationship in the market, series are co-integrated. This way, in spite individual prices showed a positive linear trend, it is not probable to eliminate price shocks in the short run and any deviation from equilibrium can be eliminated. Concretely, results indicated two possible co-integration relationships among the five price indexes, which jeans that in the long run, eventual shocks to the prices of the mentioned products would be temporal and in the long run they would tend to adjust. Even more, any alteration over the price of any of the goods would cause the others to move in the same direction to equilibrium. Second part of the exercise consisted in an estimation of the magnitude of the substitution effect among these products, so a determination model for wheat and its substitutes was calculated. The exercise was made using an annual series for the 1961-2001 period for the following variables: Per capita consumptions and real prices of: white rice, maize for human consumption, and potatoes in kilos and constant 2001 pesos per kilo. Per capita GDP (2001 prices) as a measure of income

47

Real Foreign Exchange Index weighted by the Producer Price Index with base 2001 as a measure for foreign trade policy. A model of long run demand equations was designed that allowed to calculate the average elasticities of consumption with respect to the imported wheat price, products prices, income and foreign exchange policy. These equations are approximated to conventional demand equations that contain price of the good and its close substitutes, consumer incomes and foreign exchange policies. Equations are specified in per capita terms, relating to consumption to have into account population changes. In the same way, the income measure picks the effects of changes in the economic structure of the country. In theory, changes in income modify preferences for food, while changes in prices modify consumption decisions facing close substitutes in the budget of a typical consumer. Products in this study have a high imported component in costs and external demand, real foreign exchange affects decisions of importers and products prices. It is convenient to precise that wheat, its derivatives and substitutes conform a market which components are interrelated, in consumption distribution of each good households consume. In fact, as explained before, prices of these products have a long run relationship, they evolutes over a temporal convergent path making perturbations short lasting and transitory. This implies that demand equations are dependent one to each other and as a consequence, its estimation should be simultaneous to avoid biases in the calculus of elasticities. The purpose of this exercise was to quantify the price elasticities of rice demand, human consumption maize, and potatoes, to obtain the substitution parameters among rice and wheat. Equations were estimated linearly by Factible Generalized Least Squares Method in two stages. Results appear on Table 3.7.

48

TABLE 3.7 ESTIMATIONS OF PRICE ELASTICITIES FOR RICE, WHEAT, AND

POTATOES EXPLICATIVE VARIABLES */

DEPENDENT VARIABLE Rice Real Price

Wheat Real Price

Real PIB per

cápita

Real Exchange rote indice

Corn Real Price

Potatto Real Price

Dummy 1990 R² R²

ajust.

Per capita rice consumption

Elasticities -0.94 1.16 1.99 -0.84 0.641

0.6126

t test -4.5744 5.526 11.987-4.272 Per capita wheat consumption

Elasticities 0.87 -0.33 0.12 0.646

0.6168

Prueba "t" 6.2013 -2.3300 1.9801

Per capita potatoes consumption

Elasticities 0.41 2.09 -0.78 -0.22 0.849

0.8325

t test 6.314 11.80 -3.8920 -2.9117 */ t test shows significance at 94% level Residuals first order autocorrelation test: Equation 1: serial correlation-Q (LM)= 1.6126 Prob>Q=0.2041 Equation 2: serial correlation -Q (LM)= 0.2263 Prob>Q=0.6342 Equation 3: serial correlation -Q (LM)= 1.3525 Prob>Q=0.2448 Source: FEDEARROZ, Ministerio de Agricultura, Banco de la República, Cálculos este estudio. The advantage of simultaneous estimation lies on the Fact that parameters are jointly estimated due to the interrelation among variables. This way, perturbations on rice markets affect not only rice price but other prices, gaining statistic efficiency and confidence in results. In particular, estimated parameters resulted with expected signs and elasticities were significative at levels around 94%. Equation for rice, as awaited, showed that per capita consumption of rice was altered by the real price of wheat. In fact, the sign of the price elasticity of rice demand with respect to rice is positive, implying a reduction in the real price of wheat would cause a reduction in the per capita rice consumption and vice versa. Causality is as follows: A reduction in the price of imported wheat increases per capita consumption of wheat and, this way households reduce their rice consumption. Wheat derivatives become cheaper in relative terms. This result confirms the hypothesis of substitution among rice and wheat.

49

Elasticity of substitution among rice and wheat for 1961-2001 is about 1,16. A reduction of 1% in the price of wheat would cause a 1,16 decrease in the consumption of white rice per capita. On the other hand, and as awaited, rice demand reacts inversely to its price. Price elasticity for rice demand was -0,94 for the same period, per capita consumption of rice is elastic to price. Per capita GDP influences directly over rice demand with an elasticity around 2, rice is very elastic to changes in income. Per capita rice consumption reacts directly to real appreciation of peso and vice versa. This is explained because a reduction of foreign exchange encourages wheat imports. In summary, results show that white rice prices are conditioned to the behavior of imported wheat prices. If those prices are conditioned to subsidies, then Colombian rice prices are conditioned by rice and wheat subsidies. EFFECTS OVER COLOMBIAN RICE SECTOR As shown before, international rice and wheat prices have suffered an important reduction, partly, because of the subsidies from the OECD countries. By the time, this negative trend has been transmitted over national rice prices.20 Notwithstanding, this price diminution was reduced by the behavior of some other variables that jointly explain movements of sowed areas, production, employment, and income of rice producers. This section identifies magnitude of these effects.

20 Nacional rice price is affected by wheat price via substitution.

50

IMPACT ON RICE PRODUCTION As studied before, rice demand is explained by the combined effect of income, substitute prices, and its own price. GRAPH 3.5 PRODUCTION AND GREEN PADDY PRICE IN COLOMBIA, 1981 – 2002.

Source: Agriculture Ministry and FEDEARROZ. As Graph 3.5 shows, in general, a decrease in prices decreases production with some lag, and vice versa. In fact, the price elasticity of production in the long run fro the period 1980-2001 was calculated in 1,14, indicating that Colombian rice supply is very elastic: Production increases or decreases more than proportionally than the increase or decrease of the prices that producers receive. Production increased until 1990 as result of a set of economic measures to stimulate prices for producers, known as “selective supply”. In fact, between 1988 and 1989 prices increased extraordinarily and the effects of this increase translated into an increase in production until 1990’s. With the implementation of the economic openness, imports were liberalized, lowering prices, notwithstanding, implementation of a price fringe for white rice and wheat since mid 1991 (a Colombian fringe in 1991 and then and Andean one in 1995), avoided that international prices of rice translated to national producers.

51

Since 1991, imported wheat was subjected to a harvest absorption policy and to pay the tariffs from the price fringe system. Then, the tariff was cut to 35% levels. Rice was subjected to this system and a series of measures were applied to avoid the immediate transmission of international prices on the national prices, so this works as in a closed economy, except for the behavior of the cost of import of substitute goods and the exchange rate. Since 1993 until 1988 international prices of white rice and wheat registered an important increase in their quotations, placing even over the average of the last decades. Graph 3.6

GRAPH 3.6 MONTHLY NATIONAL RICE PRICE AND IMPORT COST OF WHEAT, 1991 – 2003.

Source: Agriculture Ministry and FEDEARROZ. International white rice quotation increase 61% between the june 1993 and july 1998 period and the internal price increased 77% in constant terms. The relationships established above show that an increase in the international prices pressed up the national prices and production increased notably. This effect was reinforced by the decrease in real

52

terms of 3,4% in the production costs, shifting supply curve to the right, as a result of Colombian peso re-appreciation in this period that made imported inputs (39% of total costs) cheaper in real terms and because productivity increases. Notwithstanding, per capita consumption also increased from 29 kg to 43 kg, in part because of imports (by 1988 imports were 19% of rice consumption). This situation of price increase was not expected because of demand elasticity of -0,94. Several studies show that the increase in the expenditure and economy’s overheating because of demand excess and the income elasticity of rice (1,99), rice consumption tended to increase. Since July 1988 and to the present (December 2002) rice prices fell sharply (-49% in constant terms), over the international rice quotations of -41%21. This price decrease increased internal demand and production. As import cost of wheat kept constant, and internal price of rice decreased, relative prices were favorable for rice, explaining the increase in demand in the short run. (Graph 3.7) With no price fringe, internal rice prices should have reacted in other way. GRAPH 3.7 RELATIVE PRICES OF WHEAT DERIVATIES AND RICE PRICES, 1991 –

2002. Source: Ministerio de Agricultura, cálculos este estudio. In spite of this, price decrease should have not induced producers to increase production. In fact the harvested areas increased in 78.000 hectares and production in 415.000 tons, a 19% and 22%, respectively. 21 Figures in US$ currency of USA.

53

This is explained because of reduction in production costs22, that shifted supply curve to the right, that was superior to the decrease in costs, increasing profitability of rice. The cost decrease more than proportional than price decrease stimulated production without having into account that some other agricultural activities were covered by national rice production. As a consequence, it can be affirmed that given the relationships among international quotations (conditioned by international subsidies) and the internal rice prices, the latter decreased stimulating per capita consumption of rice. Notwithstanding, decrease in prices was compensated more than proportionally by the cost decrease that made that profitability did not get affected, stimulating production. EMPLOYMENT GENERATED BY RICE CULTIVATION Employment generated by rice increased by 0,3% annually between 1987 and 2001. While in 1987 54.789 new employments were generated, by 2001 the amount was 69.020. In summary, by the end of the period 14.222 people were hired over the initial period. Behavior of employment is tied to the production evolution and it was conditioned to the movements in relative prices with wheat derivatives, income and exchange rate. And from supply sight, an explanation comes from costs, production and hence, employment. As long as from 1997 production increased notably, employment answered the same way. Increase in harvested areas in this period did not obey to a price increase but to a reduction in costs allowing a better profitability for the harvest and increases in employment and production. EFFECTS ON TRADE BALANCE Colombia was a rice net exporter until 1991 when 85.000 metric tons were sold. This amount represented 8,1% of apparent internal consumption. Starting the following year, situation changed and trade balance on this product has been negative since then. Highest importations were made between 1994 and 1998.Table 3.8. This last year imports were about 19% of apparent consumption. Imports come mainly from Ecuador and Venezuela, CAN members. This imports could have continued growing if no safeguard had existed in 1998.

22 Between 1998 and 2001real costs decreased 4,4% while prices 4,1%

54

TABLE 3.8 MAIN ECONOMIC INDICATORS FOR RICE IN COLOMBIA, 1991 – 2001

Year Production (Tm)

Imports (Tm)

Export (Tm)

Trade Balance

Apparent Consuption

(Tm)

Percapit Consuption

(Kg/Hab)

CAN´s Imports

(Tm)

CAN’s Export

(Tm) 1991 1,130,090 249 84,509 84,260 1,045,830 29.3 150 70,485 1992 1,127,718 68,766 450 -59,612 1,187,330 33.3 34,931 450 1993 1,033,540 31,094 3,666 -30,222 1,063,762 29.8 33,877 3,661 1994 1,077,188 210,566 2,290 -223,147 1,300,335 36.4 80,418 219 1995 1,159,995 100,817 294 -101,712 1,261,707 35.4 79,417 293 1996 1,079,857 167,9547 47 -131,700 1,211,557 34.0 113,239 10 1997 1,189,686 148,739 172 -165,896 1,355,582 38.0 138,774 160 1998 1,233,590 265,028 346 -288,198 1,521,789 42.6 88,972 295 1999 1,420,401 159,3009 15 -37,594 1,457,995 40.9 40,554 0 2000 1,485,739 105,400 38 -57,093 1,542,832 43.2 52,921 2 2001 1,503,977 165,790 149 -151,942 1,655,920 46.4 89,453 141

:

Source: Ministry of Agriculture, Dane, Dian, Imports FEDEARROZ.

The cost in terms of foreign exchange for turning into an importer has been of about US$281.353 million. Only during the last year 61.400 metric tons of rice were imported for US$18,8 million.

55

IV. ENVIRONMENTAL ISSUES RELATED WITH THE PRODUCTION OF RICE

The Colombian Environment, Households and Territorial Development Ministry has developed a series of strategies focused to the continuous improvement of productive sectors such as the “National Program for a Cleaner Production”, environmental policy for the use and management of plaguicides, sub-sector environmental guides. For the case of rice cultivation, there is a series of factor influencing environment such as agricultural frontier expansion, mono-cultivation increase, ecosystems simplification, and agrochemicals consumption. Because of the importance of rice sector as employment generator and alimentary security, it must be revised the environment conservation and the rational use of resources as a factor of sustainability. New challenges must be imposed for the rice productive system with the incorporation of new environmental variables for the industry to guarantee position of the product at the national and international markets. This is, with the purpose of prevent and diminish the negative impact of intensive agriculture through environmental friendly technologies adoption to add productivity to the sector. ENVIRONMENTAL POLICIES AND NORMATIVE FRAMEWORK Relating to the environmental matter, the country accounts a very wide legislation that is contemplated in the National Code of the Renewable Natural Resources and of the Atmosphere, Decree-Law 2811 of 1974. By this way the Colombian National Political Constitution of 1991, taking as referent the former pointed out Code, establishes a group of public rights and duties of the State, the institutions and particular individuals, in environmental matter, framed in the principles of the sustainable development. This command propitiates the expedition of the Law 99 of 1993 where the environmental national system was settled down, the Environment Ministry is created, the Public Sector in charge of the Administration and Conservation of The Environment And The Renewable Natural Resources is reorganized and the Environmental National System –SINA- is arranged. The Environmental National System accounts with four Institutes of Research, with the technical and scientific support from public and private universities and research centers. Additionally it has also 34 regional environmental authorities (Regional Autonomous Corporations) that take in charge of the protection and conservation of the environment in the different regions of the country. The Ministry of the environment has emitted political, limits and environmental strategies, aimed to guide the preservation and conservation of the natural resources and to promote the sustainable sector development, items inside of which the following ones are mentioned: • Biodiversity Policies • Policies for the Wild Fauna Administration • Policies for the Integrated Classification and Sustainable Development of the Coastal Areas

56

• Forests Policies • Policies for Integral Administration of Solid Residuals • Policies and Limits for the use and handling of plaguicides • Base for population's and environment policies • Limits of a policy for the civic participation in the environmental administration • Policies for environmental education • Policies and limits for the water integral management • Strategy for a national system of protected areas • Limits for the Territory Environmental Classification Policies • Cleaner Production Policies • Strategic plan for the Forest’s Restoration And Establishment • Environmental Research Policies • National Policies for Colombian Interior Wet Areas Additionally, the country has a series of norms and regulations for the protection of the environment and the natural resources, among which some of interest are mentioned directly related with the rice sector: With respect to the hydro resource exists the National Sanitary Code (Law 09/79), management and handling of hydrographical basins ( Decree 2857/81 and Decree 1728/02), uses of the water and liquid pouring material (Decree 1594/84), efficient use and water savings (Law 373/97), concessions of use public waters ( Decree 1541/78) and Decree for Compensation Rates. With regard to the air resource norms about atmospheric emissions have settled (Decree 02/82), about Prevention and Control of Atmospheric Contamination and the Air Quality Protection. (Law 48/95). With relationship to the resource floor, norms have been promulgated for the protection and classification of the territory through the municipal authorities (Law 388/97). In the topic of ordinary and dangerous residuals, is the Sanitary Code (Law 9ª/79), the Decree 2104/83 that regulates the administration of the solid residuals, the Law 430/98 that dictates prohibitive norms in environmental matter with respect to the dangerous waste. Likewise, the National Code of the Natural Resources, establishes general measures for the protection and use of the fauna and wild flora; the Decree 1608/78 dictate dispositions for the conservation of the wild fauna and it also defines the administrative entities for its handling, management and control; the Law 84/89 created the National Statute for the Animals Protection; the Law 491/99 establishes the ecological insurance and it reforms the Penal Code in what concerning to environmental crimes; the Law 99/93 defines the competitions to authorize and concessionaires operation bylaws for the forest uses and it defines evaluation functions, control and checking. In what has to do with the plaguicides use and handling, the competent entities for this control are the Environment Ministry, within the framework of the Law 99/93 and the Decree 1728/02; the Ministry of Health through the Law 9ª/79 and the Decree 1843/91; the Ministry of Agriculture; through the ICA, the Resolution 3079 of November of 1995 has been promulgated, with which dispositions about the industry, trade and bio inputs materials application and products are dictated, and about fertilizers, amendments, conditioners of the floor and related products :

57

chemical plaguicides, physiologic regulators, helping for agricultural use and products are also dictated. Recently the country adopted the Andean Community of Nations Decision 436 about the registration and control of chemical plaguicides for agricultural use. By other hand, Colombia is signatory of the Stockholm (Persistent Organic Pollutants) and Rotterdam (Based Previous information Consent) Agreements. IDENTIFICATION AND EVALUATION OF ENVIRONMENTAL EFFECTS With the purpose to analyze and evaluate the possible environmental effects associated to the cultivation of the rice, the different phases were valued which involve this activity, starting from the infrastructure development to the final product commercialization, in such a way that the activities in a real context were kept in mind, such as: the use and handling of inputs and teams, the wastes generation, the potential conflicts and benefits, the use of the natural resources, the policies and national regulations and the management abilities of the producers and of the community. For such effect, it was designed and obtained a matrix that incorporates the previously signal components, describing them in a subjective way according to their magnitude, frequency and intensity. The results of this evaluation are described in the Figure 4.1 and the analysis of the same ones are properly detailed in the subsequent chapters, including a proposal guided to define measures of environmental management that lead to improve the competitiveness and sustainability of the Colombian rice sector.

58

TABLE 4.1 ENVIRONMENTAL EFFECTS IDENTIFICATION MATRIX

PHASE CONSTRUCTION CROP DEVELOPMENT POST-HARVEST

ACTIVITY

ENVIRONMENTAL COMPONENT

SOIL

ADE

QUA

TIO

N

CONS

TRUC

TIÓ

N AN

D RO

AD A

DEQ

UATI

ON

DAM

CO

NSTR

UCTI

ON

IRRI

GAT

ION

CANA

L CO

NSTR

UCTI

ON

RESE

RVO

IRES

BURN

ING

S

LAND

PRE

PARA

TIO

N

SOW

CULT

URAL

LAB

ORS

PLAG

UICI

DES

FERT

ILIZ

ERS

IRRI

GAT

ION

HARV

EST

HARV

EST

RESI

DUAL

S

WAS

TE M

ANAG

EMEN

T

HULL

AND

MIL

L RE

SIDS

RESI

DUAL

WAT

ER

MAN

AGEM

ENT

DRY

MIL

L

STO

RAG

E

TRAN

SPO

RTAT

ION

ELEMENT INDICATOR 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

QUALITY

STABILITY

EROSIÓN LAND

FERTILITY

QUALITY WATER

QUANTITY

PHYS

ICAL

AIR QUALITY

VEGETATION HABITATS

LAND FAUNA

AQUATIC

BIÓ

TIC

LANDSCAPE RELATIONSHIPS IN ECOSYSTEM

COMMUNITY EMPLOYMENT

AGRÍCULTURAL

INDUSTRIAL

SOCI

OEC

ONÓ

MIC

AND

CU

LTUR

AL

PRODUCTIÓN

TRADE

59

IDENTIFICATION OF USED INPUTS IN THE RICE CULTIVATION The rice agro industrial chain is one of the more demanding sectors of agrochemicals for production, represented basically by fertilizers and plaguicides of different nature, characteristic and formulation. Rice is produced in diverse types of weather of tropical humid warm climate where the insects, plagues and the illnesses are more plentiful than in a dry atmosphere; joined to the fact that rice is cultivated during the whole year in the areas of more production, propitiating the attacks of insects and illnesses whose epidemic pressure on the cultivation is translated in bigger incidence, severity and a significant increment in the control costs. CORRECTIVES In the production of the “upland rice”, especially in the Oriental Plains, applications of lime between the 5 and the 30 days before the sowing are carried out, in dose that fluctuates among 200 to 500 Kg/ha. In the Table 4.1 one can be observed the main correctives, the dose and used applications. TABLE 4.1 DOSE AND NUMBER OF APPLICATIONS OF THE CORRECTIVES MORE USED IN THE RICE, A SEMESTER, (COLOMBIA 2000).

Corrective Dose No of Applications

AGRICULTURAL LIME 100 1

LIME DOLOMITICA 150 at 500 1

CALFOS 300 1

Source: FEDEARROZ, National Rice Survey

FERTILIZERS In the Oriental Plains 70% of the technical assistants recommend the use of the DAP (Ammonium Phosphate) and 60% of the producers generally use it in more dose to those recommended. The applications of the DAP are generally carried out in doses between 50 and 300 Kg/ha in a single application; the KCl (Chloride of Potassium) is used in dose from 50 to 200 Kg/ha; depending on the fertility of the floor, the applications can be fractioned between one and three. The Urea is used in doses from 125 to 275 Kg/ha in three applications: 90% of the assistants recommend its application and 90% of the farmers they use the urea as a basic source of nitrogen for the rice production in Colombia. In the Table 4.2 the main fertilizers are indicated they are also used in the cultivation of rice.

60

TABLE 4.2 DOSE AND NUMBER OF APPLICATIONS OF THE FERTILIZERS MORE USED IN THE RICE TILLAGE , A SEMESTER , COLOMBIA - 2000

TRADE NAME APPLIED DOSE NUMBER OF

APPLICATIONS

KG/HA

UREA 50 at 600 1. 6

DIAMONICO PHOSPHATE (DAP) 25 at 300 1. 4

CHLORIDE OF POTASSIUM (KCL) 25 at 275 1. 4

SULFATE OF AMMONIUM (SAM) 25 at 400 1. 4

RAFOS 200 at 225 1. 2

TRIPLE 15 (15-15-15) 50 at 400 1. 3

TRIPLE 18 (18-18-18) 35 at 500 1. 3

TRIPLE 14 (14-14-14) 75 at 125 1

NITROMAG 50 at 125 1

AGRIMIS 10 at 46 1

FERTIAZA 10 at 50 1

HUMUS 70 at 600 1

SULFGOMAG 125 1

SULCAMAG 25 at 75 1

VIGOR 20 at 46 1

SULFURATE ELEMENTARY 10 at 75 1. 3

FERTIMEX 20 1

SULFATE OF ZINC 10 at 30 1

Source: FEDEARROZ, Survey National Rice

USE OF PLAGUICIDES As it was mentioned, a range of products of generalized use in the cultivation of rice exists, among which are herbicides, insecticides, fungicides, nematicides and raticides that indicate the importance that they have acquired inside the production process and in turn the high dependence degree on them generated in the farmers of rice of Colombia and in general in the world. The herbicides used for the control of weeds are applied as practice, before, at the moment and after the cultivation sewing. The practice of the control of weeds in rice begins with a work named “the burning” that is carried out after concluding the preparation of the floor, to

61

control mainly the wide and leaf weeds, the gramineous ones and especially the red rice. In the Table 4.3 the main herbicides are introduced and presented which are used in the cultivation.

TABLE 4.3 DOSE AND NUMBER OF APPLICATIONS OF THE MORE USED HERBICIDES IN THE CULTIVATION OF THE RICE FOR “ BURNINGS”, A SEMESTER 2000.

COMMERCIAL NAME APPLIED DOSE APPLICATIONS

LTS/HA NUMBER

ROUND UP 3 1 – 2

FAENA 3,5 1

GLYFOSAN 3 1 – 3

GLYFOSOL 3 1 – 2

COLOSO 4 1

SOCAR SL 3 1

GLIFOSATO 324 3 1

GRAMOXONE 3 1 – 2

GOAL 1,5 1 – 2

Source : FEDEARROZ, National Rice Survey.

The control of weeds in the pre-emergency stage is carried out by 100% of the farmers, especially in the system of rice watering; in the system of favored “upland variety”, the control is carried out by approximately the 60% of the farmers. The third phase of weeds control identified as post-emergency, is carried out by means of the use of products with certain specificity, because its efficiency and selectivity. In the Table 4.4 the herbicides used in the pre phases and post emergency are observed.

62

TABLE 4.4 DOSES AND NUMBER OF APPLICATIONS OF THE HERBICIDES" PRE AND POST EMERGENT" MORE COMMONLY USED IN RICE, A SEMESTER, COLOMBIA 2000.

TRADE NAME APPLIED DOSE

NUMBER OF

APPLICATIONS

LTS/KG/HA

STAMPIR 4,0 1

CLASSIC 0,075 1

PROWL 400 3,2 1

PROWL 330 3,5 1

NOMINEE 100 SC 0,4 1

COMMAND 1,2 1

SKOL 60 WG 0,5 1

FACET 250 SC 1,0 1

BOLERO 3,2 1

FURORE 12 CE 0,8 1. 2

CLINCHER 1,2 1. 2

FORESITE 2,5 1

RAFT 400 YOU 0,7 1

MACHETE 3,2 1. 2

RIFIT 500 2,3 1

ROSTAR F.M.B. 38 2,0 1

PROPANIL 500 4,5 1

Source: FEDEARROZ, Survey National Rice

The existent information permits to determine the farmer’s preference for the weed control in rice by means of the use of chemical plaguicides, with the consequent risks on the atmosphere, the health and the rice agro ecosystems of the country. An aspect to keep in mind regarding its effects is that the populations of predominant weeds have come to change in their proportions and aggressiveness for the selection effects, induced by the dependence of chemical molecules for its control and additionally it can have been generated resistance to the mentioned products. The same as in the control of weeds, the chemical control of the different insects´ plague includes a range of active ingredients of different characteristic, as much for its synthesis as for its action, and whose application dose oscillates among 1 and 2 L/Ha, and the number of applications they have are between 1 and 3 for the cultivation cycle. In the Table 4.5 the main insecticides used in the cultivation of the rice are shown.

63

The insecticides that are applied in a persistent and continued way in the cultivation, means, as the same for the case of the herbicides, a tool of artificial selection for the plagues and in turn the simplification of the entomofauna biodiversity and of other vertebrates, around the agro-rice ecosystems of the different producer regions. TABLE 4.5 DOSE AND NUMBER OF APPLICATIONS OF THE MOST USED INSECTICIDES IN RICE, A SEMESTER, COLOMBIA. 2000.

TRADE NAME APPLIED DOSE NUMBER OF

APPLICATIONS

LTS/KG/HA

AZODRIN 600 1,0 1

BASUDIN 600 EC 1,0 1

BIDRIN 48 CMA 1,0 1

CURATER 3G 1,5 1

YOU SAY 0,25 1. 2

DIPTEREX 80 SP 1,0 1

FURADAN 3D 0,8 1

FURADAN 3G 20,0 1

KARATE 0,3 1. 2

LORSBAN 4E 1,0 1. 2

MONOCROTOFOS 600 0,8 1

NUVACRON 60 SCW 0,75 1

POLYTRIN 200 EC 0,25 1

ROXION 0,5 1

SISTEMIN 40 1,0 1. 2

TRICHOGRAMA * 30,0 1. 2

BULDOCK 0,25 1

TREBON 0,50 1

INSECTRINA 200 0,30 1

HIPERKILL 0,30 1

HUNTER 80 0,06 1

INISAN 60 CO.LTD. 0,70 1

ACTARA 0,10 1

WHIP EC 0,30 1

Source: FEDEARROZ, National Rice Survey.

Illnesses that affect rice cultivation in Colombia depend on variables such as temperature,

relative humidity, precipitation, that influence the epidemic development. Rice ecosystems

64

are affected by a sort of illnesses that are controlled with a variety of products as Table 4.6

shows.

TABLE 4.6 MORE COMMONLY USED FUNGICIDES FOR RICE, SEMESTRE A,

COLOMBIA – 2000. COMERCIAL BRAND DOSIS APPLICATIONS LTS/KG/HA NUMBER BAYLETON 25 EC 0,4 1 BENLATE 0,3 1 BIM 75 0,3 1 CARBENDAZIM 0,5 1 CALIDAN 1,0 1 DEROSAL 500 0,6 1 DITHANE FMB 4,0 1 – 2 DITHANE M45 3,0 1 ELOSAL DISPERSION 0,8 1 FUDIOLAN 1,0 1 HINOSAN 0,8 1 KASUMIN 2% 1,25 1 KITASIN 48 CE 1,20 1 MANZATE 200 0,5 1 MERTEC 450 FW 0,5 1 MONCEREN 1,0 1 MONCUT 1,30 1 TILT 250 EC 0,50 1 VALIDACIN 3% 1,50 1 TIFON 50 WP 1,00 1 INDAR 0,50 1 TASPA 500 EC 0,40 1 BRODIONE SC 0,60 1 AMISTAR 250 SC 0,60 1 SOURCE: FEDEARROZ, Rice National Survey ENVIRONMENTAL EFFECTS ASSOCIATED TO RICE CULTIVATION Some important considerations to study are: WATER RESOURCE Rice cultivation is characterized by a high water consumption, independent of the production system. Irrigated rice requires between 8.500 to 9.000 m3/ha, in terrains leveled with laser while conventional land requires between 15.000 to 16.000 m3/ha. To obtain high quality rice it is important to account with clean water, referring to the deep waters, the quantities of traces elements that can be percolated through the floor, due to the continuous agrochemical applications and fertilizers, are not very well-known at the moment; however, it should be considered as a process in development of magnitude maybe low but potentially high, in case of not mediate alternatives of rationalization in the application of such practices.

65

Some loses can be presented in the construction of preys, watering districts that increase the permeability of the floors and by chemical contamination, especially in areas of transitory flood, occupied by the system of automated watering and improved dry process and, as well as of the watering areas. The magnitude of the effect is very low but of high occurrence probability and of high tendency to mitigate, just as it has been discussed previously. Given the little use of deep waters in the handling of the watering systems in rice, one cannot even speak of conflicts between the load availability and the degree of recovery of the same ones. For the time an effect of low magnitude is considered, growing, of permanent duration and from mean to high mitigate capacity, being included in these measures a bigger rationality in their consumption, a bigger knowledge of the requirements and consuming use of the cultivation, reuse of the served waters, better use of rain and superficial waters. RESOURCE LAND

Land is affected through construction, land preparation, causing compactation and erosion. In production phase, land is affected by flooding, plaguicides among others. Additionally, run off of solid into land generates negative impacts. Could be loses in land fertility for inadequate practices of preparation, deforestation of basins, and mono-cultivation, influencing the diminishing of organic matter. Land compactation alters its physical and chemical characteristics because of lack of permeability and infiltration reducing internal drain and making easier the superficial erosion nutrient wash and contamination of water bodies. Inadequate application of fertilizers ends in an increase of costs and a reduction of yields. Effects of inadequate land are evident in the pH increase and the reduction of interchangeable aluminum. Fertility can be improved through and adequate incorporation of cultivation residuals, which are rich in nitrogen and potassium. Currently, the inadequate location of these residuals causes contamination in some zones. Lands for rice cultivation are located in adequate zones, but complementary processes generate environmental effects. Land can be affected by the physical and chemical characteristics of land. In the production stage, land could be affected by the inadequate and excessive use for agrochemicals that change its characteristics. To prevent these effects alternative methods for pest control are suggested such as Integrated Pest Management (actually in use), dose reduction, and application frequency. On the other hand, anaerobic conditions of flooded lands in irrigated systems alter soil characteristics. In spite soil microorganisms are a very small fraction of its total mass, they are essential for its fertility and hence production of cultivations. In general, flooded soil can be recovered through management practices reaching the positive microbial activity and adequate physical and chemical conditions. When wastes are managed through burning practices, not controlled, soil gets affected mainly through the destruction of vegetal covering and the alteration of micro-fauna. Deficient disposition of organic matter from harvest and industrial process have negative

66

impacts over the soil. This can be reduced through direct application or alternative techniques, improving its organic content, fertility and biodiversity of its surface. To improve soil quality, implementation of technical management activities such as biological pest control, minimum sow and adequate disposition of organic residuals is clue.

RESOURCE AIR

The atmospheric component can be affected during the construction stage by the noise of the machinery and the liberation of solid particles (powder, during the operation stage); likewise, in the stages of the cultivation and mill process, the emissions consist of ashy and volatile particles by effect of the sky open burning of the husk coming from the mill industries and the resulting fuzz of the crops, which are transported by the winds and deposited in the floors and superficial waters. Affectations are also presented by the air aspersions of herbicides and fertilizers whose liberation of gases and annoying and nasty scents constitute polluting factors; likewise, are present contamination around the centers of transformation of the rice paddy in white grain in the mills. The microclimatic elements atmospheric are affected by the simplification of the landscape, because of the widespread deforestation of the areas, to facilitate the air fumigations, the watering works and drainage. These practices on the whole have got cause a regional climatic deterioration, with a significant alteration of the precipitation, which unchains other environmental changes, like in the periods of drought that were only during the months of January and February in a beginning, today they stand in some critical years for periods from December until May, with the effects over production, yield and the sanitary level of the cultivations. In the same way, atmosphere is affected by the residuals of solid products resulting of the mill of paddy rice, as well as for the final disposition of the residuals mainly by the burning of the husk whose big volumes of slow decomposition constitute an impact on the atmosphere and the landscape. BIOTIC ELEMENTS AFFECTED In general, the development of the mono cultivation in the agricultural practice, has generated a succession of changes and alterations in the biological diversity of the natural ecosystems. Although they are considered necessary in many cases to maintain the purposes of the agriculture, they are required to achieve a significant and profitable production and to offer alimentary security. The elements like the flora and the fauna are affected by the different adaptation practices and preparation of the land, the preparation of the floors and the infrastructure construction, for the watering works and drainage, the application of fertilizers and plaguicides, as well as the disposition of solid waste to open field and the package deposit preparation.

Evaluated in edaphic terms, it is affected especially in the different activities that imply a modification, change or alteration of the such floor as watering place and the use of agrochemical products of herbicides, compound in those that the cultivation of rice is the first national consumer product; the effects can be considered of regional type when the

67

rice production settles down through watering districts, by the direct contamination of the bodies of water, during the application, as later by laundry. The terrestrial fauna, in what concerns to small vertebrates is affected by the pesticides application and for the disposition of waste in dumps, as well as for the continuous contribution of residuals to the channels and natural sources of water; equally an effect of local type is presented around the mills by big volumes of waste (husk) and of which is not carried out an appropriate use, contributing to the increase of rats and other non desirable species, in urban or very near areas to them. In the concerning to the spineless ones, mainly associated to the floor, the fauna is affected by the works that modify this such element as application of insecticides, and mainly for the conditions of saturation of water that undergo the floors rice, specially under watering, anaerobic condition that affects in a significant way the biological population, physics, chemistry and biochemistry of the same one. The aquatic biota is affected by the activities that alter the quality of the waters, such as the use of pesticides and fertilizers. The quality of the landscape is affected by the activities of the construction of the different works of adaptation of lands, for the processes of massive deforestation, the inadequate disposition of residuals and plaguicides containers and fertilizers along roads, channels and cultivation lots, and at level of mills, for the accumulation and burning at open field of the husk product of the mill process.

AFFECTED SOCIOECONOMIC ELEMENTS The cultivation of the rice contributes favorably on the communities for the employment generation and labor force absorption in the group of the activities of the agro industry. By the way, the cultivation energizes the local and regional development for the other related activities to the chain along the Colombian geography; being observed in many cases a high consumption of family labor force in the systems of dry and favored dry land system characterized by the little hectares distribution that is managed by producer and a smaller mechanization degree. The community organization is reflected in the development of the activities that imply workers' concentration as the preparation of the floor, infrastructure construction and civil works of systems of drainages and transport of materials in field; however, in aspects of civic participation for the taking of decisions in the development of the agro industry with a sustained approach that includes aspects related to protection and conservation of the atmosphere and the natural resources, it has not consolidated still as a strategy that guarantees said participation. The production is affected mainly by the way of prices of the land and uses of the same one, in the activities of adaptation of the land and handling of watering districts, settling down a differential appraisement of the land (measured by the value of the lease by hectare), depending on the production area and cultivation system. Another indicator of importance, has to do with proprietor's, or renter’s condition that assumes the producing of the rice in Colombia, prevailing the lease about the property, factor of great negative impact, on most of variables and considered environmental indicators.

68

The industrial and commercial activity implies employment generation, consumption of construction materials, benefit of services, distribution of the products, provision of raw materials, teams and inputs that correspond to the development of the production of the cultivation and they are considered of great importance. In the main rice areas of the watering system in Colombia, it is possible to detect conflicts of use of the hydric resource, due to the human establishments and municipal heads that are located close to the bodies of water that they are used as flowing of the watering districts, being observed high levels of contamination by organic matters, besides the waste and pour process of domestic residual waters and garbage, increased by the growing contribution of waste, biological residuals and of agrochemical materials coming from the rice agro industry.

ADVANCES RELATED TO ENVIRONMENTAL HANDLING Because of its importance in the National Gross Internal Product –GDP-, the employment generation and the alimentary security, the Colombian rice sector, has been propitiating strategies of environmental character through its guilds Fedearroz and Induarroz: Integrated Pest Management As it has been mentioned along the document, the irrational use of agrochemicals in the cultivation of the rice, impacts significantly public health, the environment and the competitiveness of the cultivation, for that reason FEDEARROZ with the support of universities and research institutes, has developed through educational strategies and material divulgation, conducive programs to an integrated pest management for rice. However, the strategies and used mechanisms have not allowed an effective implementation that offers a cost-effective solution to this problem. Agreement for a Cleaner Production with Rice Mill Subsector As an application strategy of national cleaner production policy, the sub sector of rice mill subscribed an agreement between its associates and the regional environmental authority of the Tolima department, with the purpose of improving its productive processes, to reduce and control the generation of residuals, to strengthen the institutional capacity and to develop environment-friendlier technologies. In spite of that, the generation of husk coming from rice mill reaches the figure of 400.000 tons a year approximately. It still does not account with alternative handling, using as only alternative open sky burning, which bears to a problem of environmental contamination that has generated conflicts with the communities beside to these industries. Environmental Guide for the Rice Subsector The Environment Ministry, Fedearroz and the Society of Farmers from Colombia, developed an Environmental Guide for the Rice Cultivation, with the purpose of promoting a technical reference of good agricultural practices for the continuous improvement of the sub sector, which is oriented to big, medium and small farmers. Competitiveness Agreement of the Rice Chain This mechanism has been propitiated by the Agriculture and Rural Development Ministry and the industrial companies and CEOs of rice producing companies, with the support of the different institutions and actors that have influence in the productive chain of this cereal and it seeks to guarantee the sustainable economic and social development and its capacity of competing internationally, by means of the coordinated execution of actions of short, medium and

69

long run aimed to improve the conditions of the competitive environment of the chain and firms. Programs of Efficient Usage and Reusage of Water The production of rice in Colombia requires the development of an infrastructure that assures the readiness and quality of the hydric resource, by means of the development of watering districts and drainage that incorporate the environmental variables. Reality shows that during the last decades investment has not been focused to improve, maintain and modernize an infrastructure to guarantee the readiness of the resource to the cultivated area and their possible expansions. Joined to the above-mentioned, a quick deforestation rate is evidenced in the basins and the development of inadequate cultivation practices has impacted water flows. This situation has propitiated the definition of actions from the sector to protect basins; such as the project with the Association of Users of the Chipalo River, the Ibagué Municipality, the Public Services Company of Ibagué and Cortolima, with the purpose of evaluating the benefits of the reuse of the water, using the treated municipal waters (reuse treatment) of the Ibagué city, in the irrigation of lots cultivated with rice and starting from its results to reply the pattern in other municipalities of the country. PROPOSALS FOR AN APPROPRIATE ENVIRONMENTAL HANDLING IN THE CULTIVATION OF THE RICE Rice sector plays an important role in the Colombian corporate level and must implement measures to mitigate environmental effects for the short, medium, and long term, to conserve and take advantage of the natural resources, taking into account technical, social, environmental and economic aspects to improve competitiveness and sustainability. Here are presented some alternatives to improve environmental management of Colombian rice sector, assuming current cultivated areas under a sustainability approach: • Rice Sector Environmental Evaluation: it can be constituted in a complement for the

environmental evaluations (EA) of specific projects (watering and drainage districts, mills plants, agro chemicals plants, landing fields of air fumigation, among others) in the development planning programs.

• Planning and design of new irrigation works under the environmental norms and by means of programs of conservation of hydrographic basins.

• Application of the environmental guides for the construction and operation of Watering Districts.

• Rationalization of the plaguicides and fertilizers consumption, by means of the development of technological packages of Integrated Pest Management and Integral floor and water Management.

• Foment of programs to efficient use and reuse of waters. • Strengthening of training programs about the secure use of plaguicides and their

residuals (expired plaguicides and expired packs). • Development of programs for the use of by-products generated in the cultivation and the

mill process, such as crop residuals and husk. • Optimization of water irrigation, drainage and handling of fertilizers. • Development of community reforestation programs areas affected by cultivation having

high-priority on basins. • Optimization of mechanization and farming and plowing programs for the floor

preparation.

70

• Promotion of living fences to diminish the impact on the landscape. • Establishment of plans and programs of integral development at municipal level (urban

and suburban) to contribute to the satisfaction of the basic necessities of housing, drinkable water, health, education and recreation for low-income sectors (operators).

Considerations for the Implementation of an Environmental Action Plan • Conformation of an Operative Committee with different institutions from the Public and

Private Sector. • Identification of each problem and its possible solution alternatives. • Definition of commitments and responsibilities. • Definition of work methodologies at every level, either by means of campaigns or transfer

programs. • Design of mechanisms to follow-up the Action Plan. • Interdisciplinary Inter Institutional Committees for Surveillance must be established to

study, analyze and propose the more suitable solutions to the problems generated in the production at the decisive level.

• Prioritize short and medium term actions; as well as to formulate and to implement Pilot Projects.

• Define an information system to maintain a permanent update of the current situation diagnosis at the cultivation level.

• Promotion of research tools, transfer and validation of alternatives that grant contributions to the IPM systems to decrease plaguicides usage or the substitution of highly toxic categories, as well as the reduction in the diversity of formulations and presentations of common active ingredients.

71

V. A MEASUREMENT OF PRODUCTIVITY AND TECHNICAL CHANGE ON

AGROCHEMICAL: THE CASE OF COLOMBIAN RICE AFTER GREEN REVOLUTION

Productivity analysis is one of the most important topics in agricultural economics. There are several works on the topic at the international level, but for Colombian agriculture there is a lack of documents, mainly because there are no data sources for long run series. Private entities such as Fedearroz have compiled entire data sets relating national cost matrices, product prices, yields, cultivated areas, and other useful data to develop sector studies. For this case, rice sector productivity and technical changes are to be analyzed to derive conclusions about factor substitution, productivity changes, and technical change effects on production. This essay is divided into three sections, the first introduces and formalizes analysis of cost functions and their link with productivity. Second section shows results separately for rain-fed and irrigated-mechanized rice, linking technical change with productivity to generate policies to reduce use for agrochemicals. Last section concludes. PRODUCTION, COSTS, AND PRODUCTIVITY: THEORY AND BACKGROUND FOR ESTIMATION This research uses the link between technical change, factor substitution, and productivity measures, to evaluate the behavior of inputs and total production of Colombian rice crop for both production systems. The main hypothesis to be proved is that after the Green Revolution and despite of the introduction of improved rice varieties, productivity of agrochemicals in Colombian rice production has been lowering through time. To prove this first hypothesis it is mandatory to calculate a superlative productivity index, in principle a Törnqvist productivity index. To support this main hypothesis other questions have to be answered, the first is related to technical change biases for the associated technology. It would be interesting to prove if there has been place for biases towards the intensive use of agrochemicals in Colombian rice production. If so, it would be interesting to observe how these changes have evolved through time. An important point to analyze would be the elasticities of factor substitution, just to answer the existing relationships between inputs. All of these analyses are to be done in an interregional context or at least in a comparison between both of the production systems because there are implicit regional differences. In order to achieve this whole analysis it would be necessary to calculate either a production or a cost function. Transcendental Logarithmic Cost Function approach is very useful to calculate all the mentioned measures, given the aforementioned relationship that exists between that functional form and the Törnqvist superlative index. The reasons to use a cost function instead of a production function are clear. As mentioned by Binwanger (1974), homogeneity of degree one is not necessary because cost functions are homogeneous in prices. This happens because a change in prices will not affect the factor ratios. Furthermore, no inversion factor coefficients matrix is necessary avoiding overestimation of errors. Finally, there is usually little multicollinearity among prices.

72

Now, translog cost functions let handle neutral and non-neutral efficiency and neutral and non-neutral economies of scale among years in the time series. Translog coefficients are often not biased estimates. Another important point is that shares are log-linear, making estimation easy. The final reason to use a translog function is that it allows a direct link to productivity analysis through rate of cost diminution concept by Chambers (1994). On the other hand, the use of other functions such as Generalized Leontieff functions would lack of a clear-cut link with productivity measures despite of important distinction between long and short run inputs. The latter fact, in spite of gains in efficiency to estimate aggregated data. Production and Productivity Issues: The use of a Törnqvist index to measure productivity and a Translog Cost function would be the best options to achieve the objectives of this research. In order to introduce some definitions to approach the methods and models to be used, it is mandatory to make a little introduction as a theoretical framework for whole of the study. A continuously twice-differentiable production function has to be assumed for each rice production system, depending on factors and under the recognition that input utilization depends on time:

( ) ( ( ), ); :1,...,jY t F X t t j n= (1)

Where X(t) stands for an input vector containing N factors for rice production, and t stands for a time trend to approximate technical change. It has to be assumed also that the function is quasi-concave in the inputs, and that it has returns to scale. Furthermore, it has to be assumed that the function is well behaved in economic and mathematic sense. For that production function there may exist a dual cost function that depends on a vector of price factors (W(t)), production level (Y(t)), and a time trend (Chambers, 1994). That cost function stands for the minimum cost achievable by the rice producer:

( ( ), ( ), ) min '( ) ( ) (2)C W t Y t t W t X t=

This cost function is assumed to be positive linearly homogeneous in prices of factors, concave, and increasing. No fixed costs are assumed. It is also a well-behaved function: Twice differentiable in arguments and continuous. This assumption is not far from reality, despite of relative cost problems mentioned above, rice producers are rational in the economic sense and they minimize costs. It is also assumed a Total Productivity Factor (TFP), representing the “ratio of the output to an index of inputs” (Good et. Al., 1997), or more simply, it could be interpreted as the average product of all inputs.

( )( ) (3)

( )Y t

TPF tM t

=

Where M stands for an input index that could be calculated either using a comparison of expense on inputs with total output value or geometric weights to approximate input weights. Taking logarithmic derivatives at both sides of (3):

73

. . .

(4)TPF Y M= − Following Chambers (1994), the expression (4) can be made operational specifying a form for the time rate of change of the aggregate input. So it is assumed that:

. .

; 1,..., (5)j jjM S X j N= =∑

Where Sj stands for the share of input j in the production, it is:

/ (6)j j jS W X C=

Replacing (5) into (4):

. . .

(7)jjj

TPF Y S X= − ∑

As was mentioned above, this kind of expressions applies to continuous data, so an approximation to Törnqvist index can be useful at this point (Good, Nadiri, Sickles, 1997):

,, , 1

1 1 , 1

1log log ( ) log (8)

2j tt t

j t j tjt t j t

XTFP YS S

TFP Y X−− − −

= − +

∑

Following Jorgenson and Griliches (1972), productivity can de defined as:

,,

1 , 1

log log (9)i tti t

t i t

XYFP

Y X− −

= −

And finally using the two last expressions it is possible to come up with a useful expression to be implemented empirically, having into account that the sum of shares equals 1 :

, ,, ,

1 , 1 , 1

log log log ; (10)j t i tti t j tj

t j t i t

X XTFPFP S i j

TFP X X− − −

= + − ≠

∑

Chambers proposes his rate of cost diminution as:

.ln

(11)C

TFPt

θ∂

= − =∂

This expression can be approximated to enter into (10)

, ,, ,

, 1 , 1

log log ; (12)j t i ti t j tj

j t i t

X XFP S i j

X Xθ

− −

= − + − ≠

∑

74

It is valuable to mention that the perspective found in Good et Al (1997) mentioned at the Theoretical Frame, will not be applied here because of sample dependency problems and because for this case it has no economic sense. In some way mechanized and Mechanized Rain-fed Rice are independent and a panel data analysis containing both production systems is not clearly the best option. Furthermore, as mentioned above mechanized accounts for 98% of whole production and Mechanized Rain-fed Rice only for 2%, despite the latter contains a larger proportion of farmers. The empirical continous approximation of the above theory can be done using a Translog cost function since the data are easily obtainable: Shares and prices for j factors of production, total output (Y), a time trend (t), and a dummy variable (D) to take into account the introduction of FEDEARROZ-50 seed since the second semester of 1997:

20

2

1 1ln ln ln ln ln ln ln

2 2

1ln ln (13)

2

j j ij j i Y YY jY jj j i j

tt t jt j d Dt Dy j tj j

C W W W Y Y W Y

t t W t D Dt Dy W D

β β β β β β

β β β γ γ γ ε

= + + + + +

+ + + + + + + +

∑ ∑ ∑ ∑

∑ ∑

It is usual to impose some restrictions to the functional form to be well behaved, as was expressed above. This is to ensure homogeneity of degree 1 in prices and concavity:

∑ =j

j 1β

∑ =j

jt 0β

0=Yβ

0=YYβ

0ln =∑j

j DW

∑ =j

jY 0β

0== ∑∑∑∑j

ijii

jij

ββ

∑ =j

jY 0β

To avoid efficiency complications for these extended forms it is useful to use duality theorems:

ln

ln ln (15)ln

j jj j j ij j jy jt jD

ijj j

W WC CX S W Y t D

W W C Cβ β β β β

∂ ∂= = = = + + + +

∂ ∂ ∑

Where Sj stands for the cost share of factor j. Now estimation of parameters is straightforward as:

75

ln ln (16)j j ij j iy it iD tij

S W Y t Dβ β β β β ε= + + + + +∑

Where 1=∑

jjS

Trend coefficients are used as measures for non-neutral technical change. If coefficient appears to be positive in equation for share j, it can be said that technology is j-using or j-intensive. Alternatively, if coefficient is negative, then there is j-saving technical change. In this context, trends appear to be really useful: Following Lekakis (1994) it is easy to prove that the rate of cost diminution is obtained differentiating (13) with respect to time accordingly with Chambers (1994): This equation is well behaved if it is concave in prices and monotollically increasing, but

Chambers (1994) shows that if there are no concave inputs the conditions for cost minimization will not be affected. To prove these requirements, it is mandatory for monotonicity that “fitted shares all be positive”, and for strict quasiconcavity in prices that “n x n matrix of substitution elasticities must be negative semidefinite” (Berndt, 1993). The equation works as a continuous estimation for Total Factor Productivity, a concept that results depending on time since actions on land and environment are intertemporal. This result is going to be used for empirical estimation of equation (12) above. Equations from (16) and (17) can be estimated dropping out one of the share equations and taking the price of the dropped input as numeraire on the rest of share estimations. The estimation method is usually ISUR (Iterative Seemingly Unrelated Equations) as shown in Berndt (1991). After estimating equations, it is necessary to estimate elasticities of substitution to evaluate the way factors substituted through time period. Following Mao and Koo (1997) and Frondell (2001), Allen partial elasticities of substitution, Morishima elasticities of substitution, changes in factor shares caused by changes in relative prices, and crossed price elasticities can be calculated. For this paper, Morishima elasticities of substitution are estimated with their standard errors. Frondell (1999) suggests the use of Morishima elasticity of substitution (MES) as a more exact measure for substitution. MES measures percentage change in the ratio of input j to input i when the price of input i varies and all other input prices are held constant. It is calculated as:

2

2

ln(18)

ln

j

i ji j jj ji j i jj j jiji ji jj

i j j j i j

XW C W C S S S SX

MES AES AESW C C S S S

β β

∂ + + − = = − = − = −∂

Two inputs are MES-substitutes if MES>0 and MES-complements if MES<0. This is an asymmetric elasticity of substitution because of its composition. MES is appropriate if

)17(ln

DWTtC

Dtj

jjtttt γβββ +++=∂

∂ ∑

76

technology is homothetic, and there is some evidence to show that it captures the net substitution effects if scale effects are ignored. AES and MES are equal if production structure is CES or Cobb-Douglas, and in the case of two inputs. For the cases of Translog and Leontieff functional forms, MES and AES are different. Even more, it could happen that two inputs that are AES-complements might become MES-substitutes. MES is the base to calculate effects of factor price changes on relative cost shares with a little transformation from Mao and Koo (1997) and Frondell (1999). Mao and Koo reference this measure as coming from Huang (1991):

ln

1 (19)ln

i i

j jji ji

i

W XW X

MESW

φ

∂

= = −∂

A relative cost share is increasing if it is less than one, alternatively if relative cost share is greater than one it is decreasing. Another important fact to have into account is that standard errors for AES, cross-price, and MES elasticities of substitution can be calculated asymptotically under the assumptions of constant cost shares and equality with their own estimated values (Frondell, 1999; Binswanger, 1974).

EMPIRICAL IMPLEMENTATION OF THE MODEL: Data were taken mainly from FEDEARROZ and Colombian Agriculture Ministry. Cost shares, input prices, and total costs for production of one hectare of rice were obtained from FEDEARROZ cost matrices for 1981:1 to 2001:2 period in a semiannual base for mechanized rice and annual base for Mechanized Rain-fed Rice. All values were deflected using Producer Price Index (June 1999=100) taken from Agriculture Ministry of Colombia. Data on input quantities were selected as the average inputs used for production of one hectare of rice for each production system, also taken from FEDEARROZ surveys. A trend starting 1981:1 was used to take into account technical changes along time, and a dummy variable was introduced into the cost function to control for the change of seeds starting 1997:2. Finally, data on production were taken from “Arroz en Colombia, 1981-2001” (FEDEARROZ, 2001). Five inputs were selected for the construction of the cost function and the productivity study. As it is common in agricultural studies capital (K) and labor (L) were used for econometric estimations. Capital prices were approached by the price of machinery use for production of one hectare during a semester. Labor prices are the cost of journals (cost of non-qualified labor) along the semester for one hectare. Capital and labor quantities were measured as the total use time of machinery and non-qualified labor along the semester for one hectare of rice production, respectively. It is to mention that water quantities and prices are included into capital for the estimation of Mechanized rice. Estimations with this input as separated from capital showed no significance for water prices and coefficients tended to be statistically unstable. To measure agrochemicals

2 2var( ) var( ) / var( ) / 2cov( , ) /( ) (20)ji ji i ii j ji ii j iMES S S S Sβ β β β= + −

77

behavior, data on fertilizers (F) and other chemicals (Q) were used. Other chemical (Q) data included herbicides, insecticides, and fungi control chemicals. Prices from fertilizers and from a bundle of chemicals were used for cost function estimation. To measure use of chemicals, effective active ingredient in equivalent units during a semester for a hectare was used for both systems of production. Finally, the last input used was land (R), its prices were measured as the price of one hectare rental for the productive time line of each semester. It is important to mention that taxes and other administrative costs were excluded from total cost in order to obtain results relating only to production activity. All prices and costs are calculated in constant thousands of pesos of 1999. Because of different production technology, geographical zones, times of harvest, and mainly because of lack of data, Mechanized Rain-fed Rice analysis is to be made separated from the one for mechanized rice. The estimation procedure of the cost function and the five share equations for both production systems was Iterative Seemingly Unrelated Regression (ISUR) from Eviews 4.0. In order to avoid estimation problems, one of the five share equations was dropped from the system, and its price was taken as numeraire. After this transformation, the final resulting system is cost and share equations with an error term as was expressed above. Restrictions for concavity and homogeneity imposed above are kept for the estimation. The relationships between estimated parameters allow the calculus of the entire dropped equation mainly based on restrictions imposed. Mechanized Rain-fed Rice: The estimation results for the case of Mechanized Rain-fed Rice input share equations are listed at Table 5.1. Most commonly, the cost function is not estimated, but for this case it was important to calculate it because it is the base for the rate of cost diminution equation that is used for productivity analysis. Relevant results from cost equation will be exposed later at the productivity analysis. As was mentioned before, data on prices and cost shares were taken from 1981 to 2001 anually. Because of space problems, all tables are made to be read in columns. For example, coefficients for capital demand function are to be read on the first column of Table 5.1.

78

Table 5.1: Parameter Estimates of the Translog input Shares for Colombian

Mechanized Rain-fed Rice 1981-2001

Cost Share of K L F Q R Constant 0,4587* 0,1012* 0,1649* 0,1614* 0,1138 (0,0205) (0,0133) (0,0120) (0,0198) Price of K 0,0871* (Symmetric) (0,0109) L -0,0293* 0,0233*** (0,0109) (0,0125) F -0,0607* -0,0085* 0,1056* (0,0024) (0,0028) (0,0013) Q 0,0262* 0,0226* -0,0238* -0,0111* (0,0075) (0,0051) (0,0013) (0,0106) R -0,0233 -0,0082 -0,0126 -0,0140 0,0634 Trend -0,0011* -0,0012* 0,0004* 0,0017* 0,0002 (0,0003) (0,0002) (0,0002) (0,0004) Dummy 0,0049 0,0011 -0,0047** 0,0014 -0,0028 (0,0034) (0,0021) (0,0023) (0,0032) Y 0,0022 0,0073 -0,0097* 0,0113 -0,0111 (0,0093) (0,0064) (0,0058) (0,0090)

NOTE: Standard errors in parenthesis. (*) denotes parameter statistically significant at 1%

level, (**) at 5% level, and (***) at 10% level. These results were calculated imposing the aforementioned restrictions for concavity and first-degree homogeneity in prices. The results indicate that concavity conditions are accomplishing, and that after proving Cobb-Douglas hypothesis, translog cost functional form fits quite fine the data. Only the dummy associated to fertilizers resulted statistically significant, indicating that the use of the new seed had an influence on their use, not the same for the rest of factors. Estimation gives some indication of capital and labor-saving technical change for the period. The significant time trend variable at any equation means that at constant factor prices, the factor shares would have changed, implying non-neutral technical change during the period. On the other hand, there are agrochemicals and land-using technical changes. This fact can be seen through the associated coefficients for trend variable, which are statistically significant and positive. This shows that at least for the case of Mechanized Rain-fed Rice, some more pesticides, herbicides, and fungi control chemicals have been being used. Results of the elasticities of substitution for the case of Mechanized Rain-fed Rice are exposed at Table 5.2. All values are calculated for the second semester of 2002, but there is going to be some mention about how elasticities evolved through time. This is to

79

evaluate how elasticities are behaving now to take policy measures. Elasticities through the period and for the whole period were also calculated.

Table 5.2 Morishima Elasticities of Substitution for Colombian Mechanized Rain-fed

Rice 1981-2001

K L F Q R K 0,6713 0,0099 1,0989 -0,0393 (0,1415) (0,0857) (0,047053)

L 0,4701 0,0561 1,2909 -0,0762 (0,1483) (0,1215) (0,0813)

F 0,3095 0,6710 0,8467 -0,0864 (0,0190) (0,0304) (0,0108)

Q 0,8871 0,8173 0,0695 -0,0330 (0,0486) (0,1303) (0,0866)

R 0,4354 0,6127 -0,0447 0,8209

Note: Standard Errors in parentheses

Table 5.3 Change in Relative Cost Shares in Response to a Change in Relative Prices for Colombian Mechanized Rain-fed Rice: 1981-2001

K L F Q R

K 0,3287 0,9901 -0,0989 1,0393 L 0,5299 0,9439 -0,2909 1,0762 F 0,6905 0,3290 0,1533 1,0864 Q 0,1129 0,1827 0,9305 0,1033 R 0,5646 0,3873 1,0447 0,1791

Note: Data shall be read in columns. For the case of upland rice, capital becomes MES-substitute for all inputs, including fertilizers, which were complements for Allen and cross price elasticities of substitution. The highest susbstitution relationship for the case of capital appears to be that one with respect to chemical prices: A 1% increase on the price of chemicals would lead to a 0.88% change on the ratio of capital quantities with respect to chemicals. It is important to point out that the weakest MES-substitution relationship for capital is that one with respect to fertilizers (0.30). Labor is also MES-substitute with respect to all other inputs. Relationships with capital (0.67), fertilizers (0.67), and chemicals (0.81) are strong. As for fertilizers, there is a MES-complement relationship with land (-0.045), and they are substitutes for all other inputs. It is important to point out that all those relationships are not so strong since all values are around zero. On the other hand, it is clear that chemicals have the strongest MES-substitution relationships, values with respect to prices of capital and labor are over unity. For example, a 1% increase of the price of labor leads to a 1.29% increase in the ratio of chemicals with respect to capital. The weakest relationship for chemicals occurs with land, but in some way the value is high (0.82). Finally, land is complement for all other inputs, but just as in the case of fertilizers, relationship is not so strong with any of the inputs.

80

The measure of changes of relative input shares of input j and i with respect to changes in prices of factor j, show the following results. A change in the price of any of the inputs would increase the share of capital, the “weakest” relationship occurs again with respect to fertilizers prices. For labor, things are quite similar to capital, an increase in the price of any of the inputs increases the share of this factor. For the case of fertilizers, things change a little, an increase on the price of land decreases the cost share of fertilizers, which sounds kind of logical. With respect to the prices of all other inputs, fertilizers share would increase. The case of chemicals shows that an increase on the price of capital or labor would induce a high increase on the share of that factor. Finally, share of land decreases with increases of capital, labor, and fertilizers, and increases with an increase on the price of chemicals. Table 5.3 Now, it is time to turn to the productivity analysis for each of the inputs and the total output. As can be seen in equation (12), it is necessary to estimate the rate of cost diminution (11) from the total Translog cost function. For these cases, and because of the model specification, it is possible to express an explicit expression for the rate of cost diminution (17) that works out as an approximation for Total Factor Productivity. Table 5.4. Differentiation of Cost Function for rain-fed rice drops the result on equation (17). For Mechanized Rain-fed Rice from 1981 until 1990, almost all of the values of rate of cost diminution are negative and around zero, the mean is –0.024. For this period trend is always positive. From 1990 the positive trend continues and on average value for rate of cost diminution turns to be positive in 0.01355. See Graph 5.1 for more information. This means that output productivity has been lowering through time, a positive value means that cost is not lowering. Even more, it is easy to see that in the long run there exists an increasing trend.

Graph 5.1 Rate of Cost Diminution for Colombian Mechanized Rain-fed Rice: 1981-

2001

-.06

-.04

-.02

.00

.02

.04

.06

.08

82 84 86 88 90 92 94 96 98 00

Rate of Cost Diminution: 1981-2001

81

Table 5.4 Coefficients for rate of cost diminution, 1981-2001

Coefficient Std. Error t-Statistic

tβ -0.005961 0.001314 -4.535705

ttβ 4.83E-05 1.37E-05 3.531172

tkβ -0.001110 0.000304 -3.658769

tlβ -0.001200 0.000212 -5.669723

tfβ 0.000406 0.000150 2.708037

tqβ 0.001669 0.000391 4.272302

tfedβ -0.001095 0.000223 -4.923057

trβ 0.00236

All coefficients resulted significant at 1% level Using rate of cost diminution, estimation of individual input productivity indexes is straightforward. These measures are composite indicators of efficient use for each input. It is important to point out that these indexes are very complete, because they have into account productivity changes in the whole production, and changes of each input with respect to all other inputs. Applying equation (12) for the case of fertilizers drops interesting results. Base period for Mechanized Rain-fed Rice inputs productivity indexes estimation was 1982=100, first year of the sample. Productivity index for fertilizers for the 1982-2002 period shows a constant decreasing trend. Productivity for this input never surpasses the initial level of 100. Furthermore, the index has a minimum of -86,2 for 2000, and an average of 7,44 approximately for the whole sample (Graph 5.2). This shows that fertilizers are not productive inputs for the case of this production system. It is important to remember at this point that cost function showed that production is fertilizers-intensive in this case. Substitution of these inputs for productive ones or for inputs that show a positive trend towards productivity is mandatory.

82

Graph 5.2: Productivity index for Fertilizers: 1981-2001

Line shows base level of 1982=100 For chemicals, it is the same old story as with fertilizers, there is a downward trend in their productivity index. Furthermore, trend is very similar to the one for fertilizers over the initial periods, and then it turns somehow different but as negative and decreasing as for the latter inputs. On average, productivity index for chemicals (Graph 5.3) is lower than the one for fertilizers: 5,01, with a minimum of –72,02 and a maximum of 100 for 1982. Again, time trend for technical change is positive, there is an intensive use for this inputs in this production system. Conclusion is similar to fertilizers, it would be ideal to fin substitute inputs with a more productive behavior.

Graph 5.3: Productivity index for Chemicals: 1981-2001

Line shows base level of 1982=100

-120

-80

-40

0

40

80

120

82 84 86 88 90 92 94 96 98 00

Productivity Index for Fertilizers: 1982-2001

-80

-40

0

40

80

120

82 84 86 88 90 92 94 96 98 00

Productivity Index for Chemicals: 1982-2001

83

Capital should be an ideal substitute for chemicals and fertilizers, but story is similar to those inputs. There is a constant decreasing trend for this input, that reached a minimum of –341,87 for 1999, having an average value of –38,83 for the whole period. In spite of these facts, capital seems to be starting a recovery in its productivity index during 2001. These declines in productivity should be explained because of obsolescence of machinery all along the zones where rain-fed rice is sowed. It is important to remember that production in this system is capital-saving, which could in turn explain a portion of the productivity index trend. Capital is substitute of all inputs, and relationship with chemicals and fertilizers is kind of strong, but as long as it low productivity index levels, it does not matter at all. On the other hand, labor shows a U-shaped trend (Graph 5.4) for productivity index over the period of analysis. Furthermore, for the initial periods productivity index is positive reaching a 173,95 level during 1983. Starting this year, trend turns decreasing with a minimum of –216 and then productivity index recovers with a positive trend reaching a final level of 99.11 in 2001. This input may continue increasing its level to turn into a productive input. There is a technical bias towards saving this input, trend time coefficient sign is negative. Labor appears as an option to substitute agrochemicals in this production system because it is turning productive through time, and because results from MES show substitution for these inputs. Furthermore, it could even substitute capital.

Graph 5.4 Productivity index for Labor: 1981-2001 Finally, land appears as another unproductive factor for this production system, and it is not an alternative for substitution among the rest of factors. Rice production under rain-fed scheme is land-intensive. As conclusion, three unproductive inputs have positive and statistically significant time trends indicating that production is factor-intensive: Fertilizers, chemicals, and land. Land appears to be the most possible factor for substitution to recover productivity under rain-fed production.

-300

-200

-100

0

100

200

82 84 86 88 90 92 94 96 98 00

Productivity Index for Labor: 1982-2001

84

Estimation for mechanized rice: For the case of mechanized rice, data start from 1981:1 to 2001:2 on a semiannual base. Table 5.5 y 5.6 shows results.

Table 5.5: Parameter Estimates of the Translog input Shares for Colombian Mechanized Rice 1981:1 – 2001:2

Cost Share of

K L F Q R Constant 0,3590* 0,1039* 0,1684* 0,1879* 0,1807 (0,0129) (0,0079) (0,0057) (0,0194) Price of K 0,1425* (Symmetric) (0,0154) L 0,0023 0,0377* (0,0118) (0,0125) F -0,0496* -0,0157* 0,1176* (0,0019) (0,0020) (0,0014) Q -0,0390* -0,0010 -0,0261* 0,1041* (0,0113) (0,0065) (0,0010) (0,0127) R -0,0562 -0,0234 -0,0262 -0,0380 0,1303 Trend -0,0012* -0,0007* -0,0001*** 0,0022* -0,0002 (0,0002) (0,0002) (0,0001) (0,0003) Dummy 0,0406* 0,0068 -0,0090* -0,0269* -0,0115 (0,0070) (0,0044) (0,0023) (0,0092) Y 0,0136** 0,0233* -0,0069** -0,0206** -0,0094 (0,0075) (0,0045) (0,0033) (0,0113)

Coefficients (*) denote statistical significance at 1%, (**) denotes significance at 5%, and

(***) at 10%. The same as the case for Mechanized Rain-fed Rice, tests on concavity and Cobb-Douglas hypothesis were done. Both tests showed that data fitted quite well translog functional form. All dummy variables resulted significant at 1% level with the exception of the one related to labor share. This shows that introduction of Fedearroz-50 seed variety influenced input behavior. On the other hand, trends indicate that there was a bias towards the use of chemicals and less intensity for the other four factors. This fact is not consistent with the case of Mechanized Rain-fed Rice, where fertilizer and land-using technical changes occurred. Elasticities of substitution may shed some light on these differences and the way factors interact.

85

Table 5.6: Morishima Elasticities of Substitution for Colombian Mechanized Rice

1981-2001

K L F Q R K 0,7102 -0,0846 0,4522 0,0099 (0,1457) (0,1232) (0,088597)

L 0,6156 -0,0676 0,5547 -0,0151 (0,1239) (0,1047) (0,0836)

F 0,1988 0,5778 0,3527 -0,0416 (0,0164) (0,0218) (0,0096)

Q 0,4334 0,6995 -0,0541 0,0090 (0,0806) (0,1149) (0,1016)

R 0,2386 0,5545 -0,1115 0,3197 Note: Standard Errors in parentheses

Table 5.7: Change in Relative Cost Shares in Response to a Change in Relative Prices for Colombian Mechanized Rice 1981-2001

K L F Q R

K 0,2898 1,0846 0,5478 0,9901 L 0,3844 1,0676 0,4453 1,0151 F 0,8012 0,4222 0,6473 1,0416 Q 0,5666 0,3005 1,0541 0,9910 R 0,7614 0,4455 1,1115 0,6803

MES on Table 5.6 drops some interesting findings for all factor substitution patterns. Capital is MES-substitute for all factors and, as waited, strong relationships occur with respect to labor (0.61) and chemicals (0.43). As for labor, estimations show that it is MES-substitutes for all factors, just as capital. For this factor, there are strong patterns with capital (0.71) and chemicals (0.69), but in some extent it happens the same with respect to fertilizers and land. Surprisingly, fertilizers appear to be MES-complements for all factors, but again values are around zero with the exception of land. Note that these patterns are different from those coming from Mechanized Rain-fed Rice analysis, where fertilizers were complements only for land. Another interesting issue lies on the fact that chemicals are stronger MES-substitutes for labor (0.55) than for capital (0.45). As it is evident, all substitution relationships for chemicals are somehow strong. Finally, land appears to be MES-substitute for capital and land, and complement for fertilizers. Mechanized Rain-fed Rice case showed that land was AES-complement for all factors as can be seen above. Table 5.7 and 5.8. For the case of mechanized rice there is more data (from 1991:2 to 2001:2), so estimation of productivity measures seems to be better. Rate of cost diminution estimation shows that all values for the period are negative. On average this measure was –0,13 with a maximum value of nearly -0,08 and a minimum of –0.19. In spite of the positive trend, standard deviation is not important (0,02), which shows that in absolute terms rate of cost diminution has not changed significantly along the period. Anyway, it is important to note that total productivity is lowering. Graph 5.5 shows this result in detail.

86

Table 5.8: Coefficients for rate of cost diminution: 1981:01-2001:02

Coefficient Std. Error t-Statistic

tβ 0.000262 0.000665 0.394471

ttβ 2.79E-05 1.53E-05 1.825667

tkβ -0.001168 0.000212 -5.518568

tlβ -0.000748 0.000155 -4.831234

tfβ -0.000137 7.70E-05 -1.776893

tqβ 0.002214 0.000292 7.580034

tfedβ -0.001414 0.000556 -2.540661

trβ -0.000161

All coefficients resulted significant at 1% level

Graph 5.5: Rate of Cost Diminution for Colombian Mechanized Rice: 1981-2001 Individual input productivity indexes show interesting results if compared to those coming from rain-fed rice. Graph 5.6. In this case, fertilizers show a volatile series with some indications of a positive trend starting in 1997:02 and then a turn down from 1999:02 to 2001:02. As a general result, productivity index for this input had a mean of 116,02, which shows this one as a productive input. Minimum for this series is 54,85 and maximum is 245.41 doubling productivity from the base period. It is important to remember that production under mechanized rice is fertilizer-saving, which is different to the case of rain-fed rice, as estimations showed. Anyway, this is not showing that fertilizers are not damaging surrounding environment.

-.20

-.18

-.16

-.14

-.12

-.10

-.08

-.06

82 84 86 88 90 92 94 96 98 00

Rate of Cost Diminution: 1981:01-2001:02

87

Graph 5.6: Productivity Index and Cost Share for Fertilizers for Colombian Mechanized Rice: 1982-2001 Chemicals seem to be productive factors just as fertilizers. As it happened with estimations for rain-fed rice, behavior of fertilizers and chemicals follow a very similar pattern along the period. On average, productivity index for chemicals was 114,7 with a minimum of 56,09 and a maximum of 239,55. Unlike fertilizers, and as in rain-fed rice, production is chemical-intensive in this production system. Graph 5.7. Graph 5.7: Productivity Index and Cost Share for Chemicals for Colombian Mechanized Rice: 1982-2001 Not surprisingly because of obsolescence of machinery, capital shows a U-shaped pattern along the period 1981:02-2001:02. Graph 5.8. Productivity index starts with values around base of 100, but then it turns to be unproductive from 1985:2 to reach a minimum of 71,95

50

100

150

200

250

82 84 86 88 90 92 94 96 98 00

Productivity Index for Fertilizers: 1982:02-2001:02

50

100

150

200

250

82 84 86 88 90 92 94 96 98 00

Productivity Index for Chemicals: 1981:02-2001:02

88

in 1989:2. Starting the latter period, productivity index fluctuates between 80 and 90 to finally turn increasing from 1996:2 with a little down in 1998:2, to start going down from 2000:1 to the end of the sample. Time trend in the estimation showed that this production system is capital-saving, the same as rain-fed rice. On the other hand, land appears to be the most productive input for this production system. Only in one year the measure is under the base year level of 100. Furthermore, it is always fluctuating between 100 and 200, appearing to be a very volatile series. On average, productivity for this input is 138,01, indicating a high productivity along he period of analysis. Furthermore, the maximum for this series reaches 223,80, and the minimum is about 70,39. As capital, rice production under mechanized system is labor-saving. This fact has interesting implications in terms of environmental issues. It is now known that chemicals are being used intensively by rice production under this system, and this does not imply that those inputs are not affecting environment. Furthermore, high productivity may incentive use for chemicals because of the excellent results to control pests and fungi. Even more, labor is a substitute for chemicals, and the best option to reduce use of this factor. Graph 5.8. Productivity Index and Cost Share for Labor for Colombian Mechanized Rice: 1982-2001 Finally, land appears to be the less productive input. It has a mean of 110,30, a minimum value of 65,5, and a maximum of 161,01, in its productivity index. MES estimation showed that labor is a substitute for land, so it would be a great opportunity to substitute the former for the latter to stop the loss of productivity showed in the upward trend of the rate of cost diminution.

40

80

120

160

200

240

82 84 86 88 90 92 94 96 98 00

Productivity Index for Labor: 1981:02-2001:02

89

CONCLUSIONS It is clear now that production under rain-fed scheme is internally different to that under mechanized rice. Water, which is the main difference between the related production systems, may play an important role in determining not only technical biases but also input and total product productivity. For the case of rain-fed rice production, empirical estimation of a translog cost function and input shares controlling for introduction of variety Fedearroz-50 showed that:

1. Production is capital and labor saving, meaning that use for those factors is decreasing through time in favor of the use for agrochemicals, which are used intensively.

2. Introduction of Fedearroz-50 only affected use for fertilizers in a negative way. This means that use for fertilizers was reduced with the introduction of this variety.

3. An opportunity to reduce use of agrochemicals appears to be the use of labor, which is a substitute for fertilizers and chemicals. Labor is the only factor that is increasing productivity through the period of analysis. Integrated Pest Management may encourage this substitution, and incentive producers to reduce costs via

For the case of irrigated mechanized rice, estimation showed that:

1. Production is only chemical-using and all other factors-saving. This means that use for chemicals has been increasing through time while uses for all other factors has been lowering.

2. Introduction of Fedearroz-50 increased use for capital and reduced use for fertilizers and chemicals. The objective of that variety was accomplished in terms of environmental benefits reducing use of agrochemicals.

3. Again, an opportunity to reduce use for agrochemicals appears to be use for labor. Labor is again a substitute for chemicals, which appear to be improductive. Suggestions are the same that those for mechanized rain-fed rice.

A comparison among production systems shows that irrigated mechanized rice is much more productive than rain-fed rice. A partial explanation for this situation may be the availability of water for production of the former production system that makes it independent from randomness of rain. This may explain also why inputs are more productive in irrigated rice in spite of the similar conditions of factor availability among production zones. It is also clear that use of agrochemicals as a set is intensive in both production systems, showing that some policies such as permits to use generic products, and a lack of control is making rice production a potential negative externality to surrounding environment. Policies to increase the use of labor as a substitute for fertilizers and chemicals may overcome as an opportunity to reduce risk of environmental damage, and to reactivate economy all along Colombian rice production zones. It is also clear that a more active role from The Colombian Ministry of Environment and even Fedearroz is required to induce the rice producers to cleaner practices. Implementation of Integrated Pest Management (IPM) may be an alternative that results

90

consistent with the objective of increasing level of labor used for rice production substituting agrochemicals. Actions from government and private entities could be focused at producer and technical assistance level to create a culture around the use of IPM.

91

BIBLIOGRAPHY Abreau y E. Ablan, 25 años de cambios alimentarios en Venezuela (1974-1994), Fundación Polar, Venezuela. GRADE, Evaluación del sistema de sobretasas a la importación de trigo y harina de trigo en Perú, Mimeo, Lima, 1993. Antle, J.; Pingali, P.: “Pesticides, Productivity, and Farmer Health: A Philippine Case Study”. American Journal of Agricultural Economics 76, Aug. 1994. Ball, E.: “Output, Input, and Productivity Measurement in US Agriculture, 1948-1979”, American Journal of Agricultural Economics 57(3), 1985. Ball, E; Färe, R.; Grosskopf, S.; Nehring, R.: “Productivity of US Agricultural Sector: The case of Undesirable Outputs” in: National Bureau of Economic Research: “New Developments in productivity Analysis”, Edited by Hulten, C. et Al. The University of Chicago Press, 2001. Banco de la República: www.banrep.gov.co Binswanger, H.: “The Measurement of Technical Change Biases With Many Factors of Production”. The American Economic Review 64(6), Dec. 1974. pp. 964-976. BERNDT, E.: “The Practice of Econometrics: Classic and Contemporary”, Addison-Wesley Publishing Company, 1991. Caves, D.; Christensen, L.; Diewert E.: “The economic Theory of index numbers and the Measurement of Input, Output, and Productivity”. Econometrica 50(6), Nov. 1982. pp. 1392-1414. Coelli T.: “Measurement of Total Factor Productivity Growth and Biases and Technological Change in Western Australian Agriculture”. Journal of Applied Econometrics 11(1), Jan-Feb 1996. pp. 77-91. Chambers, R.: “Applied Production Analysis: A Dual Approach”. Cambridge University Press, London, 1994. Chambers, R.; Färe, R; Grosskopf S.: “Efficiency, Quantity Indexes, And Productivity Indexes: A Synthesis”. Bulletin Of Economic research 64(1), 1994. pp.3-21. Chambers, R.; Lichtenberg, E.: “Simple Econometrics of Pesticide Productivity”. American Journal of Agricultural Economics 76, Aug. 1994. pp. 407-17. Christensen, L.; Jorgenson, D.; Lau L.: “Transcendental Logarithmic Utility Function”. The American Economic Review 65(3), Jun. 1975. pp. 367-83. DANE, Encuesta Nacional de Hogares. Derek Byerlee, Ventaja comparativa y política agrícola para la producción de trigo en Ecuador, CIMMYT, México, 1976.

92

Federación Nacional de Arroceros, FEDEARROZ: “Arroz en Colombia: 1981:2001”, Bogotá D.C, 2001. Federación Nacional de Arroceros, FEDEARROZ: “I y II Censo Nacional Arrocero”, Bogotá D.C, 2001. Federación Nacional de Arroceros, SAC y Ministerio del Medio Ambiente, Guía Ambiental para el Cultivo del Arroz. 2002 Federación Nacional de Arroceros, Guía de reconocimiento y Manejo de las Principales Enfermedades del Arroz. 2000. Federación Nacional de Arroceros - FEDEARROZ, Tecnología para el arroz en los 90’s. 1999. Frondell, M: “Interpreting Allen, Morishima and Technical Elasticities of substitution: A theoretical and empirical comparison”, Discussion Paper No. 292, Universität Heidelberg, 1999. Fundación Polar Venezuela, Fedearroz Colombia, Fondo Latinoamericano de Arroz de Riego, Centro Internacional de Agricultura Tropical. Manejo Integrado de Plagas, 1997. Good, D.; Nadiri, M.; Sickles, R.: “Index Number and Factor Demand Approaches to the Estimation of Productivity” in: “Handbook of Applied Econometrics-Volume II: Microeconomics”. Edited by: M.H. Pesaran and P. Schmidt, 1997. Griliches Zvi: “Hybrid Corn: An Exploration In The Economics Of Technical Change”. Econometrica 25(4), Oct. 1957. pp. 501-22. Induarroz Cortolima, Convenio de concertación para una producción más limpia con el subsector de molinería de arroz, 2000. Instituto de Investigaciones del Arroz. Guía para el trabajo de campo en el manejo integrado de plagas. Cuba 1996. Integrated Assessment of Trade Liberalization and Trade-Related Policies. UNEP 2002 International Food Policy Research Institute, “Agricultural Protection in OECD Countries: its costs to less developed countries”, Research Report No 21, 1980. Kawagoe, T.; Otsuka, K.; Hayami Yujiro: “Induced Bias of Technical Change in Agriculture: The United States and Japan, 1880-1980”. The Journal of Political Economy 94(3), Jun. 1986. pp 523-544. Lekakis, Joseph: “The Productivity of Agrochemicals in Greece” in: “Agriculture, Trade and the Environment: The Impact of Liberalization on Sustainable Development”, Edgar Elgar Publishing Ltd.,1994, pp.208-223. Lichtenberg, E.; K. Thi Ngyen: “Pesticide Productivity in Green Revolution Rice Production: A Case Study of Vietnam”. Selected Paper for presentation at the 2001 AAEA Annual Meetings in Chicago, Illinois, USA, 2001.

93

Mao W.; Won, W.: “Regional Factor Demand in US Wheat Production”. Selected Paper for presentation at the 1997 AAEA Annual Meetings, in Toronto, Canada, 1997. María Helena Ramírez Tolosa, ¿Porqué no se deben reducir los aranceles para la importación de trigo en Colombia?, Comité de Seguimiento a la Competitividad de Arroz de Colombia, Informe Final, Abril 2002. Martínez Covaleda, Héctor, “El efecto sustitución de consumo entre trigo y arroz en Colombia: una medición econométrica”, Revista Induarroz, No.3, Bogotá, Agosto 1999. Agriculture Ministry and Desarrollo Rural: “Anuario Estadístico 2001, Bogotá, D.C, 2001. Ministerio de Agricultura – IICA . Convenio regional para la competitividad de la cadena del arroz – Molinería. 2000 Ministerio de Medio Ambiente . Lineamientos de política ambiental para el uso y manejo de plaguicidas en Colombia, 1998 Ministerio de Medio Ambiente - Corpoica. Diagnóstico de la situación ambiental del cultivo del arroz en Colombia. 1997. Ministerio de Medio Ambiente – SENA. Manual técnico para el manejo integral de cuencas hidrográficas. 1997. Ministerio de Medio Ambiente. Política Nacional de Producción Más Limpia, 1995. Ministerio de Medio Ambiente-United Nations Envorinmental Programme (UNEP): “Informe Nacional Sobre Reducción del Escurrimiento de Plaguicidas al Mar caribe”, Proyecto PNUMA/UCR/CAR-GEF, República de Colombia, 2001. Ley 99 de 1993 Nghiep, L. T : “The Structure and changes of Technology in Prewar Japanese Agriculture”. American Journal of Agricultural Economics 61(4), Nov. 1979. pp.687-693. PNUMA, Manual de referencia para la evaluación integrada de políticas relacionadas con el comercio. 2002 PNUMA – Ministerio del Medio Ambiente. Reducción del escurrimiento de plaguicidas al Mar Caribe, 2001. Srivastava, U; Heady E.: “Technological Change and Relative Factor Zares in Indian Agriculture: An Empirical Análisis”. American Journal of Agricultural Economics 55(3), Aug. 1953. pp. 509-14. Törnqvist, L: “The Bank of Finland’s Consumption Price Index” Bank Of Finland Monthly Bulletin 10, 1936, pp. 1-8. Valdez, Alberto y Seitz, Joachim, “Distortions in world food markets in the wake of GATT: Evidence and policy Implications” In World development.

94

World Bank, “The Costs of protectionism to Developing Countries: an analysis for selected agricultural products”, Staff Working Paper No. 769, 1986.